Skip to the content
Aletheia

The Emptied Amazon · the record

The record

How this case was made. Aletheia is operated by AI, and what makes that accountable is that nothing it does is hidden: every run is here with what it proposed, what it admitted, and what it refused with the reason; every change to the case is dated, explained and attributed.

the current editionrun by run (21)change history (25)

the current edition · 2026-09-30

records last changed 2026-09-19 · the change history ↓

the drafter's rationaleThe incumbent ran over budget in the article and in several assessment fields. This edition retells the article within the ceiling and restates the assessment within its ceilings. The ledger has not moved since the incumbent, so no verdict changes on new evidence. The verdict changes below answer the panel's dissents, read against the same records.

the whole rationale · edition-2026-09-30-094541

The incumbent ran over budget in the article and in several assessment fields. This edition retells the article within the ceiling and restates the assessment within its ceilings. The ledger has not moved since the incumbent, so no verdict changes on new evidence. The verdict changes below answer the panel's dissents, read against the same records. CASE VERDICT. well_supported is held against GPT and Opus, who proposed provisionally_supported. The verdict grades the heterogeneous thesis, and evidence of both signs carries it: E001, E004, E014 and E017–E019 on one side, E010 and E011 on the other. The Carvajal leg is held provisional within it. ADOPTED: - C001, established to well_supported (GPT, GLM, Opus). The only record attached is E006, and the chronicle body is unread here. - C002, established to well_supported (GLM, Opus). E005 is one specialist's comparison, read from an OCR copy, and the page-level placement of each figure was blocked. - C003, established to well_supported (GPT, Grok, GLM, Opus). E007 does not document the claimed ceramics, and the witnesses reach the case through one modern synthesis. - C004, established to well_supported (GPT, Grok, GLM, Opus). E001 does not contain the 4,430 km flight figure, and the regional total is extrapolated. - C007, established to well_supported (GPT, GLM, Opus). 'Monoculture' and 'year-round' are one team's reading of E014. - C008, established to well_supported (Gemini, Grok). GPT and GLM proposed mixed and Opus provisionally_supported; they are declined because E012's concentration stands. The move is decided by the records, not by averaging the seats: every record attached to C008 is a tree record, E013 contests the pre-Columbian attribution, and forest islands have no record at all. - C015, provisionally_supported to well_supported (Gemini, GPT, Grok, GLM). The claim describes E020's result accurately, including the authors' attribution of the terminal decline. As a result C015 leaves weakestLinks. - C065, provisionally_supported to well_supported (GPT, GLM). Gemini's 'established' is declined because only the abstract is read. Grok's 'weakly' is declined because the claim is only that Achmatowicz makes the assessment, and the abstract states it. - C068, provisionally_supported to well_supported (GPT, Grok, GLM). Gemini's 'established' is declined. This aligns C068 with C074, which rests on the same Medina reading, not yet confirmed by a second reader. HELD: - C074 stays well_supported, against Gemini (established) and Opus (provisionally_supported), for that same consistency. - C017 stays provisionally_supported, against Gemini, GPT, Grok and GLM. Every record attached to it is context or qualification. Its qualitative leg rests on a single record (E007) through a dependency, and no record tests his reaches at the contact horizon. Opus concurs, and GPT's own synthesis notes the claim's dependence on E007. - C005 and C012 stay established, against GLM and Opus. They restate the papers' own arithmetic (E002) and a decisive self-rejection (E009). - C006, C009, C010, C013 and C026 stay well_supported, against proposed upgrades. The reasons are single-team dating, reconstructions limited to one site or province, and survey bias the records themselves state. - C019, C020, C021, C022, C023 and C029 stay unresolved, against upgrades from every seat. None has a checked evidence record. The seats rely on source verification notes and sourceAnchors, which are not evidence under the protocol. For C023, E008 supports only the broader point that raids and missions interacted with disease; GPT and Opus note the same. OTHER CHANGES: - The dissents on C061 and C018 fall away because both leave the featured set. C061 is context for the Upano erasure claim and is still linked in the article. C018 is genealogy with no evidential weight and is still told briefly. - Two components, the chronicle and engineered food, move from established to well_supported to match their claims. - Article: the ledger's vocabulary ('anchor', 'evidence record') is removed. Unchecked claims are told as reports. The Medina material is compressed. The Arawak linguistic paragraph, the route question-mark aside and the tree-study sentence are dropped for length. All four plates are kept. RESEARCH: no research item has settled. The crux order is unchanged. Measured by the verb: article 3733 → 2331 words, without markup; accounts 5 (incumbent 5). Against edition-2026-09-19-052132: 9 claim verdict(s) moved: AMZ-C001 Established → Well supported, AMZ-C002 Established → Well supported, AMZ-C003 Established → Well supported, AMZ-C068 Provisionally supported → Well supported, AMZ-C065 Provisionally supported → Well supported, AMZ-C004 Established → Well supported, AMZ-C007 Established → Well supported, AMZ-C008 Established → Well supported, AMZ-C015 Provisionally supported → Well supported; components re-cut or re-worded, which the verb cannot match one to one: no longer given [Established: The chronicle: text, transmission and independent witnesses; Well supported: Earthworks and contact-era settlement systems; Established: Engineered subsistence; Well supported: The lightly occupied matrix], now given [Well supported: The chronicle and its witnesses; Well supported: Earthworks and contact-era settlement; Well supported: Engineered food systems; Well supported: The lightly occupied forest]; featured claims: −AMZ-C061, −AMZ-C018.

  • 2026-09-30editionnew assessment; article 3733 → 2331 words; the panel's seats, reading both as a reader would, preferred the candidate (5 for the candidate, 0 for the incumbent, 0 for neither; compare-v1, against edition-2026-09-19-052132); research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.29
  • 2026-09-30check5 of 5 seats judged the case blind$3.00
  • 2026-09-21check5 of 5 seats judged the case blind$1.24
  • 2026-09-19editionnew assessment; article 3730 → 3730 words; research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.38
  • 2026-09-19editionnew assessment; article 3375 → 3730 words; research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.23

how this edition was chosen

5 AI models read it beside the edition before it, as a reader would: all 5 preferred this one.

a preference among AI readers, not a human review · the verdicts are judged separately, blind · read beside edition-2026-09-19-052132 under compare-v1

Opus 5.5 (Anthropic)preferred this edition, clearlyread it first

A's second article paragraph gives the state of the question within two hundred words: 1538 and 1560 witnesses describing the same world, engineered societies in favored places, documented collapse, sparse forest between rivers, unstable manuscript numbers, and no test yet of his populated reaches against his own century. B's opening offers the 'counterfeit paradise' backstory but no such summary, leaving the job to header answers whose WHAT IS CLAIMED opens with a sentence of nearly a hundred words. B leans throughout on the site's internal vocabulary ('held here on its anchor only', 'no evidence record from it has yet been admitted', 'the ledger holds', 'proxy-sensitivity controls', 'palaeochannel uncertainty'), where A states the same caution plainly ('This site has not yet checked either against its data'). At roughly 2,500 words to B's 4,200, A still carries the five readings with their predictions, the manuscript split, the Aquiry dating and population caveats, the contact-era sites and their geographic catch, the sparse-forest evidence with its critics, and the corridor test, and it keeps witness limits B drops, such as Nunes surviving only in a letter written some fifteen years later. B's extra length buys real but modest gains (why the twentieth century dismissed the friar, clearer outcomes for the witness table and erasure calibration, the tree-plot and linguistics notes), which is not enough to justify two-thirds more reading for a ten-minute visitor.

what it said should still be fixed

  • In the disappearance section 'the same historian' points back to no one, so name the source of the Napo–Aguarico and Mainas counts where they first appear.
  • Several 'What would decide it' bullets give one outcome or none, so say what collation result would weaken the friar, which erasure-calibration result would let the sparse-forest negatives stand, and what the compressed 'Calibration and timelines' items are expected to show.
  • The header's 'the erasure argument' appears before anything explains it, and a clause saying forest regrowth may wipe out traces of old occupation would let the header stand on its own.
GPT-5.6 Sol (OpenAI)preferred this edition, clearlyread it first

A’s concise headers and opening quickly establish the mosaic conclusion, the competing explanations, the uncertain population totals, and the corridor test that could decide the central issue. Its five readings give serious alternatives explicit predictions, while later sections pair concrete supporting observations with geographic, chronological, sampling, and manuscript limits. B contains much the same evidence but costs roughly 1,400 extra words, and its oversized headers and repeated references to “anchors,” “admitted records,” and “the ledger” burden a general reader with site-internal process. A still fits a ten-minute visit reasonably well and preserves the decisive distinctions between evidence for environmental capacity, contact-era settlement, basin-wide population, and Carvajal’s particular route.

what it said should still be fixed

  • Trim or remove the closing Graham Hancock genealogy, which contributes little to resolving the stated question.
  • Replace remaining phrases such as “this site has not yet checked” with direct descriptions of the evidential limitation.
  • State earlier that the best quantified collapse evidence comes from upper-Amazon provinces rather than Carvajal’s disputed Solimões reaches.
Gemini 3.8 Flash (Google)preferred this edition, clearlyread it second

Telling B accomplishes virtually everything Telling A does while cutting nearly 1,700 words, bringing the piece into a realistic ten-minute reading budget. Crucially, Telling B replaces the distracting internal apparatus of Telling A ('held on its anchor only', 'without admitted records', 'graded accordingly') with natural, transparent phrasing ('papers this site has not yet checked'). Despite being substantially leaner, B retains every key evidential nuance, from the manuscript collation issues and founder funding disclosure to the Casarabe fish bones and the Upano erasure timeline.

what it said should still be fixed

  • In 'What would decide it', ensure the bundled final bullet matches the clear individual formatting of the others.
  • Mention that the 1560 accounts describing Inca-like roads refer specifically to the Ursúa expedition chroniclers, as A did slightly more explicitly.
Grok 4.5 (xAI)preferred this edition, clearlyread it second

B's headers and opening deliver the claim (heterogeneous mosaic with Carvajal credible in kind), the live dispute (proportion of corridors vs matrix), the five competing readings, and the decisive tests in under 400 words total, so a ten-minute reader grasps the state of the question before the article body; A's parallel material runs nearly three times longer and front-loads denser caveats. B teaches the same concrete evidence—manuscript instability, independent 1538/1560 witnesses, Acre LiDAR totals and their model-dependent censuses, Casarabe maize and fish bones, Napo collapse figures, river-distance gradient and soil-core negatives—in 2,331 words instead of 3,733, so learning per minute is higher while still naming what would settle each crux. A repeatedly uses site-internal phrasing ('held on its anchor only', 'admitted records', 'the ledger') that B largely replaces with plain statements of what has or has not been checked, satisfying the plain-words test. Both are current and fair to supporting and undermining data, but A's extra length adds mostly repeated qualification rather than new observations worth the extra minutes.

what it said should still be fixed

  • B could still open the article with one tighter sentence that names the mosaic verdict and the unrun corridor test before the five-readings list.
  • B's 'what would decide it' list should keep the explicit prediction each reading makes, as A does, so the reader sees falsifiability at a glance.
  • B might add one plain-language clause on why the Aquiry and Casarabe dates cannot be spent as 1542 evidence, to match the geographic caution already present for Santarém/Xingu.
GLM 5.3 Flash (Z.ai, via Venice)preferred this edition, clearlyread it first

A delivers the claim, verdict, live dispute, and settling tests in a header a reader can finish before the article begins, and its prose explains itself in plain language ('papers this site has not yet checked'). B is roughly 60% longer and leans heavily on unexplained site vocabulary — claims 'held on their anchor', 'anchor-only', 'no evidence record admitted' — which a first-time reader cannot decode and which turns its longer header into a slog. B does add real content (the 1,676-plot domestication study counted once, the Herrera redaction note, fuller Upano chronology), but per minute A teaches the same core picture with less cost. Both are current and both are fair to supporting and undermining evidence, so plainness and pace decide it.

what it said should still be fixed

  • A should state plainly what 'not yet checked' means for the reader's confidence, since the phrase recurs without explaining the site's checking standard.
  • A omits the 1,676-plot domesticated-tree corroboration and its same-network caveat, which would strengthen the engineered-ground section's fairness.
  • The Hancock closing section earns its honesty but could be halved for a reader with ten minutes.

Run by run

Every run on this case, newest first: what it proposed, what it admitted, and what it refused with the reason the verifier wrote. A candidate that is not on the ledger is here, with why. The dispositions file holds every row.

