# DARVO Article — Evidence and Review Ledger

**Version:** 1.3 · **Date:** 2026-09-15  
**Manuscript:** [The DARVO Family: Self-Protection and the Logic of Responsibility Distortion](DARVO_ARTICLE.md), draft 0.4.  
**Purpose:** Identify the current theoretical argument's sources and formal checks, with historical case and implementation records preserved separately. This ledger is working documentation, not additional manuscript prose.

## 1. Authority and completed-work boundary

The manuscript follows [DARVO_ARTICLE_PLAN.md](DARVO_ARTICLE_PLAN.md). Its conceptual authority is [canon v1.0](DARVO_FAMILY_CANON.md); its complete-response predicates follow [DARVO v1.2](DARVO_Accountability_Deflection_Self_Protection_Logical_Framework_v1_1.md). The PAST introduction uses [conceptual v1.1](PAST_Maneuver_Conceptual_Framework.md) and [extended v2.1](PAST_Maneuver_Extended_Logical_Framework_v2.md); the FATE introduction uses [FATE v1.1](FATE_Forecast_Actualization_Blame_Transfer_Logical_Framework_v1_0.md). The [writing principles](PATTERNA_WRITING_PRINCIPLES.md) and [scientific voice guide](RAFAEL_CAFERATI_SCIENTIFIC_VOICE.md) govern expression.

The current manuscript develops the conceptual and formal argument without a case study. It retains the canonical diagram, complete-response specification, five propositions, constructed countermodels, general incident derivation, and recurrence rule. Published research supplies the scholarly foundation; no original empirical analysis or implementation-performance result is presented in draft 0.4.

At the drafting-pass check on 13 September 2026, [the corpus folder](corpus/README.md) contained only the intake README. DR001 arrived later; its intake is recorded in the addendum below. The author's reported real-world observations and internal agent work are not silently converted into deposited manuscript evidence. Existing fictional Figma exchanges and application fixtures remain fictional.

**Preserved for later incorporation:** [Draft 0.2](DARVO_ARTICLE_DRAFT_0_2.md) retains the complete DR001 case and its within-case countergrievance contrast unchanged. The [DR001 corpus and analysis](corpus/DR001-domain-renewal/README.md) remain intact. The current omission is an editorial sequencing decision, not a retraction of that analysis. Sections 7–9 preserve the drafting, intake, and incorporation records; section 10 records the theoretical revision and section 11 its full correctness review. Fabricated urgency and the withdrawn technician-availability premise remain excluded from the preserved case derivation.

## 2. Claim-to-source map

| Claim | Manuscript location | Grounds and review obligation |
|---|---|---|
| Exact family definition and member inclusion | 3.1; abstract; conclusion | Canon v1.0. Adopted family taxonomy, with common incident scopes and fixed versions. |
| Shared operations and different objects of reversal | 3.2, Table 1 | Canon operation table. General responsibility reversal does not supply paired victim/offender roles. |
| Functional self-protection versus particular interests and mental states | 3.3; 5.5 | Canon section 4; DARVO source sections 11, 19–21. |
| Responsibility predicate and obligation basis | 5.1 | Coordinated plan section 2; DARVO source sections 5–6. Distinguish normative grounds from descriptive blame judgments. |
| Source attribution, accepted premises, and logical closure | 5.2 | Coordinated plan section 2.2; direct stated semantics; bounded pragmatic comparison in R09. |
| Complete-response form, support, and multi-person linkage | 5.3 | DARVO source sections 20.2–20.3; actual attack targets and principal counterpart retained separately. |
| Three-valued aggregation and review status | 5.4 | Explicit table and dependency rules; DARVO source section 21. Review precedes aggregation. |
| Particular protection, deliberate qualification, anticipation, enactment | 5.5 | DARVO source sections 20.4–20.7. Preserve L→Q, K→M, B→E and correctly bound target/J records. |
| D1–D5 | 6.1–6.5 | Constructed countermodels and deductions reviewed in section 4 of this ledger. |
| PAST feasibility and FATE non-entailment anchors | 3.2 | Corrected mechanism specifications and explicitly stated elementary deductions. R13/R14 supply bounded antecedents, not missing case evidence. |
| Scholarly origin and empirical findings | 2; 4 | Individually verified passages in section 3 below. |
| Complete-response derivation | 7.1 | Conjunction introduction from eight separately established predicates under common bindings, with consistency and dependency review. |
| Concurrent responsibilities and mechanism intersection | 7.2 | D1; canon inclusion and same-incident set semantics. Shared operations do not supply both complete mechanism definitions. |
| Recurrence and overlap | 7.2–7.3 | Canon and coordinated plan section 2.5. Definition and witness derivation only; no observed recurrence claim. |
| Research implications | 8 | Questions and evaluation requirements derived from the formal decomposition. No measured performance or effects asserted. |

