KW_Conservativity_v14: version correction and review history
September 30, 2026

Current source SHA256: C9894E37B4FC558BCAE1801D37DD98E3309BAA6BA3A5EEBF58084FF1F285C0D3

The restored-abstract/context revision was initially delivered under v13 in
error. It is now v14. The only source change since that editorial review is
the PDF subject's version number. Exact source comparison, all-page PDF text
comparison, and all-page rendered image comparison confirm preservation of
the reviewed content and layout. These are version/preservation checks, not
a new mathematical audit. The original v13 has been restored byte-for-byte.

The historical reviews below retain their original names and hash bindings.
References there to the current editorial source under v13 describe the
misnumbered delivery now corrected to v14; earlier mathematical reviews concern
the original v13. No historical review is rebound to the v14 source hash.

PRESERVED REVIEW HISTORY

KW_Conservativity_v13: current editorial review and prior mathematical reviews
September 30, 2026

Current source SHA256: 8A72CBA9B506F438A543774A1E1303EAE87FC1F2F2A8712A0476B67FAB7747DB

The current revision restores the v12 abstract and reduces the new Chang
result to introductory context, moving its formal statement to Appendix D.
The current review below verifies editorial preservation and cross-reference
integrity. It is not a new mathematical audit. The three earlier mathematical
reviews that follow are historical records bound to the prior source hash
307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA.
Their theorem/page numbering refers to that earlier source. The theorem is
now D.1 and the open question is D.5. The proof and open question are unchanged.

CURRENT EDITORIAL REVIEW
Review SHA256: FB3B14793FFF946D8CB8DFACCC9DB5BB7DBF38621D3DAE32A36D65F33FB41209

EDITORIAL PRESERVATION REVIEW: PASS

Reviewed source: outputs/KW_Conservativity_v13.tex
SHA256: 8A72CBA9B506F438A543774A1E1303EAE87FC1F2F2A8712A0476B67FAB7747DB
Comparison source: work/v13/before_context_revision/KW_Conservativity_v13.tex
Abstract reference: C:/Users/glaze/Documents/Codex/2026-09-25/st/outputs/KW_Conservativity_v12.tex

Scope: editorial placement, wording, and exact preservation only. This review
does not independently audit the mathematical proof or its cited inputs.

1. The abstract is restored exactly to v12. Its original concluding sentence
   is preserved, and the added Chang paragraph is removed.
2. Sections 2 onward through Appendices A-C are exactly unchanged.
3. The formal definitions of E_1^C, CS, B_str, and sigma_str, the guard
   explanation, the labeled separation theorem, and the subsequent
   consistency-bound explanation are moved intact into Appendix D.
4. The entire prior Appendix D content from the imported mouse construction
   through its proof and anti-inner-model question is exactly unchanged.
5. The introduction gives the new result in one contextual paragraph. It
   retains the precise relative consistency assumption, identifies the
   distinction between theories (1) and (3), explains its relevance to the
   choice of HOD_{Ord^omega}, and leaves anti-inner-model restoration open
   with the corrected Mitchell-order threshold of at least two.
6. The formal theorem, equation, and question labels remain available for
   cross-references. Their numerical values may change when compiled.
7. The original keywords are restored, and the PDF subject now emphasizes
   external HOD and local arbitrariness.

Checks 1-4 were confirmed by exact text comparisons as well as inspection
of the complete editorial diff. No manuscript edits were made by this reviewer.

PRIOR MATHEMATICAL REVIEWS, PRESERVED VERBATIM

Independent reviews of KW_Conservativity_v13
September 30, 2026

Source SHA256: 307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA

These reviews cover the new separation result and its integration into the existing paper. They are mathematical and source reviews, not proof-assistant verification or a new independent audit of every unchanged theorem in v12.


============================================================
Review: final_separation_audit.txt
Review SHA256: 2CC737AF14CE382547C10B20B8E90AD0338E94E4B9435B56090E988F29975DC9
============================================================

Final v13 separation integration audit
September 30, 2026

Verdict: PASS for the strong-cardinal separation and its integration.
No mathematical correction is required within the reviewed scope.

