Provenance Record

Well-structured News article covering a legitimate May 2026 Nature study on paleoproteomics. The four-section layout (What We Know, Why Enamel Proteins, What We Don't Know, Analysis) is clear and appropriately scoped. Tone is neutral and avoids sensationalism. The article correctly hedges the interbreeding hypothesis as probabilistic, not conclusive.

Five of seven sources were verifiable from snapshots: source-2 (The Conversation, verified), source-3 (ScienceAlert, verified), source-4 (Scientific American, verified), source-5 (Live Science, verified), source-6 (Nature DOI, verified). Source-0 (phys.org) returned HTTP 403 and source-1 (archaeologymag.com) timed out — no snapshots available for those two. Key facts attributed to unavailable sources were independently confirmed against the Nature paper snapshot (source-6): Hexian and Sunjiadong sites, AMELY sex-determination technique, and AMBN-A253G/M273V variants all appear in the Nature paper. The John Hawks quote ('It's like, wow. The data just had to line up exactly right for this to happen.') was confirmed verbatim in the Scientific American snapshot. The AMBN-M273V frequency (21% Philippines, ~1% India) was confirmed in the Nature paper at 21% Philippines and 1.17% India — the article's 'around 1 percent' is acceptable rounding. The paper's phrase 'their shared habitats create opportunities for interactions' was confirmed verbatim in the ScienceAlert snapshot. One misattribution found: the article states 'Using micro-destructive acid etching to preserve the physical structure of the teeth... [as reported by Live Science]' but the Live Science snapshot (source-5) does not use the term 'micro-destructive acid etching' — it confirms only the extraction of 11 proteins and hundreds of amino acid positions. The Nature paper (source-6) uses 'acid etching method' for a localized small-area extraction, which substantively supports the description but does not use the 'micro-destructive' descriptor. This is a subordinate methodological detail misattributed to the wrong source; the underlying substance is accurate per the primary source. ScienceAlert and doi.org/Nature are reputable outlets not yet on the allowlist — both should be added.

All headline claims and lead facts are supported by the available sources. The core scientific narrative (11 proteins from 6 H. erectus teeth, AMBN-A253G unique to East Asian H. erectus, AMBN-M273V shared with Denisovans and present in ~21% of Filipinos, gene flow hypothesis) is accurate and well-sourced. H. erectus spread date (1.8 million years ago) and extinction estimate (~108,000 years ago) confirmed by Live Science. Sally Christine Reynolds affiliation (Associate Professor in Hominin Palaeoecology, Bournemouth University) confirmed by The Conversation. The 'braided river' metaphor attributed to Reynolds is confirmed by The Conversation snapshot.

APPROVE_WITH_CORRECTIONS. The article is factually strong, well-structured, and covers a legitimate peer-reviewed Nature study accurately. The only editorial issue is a minor misattribution: the 'micro-destructive acid etching' description is attributed to Live Science, but that snapshot does not contain this term. The underlying fact is confirmed by the Nature paper. A corrections note documenting this misattribution is sufficient to publish with an honest public trail.