The article's proposition numbering follows the dedicated article plan. It does not reuse the source document's numbering blindly: article D1 maps to source section 6.4; D2 to sections 20.3 and 24; D3 to source proposition D4 in section 20.7; D4 to sections 16 and 20.5; D5 to source propositions D2–D3 in section 17.

## 3. Verified scholarly passages

Reference IDs retain the coordinated plan's register. Full bibliographic entries appear in the manuscript's **References**. Verification below was performed on 13 September 2026. **Selected full text** means the identified passages were read; it does not claim a whole-publication critical review. **Abstract** means the manuscript claim is limited to the inspected abstract. Short excerpts locate the support; the article's account is paraphrased.

### R01 — Freyd (1997)

**Source/version:** [Author-hosted original journal scan](https://dynamic.uoregon.edu/jjf/articles/freyd97r.pdf), *Feminism & Psychology*, 7(1), 22–32. **Read:** selected full text; title/bibliographic page 22 and printed pp. 29–30, visually inspected. PDF pages 5–6 contain the founding passage. **Locator excerpt:** “Figure and ground are completely reversed.” **Use:** origin, linked operations, and denial/guilt distinction in manuscript section 2. The paper's family expansion is not attributed to this source.

### R02 — Harsey, Zurbriggen, and Freyd (2017)

**Source/version:** [Author-lab publication record and reproduced abstract](https://dynamic.uoregon.edu/jjf/traumapapers.html), 2017 entry; [publication DOI](https://doi.org/10.1080/10926771.2017.1320777). **Read:** abstract; publisher full text could not be retrieved. **Locator excerpt:** “increased perceptions of self-blame.” **Use:** section 4.1's sample and reported associations only. No effect size, item-level analysis, or full-method conclusion is drawn from this access. Original full-text methods/results remain a reading task.

### R03 — Harsey and Freyd (2020)

**Source/version:** [Author-hosted online-first PDF](https://dynamic.uoregon.edu/jjf/articles/hf2020.pdf); final citation 29(8), 897–916. **Read:** methods/results; numbered pp. 4–6 and 9–13 (PDF pp. 5–7, 10–14), including Table 1. **Locator excerpt:** “perpetrator as less believable.” **Use:** section 4.1's outcome directions, admission/remorse control, and education manipulation. Internal PDF pagination is not final journal pagination. No education effect size is reported.

### R04 — Harsey and Freyd (2023)

**Source/version:** [Author-hosted HTML-print PDF](https://dynamic.uoregon.edu/jjf/articles/hfjiv2023.pdf); final citation 38(17–18), 9985–10008. **Read:** PDF pp. 5–9, 12, methods/results/comparison. **Excerpt:** “rated the perpetrator as more believable.” **Use:** section 4.1. **Discrepancy:** abstract N=230; Method and randomized cells N=235; gender-restricted complete-case analysis N=227. The article omits N. PDF pages differ from journal pagination. Opposite credibility directions are not evidence isolating their cause.

### R05 — Harsey, Adams-Clark, and Freyd (2024)

**Source/version:** [PLOS ONE version of record](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0313642), 19(12), e0313642. **Read:** Method/Participants, Correlations, Discussion, and Limitations and future directions. **Locator excerpt:** “findings are based on correlational data.” **Use:** section 4.1's total-score associations and unresolved directionality. The manuscript does not claim that every subscale has identical associations or that the regressions establish causal direction.

### R06 — Durland, Harsey, and Freyd (2026)

**Source/version:** [Author-hosted journal PDF](https://dynamic.uoregon.edu/jjf/articles/dhf2026.pdf), 27(2), 221–237; online 11 December 2025, issue year 2026. **Read:** abstract; measurement pp. 222–223; Results pp. 230–232; Table 2 p. 231; limitations p. 235. **Locator excerpt:** “DARVO-SF consists of 3 Deny items.” **Use:** section 4.1's instrument and symptom-association description. Distinction from transcript certification is this article's methodological comparison.