Exact reviewed source:
outputs/KW_Conservativity_v13.tex
SHA256 307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA

Comparison sources:
outputs/Chang_Rank_Separation.tex
SHA256 3FA336AF505F4B3B86A8C9451E8E948866B194176DC6737FAE46A3E2E2B12AB5

C:/Users/glaze/Documents/Codex/2026-09-25/st/outputs/KW_Conservativity_v12.tex
SHA256 69BDAF03C3201DBEAEA465059875C32062117EA6F4D05C8FB10C707BAD435621

Checks

1. The abstract, introduction theorem, and Appendix D agree on the
   conclusion. No full Chang rank in a ZFC ambient universe satisfies
   ZFC+CS. The fixed sentence makes E_1^C inconsistent. Consistency of
   E_3 for that sentence is asserted only relative to B_str, consisting
   of ZFC, unbounded strong cardinals, and an inaccessible limit of
   strong cardinals. The draft expressly distinguishes this bound from
   ZFC+CM. The ambient CM assumption is unused in the prohibition.

2. The notation is consistent. E_1^C is newly defined with its witnessing
   rank computed in Chang. E_3 retains its existing ZF+DC+CM definition
   and ambient witnessing rank. The proof consistently uses these names
   rather than accidentally modifying the original external-HOD E_1.

3. The displayed sentence requires ordinal lambda and nu, lambda<nu,
   and nu limit. Its hierarchy and set-satisfaction interpretation is
   retained, including all recursive ZFC axioms. The guard provides
   lambda+omega<=nu<kappa in the outer witnessing rank. The realization
   remains V_(delta+omega*2), with lambda=delta and nu=delta+omega.

4. Appendix D preserves the audited mathematical proof. It retains the
   complete Mitchell datum, the bounded ambient-countable closure of W,
   equality of reals and omega_1, the countable-hull linear-iterability
   transfer, and the external height-cofinality check. The forbidden-
   datum argument retains internal set-stage term evaluation, the
   permitted indiscernible shift, and parameter-definability over Chang
   needed for Suzuki's ZF theorem. The bounded-extender argument for
   strongness inside V_delta retains its truncated representatives and
   target-rank factor-map check. Relabeled references resolve to the
   corresponding new lemmas and theorem.

5. The anti-inner-model discussion is an explicit open question, with
   A_2 appended to both ambient theories. It uses ordinary normal-measure
   Mitchell order, spells out the excluded two-measure configuration,
   and distinguishes the semantic formulation from a future fixed
   mouse or scheme formalization. It does not claim restoration, infer
   no Kukri from A_2, or assert separation from ordinary measurables.

6. The full comparison against v12 shows no edits to its existing
   mathematical proof sections or its first three appendices. Changes
   are confined to the declared metadata and notation, the abstract,
   the Chang discussion and proof-outline cross-reference, new Appendix
   D, and the relevant bibliography. This is an edit-scope check, not a
   new audit of the unchanged v12 arguments.

The substantive theorem review and primary-source checks are recorded
in work/v13/strong_separation_review.txt. This final audit is bound to
the exact v13 source hash above. Compilation and layout verification
are separate from this mathematical integration review.

Final editorial refresh

The three final edits remove the preliminary-PDF comparison, rephrase
the appendix opening to state the B_str assumption directly, and remove
the unused omega_2-shortcut paragraph. All are mathematically harmless.
Reversing exactly these edits in memory reproduces the previously
audited source hash
8286755E907EED34EF462AF1C752E1A8632B834BF1515D43C04325130AC39195.
Thus no proof step, theorem, bound, or other text changed in this refresh.


============================================================
Review: final_sources_audit.txt
Review SHA256: E0C82C83573EA9FE88B38567A1947C844B9E6E2DE180B86F3819BAB7B432AF49
============================================================

Final v13 source and integration audit
September 30, 2026

Reviewed file: outputs/KW_Conservativity_v13.tex
SHA256: 307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA

Verdict: PASS for the new strong-cardinal Chang appendix, its integration and references, and the stated open status of anti-inner-model restoration. This is not a new audit of every inherited theorem in the full paper. No manuscript edits were made.

1. Exact integration

The appendix proof, from the imported mouse construction through the separation proof, matches outputs/Chang_Rank_Separation.tex at SHA256 3FA336AF505F4B3B86A8C9451E8E948866B194176DC6737FAE46A3E2E2B12AB5 after the documented changes of section levels, citation keys, labels, real-set notation, B to B_str, and T_i to E_i, together with two editorial changes. The source-introduction sentence was shortened and the paragraph warning against an omega_2 shortcut was removed. A whitespace-normalized comparison allowing precisely these changes returned equality, with 10,556 characters on each side. The separate appendix-opening revision states directly that the relative consistency direction assumes B_str. These changes alter no theorem or proof step.