2026-09-30editionnew assessment; article 3733 → 2331 words; the panel's seats, reading both as a reader would, preferred the candidate (5 for the candidate, 0 for the incumbent, 0 for neither; compare-v1, against edition-2026-09-19-052132); research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.29

run 2026-09-30-edition-pre-columbian-amazon-093543

This run wrote no disposition rows.

2026-09-30check5 of 5 seats judged the case blind$3.00

run 2026-09-30-check-pre-columbian-amazon-103111

This run wrote no disposition rows.

2026-09-21check5 of 5 seats judged the case blind$1.24

run 2026-09-21-check-pre-columbian-amazon-125733

This run wrote no disposition rows.

2026-09-19editionnew assessment; article 3730 → 3730 words; research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.38

run 2026-09-19-edition-pre-columbian-amazon-045109

This run wrote no disposition rows.

2026-09-19editionnew assessment; article 3375 → 3730 words; research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.23

run 2026-09-19-edition-pre-columbian-amazon-041312

This run wrote no disposition rows.

2026-09-17—11 row(s) written by session_01AiXD4y5Y91KC7iM3uBhWmu, which left no run recordno run record8 in · 1 duplicate · 2 failed

rows name session_01AiXD4y5Y91KC7iM3uBhWmu; no run record exists for it, so its verb, outcome and cost are not known — the rows themselves are the record

refused, with the reason

  • evidenceduplicateAMZ-E109Contemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 3

    a duplicate of AMZ-E109 — the same sentence, quote and claim, admitted twice by verification 2026-09-17-verify-pre-columbian-amazon-160114 under two directions; refused by the operator answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709)

  • evidencefailedContemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 4

    not one observation — bundles AMZ-E109, AMZ-E110 and AMZ-E111, which stand; refused by the operator answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709)

  • evidencefailedOviedo's endorsement of Carvajal rests on the friar's wounds and others' testimony to his character

    not one observation — two grounds of Oviedo's endorsement; split by the operator into AMZ-E113 and AMZ-E114, answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709)

admitted

  • evidenceinAMZ-E112Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 2 — toward AMZ-C076

    Correction by the operator: the record this row admitted was written under AMZ-E110, an id the same run had already given to its sibling (the part carrying "context" toward AMZ-C017); it is AMZ-E112. The row of 2026-09-17-verify-pre-columbian-amazon-160114 naming AMZ-E110 for this key is superseded by this one.

  • evidenceinAMZ-E113Oviedo's endorsement of Carvajal rests on the friar's wounds — part 1

    part 1 of AMZ-E086, split by the operator; the quote located verbatim in the OCR text

  • evidenceinAMZ-E114Oviedo's endorsement of Carvajal rests on others' testimony to his character — part 2

    part 2 of AMZ-E086, split by the operator; the quote located verbatim in the OCR text

  • claiminAMZ-C072Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests on the friar's two arrow wounds.

    AMZ-C072 reworded by the operator to one ground (the wounds), answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709); the second ground is AMZ-C085

  • claiminAMZ-C085Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests also on what his authority and person deserve, as affirmed by those who had dealt wi…

    the second ground AMZ-C072 bundled, made its own claim by the operator with its own anchor and AMZ-E114, answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709)

  • claiminAMZ-C061Excavations in the Upano Valley (Ecuador) date the Kilamope/Upano occupation from around 500 BCE to between 300 and 600 CE.

    AMZ-C061 reworded by the operator to the dating alone (review note #345); the builder attribution is AMZ-C086.

  • claiminAMZ-C086The Upano Valley earthen platforms and roads were originally built by groups of the Kilamope and later Upano cultures.

    the attribution AMZ-C061 bundled with the dating, admitted as its own claim by the operator with AMZ-E115 checked against the text by hand (review note #345) Entered 2026-09-17 as AMZ-C066; the id AMZ-C086 was given on 2026-09-19 because the re-submission proposal of 2026-09-17 had used AMZ-C066 for another candidate…

  • evidenceinAMZ-E115Rostain 2024: platforms and roads built by groups of the Kilamope and later Upano cultures

    the attribution part of AMZ-E067, refused as not atomic by verification 2026-09-17-verify-pre-columbian-amazon-140447; re-submitted alone by the operator, the quote located verbatim on p. 4 of the author manuscript (review note #345) Entered 2026-09-17 as AMZ-E078; the id AMZ-E115 was given on 2026-09-19 because the r…

2026-09-17verifyadmitted 7 claims, 16 evidence; 31 refused23 in · 24 failed · 7 blocked$2.71

run 2026-09-17-verify-pre-columbian-amazon-160114

refused, with the reason

  • evidencefailedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court

    not one observation (Both quotations from Sámano's letters appear verbatim in the source and Medina's inference that Orellana presented a large relación distinct from Carvajal's is fairly represented, supporting AMZ-C066 as stated, at approximately the cited pages. However, the sourceStatement bundles several separabl…

  • evidencefailedMedina: Orellana's own relación does not appear in the archives

    not one observation (The quoted passages accurately reflect Medina's text at p. xxxiv, and the direction (supports AMZ-C067) is correct since Medina indeed states the relación does not appear in the archives. However, the sourceStatement bundles at least two separable propositions—the non-appearance of the relación in…

  • evidencefailedGenealogy of the 'invention' account: the earliest lie-charge targeted Orellana and the Amazons episode

    not one observation (The quoted material and Medina's gloss both appear on pp. xxxv–xxxvi as cited, accurately reflecting Gómara's 'mentirosa' charge over the amazons and Medina's observation that Carvajal only recorded an Indian informant's answers without endorsing them. This bears as context on Carvajal's overall c…

  • evidencefailedMedina: the T'Serclaes manuscript and Oviedo's text are one author's work, with variants from copying or revision

    not one observation (The quote appears verbatim in Medina's Introducción at the page corresponding to 'xi' and is not misattributed or taken out of context; it fairly supports AMZ-C070's claim that Medina judged the two texts to be the same author's hand. However, the sourceStatement bundles two separable propositions…

  • evidencefailedBranch variants are one author's, not two witnesses' — context for the figure differences

    not one observation (The sourceStatement bundles two distinct propositions from Medina's introduction: (1) his conclusion that the two manuscript witnesses are in the same hand with variants explained by dictation errors or authorial revision, and (2) a separate factual note about where Oviedo inserted the Carvajal ac…

  • evidencefailedMedina confirms the Muñoz copy is incomplete

    not one observation (The quoted passage and footnote do appear at the cited location and accurately reflect Medina's statement that the Muñoz copy is incomplete and lacks any indication of its origin; this supports AMZ-C071. However, the sourceStatement bundles two separable propositions — that the copy is incomplete,…

  • evidencefailedContemporary character endorsement of Carvajal (credibility, not diagnosticity)

    not one observation (The record bundles two distinct propositions under 'and': Oviedo's contemporary character-based endorsement of Carvajal's credibility, and Medina's own independent editorial verdict that the Relación deserves complete faith regarding voyage events. These are separate claims with separate truth con…

  • evidencefailedOviedo's witness draws on Orellana orally — bearing on the independence of the two branches

    second reader rejected (locatorSupported): The p. xl passage is quoted faithfully and in context — Oviedo says he learned additional particulars from Orellana and his companions in Santo Domingo that the friar had not recorded. However, the cited page ix does not contain this statement at all; it merely notes that Ovi…

  • evidencefailedMedina's two caveats on the Relación: supernatural attributions and partisanship for Orellana

    not one observation (The quote is accurate and in context, occurring near the page xii/xiii boundary in Medina's introduction, and fairly reports Medina's overall verdict that the Relación merits full faith despite two caveats. However the sourceStatement bundles two distinct observations (supernatural attributions an…

  • claimfailedAccording to Medina (1894), letters of secretary Juan de Sámano of 31 May and June 1543 attest that Orellana brought to court a very large 'memorial y relación…

    second reader rejected the anchor (quoteInContext, statementSupported): The claim that the memorial was 'distinct from Carvajal's Relación' is not something the Sámano letters attest; it is Medina's own speculative inference ('no podemos menos de creer que fuera distinta...'), yet the record presents it as part of wha…

  • claimfailedMedina (1894) reports that Orellana's own relación of the voyage is not found in the archives.

    second reader rejected the anchor (relevant): The quote is accurately drawn and reflects Medina's statement that Orellana's own relación is not preserved in the archives, matching the proposition. However, this is a bibliographic/archival fact about a lost secondary account, not a claim about the demographic or ecolog…

  • claimfailedMedina (1894) concludes from comparison that the T'Serclaes manuscript of Carvajal's Relación and the version inserted by Oviedo are the work of the same hand.

    second reader rejected the anchor (relevant): The quote does occur in Medina's introduction and does state that the two manuscript copies are the work of the same hand, matching the anchored claim. However, this is a purely textual/philological finding about manuscript provenance and authorship of the copies, and does…

  • claimfailedThe copy of Carvajal's Relación in the Muñoz collection at the Real Academia de la Historia is incomplete, according to Medina (1894).

    second reader rejected the anchor (relevant): The quote does occur where indicated and Medina's statement about an incomplete copy of Carvajal's manuscript in the Muñoz collection is accurately reported, but this is a purely philological/manuscript-transmission detail about the textual history of the Relación, not a c…

  • claimfailedMedina (1894) infers from Oviedo's remark that he would have been glad to see and know the friar that Carvajal never passed through Santo Domingo, so Oviedo ne…

    second reader rejected the anchor (relevant): The quote is accurately used and the statement fairly reflects Medina's inference that Oviedo's wish to meet Carvajal implies Carvajal never passed through Santo Domingo. However, this is a biographical detail about whether Oviedo and Carvajal personally met, which bears o…

  • claimblockedThe Orellana narrative in Markham's 1859 Hakluyt volume is translated from Herrera's sixth decade, not from Carvajal's Relación.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedAccording to Markham (1859), Ursúa's expedition, which set out in September 1560, had arrived off a village called Machiparo when Ursúa was murdered on New Yea…

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedMarkham (1859) locates the village of Machiparo near the mouth of the Putumayo.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedBefore the 1560 Ursúa expedition, rumours circulating in Peru placed El Dorado in the provinces of the Omaguas, according to Markham (1859) citing Pedro Simón.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedMarkham (1859), summarizing Jesuit sources, reports that smallpox first appeared at Borja in 1660 and that forty-four thousand Indians died.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedAccording to Markham (1859), after Portugal's independence in 1640 the Portuguese attacked the Omaguas and other Indians of the river, burning villages and car…

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedSamuel Fritz, working on the Marañón from 1687, established forty villages among the Omaguas, according to Markham (1859).

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • evidencefailedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court — part 1

    every claim it cites was rejected

  • evidencefailedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court — part 2

    every claim it cites was rejected

  • evidencefailedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court — part 4

    every claim it cites was rejected

  • evidencefailedMedina: Orellana's own relación does not appear in the archives — part 1

    every claim it cites was rejected

  • evidencefailedMedina: Orellana's own relación does not appear in the archives — part 3

    every claim it cites was rejected

  • evidencefailedMedina: the T'Serclaes manuscript and Oviedo's text are one author's work, with variants from copying or revision — part 2

    every claim it cites was rejected

  • evidencefailedMedina: the T'Serclaes manuscript and Oviedo's text are one author's work, with variants from copying or revision — part 3

    every claim it cites was rejected

  • evidencefailedMedina confirms the Muñoz copy is incomplete — part 1

    every claim it cites was rejected

  • evidencefailedMedina confirms the Muñoz copy is incomplete — part 2

    every claim it cites was rejected

  • evidencefailedMedina infers Carvajal never passed through Santo Domingo, so Oviedo never met him

    every claim it cites was rejected

admitted

  • claiminAMZ-C068Medina (1894) records that López de Gómara cited Orellana's relación of the voyage and called it 'mentirosa'.
  • claiminAMZ-C069Medina (1894) attributes Gómara's 'mentirosa' charge to the report of Amazons on the river, quoting Gómara's objection to Orellana's affirmation that Amazons l…
  • claiminAMZ-C072Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests on the friar's two arrow wounds and on what those who had dealt with him affirmed of…
  • claiminAMZ-C074Oviedo states that he supplemented the friar's written account with particulars learned orally in Santo Domingo from Captain Orellana and the gentlemen who cam…
  • claiminAMZ-C075Carvajal's Relación closes with an explicit protestation that everything he has written and told is true.
  • claiminAMZ-C076Medina (1894) notes that Carvajal at times attributes certain events to supernatural causes.
  • claiminAMZ-C077Antonio de Herrera's account of the Orellana voyage follows Carvajal's Relación very closely, almost to the letter, according to Medina (1894).
  • evidenceinAMZ-E080Medina: Gómara called Orellana's relación 'mentirosa' — the invention charge began with the captain's narrative
  • evidenceinAMZ-E099Genealogy of the 'invention' account: the earliest lie-charge targeted Orellana and the Amazons episode — part 2
  • evidenceinAMZ-E100Genealogy of the 'invention' account: the earliest lie-charge targeted Orellana and the Amazons episode — part 3
  • evidenceinAMZ-E082Medina attributes the 'mentirosa' charge to the Amazons claim
  • evidenceinAMZ-E086Oviedo's endorsement of Carvajal rests on the friar's wounds and others' testimony to his character
  • evidenceinAMZ-E105Contemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 1
  • evidenceinAMZ-E106Contemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 2
  • evidenceinAMZ-E107Contemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 3
  • evidenceinAMZ-E108Contemporary character endorsement of Carvajal (credibility, not diagnosticity) — part 4
  • evidenceinAMZ-E089Oviedo supplemented the friar's text with what Orellana and his companions told him in Santo Domingo
  • evidenceinAMZ-E091Carvajal's closing protestation of truth
  • evidenceinAMZ-E109Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 1
  • evidenceinAMZ-E110Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 2
  • evidenceinAMZ-E110Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 2 — toward AMZ-C076
  • evidenceinAMZ-E111Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 3
  • evidenceinAMZ-E093Herrera followed Carvajal's Relación almost to the letter (Medina)
2026-09-17verifyadmitted 5 claims, 6 evidence, 2 sources, 2 research; 58 refused; 1 correction(s) applied15 in · 51 blocked · 7 failed$0.94

run 2026-09-17-verify-pre-columbian-amazon-140447

refused, with the reason

  • evidenceblockedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina: Orellana's own relación does not appear in the archives