### R09 — Grice (1975)

**Source/version:** [Stanford-hosted chapter scan](https://web.stanford.edu/class/psych205/papers/Grice-1975.pdf), *Syntax and semantics*, volume 3, pp. 41–58. **Read:** visually inspected pp. 41–45; implicature discussion on pp. 43–45. **Locator excerpt:** “what he had said.” **Use:** section 4.3's said/implicated and contextual-reference distinction. The linked DOI in References identifies the chapter's later hosting; manuscript locators use the scanned original pagination.

### R11 — Bandura (1999)

**Source/version:** [Full-text manuscript reproduction](https://www.antoniocasella.eu/nume/Bandura_1999.pdf), checked against the [abstract record](https://pubmed.ncbi.nlm.nih.gov/15661671/). **Read:** manuscript pp. 5–7 and 11–12, **Displacement of Responsibility** and **Attribution of Blame**. **Excerpt:** “faultless victims driven to injurious conduct.” **Use:** section 4.2's theoretical comparison. Manuscript pagination is not journal pp. 193–209; no individual participant's mental state is inferred from the theory.

### R12 — Malle, Guglielmo, and Monroe (2014)

**Source/version:** [Author-hosted journal PDF](https://research.clps.brown.edu/SocCogSci/Publications/Pubs/Malle_Guglielmo_Monroe_2014_Blame_theory.pdf), 25(2), 147–186. **Read:** pp. 151–152, 155–156, 158–159, 175–176, Path Model and blame management. **Excerpt:** “causality, intentionality, and obligation.” **Use:** section 4.2's conceptual distinctions. Their model concerns blame judgments; the manuscript supplies its own explicit grounds for responsibility assessment.

### R13 — Merton (1948)

**Source/version:** [University-hosted original journal scan](https://eclass.uowm.gr/modules/document/file.php/NURED263/Merton%20The%20Self-Fulfilling%20Prophecy%201948%20Antioch%20Review.pdf), 8(2), 193–210. **Read:** pp. 194–196 and 199–200; definition and apparent vindication together on p. 195. **Excerpt:** “proof that he was right.” **Use:** section 4.2 credits both self-fulfillment and vindication. FATE's explicit conditions remain the project's contribution.

### R14 — Halpern and Pearl (2005)

**Source/version:** [Author-hosted manuscript, 24 October 2005](https://www.cs.cornell.edu/home/halpern/papers/actcaus.pdf), linked to the [journal record](https://doi.org/10.1093/bjps/axi147), 56(4), 843–887. **Read:** section 2.1, manuscript pp. 3–4, equations/variables/interventions. **Excerpt:** “support a counterfactual interpretation.” **Use:** section 3.2's bounded causal-model statement. The article adopts no complete actual-cause decision procedure; conference and journal formulations are not silently equated.

### R18 — Sykes and Matza (1957)

**Source/version:** [Mirrored original journal scan](https://www.antoniocasella.eu/nume/sykes%26matza_1957.pdf), 22(6), 664–670. **Read:** visually inspected pp. 666–668 and bibliographic cover. **Excerpt:** “precede deviant behavior.” **Locators:** anticipation, pp. 666–667; Denial of the Victim and Condemnation of the Condemners, p. 668. **Use:** section 4.2's historical antecedents. Theoretical formulation is distinguished from experimental proof of DARVO-specific prior reliance.

### 3.1 Comparison and nonretained references

The 2020/2023 comparison is worded as an observed difference under unmatched comparisons, not proof that the controls or violence type caused the directional difference. This was checked independently and the causal-sounding wording was corrected. The manuscript does not report a combined effect size or use a single sample count across different analyses.

The original register's R07 (von Wright), R08 (Kripke), R10 (McCornack), R15 (Halpern/Pearl Part II), R16 (Walton), and R17 (Kleene) are not cited in this draft. Its current claims do not require their unexamined passages. The declared evidence-state table and elementary deductions are stated explicitly; no unverified historical attribution is made. These entries remain available in the planning bibliography for future argument-specific reading, not as decorative manuscript references.

