{
  "title": "When should an anomaly change an explanation?",
  "evidence_type": "Reviewed general reading path — English editorial slice",
  "prepared_at": "2026-10-07",
  "question": "A prediction fails. Should we question the observation, revise the model, or change the research programme?",
  "answers": [
    {
      "author": "Karl Popper",
      "answer": "A claim must risk an observable failure. Accepting a falsification also requires a testable refuting effect; a stray contradictory observation alone does not settle the case.",
      "source_title": "The Logic of Scientific Discovery, chapter 4, sections 19–22",
      "locator": "Chapter 4, especially section 22; excerpt printed pages 57–66",
      "url": "https://fenix.ciencias.ulisboa.pt/downloadFile/1688987299217625/Falsiability.pdf",
      "source_nature": "University-hosted chapter excerpt; underlying excerpt edition not established",
      "anonymous_check": "Opened without authenticated repository access on 2026-10-07"
    },
    {
      "author": "Imre Lakatos",
      "answer": "Evaluate the sequence of revisions: does it lead to new predictions that later receive support, or only accommodate facts already known? A young programme may need time; that is not unlimited protection.",
      "source_title": "Science and Pseudoscience, 1973 talk",
      "locator": "Four-page transcript, pages 2–4; editorial bracketed additions distinguished",
      "url": "https://www.inf.fu-berlin.de/lehre/pmo/eng/Lakatos-Science.pdf",
      "source_nature": "University-hosted transcript of the 1973 Open University talk",
      "anonymous_check": "Opened without authenticated repository access on 2026-10-07"
    },
    {
      "author": "Thomas S. Kuhn",
      "answer": "Researchers can share values such as accuracy and scope while applying and weighting them differently. The absence of a compulsory choice algorithm does not make every choice equally defensible.",
      "source_title": "Objectivity, Value Judgment, and Theory Choice",
      "locator": "Collection chapter 13; existing collection read checked anthology pages 357–358 and 362–363",
      "url": "https://press.uchicago.edu/ucp/books/book/chicago/E/bo5970650.html",
      "source_nature": "Publisher catalogue verifies chapter identity; the catalogue is not the essay text",
      "anonymous_check": "Publisher catalogue opened on 2026-10-07; this round did not reverify the course scan text"
    }
  ],
  "disagreement": "These are different units and purposes of evaluation: a testable refutation, the progress of a programme, and reasoned choices within a community. The reading path does not fuse them into one verdict procedure.",
  "boundary": "An editorial comparison of bounded source readings, not a new historical study or a validated personal decision method. Original texts remain copyrighted; no scans, source bodies, long quotations or private personal material are distributed.",
  "next_question": "What dated record would distinguish a new prediction from a revision made after the result?",
  "provenance": {
    "collection": "The Ancestors",
    "repository_access": "Private; repository access is not a visitor-facing link",
    "publication_basis": "The owner requested a public general collection slice. This export rewrites only the reviewed general reading path in English and links to external sources. Private revision, user archives, chats and relationship material are excluded.",
    "execution": "Bounded authenticated file reading and anonymous source-link checks. No reader study was executed."
  },
  "collection_snapshot": {
    "as_of": "2026-10-06",
    "structured_records": 503,
    "exhibits": 67,
    "reviewed_relationships": 381,
    "source": "Internal collection audit, 6 October 2026",
    "publication_basis": "Owner-approved aggregate counts only; the private audit is not distributed.",
    "boundary": "Historical collection scale and review status, not a claim of learning outcomes or that all records were reviewed."
  }
}