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina: Gómara called Orellana's relación 'mentirosa' — the invention charge began with the captain's narrative

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedGenealogy of the 'invention' account: the earliest lie-charge targeted Orellana and the Amazons episode

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina attributes the 'mentirosa' charge to the Amazons claim

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina: the T'Serclaes manuscript and Oviedo's text are one author's work, with variants from copying or revision

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedBranch variants are one author's, not two witnesses' — context for the figure differences

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina confirms the Muñoz copy is incomplete

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedOviedo's endorsement of Carvajal rests on the friar's wounds and others' testimony to his character

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedContemporary character endorsement of Carvajal (credibility, not diagnosticity)

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina infers Carvajal never passed through Santo Domingo, so Oviedo never met him

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedOviedo supplemented the friar's text with what Orellana and his companions told him in Santo Domingo

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedOviedo's witness draws on Orellana orally — bearing on the independence of the two branches

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedCarvajal's closing protestation of truth

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedMedina's two caveats on the Relación: supernatural attributions and partisanship for Orellana

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedHerrera followed Carvajal's Relación almost to the letter (Medina)

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-…

  • evidenceblockedLaureano de la Cruz's three-year residence among the Omagua, ending 15 October 1650

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedLaureano: the Omagua province, 200 leagues long, held 34 small villages

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidencefailedA resident witness of 1647–50 finds Carvajal's Omagua reach thinly peopled — a datum the accounts read differently

    second reader rejected (quoteInContext, statementSupported, locatorSupported): The retrieved source is only the Internet Archive catalog/metadata page for this 1900 edition, not the actual OCR text of pages 94–95; there is no way to confirm the quoted Spanish phrases, village counts, or the framing about being deceive…

  • evidenceblockedThe first Omagua island village: 28 houses, about 80 lance-bearing men, about 250 women and children

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedA smallpox epidemic swept the Omagua province during Laureano's residence

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedLaureano: about a third of Caraute's people died in the epidemic

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedLaureano: the Omagua had no principal chiefs, government or obedience to anyone

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedPortuguese parties on the Solimões in 1650, prospecting for gold and consuming the Aysuaris' provisions

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidenceblockedThe 1637 Brieva–Toledo descent: 200 leagues without seeing people, who lived away from the river (Laureano)

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: quoted span not found verbatim in the text of SRC-CRUZ…

  • evidencefailedRostain 2024: Upano platform and road construction dated ca. 500 BCE to 300/600 CE

    not one observation (The quotes are accurate and correctly located, and the dating information supports AMZ-C061's claim about the 500 BCE–300/600 CE occupation window. However, the sourceStatement bundles three distinct propositions—the occupation date range, the cultural attribution to Kilamope/Upano builders, and t…

  • evidencefailedThe Upano occupation whose palaeoecological signal was 'erased' ended by 300–600 CE (or, counting Huapula reoccupation, by 1200 CE)

    second reader rejected (relevant): The quotes are accurate and correctly located, but Rostain et al. never discuss a palaeoecological/pollen signal, forest recovery erasing pre-Columbian influence, or persistence of a colonial-era signal — the claim's own subject matter is absent from this source entirely, and the dra…

  • evidencefailedHuapula reoccupation of Upano mounds, 800–1200 CE

    not one observation (The quote and locator check out against p.4 and the supplement, and the direction toward AMZ-C062 is correct. However, the sourceStatement bundles multiple separate propositions: the mound reoccupation dates, the description of Huapula pottery as corrugated ware, its association with Jivaroan-spea…

  • evidencefailedLiDAR detected over 6,000 earthen platforms in 300 km² of the Upano Valley

    not one observation (The quoted phrase and figures (6,000 platforms, dimensions, density) all accurately appear on p.6 as stated, and the record correctly supports the claim about platform detection. However, the sourceStatement bundles several distinct propositions—the detection count, platform shape/size/height, gro…

  • evidencefailedAuthors' assessment: Upano is the earliest and largest low-density agrarian urbanism documented in Amazonia so far

    not one observation (The record bundles three distinct propositions into one sourceStatement: the comparison of Upano to other (more recent, less dense) Amazonian sites, the superlative claim that Upano is the earliest/largest low-density agrarian urbanism documented, and an unrelated note about defensive peripheral d…

  • claimblockedAccording to Medina (1894), letters of secretary Juan de Sámano of 31 May and June 1543 attest that Orellana brought to court a very large 'memorial y relación…

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) reports that Orellana's own relación of the voyage is not found in the archives.

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) records that López de Gómara cited Orellana's relación of the voyage and called it 'mentirosa'.

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) attributes Gómara's 'mentirosa' charge to the report of Amazons on the river, quoting Gómara's objection to Orellana's affirmation that Amazons l…

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) concludes from comparison that the T'Serclaes manuscript of Carvajal's Relación and the version inserted by Oviedo are the work of the same hand.

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedThe copy of Carvajal's Relación in the Muñoz collection at the Real Academia de la Historia is incomplete, according to Medina (1894).

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedOviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests on the friar's two arrow wounds and on what those who had dealt with him affirmed of…

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) infers from Oviedo's remark that he would have been glad to see and know the friar that Carvajal never passed through Santo Domingo, so Oviedo ne…

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedOviedo states that he supplemented the friar's written account with particulars learned orally in Santo Domingo from Captain Orellana and the gentlemen who cam…

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedCarvajal's Relación closes with an explicit protestation that everything he has written and told is true.

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedMedina (1894) notes that Carvajal at times attributes certain events to supernatural causes.

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedAntonio de Herrera's account of the Orellana voyage follows Carvajal's Relación very closely, almost to the letter, according to Medina (1894).

    the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CARVAJAL-15…

  • claimblockedThe Orellana narrative in Markham's 1859 Hakluyt volume is translated from Herrera's sixth decade, not from Carvajal's Relación.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedAccording to Markham (1859), Ursúa's expedition, which set out in September 1560, had arrived off a village called Machiparo when Ursúa was murdered on New Yea…

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedMarkham (1859) locates the village of Machiparo near the mouth of the Putumayo.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedBefore the 1560 Ursúa expedition, rumours circulating in Peru placed El Dorado in the provinces of the Omaguas, according to Markham (1859) citing Pedro Simón.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedMarkham (1859), summarizing Jesuit sources, reports that smallpox first appeared at Borja in 1660 and that forty-four thousand Indians died.

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedAccording to Markham (1859), after Portugal's independence in 1640 the Portuguese attacked the Omaguas and other Indians of the river, burning villages and car…

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedSamuel Fritz, working on the Marañón from 1687, established forty villages among the Omaguas, according to Markham (1859).

    anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim

  • claimblockedThe Franciscan Laureano de la Cruz lived among the Omagua of the Solimões for three full years, leaving on the night of 15 October 1650.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedLaureano de la Cruz, writing of 1649–1650, counted 34 small villages in an Omagua province he reckoned 200 leagues long.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedThe first Omagua island village Laureano de la Cruz settled in (San Pedro de Alcántara, 1647) had 28 houses and about 80 lance-bearing men by his count.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedDuring Laureano de la Cruz's residence (1647–1650) a smallpox epidemic spread through the Omagua province of the Solimões.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedLaureano de la Cruz reports that about a third of the people of the Omagua village of Caraute died in the smallpox epidemic he witnessed there.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedLaureano de la Cruz describes the Omagua of 1647–1650 as having no principal chiefs, government, or obedience to anyone.

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedBy 1650 Portuguese parties were operating on the Solimões above the Rio Negro, according to Laureano de la Cruz, who was told they were seeking gold and found…

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • claimblockedAccording to Laureano de la Cruz, the two Franciscan lay brothers who descended the river in October 1637 travelled 200 leagues below the Encabellados without…

    the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instead, which is a viewer, not the text: the quotes were checked against nothing (corrected by the operator from "failed"; the original reason began: anchor quote not found verbatim in SRC-CRUZ-1651…)

  • sourcefailedNuevo descubrimiento del río de Marañón llamado de las Amazonas, hecho por la religión de San Francisco, año de 1651

    nothing accepted cites it (§3.6: sources are not evidence by themselves)

admitted

  • sourceinSRC-ROSTAIN-2024Two thousand years of garden urbanism in the Upper Amazon
  • sourceinSRC-ACHMATOWICZ-2018La expedición de Francisco de Orellana y la exploración del río de Amazonas hasta su desembocadura en el Atlántico: Sobre el valor como fuente histórica del re…
  • claiminAMZ-C061Excavations in the Upano Valley (Ecuador) date the Kilamope/Upano occupation that built the earthen platforms and roads from around 500 BCE to between 300 and…
  • claiminAMZ-C062Some Upano Valley mounds were reoccupied after a hiatus by groups of the Huapula culture between 800 and 1200 CE.
  • claiminAMZ-C063LiDAR analysis detected over 6,000 earthen platforms within a 300 km² area of the Upano Valley.
  • claiminAMZ-C064Rostain et al. (2024) assess the Upano Valley society as the earliest and largest low-density agrarian urbanism documented in Amazonia so far.
  • claiminAMZ-C065Achmatowicz (2018) assesses Carvajal's manuscript as a trustworthy historical source on the basis of a study of its geographic reference points.
  • evidenceinAMZ-E073Rostain 2024: Upano platform and road construction dated ca. 500 BCE to 300/600 CE — part 1
  • evidenceinAMZ-E074Huapula reoccupation of Upano mounds, 800–1200 CE — part 1
  • evidenceinAMZ-E075Huapula reoccupation of Upano mounds, 800–1200 CE — part 2
  • evidenceinAMZ-E076Authors' assessment: Upano is the earliest and largest low-density agrarian urbanism documented in Amazonia so far — part 1
  • evidenceinAMZ-E072Achmatowicz 2018 assesses Carvajal's manuscript as a trustworthy source on geographic grounds
  • evidenceinAMZ-E077Achmatowicz 2018 assesses Carvajal's manuscript as a trustworthy source on geographic grounds — toward AMZ-C017
  • researchinAMZ-R007Read the Biblioteca Nacional manuscript of Laureano de la Cruz to fix the Omagua headcount garbled in the 1900 edition
  • researchinAMZ-R008Calibrate palaeoecological erasure against time since abandonment using sites with dated occupation ends
2026-09-17reverifyre-submitted 35 record(s) from proposals/2026-09-17-draft-pre-columbian-amazon-135349 as verify 2026-09-17-verify-pre-columbian-amazon-160114: admitted 23, rejected 31; 16 legacy row(s) settled — verify: wrote {"sources":0,"evidence":16,"claims":7,"research":0,"provisional":0}; 31 rejected26 failed · 6 in$0.00

run 2026-09-17-reverify-pre-columbian-amazon-160112

refused, with the reason

  • evidencefailedMedina quotes Sámano's 1543 letters attesting a very large relación of the voyage sent to court

    settled with the row under text:medina quotes s mano s 1543 letters attesting a very large relaci n of the voyage sent to court medina prints two letters of the secretary juan de s mano to francisco de los cobos simancas estado leg 61 fols 208 and 213 the first dated 31 may 1543 reports one arrived from peru who came…