## 4. Formal proof and counterexample review

| Result | Proof check | Counterexample or boundary check |
|---|---|---|
| Family inclusion | All sets share the fixed incident universe, evidence, and versions. Inclusion is adopted taxonomy. | Membership never supplies complete-response predicates; overlap must be the same incident. |
| D1 | An interpretation with both independently grounded responsibilities falsifies automatic exoneration. | An additional valid discharge or dependency premise is not ruled out. Original answerability is not inferred from accusation. |
| D2 | False-allegation model has Q/D/A/V/O/L true, W/M false; accurate attack grounds and target linkage are stipulated. | A second model has the form and W/M true. Same configuration does not require identical transcripts. |
| D3 | No audience-outcome variable occurs in the conjunctions or fixed validation inputs. | New evidence changing a predicate is outside the fixed-input comparison; rejection alone changes no required premise. |
| D4 | Identical later response is compatible with different earlier expectation/influence histories. | Later conduct can contribute evidence but does not by itself establish earlier reliance. Later source recording can establish an earlier state when grounded. |
| D5 | Algebra cancels fixed b/c terms and leaves positive (p0−p1)ell. Strict choice result compares against the best fixed alternative. | Zero loss, unchanged probability, increase below best alternative, equality/tie, and changed decision assumptions are separated. |
| PAST anchor | Feasible witness contradicts universal absence in the same domain/time/objective/constraints. | An alternative procedure outside the asserted scope is not a counterexample to truthful narrow closure. |
| FATE anchor | Same forecast/outcome observations permit different stipulated authorship and responsibility accounts. | The anchor alone supplies neither all FATE predicates nor causal-to-normative responsibility allocation. |
| Recurrence | Two distinct, supported witnesses establish the existential claim. | Duplicate exports and same-incident overlap do not create recurrence. Unresolved inventory preserves lower bounds. |

### 4.1 Dependencies and assessment integrity

L→Q, K→M, and B→E are reviewed before affected conjunctions. A positive dependent with a false prerequisite is inconsistent; with an unresolved prerequisite it requires review. Review status is distinct from T/F/U. Valid source-grounded implication and contraposition establish their respective consequences after review; an unreviewed flag cannot promote an assessment automatically. Draft 0.4 makes the negative consequence ¬Q→¬L explicit without changing the classifiers.

A positive A requires an exact, supported principal/supporter target witness. A correctly bound J=false can be a legitimate negative paired finding. A join to different supplied records is an integrity defect. Missing evidence does not establish nonoccurrence. Case-specific S requires a named position and a grounded relationship. Generic self-protection, S, knowledge, purpose, prior influence, and effects remain separate assessments.

### 4.2 Review record

- **Argument:** nine sections preserve the family-led thesis, complete-response specification, and return to the original accountability issue.
- **Demonstration, original draft 0.1 review:** all five propositions and both family anchors include explicit scope and constructed reasoning; the case section then contained only its acknowledged scaffold. Draft 0.2's case incorporation is recorded separately below.
- **Source correctness:** original Freyd scan inspected; 2020/2023 methods compared; antecedents and precise reading scopes retained above.
- **Expression:** one authorial narrative moves from the accountability problem to its relations and deductions. Established findings use direct language. Source availability is kept in editorial notes and this ledger.
- **Independent reviews:** mathematical/evidence-boundary review and a separate 2020/2023 source check are recorded on completion below.

## 5. Historical implementation inspection and boundaries

The inspection below supported implementation descriptions in drafts 0.1–0.2. Draft 0.3 replaces those descriptions with theoretical evaluation requirements; the original inspection and fingerprints remain a dated provenance record, not a new verification of the current code.

Read-only inspection occurred on 13 September 2026. No application tests, model calls, corpus runs, deployment, or remote configuration checks were performed. Paths below are relative to `patterna-core/vrdx-fire/`; links resolve from this ledger's location.

