# Atlas authoring Batch 23: corrected-release promotion

**Date:** 2026-08-28

**Status:** authorized from immutable `second_pass_v6_675`; ordinary authoring
batch; no new statistical judgment and no recursive review

## Scope

This batch makes the public atlas inherit the completed post-v5 correction
cycle. It refreshes three already visible cases whose public prose or cards
were corrected, authors six previously unrendered active v6 events, and closes
the old public-authoring hold for the overbroad Indo-Aryan wrapper because v6
has retired that wrapper from the model.

Refreshed visible cases:

1. `prehistory-iberia-steppe-ancestry-turnover`
2. `ancient-egypt-thutmose-iii-megiddo-captive-communities`
3. `prospective-amenhotep-ii-year-7-campaign-captive-removal`

New visible cases:

1. `prospective-amenhotep-ii-year-9-campaign-captive-removal`
2. `ancient-hittite-mursili-arzawa-deportation`
3. `early-neo-assyrian-ashurnasirpal-nipur-pasate`
4. `early-neo-assyrian-ashurnasirpal-tumme-habhu`
5. `early-neo-assyrian-ashurnasirpal-upper-tigris-nirbu-882`
6. `early-neo-assyrian-ashurnasirpal-zamua-881`

## Binding authority

- Active event identities, families, and eight-trait cards come from immutable
  release `second_pass_v6_675`.
- Public factual claims and caveats are bounded by each event's `summary`,
  `post_v5_source_calibration`, `mechanism`, and source list in
  `events_enriched.json`.
- The final three-stage Editorial Suite disposition is `PASS AFTER REVISION /
  PROMOTE WITH QUALIFICATIONS`.
- Earlier Batch 19 and 20 source reviews remain the authoring authorities for
  the three refreshed public cases.

## Visual and editorial rulings

1. Every new event receives one consolidated playback stop. Source-internal
   phases may be described within the stop, but they are not duplicated as
   adjacent phase cards.
2. Amenhotep II Year 9 shows forced movement from a broad Retjenu/Aphek
   campaign region into Egyptian custody. Memphis is a context marker for the
   king's recorded return, not a claimed final destination for every captive.
3. Muršili II shows a broad forced-removal route from Arzawa into Hatti and a
   simultaneous political-incorporation outcome. The royal aggregate of about
   65,000–66,000 is not added to interim captures or returned Hittite subjects.
4. Nipur-Pasate, Tumme-Habhu, and Zamua show an Assyrian operational approach
   to the affected source region. They do not draw captives to Assyria because
   the source does not identify a captive destination.
5. Upper Tigris/Nirbu keeps its different cohorts separate: prisoners of
   unknown destination, Assyrian famine returnees placed at Tushha, local
   returnees to their own towns, hostages, and later fugitives. Schematic
   internal routes are labeled as such and do not become a single stream.
6. Royal military figures remain attributed claims rather than civilian
   censuses. Unknown resident denominators, return, mortality, descendants,
   and labor assignments remain unknown.
7. Approximate ancient place anchors orient the reader; labels explicitly say
   when a point is a broad or schematic regional anchor rather than an exact
   excavated location.

## Indo-Aryan wrapper

`atlas-indo-aryan-migration` is no longer an active event in v6. Its earlier
hold, sources, and decision trail remain in the audit files, but the active
atlas-authoring register no longer treats it as a pending visible case. No
replacement redirect is created because no single bounded successor has been
admitted as its public substitute.

## Expected public effect

If validation confirms one stop per new case and no unrelated roster change,
the atlas moves from 307 cases / 317 stops to 313 cases / 323 stops. The builder
and inventory, not this expectation, are authoritative for the final count.