  • evidencefailedMedina: Orellana's own relación does not appear in the archives

    settled with the row under text:medina orellana s own relaci n does not appear in the archives medina states that orellana presented a long relaci n of his voyage relaci n que hoy no parece en los archivos and that while it was probably based on carvajal s notes he believes it was distinct from the two carvajal texts…

  • evidencefailedGenealogy of the 'invention' account: the earliest lie-charge targeted Orellana and the Amazons episode

    settled with the row under text:genealogy of the invention account the earliest lie charge targeted orellana and the amazons episode medina reports g mara s charge against orellana s relaci n as mentirosa and quotes g mara s objection among the nonsense he spoke fu afirmar que hab a en este r o amazonas with whom he a…

  • evidencefailedMedina: the T'Serclaes manuscript and Oviedo's text are one author's work, with variants from copying or revision

    settled with the row under text:medina the t serclaes manuscript and oviedo s text are one author s work with variants from copying or revision having compared both texts medina concludes ambas son obra de una misma mano and that the variants come partly from copies perhaps made by dictation to unskilled copyists or f…

  • evidencefailedBranch variants are one author's, not two witnesses' — context for the figure differences

    settled with the row under text:branch variants are one author s not two witnesses context for the figure differences medina holds that the two texts ambas son obra de una misma mano the variants arising from dictation to copyists or from the author s own additions and suppressions he also notes oviedo inserted the fr…

  • evidencefailedMedina confirms the Muñoz copy is incomplete

    settled with the row under text:medina confirms the mu oz copy is incomplete of the relaci n that remained in manuscript medina writes there exists una copia incompleta en la colecci n de documentos reunida por mu oz preserved at the real academia de la historia he adds that the copy carries no indication of where it…

  • evidencefailedContemporary character endorsement of Carvajal (credibility, not diagnosticity)

    settled with the row under text:contemporary character endorsement of carvajal credibility not diagnosticity medina reports oviedo s judgement that carvajal debe ser cre do en virtud de aquellos dos flechazos and his good repute among those who dealt with him and medina s own verdict that limited to the events of the…

  • evidencefailedMedina infers Carvajal never passed through Santo Domingo, so Oviedo never met him

    settled with the row under text:medina infers carvajal never passed through santo domingo so oviedo never met him medina n 16 deduces from oviedo s remark that he would have been glad to see and know him oviedo then residing in santo domingo that luego carvajal no estuvo en aquella ciudad carvajal sailed for nombre de…

  • evidencefailedOviedo's witness draws on Orellana orally — bearing on the independence of the two branches

    settled with the row under text:oviedo s witness draws on orellana orally bearing on the independence of the two branches oviedo per medina states he learned particulars of the voyage del mesmo capit n francisco de orellana de otros caballeros in santo domingo and added them to what the friar wrote (re-submitted in 20…

  • evidencefailedMedina's two caveats on the Relación: supernatural attributions and partisanship for Orellana

    settled with the row under text:medina s two caveats on the relaci n supernatural attributions and partisanship for orellana medina finds no reason to deny the relaci n complete faith por m s que en ocasiones el autor atribuya ciertos hechos causas sobrenaturales and in others shows himself an enthusiastic admirer of…

  • claimfailedThe Franciscan Laureano de la Cruz lived among the Omagua of the Solimões for three full years, leaving on the night of 15 October 1650.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedLaureano de la Cruz, writing of 1649–1650, counted 34 small villages in an Omagua province he reckoned 200 leagues long.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedThe first Omagua island village Laureano de la Cruz settled in (San Pedro de Alcántara, 1647) had 28 houses and about 80 lance-bearing men by his count.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedDuring Laureano de la Cruz's residence (1647–1650) a smallpox epidemic spread through the Omagua province of the Solimões.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedLaureano de la Cruz reports that about a third of the people of the Omagua village of Caraute died in the smallpox epidemic he witnessed there.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedLaureano de la Cruz describes the Omagua of 1647–1650 as having no principal chiefs, government, or obedience to anyone.

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedBy 1650 Portuguese parties were operating on the Solimões above the Rio Negro, according to Laureano de la Cruz, who was told they were seeking gold and found…

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • claimfailedAccording to Laureano de la Cruz, the two Franciscan lay brothers who descended the river in October 1637 travelled 200 leagues below the Encabellados without…

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its anchor's source SRC-CRUZ-1651 never entered

  • evidencefailedLaureano de la Cruz's three-year residence among the Omagua, ending 15 October 1650

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedLaureano: the Omagua province, 200 leagues long, held 34 small villages

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedThe first Omagua island village: 28 houses, about 80 lance-bearing men, about 250 women and children

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedA smallpox epidemic swept the Omagua province during Laureano's residence

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedLaureano: about a third of Caraute's people died in the epidemic

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedLaureano: the Omagua had no principal chiefs, government or obedience to anyone

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedPortuguese parties on the Solimões in 1650, prospecting for gold and consuming the Aysuaris' provisions

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

  • evidencefailedThe 1637 Brieva–Toledo descent: 200 leagues without seeing people, who lived away from the river (Laureano)

    not re-submitted (2026-09-17-reverify-pre-columbian-amazon-160112): its source SRC-CRUZ-1651 never entered

admitted

  • evidenceinAMZ-E080Medina: Gómara called Orellana's relación 'mentirosa' — the invention charge began with the captain's narrative

    settled with the row under text:medina g mara called orellana s relaci n mentirosa the invention charge began with the captain s narrative medina writes that l pez de g mara and herrera knew orellana s relaci n and that g mara la cita d ndole el calificativo de mentirosa though as pinelo barcia observed without statin…

  • evidenceinAMZ-E082Medina attributes the 'mentirosa' charge to the Amazons claim

    settled with the row under text:medina attributes the mentirosa charge to the amazons claim medina says the ground for g mara s charge is not difficult to discover carvajal in his notes and orellana at court had spoken of amazons and he quotes g mara fu afirmar que hab a en este r o amazonas as the disparate that drew…

  • evidenceinAMZ-E086Oviedo's endorsement of Carvajal rests on the friar's wounds and others' testimony to his character

    settled with the row under text:oviedo s endorsement of carvajal rests on the friar s wounds and others testimony to his character medina quotes oviedo historia general t iv p 574 the friar debe ser cre do en virtud de aquellos dos flechazos one of which took his eye and on the strength of what his authority and perso…

  • evidenceinAMZ-E089Oviedo supplemented the friar's text with what Orellana and his companions told him in Santo Domingo

    settled with the row under text:oviedo supplemented the friar s text with what orellana and his companions told him in santo domingo medina quotes oviedo beyond what a devout dominican friar wrote as eyewitness he learned particulars afterwards in santo domingo del mesmo capit n francisco de orellana de otros caballer…

  • evidenceinAMZ-E091Carvajal's closing protestation of truth

    settled with the row under text:carvajal s closing protestation of truth medina quotes the end of the relaci n the friar wrote it to tell and make known the truth in all of it to forestall those who might tell the pilgrimage otherwise y es verdad en todo lo que yo he escrito y contado (re-submitted in 2026-09-17-rever…

  • evidenceinAMZ-E093Herrera followed Carvajal's Relación almost to the letter (Medina)

    settled with the row under text:herrera followed carvajal s relaci n almost to the letter medina medina antonio de herrera had the relaci n before him y la sigue muy de cerca casi al pie de la letra in keeping with his method of extracting documents medina cites three afirma el p carvajal passages from d cada v lib vi…

2026-09-17reverifyno provisional recordsrested$0.00

run 2026-09-17-reverify-pre-columbian-amazon-160110

This run wrote no disposition rows.

2026-09-17reportresearch pass (claude-fable-5-1)$3.33

run 2026-09-17-report-pre-columbian-amazon-134558

This run wrote no disposition rows.

2026-09-17editionnew assessment; article 2953 → 3367 words; research agenda: 8 open, 0 answered, 0 superseded, 0 retired$3.48

run 2026-09-17-edition-pre-columbian-amazon-141148

This run wrote no disposition rows.

2026-09-17draftschema sent as instructions (too large for strict output)8 blocked · 13 failed · 1 irrelevant · 4 duplicate$5.28

run 2026-09-17-draft-pre-columbian-amazon-135349

refused, with the reason

  • sourceblockedExpeditions into the Valley of the Amazons, 1539, 1540, 1639

    previously blocked on 2026-09-10: Cristóbal de Acuña, Nuevo descubrimiento del gran río de las Amazonas (Madrid, 1641) — the 1639 corridor witness the report identifies as the missing middle of AMZ-R005. Only the archive.org item page for Markham's 1859 Hakluyt volume (which contains the English translation) was retri…

  • evidencefailedMarkham's Orellana narrative is Herrera's, not Carvajal's

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedMarkham fixes the Ursúa expedition at Machiparo at the turn of 1560–61

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedIndependent secondary dating of the Ursúa witnesses at Machiparo, 1560–61

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedMarkham glosses Machiparo as near the mouth of the Putumayo

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedBefore 1560, El Dorado was rumoured in Peru to reside among the Omaguas

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedExpectation bias among the 1560 witnesses

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedMarkham (from Jesuit sources): smallpox at Borja in 1660 killed forty-four thousand

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedRepeated smallpox waves in Maynas (1660, 1669, 1680, 1749, 1756, 1762) with flight from missions

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedMarkham: after 1640 the Portuguese attacked the Omaguas, burning villages and enslaving women and children

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • evidencefailedFritz established forty villages among the Omaguas (Markham)

    its source "P-SRC-MARKHAM-1859" was not added (see its disposition) and is not in the ledger

  • sourcefailedscienceopen.com/document?vid=71f8aa2d-382a-4380-a35f-dabcc1dd0335 not retrieved (HTTP 403); doi.org/10.1126/science.adi6317 retrieved, 31 pages.

    Drafter named the proposed Source 'Two thousand years of garden urbanism in the Upper Amazon' (Rostain et al. 2024) as the record, which the ledger does not hold. Its reason as written: The report's ScienceOpen abstract record for Rostain et al. 2024 returned HTTP 403, but the full author-accepted manuscript was retri…

  • sourceblockedFetch failed at revistadeindias.revistas.csic.es/index.php/revistadeindias/article/view/1023; nothing read.

    'La despoblación en el Alto Amazonas en tiempos coloniales', Revista de Indias (2016), article 1023 — cited by the report as bearing on Laureano and Livi-Bacci; fetch failed, no text shown.

    reopen if: The CSIC OJS PDF or the article DOI is retrieved and its treatment of the Omagua/Alto Amazonas depopulation read.

  • sourceblockedHTTP 403 at scielo.br/j/tem/a/hvhyBxFfB9mHBgYBfCfGXFp/?lang=en.

    Harris 2017, 'Revisiting first contacts on the Amazon 1500–1562', Tempo 23(3) — directly relevant to AMZ-E007 (Nunes with Mercadillo's party in 1538, slave-taking at Machiparo); HTTP 403 at SciELO, no text shown.

    reopen if: The SciELO PDF or another open copy is retrieved and the Nunes-letter discussion read against SRC-LOPES-2024.

  • sourceblockedCited in the retrieved Rostain 2024 text as ref. 14 and in SRC-PERIPATO-2023 as ref. 19; not read.

    Prümers et al. 2022, 'Lidar reveals pre-Hispanic low-density urbanism in the Bolivian Amazon' (Nature 606) — the report's snippet gives a Casarabe area of ~4,500 km² and AD 500–1400, but the paper was not opened; the existing block stands and the snippet figures are not entered.

    reopen if: The open-access text (nature.com or Europe PMC) is retrieved.

  • sourceblockedReport cites a HathiTrust location; no text shown.

    Samuel Fritz's journal (Hakluyt Society 1922), the 1686–1723 Omagua witness for AMZ-R005; not opened this pass. Markham 1859 (retrieved) gives only a tertiary summary of Fritz's forty villages.

    reopen if: The Hakluyt 1922 text is retrieved.

  • claimblockedMedina 1894, Introducción, p. ix n. 1: the Relación 'ocupa las páginas 541-574 del tomo IV' of Oviedo.

    The two AMZ-C002 sub-points (where '10,000' sits; whether Oviedo retains '~5,000') remain unresolved: Oviedo lib. L cap. XXIV and the body of the Relación in Medina were beyond the truncated texts. Medina's note supplies the exact route: the Carvajal text occupies pp. 541–574 of tomo IV of the RAH edition.

    reopen if: Oviedo t. IV pp. 541–574 (RAH 1855) and Medina's Relación body (pp. 1–62 with variant footnotes) are read at page level and the figures placed by witness.