| Manuscript statement | Inspected code |
|---|---|
| Discovery, extraction, and later model review use original messages. | [mechanismPipeline.ts](../../vrdx-fire/functions/src/analysis/mechanismPipeline.ts), lines 165–217, 355–460. |
| Validation precedes three-valued aggregation and certificate production. | [validation.ts](../../vrdx-fire/functions/src/mechanisms/validation.ts), lines 751–809; [evaluator.ts](../../vrdx-fire/functions/src/mechanisms/evaluator.ts), lines 439–499, 531–589; [logic.ts](../../vrdx-fire/functions/src/mechanisms/logic.ts), lines 4–60. |
| Complete DARVO exists separately from its form, alongside PAST/FATE. | [registry.ts](../../vrdx-fire/functions/src/mechanisms/registry.ts), lines 27–53, 202–290; evaluator lines 611–617. |
| Family membership uses accepted base findings and merges same-incident overlap. | [family.ts](../../vrdx-fire/functions/src/mechanisms/family.ts), lines 78–150. |
| Family recurrence uses supported distinctness and lower bounds. | family.ts, lines 165–245; pipeline lines 931–939 do not certify exhaustive incident inventory. |
| Mechanism labels supply no automatic scoring points. | registry.ts, non-additive definitions; [common.ts](../../vrdx-fire/functions/src/analysis/common.ts), lines 199–220. |
| Code contains a production qualification gate. | [runtimeConfiguration.ts](../../vrdx-fire/functions/src/analysis/runtimeConfiguration.ts), lines 277–286; [prepareRuntimeBundle.ts](../../vrdx-fire/functions/src/scripts/prepareRuntimeBundle.ts), lines 38–45. |

The gate checks activation/qualification and matching definition, execution, and corpus-report records for nonlocal publication. Bundle-generation defaults are disabled/unqualified. These local properties do not establish the state of a remote deployment. [Implementation status](../../vrdx-fire/PAST_FATE_RUNTIME_IMPLEMENTATION.md), lines 164–170, records real-corpus qualification as outstanding.

The software checks referenced and reviewed assessments; reference existence and a model review do not independently guarantee empirical truth. No extraction-accuracy or psychological-effect claim follows from the source inspection. A reported evaluation would additionally require immutable input/configuration identifiers, retained outputs, expected assessments, annotation provenance, error analysis, and incident-separated development/evaluation sets.

### 5.1 Source fingerprints

| Inspected file | SHA-256 |
|---|---|
| `functions/src/mechanisms/family.ts` | `001a2dabe5cd4f7d8663c32354901f452e9c71c1b994d2ba536cf2208733ae39` |
| `functions/src/mechanisms/logic.ts` | `479795040b73391445a54913d613637da26bb1192d2376fe3cc8d3d0ab32e930` |
| `functions/src/mechanisms/evaluator.ts` | `648205a38f5b1ea760635c2bbcee1f3d4d38463e9558da465e233aa36555ba2a` |
| `functions/src/mechanisms/registry.ts` | `6456865013b888dff4a6e0527a90603ee55652990ec04fbb378bc7a0483e5b20` |
| `functions/src/mechanisms/validation.ts` | `65f5ca85b707fcb4822323f785999d3bdd6d6da6517b49581bb9bbe77e6343bc` |
| `functions/src/analysis/mechanismPipeline.ts` | `e14445bdb7a4c295a61732b1201db5af06a0109d3f98517b1089e6c19f6ae274` |
| `functions/src/analysis/runtimeConfiguration.ts` | `91b8e713af2a395b9e8a02968ade89a58be84856677f5fe1fad2d5285d49423d` |
| `functions/src/scripts/prepareRuntimeBundle.ts` | `24b8699c2fd0a31fb96f2d0546a12339ae26e375674db7650845bbe8056f32e9` |
| `functions/src/analysis/common.ts` | `0517be1684dc84bc4287e75cae86ce44bb9c772ea4a9a56e64a99cd57a5ccde6` |
| `PAST_FATE_RUNTIME_IMPLEMENTATION.md` | `79cd6155a601bc4b10d41785ca3fda22291e9496650d8cbcc2062296ba3aba6f` |

## 6. Deferred case material and outstanding research work

| Item | Current state | Required completion |
|---|---|---|
| CASE-A | Preserved in draft 0.2; excluded from current theoretical manuscript. | Later incorporation can reuse DR001-E01's eight-condition analysis and exact quotations, with source and definition versions rechecked. |
| CASE-B | Preserved as a countergrievance contrast within draft 0.2. | It remains a contrast within DR001, not a second independent case. A separate observed accurate-defense episode remains distinct future material. |
| Opening/concluding argument | Theoretical and complete. | No case placeholder or promised return is required in the current manuscript. |
| Particular protection, deliberate use, prior reliance | Formally specified; no current manuscript case assessment. | The preserved case's helper/defender finding and unresolved additional qualifications retain their original scopes in section 9. |
| Actual recurrence | No observed recurrence asserted. | Named distinct supported incidents, common operator, bounded history/cutoff/versions; exact count or supported lower bound. |
| Agent evaluation | No new performance result asserted. | Retained run evidence and independent expected assessments appropriate to the claimed output. |
| R02 full-text depth | Abstract-supported use only. | Examine original methods/results before adding claims beyond the verified abstract. |
| Final bibliographic/venue pass | Draft APA author–date references. | Journal-specific format and a literature update before submission; preserve consulted-edition locators. |