The integrated proof retains the full Mitchell datum, the external height-cofinality check, the countable-hull argument restricted to linear iterations, the stabilized set-stage evaluation of type-1 terms, and the first-order definition of the particular embedding inside Chang. The guarded sentence and the bounded-extender consistency argument are retained. Thus the integration has not silently strengthened the accepted proof or removed one of its safeguards.

2. References

The appendix uses the valid keys MitchellChang, GappoSargsyan, and Suzuki. Their entries point to the exact sources checked in work/v13/strong_sources_review.txt, SHA256 66D8C1B8D9EC1E2F0B92A82C287FF158FBC70F9EFC0F37D497226F24B91C65DE.

MitchellChang is the current arXiv 1705.00208 manuscript, with the correct published journal entry. The appendix cites Theorem 1.5, Sections 1.3 and 3, Definition 4.1, Proposition 1.4, and Definition 3.9 for their previously audited roles. It does not substitute the preliminary author draft. GappoSargsyan is cited explicitly for its introductory proper-class-of-strongs existence bound. The manuscript does not attribute that bound to their numbered determinacy theorem. Suzuki is used for a parameter-definable elementary self-embedding inside a ZF model; this matches the source's scope and does not assume Choice in Chang.

The source-wide mechanical reference check found 34 bibliography keys and 34 distinct cited keys, with no missing citation key, duplicate bibliography key, missing referenced label, or duplicate label. The new question environment and CS command are defined. The appendix refers to the existing E_3 theory, not to an undefined replacement symbol.

3. Introductory theorem and consistency scope

The abstract and Theorem chang-rank-separation agree with the appendix: no full Chang rank in an ambient ZFC universe satisfies ZFC+CS; the guarded instance of E_1^C is inconsistent; and Con(B_str) implies Con(E_3(sigma_str)). B_str includes both a proper class of strong cardinals and an inaccessible limit of strong cardinals. The text does not claim to obtain this instance from CM alone. The previously established global no-SVC obstruction is kept distinct from this rank theorem.

4. Open problem and draft-only claims

The concluding question appends the same semantic A_2 restriction to both ambient theories and explicitly requires any formal treatment to fix the same mouse or scheme formulation. It uses the standard threshold o(nu)>=2, which permits ordinary order-one measurable cardinals. It does not confuse the restriction with the absence of measurable limits of measurables.

The abstract, introductory discussion, and appendix all state that restoration is open. The text expressly asserts no equiconsistency under A_2 and leaves measurable-strength Chang separation open. No bounded-base core inversion, conditional comparison construction, unproved tail-shape claim, or other recent research-draft result has been inserted as a new restoration theorem.

5. Verification limit

This review inspected the TeX source and its integration. It does not claim a separate PDF compilation or rendered-page audit. The parent's reported successful local compilation and the built-in compiler's directory-lookup failure are independent of this reference audit.


============================================================
Review: final_integration_audit.txt
Review SHA256: B783ADEBF3A506A10FAE1B28E79D5A4E885C97704CA53F37652B11287B727FF5
============================================================

FINAL INTEGRATION AUDIT -- V13
=============================
Verdict: PASS for integration of the accepted strong-cardinal separation
into the full v12 paper. No required correction found in the audited source.
This is an integration audit, not a claim that the open A2 comparison or
any unfinished measurable-level restoration argument has been proved.

Audited file: outputs/KW_Conservativity_v13.tex
SHA256: 307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA

Baselines:
Original v12: 69BDAF03C3201DBEAEA465059875C32062117EA6F4D05C8FB10C707BAD435621
Accepted Chang source: 3FA336AF505F4B3B86A8C9451E8E948866B194176DC6737FAE46A3E2E2B12AB5

1. Statements and consistency strength
-------------------------------------
The abstract now states failure of the two-way Chang comparison and makes
the strong-cardinal consistency assumption explicit. The introduction
preserves E_1,E_2,E_3 and adds E_1^C, with no parallel T_3 notation.
The stale assertion that the unrestricted E_1/E_3 Chang comparison is
unsettled has been removed. The remaining discussion correctly distinguishes
the no-SVC proposition from the stronger full-rank obstruction.

The separating sentence at lines 426--437 quantifies ordinal lambda and
nu, requires nu to be a limit ordinal above lambda, and retains the
recursive-ZFC set-satisfaction interpretation. The height guard
lambda+omega <= nu < kappa is explicit. It has not been weakened.