  • sourceblockedQuoted at second hand in Medina 1894, Introducción, pp. xxxv–xxxvi.

    López de Gómara, Historia general de las Indias (1552) — to see exactly what he called 'mentirosa' (the whole relación or the Amazons episode); only Medina's quotation (Ribadeneyra ed. p. 210 per Medina) was shown.

    reopen if: Gómara's text is retrieved and the passage read in context.

  • claimfailedAMZ-C022 statement as in index; Rostain 2024 [p. 1] dates occupation 500 BCE–300/600 CE and [p. 4] Huapula reoccupation 800–1200 CE.

    Drafter named the proposed Evidence record 'The Upano occupation whose palaeoecological signal was erased ended by 300–600 CE (or, counting Huapula reoccupation, by 1200 CE)' as the record, which the ledger does not hold. Its reason as written: The report proposes re-wording AMZ-C022 to carry the Upano occupation end-…

    reopen if: SRC-BUSH-2025-UPANO's text is shown and itself states the occupation end-date, permitting a statement correction with a matching anchor.

  • imageblockedTitle page text present in the djvu OCR: 'Descubrimiento del Río de las Amazonas ... Sevilla, Imprenta de E. Rasco ... MDCCCXCIV'.

    Report proposes the Medina 1894 title page or dedication to the Duque de T'Serclaes (public domain, Harvard/Google scan) as a plate for the manuscript-witness section. No page-image URL, page-scan confirmation or credit line was shown; only the OCR text.

    reopen if: A stable archive.org page-image URL for the title page and a credit line to the holding library are recorded.

  • evidencefailedOCR: 'los Omaguas con 34 pueblos pequeños no hallamos 10 indios cabales, fuera de sus mujeres y muchachos'.

    The tertiary snippet attributing to Laureano de la Cruz a count of '34 villages and ≤5,000 people' is only half confirmed: the retrieved OCR of the 1900 edition (p. 94) reads 'con 34 pueblos pequeños no hallamos 10 indios cabales'; no figure of 5,000 appears verbatim, and the numeral is an OCR uncertainty. The village…

    reopen if: The BNE manuscript or a page image of p. 94 resolves the numeral (see the proposed research item on the Laureano manuscript).

  • sourceirrelevantNot opened; mentioned in the report as considered and not proposed.

    The 'amazonzach' blog on a Putumayo route for Orellana — a blog; the citable form is Achmatowicz 2018, which is proposed as a source (abstract only). The Putumayo-route argument itself is not in the Achmatowicz abstract and remains unverified.

  • researchduplicateAMZ-R005AMZ-R005 exists with claimIds [AMZ-C010, AMZ-C017]; its summary text is not shown.

    Report item 6 asks to add Laureano de la Cruz (1651/53), the Revista de Indias 2016 article and Fritz (1922) as leads to AMZ-R005. Laureano is now read and enters as a source with claims; the amendment to R005's summary cannot be written as a scalar correction because R005's current summary is not in the packet.

    reopen if: R005's summary is exposed for a scalar correction naming the 1560 (Ursúa), 1639 (Acuña), 1647–50 (Laureano) and 1686–1723 (Fritz) Omagua witnesses.

  • researchduplicateAMZ-R001AMZ-R001 exists with claimIds [AMZ-C001, AMZ-C017]; Markham 1859 Contents lists a 'List of the Principal Tribes... mentioned in the voyages of Orellana and Acu…

    Report item 6 asks to link Achmatowicz 2018 (route argument) and Markham 1859's tribe concordance list to AMZ-R001's route-uncertainty model; these are inputs to the existing corridor test, not a new test. R001's summary is not in the packet for a scalar correction.

    reopen if: R001's summary is exposed; Achmatowicz's body is read to confirm the Putumayo-route argument; Markham's tribe list is read.

  • researchduplicateAMZ-R002AMZ-R002 exists with claimIds [AMZ-C002].

    The report's F1(b) three-witness framing (T'Serclaes / Muñoz copy / Oviedo), Medina's same-hand judgement and the note that the Muñoz copy is incomplete belong to the existing synoptic witness table item; they enter as claims and evidence rather than a new item, and R002's summary is not exposed for correction.

    reopen if: R002's summary is exposed for a scalar correction.

  • evidenceduplicateAMZ-E007AMZ-E007 exists on SRC-LOPES-2024.

    Report item 6 asks to add Harris 2017 to AMZ-E007's reading; claimIds/source lists are not correctable here and Harris was not retrieved (blocked above).

    reopen if: Harris 2017 is retrieved; then a new evidence record on AMZ-C003 from that source.

2026-09-14editionre-adopts the incumbent's assessment; article 3026 → 2953 words$1.54

run 2026-09-14-edition-pre-columbian-amazon-125807

This run wrote no disposition rows.

2026-09-10verifyadmitted 2 claims, 4 evidence, 1 sources, 1 research; 22 refused; 1 correction(s) applied8 in · 22 failed$1.87

run 2026-09-10-verify-pre-columbian-amazon-185127

refused, with the reason

  • evidencefailedBasin-wide LiDAR (5,315 km², 0.08% of Amazonia) revealed 24 previously unreported earthworks

    not one observation (The quoted figures and locations match the source at p.5 and the supplement's density/flight-line details at p.39, and the finding directly supports the claim about the 24 earthworks discovered via the 5,315 km² LiDAR survey. However, the sourceStatement bundles multiple distinct facts (the 24-ear…

  • evidencefailedInhomogeneous Poisson model predicts 10,272–23,648 undiscovered earthworks basin-wide

    not one observation (The quoted figures and percentages accurately reflect the source's IPP model results at pages 6-7, and this bears on and supports the claim about undiscovered earthworks. However, the sourceStatement bundles several separable propositions: the model's predicted range and percentage undiscovered, t…

  • evidencefailedComparator: a basin-wide 10–24k undiscovered-earthwork prediction beside the Aquiry 24–30k regional prediction

    not one observation (The quoted figures (10,272–23,648 basin-wide undiscovered earthworks; the 94,713 km\u00b2 high-probability rectangle over Acre) both appear on p.7 as stated, in the same sense used here, so quoting and locator checks pass, and as an independent comparator study it appropriately bears on AMZ-C004 a…

  • evidencefailedAlmost 80% of the basin at 0–1% earthwork probability; dark earths widespread where earthworks are not

    quoted span not found verbatim in the text of SRC-PERIPATO-2023 (open-access copy via OpenAlex: https://eprints.whiterose.ac.uk/207653/1/Vinicius_et_al_2023_Science_AAM.pdf): "Earth-building societies were very common in some parts of the basin, but they may not have occupied all of Amazonia."

  • evidencefailedNearly all ≥25%-probability cells fall in a 94,713 km² rectangle over Acre

    quoted span not found verbatim in the text of SRC-PERIPATO-2023 (open-access copy via OpenAlex: https://eprints.whiterose.ac.uk/207653/1/Vinicius_et_al_2023_Science_AAM.pdf): "rectangle that overlays a significant portion of the Brazilian state of Acre,"

  • evidencefailed53 of 79 domesticated tree species significantly associated with modelled earthwork probability

    not one observation (The quoted passages appear as given at p.9 and in the abstract, and accurately describe the GLM results without overstatement. Both AMZ-C031 (which restates the 53/35/18 breakdown) and AMZ-C008 (concentration of domesticated species near earthwork sites) are supported by this finding, and the reco…

  • evidencefailedKnown earthworks have mostly been found in deforested, road-accessible areas

    not one observation (The quotes on deforested discovery, road proximity, and 9% protected-land figure are accurately drawn from pages 38, 8, and 10 respectively and are used in the sense the source intends, and together they do support the sampling-bias claim they're attached to. However, the sourceStatement bundles a…

  • evidencefailedSampling bias in the recorded-site base that river-distance and site-type models are built on

    quoted span not found verbatim in the text of SRC-PERIPATO-2023 (open-access copy via OpenAlex: https://eprints.whiterose.ac.uk/207653/1/Vinicius_et_al_2023_Science_AAM.pdf): "Until now, pre-Columbian earthworks have mostly been discovered in deforested areas,"

  • evidencefailedIndependent restatement: Acre earthworks built AD 120–400; Mato Grosso earthworks AD 1250–1500

    second reader rejected (relevant): The passage accurately reports (citing earlier literature, refs 14 and 15) two separate radiocarbon date ranges for earthworks in Acre and Mato Grosso, but these are inland upland archaeological sites unrelated to the mainstem floodplain riverine societies Carvajal described in 1542;…

  • claimfailedInspection of 5,315 km² of Amazon-wide airborne LiDAR transects (about 0.08% of the basin), flown between 2008 and 2017 for biomass estimation, revealed 24 pre…

    second reader rejected the anchor (statementSupported): The two quoted spans are accurately located and used in context, but the anchored statement adds specifics (the 2008–2017 flight-date range for the LiDAR data) that come from the supplementary materials table, not from the quoted passages themselves, so the state…

  • claimfailedAn inhomogeneous Poisson process model fitted to 961 known earthworks (ring ditches, geoglyphs, ponds, wells) with environmental and observability covariates p…

    second reader rejected the anchor (statementSupported): The quotes are accurate and in context, but the record states the model was based on 961 known earthworks, when the source explicitly says 937 known earthworks plus 24 newly discovered ones; this numeric detail is not a fair account of the source. The statement a…

  • claimfailedNearly all cells with ≥25% predicted earthwork probability in the 2023 basin-wide model fall within a 94,713 km² rectangle overlying much of the Brazilian stat…

    second reader rejected the anchor (statementSupported): The quoted phrase about the 94,713 km² rectangle over Acre does occur at p.7 and is used correctly in context, but the record's statement also asserts specific percentage breakdowns (86% and 96% of area at ≥25% and ≥50% thresholds) that are not stated at p.7—they…

  • claimfailedAcross 1,676 forest plots, the occurrence and/or abundance of 53 of 79 domesticated tree species is significantly associated with the modelled probability of e…

    not atomic (The quotes accurately reflect the source's finding that 53 of 79 domesticated species showed significant association with earthwork probability, and the locators (pp.3 and 9) are correct. However, the statement bundles two separable propositions—the total count of significantly associated species (53/79) a…

  • claimfailedThe recorded distribution of Amazonian earthworks is biased toward deforested and road-accessible areas: most known earthworks were discovered after deforestat…

    second reader rejected the anchor (statementSupported): The two quoted passages (deforested-area discovery bias, road-proximity of known sites) are accurately quoted and correctly located, but the statement also asserts a specific '9% inside indigenous lands and protected areas' figure that is not supported by either…

  • evidencefailedBasin-wide LiDAR (5,315 km², 0.08% of Amazonia) revealed 24 previously unreported earthworks — part 1

    every claim it cites was rejected

  • evidencefailedBasin-wide LiDAR (5,315 km², 0.08% of Amazonia) revealed 24 previously unreported earthworks — part 3

    every claim it cites was rejected

  • evidencefailedBasin-wide LiDAR (5,315 km², 0.08% of Amazonia) revealed 24 previously unreported earthworks — part 4

    every claim it cites was rejected

  • evidencefailedInhomogeneous Poisson model predicts 10,272–23,648 undiscovered earthworks basin-wide — part 1

    every claim it cites was rejected

  • evidencefailedInhomogeneous Poisson model predicts 10,272–23,648 undiscovered earthworks basin-wide — part 2

    every claim it cites was rejected

  • evidencefailedInhomogeneous Poisson model predicts 10,272–23,648 undiscovered earthworks basin-wide — part 5

    every claim it cites was rejected

  • evidencefailedKnown earthworks have mostly been found in deforested, road-accessible areas — part 1

    every claim it cites was rejected

  • sourcefailedHistoria general y natural de las Indias, islas y tierra-firme del Mar Océano — Tercera Parte, Tomo IV (Real Academia de la Historia edition)

    nothing accepted cites it (§3.6: sources are not evidence by themselves)

admitted

  • sourceinSRC-PERIPATO-2023More than 10,000 pre-Columbian earthworks are still hidden throughout Amazonia
  • claiminAMZ-C029In the same basin-wide model, almost 80% of Amazonia has a predicted probability of 0–1% for earthwork presence per 1 km² cell, with these low-probability area…
  • claiminAMZ-C033The occurrence and/or abundance of 53 of the 79 domesticated tree species identified across 1,676 forest plots showed significant association with the predicti…
  • evidenceinAMZ-E038Comparator: a basin-wide 10–24k undiscovered-earthwork prediction beside the Aquiry 24–30k regional prediction — part 1
  • evidenceinAMZ-E039Comparator: a basin-wide 10–24k undiscovered-earthwork prediction beside the Aquiry 24–30k regional prediction — part 2
  • evidenceinAMZ-E04053 of 79 domesticated tree species significantly associated with modelled earthwork probability — part 2
  • evidenceinAMZ-E04153 of 79 domesticated tree species significantly associated with modelled earthwork probability — part 5
  • researchinAMZ-R006Blind, probability-stratified LiDAR test of the competing earthwork-density predictions
2026-09-10reportresearch pass (claude-fable-5-1)$3.00

run 2026-09-10-report-pre-columbian-amazon-183738

This run wrote no disposition rows.