The theoretical revision replaces the case section with a general derivation, concurrent-responsibility analysis, mechanism-intersection semantics, and recurrence specification. Draft 0.3's length and verification remain in section 10; the current draft 0.4 review appears in section 11. Draft 0.2's 7,595-word inventory and case review remain in section 9.

## 7. Historical validation record — draft 0.1

**Completed 13 September 2026 for draft 0.1.** All checks below concern that version and its sources; they are preserved as historical results.

- Independent mathematical/evidence-boundary review completed with no remaining material concerns after repairs to the recipient-effect sentence, statement/content scope, and case evidence table.
- Independent 2020/2023 passage comparison completed; causal-sounding wording was replaced by an observation about unmatched comparisons. Sample and edition differences remain documented above.
- The nine manuscript sections, five named propositions, exact canonical definition, figure reference, mathematical delimiters, and local links were checked.
- All nine binary conjunction entries and all **6,561 raw eight-predicate T/F/U assignments** were checked independently of application code. Semantic admissibility was reviewed separately; the raw assignment count is not a count of admissible incidents.
- **108 exact-rational utility comparisons** checked the strict-increase identity, alongside zero-loss and choice-threshold boundaries. The general result is established by the algebra in D5; these finite checks do not replace its proof.
- Existing core documents and the diagram match their pre-draft SHA-256 snapshot. Only `DARVO_ARTICLE.md` and this ledger were added in the core folder. Inspected implementation hashes also match the read-only review packet.
- No application build, runtime test, live model operation, remote inspection, or corpus-performance evaluation was performed.

## 8. Historical corpus intake — 13 September 2026

The contributor supplied `DOMAIN_RENEWAL_ANONYMIZED_ENGLISH_STUDY.md`. [DR001's intake](corpus/DR001-domain-renewal/README.md) preserves the source, indexes all 179 entries, separates 47 reproductions from original statements, and records nine context statements with their provenance. Its [analysis](corpus/DR001-domain-renewal/ANALYSIS.md) supplies a bounded complete-response assessment, the non-cancellation application, relevant counterevidence, and explicit missing PAST/FATE premises.

At initial intake, DR001 was a candidate for CASE-A, with the discourse-level D/M inference available for review. It was not automatically an independent CASE-B, a second distinct mechanism incident, a completed PAST/FATE example, or a runtime qualification result. The assessment was performed by the assistant and had not been independently adjudicated. Manuscript incorporation remained separate work at that point. No manuscript prose or earlier scholarly verification record was changed by that intake; draft 0.2 incorporation is recorded below.

**Author-defined scenario scope — analysis v1.1:** all supplied contextual facts are true; the supplied text is the exact text being judged. Translation history and independent contextual corroboration are excluded from the assessment. The earlier intake's original-language alignment requirement is withdrawn. The nine context records are established premises; propositions inside the dialogue remain assertions to assess against them. This scope rule changes no earlier scholarly source-verification record or runtime-performance claim.

### 8.1 Historical draft 0.1 length

**Main-text count: 6,159 words.** The count covers sections 1–9, including headings, keywords, table text, captions, and the case-analysis scaffold. It excludes title/byline/status metadata, References, editorial blockquote placeholders, image alt text, and inline/displayed mathematics. Markdown links count by their visible labels. Word tokens contain letters/digits and may contain internal apostrophes or hyphens. This is an explicit manuscript counting convention, not a publisher's automated count.

| Section | Words |
|---|---:|
| 1. Abstract and keywords | 231 |
| 2. Accountability problem | 518 |
| 3. Family | 940 |
| 4. Literature and distinctions | 892 |
| 5. Formal specification | 1,513 |
| 6. Propositions | 1,021 |
| 7. Case scaffold and recurrence rule | 393 |
| 8. Research and implementation | 405 |
| 9. Conclusion | 246 |
| **Total** | **6,159** |

There are **12 scholarly references** and four grouped internal project-definition entries. The real-case section remains reserved; the draft is not represented as a completed empirical manuscript.

