W6 address pour — extract-v1 over the W6 corpus: 96 figure / 5 event / 118 place / 11 people candidates offered (existing rows untouched), 3,805 attestations
receipt:addresses-w6b · v1 · ingest_verified
sha256 75801b9c6c93de7ff6d1a69b682f88aee6b2bee4464aa89f54d4fd441c6836d1
A receipt records what the house consulted and what it did not establish. The open state is part of the record, not a defect in it.
- Perseus Digital Library — Tacitus, The Annals (Church & Brodribb translation)Alfred John Church & William Jackson Brodribb (1876); Perseus Project digital editionlicense: public-domain (the Church & Brodribb translation, 1876); Perseus digital edition CC BY-SA, attribution recorded per ops/corpus-staging/SOURCES.md patternpinned: github.com/PerseusDL/canonical-latinLit @ 76b87b04b36afe2cbcd70e285b4ceb248b103438 (fetched 2026-07-08); file data/phi1351/phi005/phi1351.phi005.perseus-eng1.xml sha256 dbb7230ed993d9e030a13cb6a5915a0e4fdae2bf925e6b7fef424e907ea5ea6c; profile tac-perseus-v1 (ops/pipeline/). The shelf's B&N 2005 reprint scan was probed and retired for serving (unusable shipped text layer; no printed chapter numbers) — recorded in tacitus.works.json
Deterministic extract-v1 (unchanged prototype logic, W6 corpus scope) over the W6 episodes. 3,805 attestation rows carry exact surface forms and per-episode counts (receipt_id = receipt:addresses-w6 on the rows themselves). Existing addresses gain W6 attestations by slug identity and have their mention_count recomputed from the full attestation ledger; their curation-owned columns (labels, kinds, notes, states, indexability) are untouched by construction — the insert is ON CONFLICT DO NOTHING and the only UPDATE touches mention_count. New candidates are born candidate/unreviewed/noindex. The W6 unclassified pool and era phrases stay local (D-LEXICON-001; era is book-gated W3 law).
What this does not prove- extract-v1 is frequency + pattern extraction — not NER, not curation; misfiled rows exist at the margins (this batch files 'Senate' as a figure candidate, the same class of margin W1's 'Roman' sat in).
- This batch's extraction scope is the Annals alone (extract-v1's gates are corpus-relative; batch 1 was poured first and stays append-only) — candidate counts are not comparable across batches.
- Attestations attach to existing addresses by surface-slug identity, not by curation: 'Nero' in the Annals attaches to the same candidate row Suetonius' Nero fed — homonym mixing is possible on any shared surface.
- mention_count is corpus-coverage arithmetic (full-ledger sum for figures/places/peoples the wave touches); event mention_counts of pre-existing events keep their original wave's scope.
- figure_kind is 'unassigned' on every new candidate — person|legendary|deity is a curation verdict.