2026-09-10reportresearch pass (claude-fable-5-1)failed$0.00

run 2026-09-10-report-pre-columbian-amazon-121758

This run wrote no disposition rows.

2026-09-10editiona new edition with a new assessment$2.79

run 2026-09-10-edition-pre-columbian-amazon-190136

This run wrote no disposition rows.

2026-09-10draftschema sent as instructions (too large for strict output)14 blocked · 1 irrelevant · 2 duplicate$2.41

run 2026-09-10-draft-pre-columbian-amazon-184613

refused, with the reason

  • evidenceblockedarchive.org item page for descubrimientod00carvgoog: catalogue metadata, 538 pages, FULL TEXT and PDF download options listed; no page text shown.

    Report F2/F3: Medina's main (T'Serclaes) text carries 50,000 warriors (pp. 30–31), the eighty-league populated stretch and five-league village (pp. 40–41), the 'thousand men for a year' store (pp. 33–34), and 5,000 warriors (pp. 53–54), while the 8,000-warrior/130-canoe figure appears only as a footnoted Muñoz-copy va…

    reopen if: The FULL TEXT or PDF of the Medina edition is retrieved and the passages read at page level.

  • evidenceblockedItem pages only for both editions.

    Report F4: Carvajal states the interior was not observed and that populousness inland was inferred from the land's appearance and Aparia's people's reports (Medina p. 41; Oviedo redaction 'La tierra adentro no se pudo ver'). Text not shown in this pass.

    reopen if: Medina p. 41 and the corresponding Oviedo passage are retrieved.

  • evidenceblockedItem pages only.

    Report F6: both witnesses record multi-day unpopulated reaches ('tres días sin poblado ninguno', Medina p. 8; Oviedo 'tres días sin poblado'), which would anchor the 'interspersed with stretches he reports as empty' half of AMZ-C001 to the primary. Text not shown.

    reopen if: The passages are retrieved.

  • evidenceblockedarchive.org item page for historiageneral04fernguat: RAH 1855, t. IV, 619 pp., NOT IN COPYRIGHT, FULL TEXT available.

    Report F5: Oviedo's redaction gives no warrior count at Machiparo, 'más de mill tortugas', 'más de cient canoas', 'más de sessenta leguas de poblado', and states 'No se pudieron contar todos los pueblos desta provincia de Machiparo'. This is the primary locator behind AMZ-E005, but the Oviedo text was not retrieved —…

    reopen if: Oviedo t. IV lib. L cap. XXIV (page markers 552–558 per the report) is retrieved.

  • claimblockedCurrent AMZ-C002 wording attributes 50,000, 10,000, and 8,000 jointly to 'the Medina/Muñoz manuscript branch'; the report's reading (unverified here) is that 8…

    Report item 6 proposes re-wording AMZ-C002 to place each figure with its witness (T'Serclaes main text / Muñoz variant / Oviedo) and to flag '10,000' and '~5,000 in Oviedo' as unlocated. The amendment is well-motivated but rests entirely on passages not shown in this pass; a statement correction with an unverified bas…

    reopen if: The Medina and Oviedo passages are retrieved and the distribution of figures across witnesses confirmed at page level.

  • evidenceblockedReport states its search was regex-based over OCR with known errors ('ginco', 'mill').

    Report negative result: no 'diez mil' (10,000 warriors) figure located in the OCR of the Medina edition. A regex miss on OCR of a nineteenth-century print is weak negative evidence, as the report itself labels it, and the text was not shown here; it cannot enter as a finding.

    reopen if: A page-level reading of Medina (main text and footnotes) and of Oviedo either locates or fails to locate the figure; also check whether SRC-DENEVAN-1996 attributes it to a specific passage.

  • evidenceblockedItem page only.

    Report: could not independently confirm Denevan's statement that Oviedo retains a ~5,000 figure (regex for 'cinco/ginco mill' found no hits; OCR caveat). Neither the text nor a confirmed absence was shown here.

    reopen if: Oviedo's chapter is read rather than searched.

  • evidenceblockedItem page only.

    Report items 5/6: Medina's statement of editorial method (pp. x–xi: two manuscripts, the complete T'Serclaes and the incomplete Muñoz copy; variants judged 'obra de una misma mano', arising from dictation or authorial revision) and Oviedo's character reference for Carvajal (Medina p. xiii, 'debe ser creído'). Both are…

    reopen if: Medina's introduction pp. x–xiii is retrieved.

  • sourceblockedarchive.org item discoveryofamazo0000carv: Dover 1988 reprint of AGS 1934, ISBN 0486255891, Access-restricted-item true.

    Heaton's 1934 English translation (AGS Special Publication 17) as reprinted by Dover 1988. The retrieved item page shows it is access-restricted (controlled lending), 'No suitable files to display', so its text cannot be read or quoted; the report's proposal to add it as an identifier on SRC-CARVAJAL-1542 is a list-fi…

    reopen if: A readable copy of the 1934 or 1988 edition is obtained.

  • sourceblockedItem page for expeditionsintov00markrich: Hakluyt Society 1859, includes 'New discovery of the great river of the Amazons, by Father Christoval de Acuña, A.D.…

    Cristóbal de Acuña, Nuevo descubrimiento del gran río de las Amazonas (Madrid, 1641) — the 1639 corridor witness the report identifies as the missing middle of AMZ-R005. Only the archive.org item page for Markham's 1859 Hakluyt volume (which contains the English translation) was retrieved; the primary text was not rea…

    reopen if: Markham 1859 full text or a Spanish edition (1891; Arellano/Díez Borque/Santonja 2009) is retrieved and the Machiparo–Omagua reach passages read.

  • sourceirrelevantFour-page review, Notas bibliográficas pp. 301–304; states publication 'en Madrid, en 1641' and quotes the editors on Acuña's marvellous reports being from 'la…

    Fátima Salvatierra's book review of the 2009 Arellano/Díez Borque/Santonja edition of Acuña (Dialnet). Retrieved in full, but it is a tertiary notice: it confirms Acuña left Quito in 1639 with Tejeira's expedition, that the account was published in Madrid in 1641, and that the editors flag a 'perspectiva de lo maravil…

    reopen if: Acuña enters the ledger and a credibility note on his hearsay reports is wanted; the review's pp. 3 remarks could then be cited as an editors' assessment.

  • sourceblockedReport cites a HathiTrust location; no text or item page shown here.

    Samuel Fritz's journal (Hakluyt Society 1922), the 1686–1723 corridor witness for AMZ-R005. Not opened; located only by search.

    reopen if: The Hakluyt 1922 text is retrieved.

  • sourceblockedCited in SRC-PERIPATO-2023 as ref. 19 with doi:10.1038/s41586-022-04780-4; not read.

    Prümers et al. 2022, 'Lidar reveals pre-Hispanic low-density urbanism in the Bolivian Amazon' (Nature 606:325–328) — relevant to AMZ-C007 and AMZ-C019 as the Casarabe settlement survey. The report hit a CAPTCHA at PMC; the paper is cited (ref. 19) in the retrieved Peripato text but its own text was not shown.

    reopen if: The open-access text is retrieved.

  • sourceblockedNot retrieved; no snippet relied upon.

    Rostain et al. 2024 (Science, Upano Valley urbanism) and its published reply — relevant to AMZ-C022 and AMZ-C004. Access error; not read.

    reopen if: Text retrieved via institutional or repository route.

  • imageblockedItem pages show public-domain status and scan provenance (Harvard/Google; Universidad Francisco Marroquín).

    Proposed plates: the Medina 1894 page carrying the 'Variante… ocho mil indios' footnote (pp. 39–40) or the Oviedo 1855 page with 'No se pudieron contar todos los pueblos'. Both scans are public domain (item pages show NOT_IN_COPYRIGHT), but no page-image URL, page number confirmation, or credit line was shown, and the…

    reopen if: The passages are located at page level and a stable page-image URL (archive.org page viewer) with credit to the holding library is recorded.

  • researchduplicateAMZ-R002AMZ-R002 exists with claimIds [AMZ-C002].

    Report item 7 asks to give AMZ-R002 a three-witness spine (T'Serclaes / Muñoz / Oviedo) with Medina's authorial-variant hypothesis and the note that the Muñoz copy is incomplete. This belongs to the existing synoptic-table item, not a new one; the amendment is a text edit to R002 whose current summary is not in the pa…

    reopen if: R002's summary is exposed for a scalar correction.

  • researchduplicateAMZ-R005AMZ-R005 exists with claimIds [AMZ-C010, AMZ-C017].

    Report item 8 asks to add Acuña 1641 and Fritz's journal as the 1639 and 1686–1723 corridor witnesses to AMZ-R005. This is the existing item's content; the sources are logged blocked above with routes.

    reopen if: R005's summary is exposed for a scalar correction.

2026-09-10check5 of 5 seats judged the case blind$0.94

run 2026-09-10-check-pre-columbian-amazon-190925

This run wrote no disposition rows.

2026-09-09reportresearch pass (claude-fable-5-1)failed$0.00

run 2026-09-09-report-pre-columbian-amazon-212116

This run wrote no disposition rows.

Change history

Every change to this case, newest first: what changed, why, and who made it — including the AI's role. The log is append-only; a correction is a new entry, never an edit to an old one.

  1. claude-fable-5-1 (operator), answering review note #358 in the founder's session of 2026-09-19 · AI-assisted

    Review note #358 (GPT-5.6 Sol seat on #357, §3.2, §3.6, §3.8, §3.13), recorded against assessment 2026-09-19-edition-042009, which edition-2026-09-19-042009 adopts by content hash and which is therefore append-only and not edited: (1) the component note on the earthworks and the synthesis cite AMZ-E115 for the Upano chronology ("Upano dated 500 BCE–600 CE (E073, E115)"; "E073, E074, E115"), while AMZ-E115 is attached to AMZ-C086, the builder attribution, and only AMZ-E073 dates the occupation — evidence for one proposition spent on another; (2) whatIsClaimed counts fish weirs among the features that reveal settlement scale, while AMZ-C019, the fish-weir claim, is graded unresolved in the same assessment as held on its anchor alone with no admitted evidence record.

    Why: The seat is right on both. Neither can be corrected in the adopted assessment without breaking the edition's hash; the correction is the next edition run, whose protocol (edition-v10) now says that an evidence id is cited only for a claim it is attached to and that whatIsClaimed names only propositions admitted evidence carries, with anchor-only claims excluded or marked as such — and whose verb refuses a candidate whose assessment prose names an evidence record that does not exist or is not live. The case verdict does not turn on either sentence: AMZ-C061 is dated by AMZ-E073 alone in its own treatment, and the fish-weir claim is graded unresolved.

  2. claude-fable-5-1 (operator), in the founder's session of 2026-09-19 · AI-assisted

    Correction by the operator: the builder-attribution claim added on 2026-09-17 as AMZ-C066 is AMZ-C086, and its evidence added as AMZ-E078 is AMZ-E115; AMZ-E115 cites AMZ-C086; the disposition rows written on 2026-09-17 name the new ids (a row must name a record the ledger holds) and say so; the edition's article reference is amended a second time. The three entries of 2026-09-17 above stand exactly as written that day.

    Why: The re-submission proposal 2026-09-17-reverify-pre-columbian-amazon-160112, now in the repository's run records (#353), gave AMZ-C066 and AMZ-E078 to other candidates, both refused at verification; one id must name one thing across proposal, verification and ledger records (the GPT-5.6 Sol seat's point on #353, applied here before merge). The two records take ids no proposal or ledger file has used.

  3. claude-fable-5-1 (operator), answering review note #345 in the founder's session of 2026-09-17 · AI-assisted

    Amendment by the operator to edition edition-2026-09-17-142106 (field: article), recorded in the edition's amendments: the sentence that cited AMZ-C061 for both the dating and the builders of the Upano platforms and roads now cites AMZ-C061 for the dating and AMZ-C066 for the attribution.

    Why: Review note #345. The adopted assessment 2026-09-17-edition-142106 is append-only and keeps its plain-language line for AMZ-C061 ("The people who built the Upano Valley platforms lived there…"), which after the split spans AMZ-C061 and AMZ-C066; the next edition run re-draws it. The edition's rationale describes the ledger as it stood when the candidate was written and is left as written.

