awesome-thought-experiments

Batch 7: Three Risk-Ranked Entries Source-Checked

Outcome: All three entries are source_checked for their declared claim-traceability scope after edits and independent read-only review. None is expert_reviewed; source checking does not establish that an argument is sound or settle a disputed interpretation.

What changed

• Split Brains and the Unity of Consciousness — The Pinto et al. (2017) summary now distinguishes what both patients did from the separate cross-hemifield comparison tested on Patient DDC only. The article names the exact Methods, Results, and Discussion passages consulted and keeps the authors’ unity claim identified as a contested interpretation. See the item evidence record.
• Goodman’s New Riddle of Induction (Grue) — The cutoff and simple-confirmation assumption are explicit; entrenchment is kept distinct from ultimate justification. Naturalness and Bayesian approaches are attributed narrowly, while the machine-learning and place-dependent examples are labelled as editorial extensions. The independent source review found no remaining material correction. See the item evidence record.
• Pereboom’s Four-Case Argument — Case 1 now describes moment-to-moment manipulation of decisions and actions. The Google Books note accurately says Chapter 4 is listed but its text is not exposed, and the unverified Internet Archive access claim was removed. The item is source_checked for the stated paraphrase scope based on full relevant passages in the SEP and Jeppsson; direct access to the 2001 chapter remains unavailable and is disclosed. See the item evidence record.

The three dataset records match their articles’ Further Reading URLs and review dates. No public review-status field was added to the dataset.

Review limits and audit count

The Four-Case item’s status does not imply direct verification against Pereboom’s 2001 Chapter 4: the Google Books preview exposes selected image-only pages, not the chapter text. A publicly hosted excerpt was encountered, but its publication container and edition identity remain unverified and it is not treated as evidence for the article’s material claims. For the Grue item, Fitelson’s manuscript text was accessible, but the publisher page for the final paper did not extract; the publication metadata was cross-checked separately. These limits are recorded in the item ledgers. No named human expert reviewed this batch.

After this batch, 29 of 57 expansion entries have claim-level audits; 28 remain outside these audits. Work remains local, uncommitted, and unpushed on audit/source-review-high-risk.

The consolidated batch evidence ledger contains the three full item records and review details.

This site is open source. Improve this page.