B_str at lines 422--423 includes ZFC, global CS, and an inaccessible limit
of strong cardinals. Theorem thm:chang-rank-separation says that no full
Chang rank is ZFC+CS in any ZFC ambient universe; E_1^C(sigma_str) is
inconsistent; and Con(B_str) implies Con(E_3(sigma_str)). It does not
assert Con(B_str), a measurable-only consistency bound, or a new
anti-inner-model theorem. The distinction between the negative theorem's
ambient assumptions and the positive consistency bound is explicit.

2. Complete proof retained
--------------------------
The new Appendix app:chang includes all four parts of the accepted proof:
full Mitchell datum and approximation system; fixed-set countable closure
and linear iterability transfer; internally definable term shift plus
Suzuki; and the guarded rank contradiction with V_delta realization.
The full reals and cofinality checks, stabilized set-stage term evaluation,
and truncated-function argument for strongness in V_delta remain present.
No proof step was replaced by the global failure of SVC or by an unproved
Chang-locality assertion.

A direct text comparison of the accepted proof from 'The imported mouse
construction' through the end of the main proof, after the documented
section-level, notation, citation-key and theorem-label substitutions, plus
the two editorial changes within that proof range described below,
returned exact equality: both normalized strings have 10,590 characters.
The research-history conclusion was replaced by a neutral mathematical
summary. The theorem is announced in the introduction, its full proof is
in a clearly identified appendix, and the roadmap points to it.

3. Original HOD paper preserved
------------------------------
A direct text comparison from the start of 'Small forcing and
generated-model presentations' through all pre-existing appendices,
stripping only final whitespace and a terminal clearpage, returned exact
equality between v12 and v13: 176,680 characters on each side.
Thus the original forcing construction, core preparation, structural
conservativity and three HOD consistency implications have not been
silently changed. No later experimental bounded-base, linked-tail,
Chang-monotonicity or decoder claim has been imported.

4. Open A2 question
-------------------
The closing anti-inner-model subsection defines A2 with the standard threshold o(nu)>=2, computed
inside an inner ZFC model. The equivalent normal-measure configuration
U_0 in Ult(M,U_1) is stated with the ultrapower using M's functions.
The semantic inner-model convention and need to fix a corresponding
formalization on both sides are explicit; no unproved choiceless coding
equivalence is asserted.

Question q:chang-anti-inner quantifies over every fixed first-order sigma
and displays exactly
 Con(E_1^C(sigma)+A2) <-> Con(E_3(sigma)+A2).
The following paragraph expressly appends A2 to both ambient theories and
identifies the two locations of the witnessing ranks. It correctly allows
order-one measurables and measurable limits of measurables. It asserts no
restoration result. The additional question about lowering the separating
consistency strength to ordinary measurables remains appropriately open.

5. References and mechanical checks
----------------------------------
The existing MitchellChang entry is updated to the final publication;
GappoSargsyan and Suzuki are added. The source uses the final Mitchell
construction rather than treating a bare long-extender mouse as the full
datum. The imported labels are namespaced and do not collide with the
original structural theorem's thm:main.

Automated source checks found:
* 93 labels, no duplicate labels, and no unresolved ref/eqref/pageref keys.
* 34 bibliography entries, no duplicate keys, and no unresolved cite keys.
* The final source hash still matches the audit target after inspection.

The introduction writes L[Ord^omega] and the appendix L(Ord^omega), but
both explicitly define the same Chang model; this is harmless typography,
not a substantive discrepancy. Normalizing it later would be optional.

Compilation and rendered-page review are separate production checks, not
claimed to have been performed by this audit. No edits to the manuscript
were made during this review.

Editorial rebind, final source
-----------------------------
The verdict remains PASS after reviewing the three editorial changes from
source 8286755E907EED34EF462AF1C752E1A8632B834BF1515D43C04325130AC39195
to the currently audited source
307BB2F6E34174488E6903616CA523FE34166137391BC96E780D7B41765ABCFA:

* The appendix opening now says that the relative consistency direction
  assumes B_str; the exact definition and theorem remain unchanged.
* The imported-construction opening now says 'We use Mitchell's Chang-sharp
  construction. The relevant results are ...', without the editorial
  contrast with the preliminary PDF. The references remain identical.
* The freestanding comment about an unsafe omega_2 comparison is deleted.
  The countable-hull iterability proof itself is unchanged.

The normalized proof equality reported above allows exactly the latter
two editorial changes within its comparison range. The first is before
that range and was read directly. The unchanged-original-body comparison
and reference-key checks were rerun successfully on the final source.
No theorem, hypothesis, or mathematical proof step was altered by this
cleanup, and no manuscript edits were made by the auditor.