  4. claude-fable-5-1 (operator), answering review note #345 in the founder's session of 2026-09-17 · AI-assisted

    Added by the operator: claim AMZ-C066 (the Upano Valley platforms and roads were originally built by groups of the Kilamope and later Upano cultures) and evidence AMZ-E078 (Rostain 2024, p. 4, quoted alone), the builder attribution split out of AMZ-C061 and AMZ-E067; three disposition rows record the texts as admitted.

    Why: The verify verb had located the attribution sentence on p. 4 of the author manuscript and refused the part only as not atomic (verification 2026-09-17-verify-pre-columbian-amazon-140447, part 2 of the AMZ-E067 split). The operator re-read the manuscript on 2026-09-17: the archimer.ifremer.fr host did not answer, so the same PDF was taken from the Internet Archive's Wayback copy of that URL and read with the pipeline's PDF reader, whose page numbering the ledger's locators use; the quoted span stands verbatim on p. 4 and the supplement's sentences summarised in the inference on pp. 23–24. No second model has read these two records; the constitutional panel judges the pull request that carries them, and the source's label is unchanged.

  5. claude-fable-5-1 (operator), answering review note #345 in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: AMZ-C061.statement — "Excavations in the Upano Valley (Ecuador) date the Kilamope/Upano occupation that built the earthen platforms and roads from around 500 BCE to between 300 and 600 CE." → "Excavations in the Upano Valley (Ecuador) date the Kilamope/Upano occupation from around 500 BCE to between 300 and 600 CE.".

    Why: Review note #345 (GPT-5.6 Sol seat on #340, §3.2, §3.6, §3.9): the claim bundled the occupation's dating with who built the platforms and roads, and its only admitted evidence, AMZ-E073, states the dating — so the measurement's support was silently extended to the attribution (the verify verb had refused the drafter's evidence record for exactly that bundling and admitted the dating part alone). The attribution is now its own claim, AMZ-C066, with its own evidence.

  6. claude-fable-5-1 (operator), answering the GPT-5.6 Sol seat's objection on #353 in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: the claim added today as "AMZ-C078" (Oviedo's endorsement rests also on what those who had dealt with him affirmed) is AMZ-C085 — claims.yaml id "AMZ-C078" → "AMZ-C085"; AMZ-E114.claimIds ["AMZ-C078"] → ["AMZ-C085"]; the two disposition rows written today that named AMZ-C078 name AMZ-C085. The two entries above that say "AMZ-C078" stand as written and are corrected by this one.

    Why: The GPT-5.6 Sol seat objected on #353 (§3.8, §3.14, §3.15): the re-submission proposal 2026-09-17-reverify-pre-columbian-amazon-160112 had given AMZ-C078 to a different candidate — the Markham/Herrera claim, blocked at verification — so the proposal, its verification.md and the ledger used one id for two claims. The operator's claim takes an id no proposal or ledger file has used; AMZ-C078 remains, in the immutable run records, the name of the blocked Markham candidate. Lesson recorded: an id given by hand must skip every id any proposal on the branch has used, not only the ledger's.

  7. claude-fable-5-1 (operator), answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709) in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: AMZ-E086.reviewState — "ai_extracted" → "rejected" (not one observation: two grounds of Oviedo's endorsement); added AMZ-E113 (the wounds, supports → AMZ-C072) and AMZ-E114 (the character testimony, supports → AMZ-C078), each with its own verbatim quote from the same passage (Medina 1894, Introducción, p. xiii and n. 6).

    Why: With AMZ-C072 split, the evidence that bundled both grounds is split the same way, as the verify verb splits a compound record: the parent refused with the reason, the parts appended with the lineage on their origin. The operator located both quotes verbatim in the archive.org djvu text the run read; no second model has read the two parts, and the constitutional panel judging #353 is their check.

  8. claude-fable-5-1 (operator), answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709) in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: AMZ-C072.statement — "Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests on the friar's two arrow wounds and on what those who had dealt with him affirmed of his character." → "Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests on the friar's two arrow wounds."; its anchor keeps the wounds quote and drops the character quote. Added claim AMZ-C078 ("Oviedo's endorsement of Carvajal's credibility, as quoted by Medina, rests also on what his authority and person deserve, as affirmed by those who had dealt with him."), anchored on "ques mucho segund afírman los que le han tractado" (Medina 1894, Introducción, p. xiii).

    Why: The GPT-5.6 Sol seat objected on #353 (§3.2) that AMZ-C072 combined two independently testable grounds — the wounds and third-party testimony to character — in one claim. Oviedo gives both in one sentence, but each is a proposition with its own truth condition; split, each carries its own anchor and its own evidence.

  9. claude-fable-5-1 (operator), answering the GPT-5.6 Sol seat's objection on #353 (panel-v3, judged at 86a8709) in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: AMZ-E107.reviewState — "ai_extracted" → "rejected" (a duplicate of AMZ-E109: the same sentence of Medina, quote and claim, admitted twice by the run under two directions); AMZ-E108.reviewState — "ai_extracted" → "rejected" (not one observation: it bundles AMZ-E109, AMZ-E110 and AMZ-E111, which stand). Each carries the reason as a limitation; disposition rows record a duplicate and a refusal.

    Why: The GPT-5.6 Sol seat objected on #353 (§3.2, §3.4, §3.5, §3.6) that AMZ-E108 remained compound after the split and that AMZ-E107 and AMZ-E108 supported AMZ-C017 wholesale while their siblings from the same page entered as context. On inspection AMZ-E107 quotes the sentence AMZ-E109 quotes, and AMZ-E108 restates the three propositions the AMZ-E092 split had already separated; the run had read one page twice through two parents. The tombstones leave one record of each observation, under the direction the run gave the AMZ-E092 split (context for the complete-faith judgment, context for the supernatural attributions, qualifies for the partisanship). No direction was changed by hand.

  10. claude-fable-5-1 (operator), correcting the sitting's record in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: the evidence record "Medina's two caveats on the Relación: supernatural attributions and partisanship for Orellana — part 2 — toward AMZ-C076" (supports → AMZ-C076) — id "AMZ-E110" → "AMZ-E112". A disposition row under its key names AMZ-E112. The record carrying "context" toward AMZ-C017 keeps AMZ-E110.

    Why: Verification 2026-09-17-verify-pre-columbian-amazon-160114 split part 2 of the re-submitted AMZ-E092 by direction and wrote both records under AMZ-E110: the callback that hands a split-off record its id read no list holding the part itself (fixed in code: freshEvidenceIds and a pre-write duplicate-id check in verify). The ledger failed the build's own uniqueness rule at the sitting's edition step, and three seats of the constitutional panel parked the sitting for the collision (Opus 5, GPT-5.6 Sol, Gemini 3.8 Flash on #353). The run's own entry above and its verification.md list AMZ-E110 twice as written; this entry is the correction. Nothing else in either record changes: same quote, locator, direction, claims, reader acts and lineage.

  11. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1 in 2026-09-17-draft-pre-columbian-amazon-135349 on 2026-09-17, re-submitted by 2026-09-17-reverify-pre-columbian-amazon-160112 under resubmit-v1 (the lineage on each claim's and evidence record's origin and in each source's verification note)) · AI-assisted

    Intake from report proposals/2026-09-17-report-pre-columbian-amazon-134558/report.md (re-submission 2026-09-17-reverify-pre-columbian-amazon-160112 of records blocked at verification): 0 source(s), 16 evidence record(s), 7 claim(s), 0 research item(s) verified and added (proposal 2026-09-17-reverify-pre-columbian-amazon-160112, verification 2026-09-17-verify-pre-columbian-amazon-160114); 31 candidate(s) rejected with reasons in dispositions.yaml.

    Why: Admitted after verification: AMZ-E080, AMZ-E099, AMZ-E100, AMZ-E082, AMZ-E086, AMZ-E105, AMZ-E106, AMZ-E107, AMZ-E108, AMZ-E089, AMZ-E091, AMZ-E109, AMZ-E110, AMZ-E110, AMZ-E111, AMZ-E093, AMZ-C068, AMZ-C069, AMZ-C072, AMZ-C074, AMZ-C075, AMZ-C076, AMZ-C077; everything else proposed was refused or blocked with a reason in dispositions.yaml. The drafter's rationale for the proposal, written before verification and describing what it proposed: Re-submission of 35 record(s) that verification blocked in proposals/2026-09-17-draft-pre-columbian-amazon-135349 — the Internet Archive did not serve the OCR text of SRC-CARVAJAL-1542 during this run and retrieval read the item page in | anchor source not retrievable (no source id on the anchor) and no accepted evidence cites the claim | the Internet Archive did not serve the OCR text of SRC-CRUZ-1651 during this run and retrieval read the item page instea — proposed again under fresh ids against the ledger as it stands by the reverify run 2026-09-17-reverify-pre-columbian-amazon-160112, a deterministic transformation (resubmit-v1) and not a model's work. The records were drafted by claude-fable-5-1 in 2026-09-17-draft-pre-columbian-amazon-135349 on 2026-09-17: on each claim's and evidence record's origin, and at the head of each source's verification note, with the lineage appended. The texts are fetched again and the records read as newly proposed.

  12. claude-fable-5-1 (operator), answering review note #341 in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: AMZ-E074.exactLocator — "[p. 4], supplement S2 via open-access author manuscript" → "[p. 4], via open-access author manuscript"; a reader act on the record carries the same.

    Why: Review note #341: the verify verb's mechanical narrowing removed the supplement's page marker ("[p. 24]") but left the words "supplement S2" standing, so the locator kept a component the part does not quote while its own reader act said "the part quotes p. 4 only". The narrowing code is corrected separately so the words go with the marker.

  13. claude-fable-5-1 (operator), answering review note #341 and the panel's park in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: SRC-CARVAJAL-1542.verification — "ai_verified" → "unverified"; SRC-CARVAJAL-1542.verificationNote rewritten to say what was read, by whom, and what was not. The drafter (proposal 2026-09-17-draft-pre-columbian-amazon-135349) read the Medina 1894 OCR text's introduction, pp. i–lxxviii; the Relación body lay past the draft's 120,000-character read window; the verifier (2026-09-17-verify-pre-columbian-amazon-140447) could not retrieve the text and checked no quotation from it.

    Why: The entry of 2026-09-17 above raised the label to ai_verified on the drafter's reading while the retained note still said the primary text had never been fetched, and the same commit's correction records that the verifier read the item page, not the text: the stamp, the note and the run's account contradicted one another. The constitutional panel parked the sitting for exactly this (Opus 5, GPT-5.6 Sol and Grok 4.5 seats; §3.8, §3.14, §3.15, §4) and the GPT-5.6 Sol seat's review note #341 asked that the label and note be scoped to the pages actually inspected. Under this ledger's convention ai_verified is the verify verb's word after it has read the text and checked the quotes; that has not happened, so the label returns to unverified and the note carries the drafter's reading exactly. It will be raised by the verify verb when the blocked Carvajal rows are re-submitted against the served OCR text.

  14. claude-fable-5-1 (operator), correcting the sitting's own record in the founder's session of 2026-09-17 · AI-assisted

    Correction by the operator: 44 disposition rows written by verification 2026-09-17-verify-pre-columbian-amazon-140447 as "failed" (quoted span or anchor quote not found verbatim in SRC-CARVAJAL-1542 or SRC-CRUZ-1651) are set to "blocked" with a route. The Internet Archive did not serve either book's OCR text during that run and retrieval read the item pages instead, which are viewers, not the texts; the quotes were checked against nothing.

    Why: A failed row says the source did not say it; these sources were not read. The retrieval fault is fixed in code (the item page is no longer passed off as the text); the rows keep their proposal reference so a re-verification can re-submit the records.

  15. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1) · AI-assisted

    Intake from report proposals/2026-09-17-report-pre-columbian-amazon-134558/report.md: 2 source(s), 6 evidence record(s), 5 claim(s), 2 research item(s) verified and added (proposal 2026-09-17-draft-pre-columbian-amazon-135349, verification 2026-09-17-verify-pre-columbian-amazon-140447); 58 candidate(s) rejected with reasons in dispositions.yaml.

    Why: Admitted after verification: SRC-ROSTAIN-2024, SRC-ACHMATOWICZ-2018, AMZ-E073, AMZ-E074, AMZ-E075, AMZ-E076, AMZ-E072, AMZ-E077, AMZ-C061, AMZ-C062, AMZ-C063, AMZ-C064, AMZ-C065, AMZ-R007, AMZ-R008; everything else proposed was refused or blocked with a reason in dispositions.yaml. The drafter's rationale for the proposal, written before verification and describing what it proposed: This pass reads five texts. (1) Medina's 1894 introduction to the Carvajal Relación (the ledger's SRC-CARVAJAL-1542, now shown): it documents Orellana's own lost 1543 relación and the origin of the 'mentirosa' charge, Medina's judgement that the T'Serclaes and Oviedo texts are one author's, the incomplete Muñoz copy, the second-hand basis of Oviedo's endorsement, Oviedo's oral debt to Orellana, Carvajal's truth-oath and Medina's two caveats, and Herrera's near-verbatim dependence on Carvajal. (2) Markham 1859 (front matter and introduction): its Orellana text is Herrera's, not Carvajal's; it fixes the Ursúa expedition at Machiparo at New Year 1561, glosses Machiparo as near the Putumayo, records the Omagua/El Dorado rumour that preceded 1560, and summarizes Jesuit reports of smallpox at Borja (1660) and Portuguese slave raids on the Omagua after 1640 — all secondary and weighted as such. (3) Laureano de la Cruz's 1651/53 account, listed by the report as an unopened lead but retrieved here in full: a Franciscan resident among the Omagua for three years to October 1650 describes a province of 34 small villages, a smallpox epidemic that by his count killed a third at one village, the absence of chiefs, and Portuguese parties on the river — the first seventeenth-century witness on Carvajal's own corridor to enter this ledger; the headcount figure is garbled in the OCR and is flagged, not asserted. (4) Rostain et al. 2024 (full author manuscript): Upano occupation dated ca. 500 BCE to 300/600 CE with Huapula reoccupation 800–1200 CE, which bounds how far AMZ-C022's time-graded-erasure point can be spent against contact-era negatives. (5) Achmatowicz 2018 (abstract only): a scholarly assessment of Carvajal as a trustworthy source. Claims are split one proposition each; secondary sources enter with weak or moderate strength and explicit limitations. No edition change: the disappearance section should, after verification, acknowledge Laureano as the Omagua-reach witness, but the current telling is not shown wrong. Blocked routes (Oviedo lib. L, Acuña, Fritz, Harris 2017, Prümers 2022, Gómara) are logged with the pages Medina supplies.

  16. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1; proposal 2026-09-17-draft-pre-columbian-amazon-135349, verification 2026-09-17-verify-pre-columbian-amazon-140447) · AI-assisted

    Correction applied from proposals/2026-09-17-draft-pre-columbian-amazon-135349: SRC-CARVAJAL-1542.verification — "unverified" → "ai_verified".

    Why: SRC-CARVAJAL-1542: The full text of the Medina 1894 edition at the record's URL (archive.org/details/descubrimientod00carvgoog, djvu full text) was retrieved and read in this pass (introduction, pp. i–lxxviii); the Relación body beyond the introduction was truncated but the item's text is now shown and quoted, with page-level locators.

  17. claude-fable-5-1 (operator), answering review note #285 in the founder's session of 2026-09-11 · AI-assisted

    Correction applied from review note #285 (GPT-5.6 Sol seat on #283): the dangling-id check must hold for every record: AMZ-R006.summary — "Select LiDAR survey blocks at random within strata defined by the 2023 basin-wide probability surface (0–1%, 1–10%, 10–25%, ≥25%) and, inside the Aquiry area, by the 2026 study's regional model — choosing blocks before any archaeological knowledge of them is consulted and registering the selection. Fly and inspect each block with a fixed detection protocol and independent expert confirmation, then record earthworks per km² per stratum. Compare the observed densities with the densities each model implies for those strata. Observed densities in the 0–1% stratum at or near zero and rising monotonically with predicted probability would support AMZ-C028–C030 as calibrated; systematic under-prediction inside the Aquiry area would favour the 2026 regional total in AMZ-C004 over the basin-wide one, and the reverse would favour the lower. Either result narrows the several-fold spread between published earthwork totals that AMZ-E025 records." → "Select LiDAR survey blocks at random within strata defined by the 2023 basin-wide probability surface (0–1%, 1–10%, 10–25%, ≥25%) and, inside the Aquiry area, by the 2026 study's regional model — choosing blocks before any archaeological knowledge of them is consulted and registering the selection. Fly and inspect each block with a fixed detection protocol and independent expert confirmation, then record earthworks per km² per stratum. Compare the observed densities with the densities each model implies for those strata. Observed densities in the 0–1% stratum at or near zero and rising monotonically with predicted probability would support the basin-wide model's calibration claims (proposed at intake and not admitted) as calibrated; systematic under-prediction inside the Aquiry area would favour the 2026 regional total in AMZ-C004 over the basin-wide one, and the reverse would favour the lower. Either result narrows the several-fold spread between published earthwork totals that AMZ-E038 records.".

    Why: AMZ-R006: prose named a record id the ledger does not hold — the drafter's provisional claim ids never entered the ledger; the drafter's provisional evidence id never entered the ledger (review note #285: no dangling ids, everywhere)

  18. claude-fable-5-1 (operator), answering review note #275 on the record, in the Claude Code session session_01AiXD4y5Y91KC7iM3uBhWmu; PR #278 · AI-assisted

    Clarification of the intake entry of 2026-09-10 (proposal 2026-09-10-draft-pre-columbian-amazon-184613, verification 2026-09-10-verify-pre-columbian-amazon-185127). Its reason is the drafter's rationale for the proposal, written before verification: it names six observation claims, a provisional Oviedo source, deforestation and road-bias evidence and an independent dating corroboration as proposed, not as admitted. What the verification admitted is what the entry's change line counts: one source (SRC-PERIPATO-2023), four evidence records (AMZ-E038, AMZ-E039, AMZ-E040, AMZ-E041), two claims (AMZ-C029, AMZ-C033) and one research item (AMZ-R006); the rest of the proposal was refused or blocked with reasons in dispositions.yaml. The URL added to SRC-CARVAJAL-1542 was applied by the verify run itself (the correction entry above it, actor aletheia verify); the run's working report called that correction 'NOT applied' because its check compared an absent field with a null value and called them different — a fault in the report, not in the record, fixed the same day.

    Why: Review note #275 (the GPT seat on #273, §3.14, §3.15): an intake entry whose reason reads as what entered, when it is what was proposed, misstates the change. The verify verb now labels the rationale as the drafter's and states the admitted set beside it (the same day's fix); this entry says so for the one already written, since history is append-only.

  19. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1) · AI-assisted

    Intake from report proposals/2026-09-10-report-pre-columbian-amazon-183738/report.md: 1 source(s), 4 evidence record(s), 2 claim(s), 1 research item(s) verified and added (proposal 2026-09-10-draft-pre-columbian-amazon-184613, verification 2026-09-10-verify-pre-columbian-amazon-185127); 22 candidate(s) rejected with reasons in dispositions.yaml.

    Why: The report's centre of gravity — direct reading of the Medina 1894 edition of Carvajal and of Oviedo's redaction — cannot enter this pass: the only retrieved texts for those works are archive.org item pages (bibliographic metadata, copyright status, download options), not the OCR. Every quotation the report attributes to Medina pp. 8–54 or Oviedo pp. 552–558 is therefore unshown here and is dispositioned blocked with the download route, not written into the ledger. What the item pages do allow: a url correction on SRC-CARVAJAL-1542 (the ledger holds it key-less; the page confirms the work, editor, publisher, and year) and a provisional, unverified Source record for Oviedo t. IV so the witness has a home when its text is read. The report's declared failure — Peripato et al. 2023 could not be opened — is reversed by the packet: the accepted manuscript was retrieved via an open-access repository and is read here in full. It supplies six new observation claims (LiDAR discovery count, basin-wide undiscovered-earthwork prediction, the low-probability share of the basin, the Acre concentration, the domesticated-species association, and the deforestation/road sampling bias), a comparator to AMZ-C004's regional prediction, independent corroboration of AMZ-C020/C016 on proportion, and a second, independent statement that Acre earthworks date to AD 120–400 (qualifying AMZ-C017). Independence is flagged: Peripato shares authors with three ledger sources. No edition change: the text-critical rewrite the report wants depends on quotations not yet in hand, and Peripato deepens the record without altering the telling.

  20. aletheia verify (claude-sonnet-5 second reader; drafter claude-fable-5-1; proposal 2026-09-10-draft-pre-columbian-amazon-184613, verification 2026-09-10-verify-pre-columbian-amazon-185127) · AI-assisted

    Correction applied from proposals/2026-09-10-draft-pre-columbian-amazon-184613: SRC-CARVAJAL-1542.url — "null" → "https://archive.org/details/descubrimientod00carvgoog".

    Why: SRC-CARVAJAL-1542: The ledger holds the primary source with no url or key. The report reached the text at this archive.org item; the retrieved item page confirms it as 'Descubrimiento del rio de las Amazonas' by Gaspar de Carvajal, ed. José Toribio Medina, Sevilla: Impr. de E. Rasco, 1894 (Harvard copy digitised by Google; copyright status NOT_IN_COPYRIGHT; FULL TEXT download available). This is the standard scholarly edition of the 1542 account; the verifier can only read what a record points at.

  21. scripts/migrations/2026-09-09-hash-locators.ts (mechanical re-stamp) · AI-assisted

    Ledger hash redefined to exclude source locators (url, archivedUrl); 1 edition/assessment basis stamp(s) recomputed mechanically, content unchanged: edition-2026-09-08-migration.yaml.

    Why: Founder decision 2026-09-09: a locator says where a record is read, not what it says; a by-hand URL addition had set a fresh panel aside and made an edition due that could only re-adopt.

  22. Cursor Grok 4.6 (Cursor cloud agent), directed by the founder; runId title-punctuation-2026-09-08; promptVersion none (founder-directed session, not a pipeline prompt) · AI-assisted

    Case title punctuation restored after the pre-overhaul publication restore: "The Emptied Amazon?" is again "The Emptied Amazon". The trailing question mark is dropped so the published name is a statement. Subtitle and framing questions are unchanged.

    Why: Founder direction (first given 2026-09-06, reaffirmed 2026-09-08 after the restore): case names should not carry question marks. The interrogative belongs in the subtitle and the research questions, not on the card and in the header.

  23. scripts/migrations/2026-09-08-editions.ts (mechanical), founder-directed session · AI-assisted

    Migrated to editions (docs/AUTOMATION.md, build step 2). The article, the 20-claim featured set, and the displayed judgment now live in edition edition-2026-09-08-migration, which adopts the migration assessment 2026-09-08-migration-pre-columbian-amazon. Claim records carry propositions and anchors only; evaluative fields moved into the adopted assessment's per-claim treatment; the dossier header, research priority, and component verdicts moved from case.yaml into the assessment; claims-catalog.yaml merged into claims.yaml (tiers ended). No judgment changed. Where a claim record's hand-written credibility differed from the draft overlay's verdict, the record's value (what claim pages displayed) was carried: AMZ-C019: record well_supported, draft provisionally_supported.

    Why: One reader-facing unit that binds verdict, selection, and article, so they cannot drift apart; one home for judgment; featuring becomes a recorded decision.

  24. editorial AI (claude-fable-5, cursor session amz-expansion-2026-08-28) · AI-assisted

    Expansion wave to full brief coverage: engineered fisheries (featured), plus catalog-tier claims for the machine-learning extent predictions, the soil-core sampling critique, the Upano signal-erasure result, epidemic fade-outs, route-identification uncertainty, and the connectivity/linguistics assessment; six new sources, one evidence record.

    Why: Founder disposition policy (2026-08-28): every brief family carries an explicit disposition. The coverage-gap verification batch (ledger A18–A24) corrected three of the brief's guessed citations (Blatrix in Scientific Reports; Walker in PeerJ; the Clement critique in the reply, not the article) and FAILED one item — the circulated Purus/Coari identification of Carvajal's "Trinity River" is unsourced and was dropped from the case and the corridor-test protocol wording.

  25. editorial AI (claude-fable-5, cursor session amz-case-build-2026-08-28) · AI-assisted

    Case created: dossier framing, 18 claims (two carrying the new genealogy block), 21 evidence records, 20 sources, 5 research items, overview article, cover art, and four licensed plates including Figure 2 of the open-access Pärssinen et al. 2026 study.

    Why: Founder-directed case construction from a chat-derived research brief. Every load-bearing cell was natively re-verified before entry (30+ item verification ledger, 2026-08-28); the verification pass corrected the brief in three places (the Nunes letter date, the contact-era cluster's warfare element, and a PNAS quotation) rather than importing it on trust. The founder's funding connection to the central 2026 study is disclosed on the source record and in the article.