### 8.2 Historical draft 0.1 fingerprint

`DARVO_ARTICLE.md` SHA-256: `6540313e6f6d325a0c4d755b081c32096b93746cd27a217ac0649912872bc077`.

## 9. Historical DR001 incorporation — draft 0.2

The record below concerns the unchanged [draft 0.2 archive](DARVO_ARTICLE_DRAFT_0_2.md), not the current theoretical manuscript. Its source references, assessments, counts, fingerprint, and completed checks retain their original scope.

**Authority and source scope.** The contributor approved use of the case for the main DARVO example, the countergrievance contrast, and the opening/concluding return. The assessment uses [source.md](corpus/DR001-domain-renewal/source.md), [entries.jsonl](corpus/DR001-domain-renewal/entries.jsonl), and the [v1.3 analysis](corpus/DR001-domain-renewal/ANALYSIS.md). Manifest intake revision 5 retains the original nine context facts. The separate responsibility clarification is active; SUP01's speculative earlier technician availability is withdrawn. Neither that claim nor fabricated urgency enters the manuscript's argument. The bounded accountability issue is authored personal denigration and the subsequent treatment of Alex's complaint.

**Witness map.** W: A038/A044/A045/A053; Q: A057; D and A: B052 tied to A057 through B069; V: B052 with A070; O: B052; L: B052/B069/B094; M: the same response framing plus A038/A041 versus B053/B054. The attack target is Alex, and Claire is the audience and supporter. Denial concerns the complaint's validity as an accountability claim, not typing the sentences. Its contextual bridge is stated before the table. The table's accepted predicates are joined by conjunction introduction at the fixed incident E01; the Boolean step does not establish the interpretation independently.

**Selection and competing evidence.** A037/A038 supply adverse antecedents and A039 closes with “Be well.” All three precede the retained A040/A041 response sequence and are absent from the Day 4 forward. A046 is the additional omission inside A040–A081. The forward retains 41 entries, with 15 authored by Alex and 26 by Helen; it includes substantial Helen content and the complaint itself. B052 promises an account without distortion, not literally every message. The article grounds material distortion in the changed beginning combined with the accompanying disqualification and defensive representation. It does not infer deception merely from omission or length.

**Countergrievance and qualifications.** Alex's A057 closing taunt, A058, and A078 support a genuine countergrievance within the same case. They instantiate D1's non-cancellation application, not an independent CASE-B or an observed full accurate-defense model for D2. Helen's help offers and possible sincere concern can coexist with the documented response. The particular helper/defender position is supported; knowledge of the discrepancy, purposeful obstruction, prior reliance, and Claire's final uptake remain unresolved. Alex's rejection does not negate the response. One supported witness establishes a family-occurrence lower bound of one, not recurrence.

**Review.** A separate agent reviewed the source-to-predicate mapping and then the revised prose. The review confirmed the quotations, counts, scope, and contextual D/M bridge; the final equation was tightened to show each predicate's incident argument explicitly. This is an internal assistant review, not independent empirical adjudication or a provider-evaluation result. Existing scholarly claims, references, five propositions, framework definitions, and the diagram remain unchanged. No literature update, model analysis, or deployment is claimed by this pass.

**Reader packaging.** The manuscript is surfaced locally as article version 0.2, with a reviewed annotation hash and regenerated content. Earlier 0.1 downloads remain unchanged. Only approved manuscript excerpts enter the article; the full study, entry index, and temporary reader annotations remain outside public source assets. The ledger's corpus links are repository provenance, not public study downloads.

### 9.1 Draft 0.2 length

The count includes sections 1–9, headings, keywords, quoted case evidence, visible link labels, table text, and captions. It excludes front matter, references, image alt text, and inline/display mathematics. Tokens contain Unicode letters or digits and may contain internal apostrophes or hyphens. The historical draft 0.1 count above is retained as reported; this pass counts the current Markdown syntax tree.

| Section | Words |
|---|---:|
| 1. Abstract and keywords | 230 |
| 2. Accountability problem and opening case | 668 |
| 3. Family | 940 |
| 4. Literature and distinctions | 894 |
| 5. Formal specification | 1,513 |
| 6. Propositions | 1,021 |
| 7. Case analysis, contrast, and recurrence scope | 1,550 |
| 8. Research and implementation | 460 |
| 9. Conclusion | 319 |
| **Total** | **7,595** |

The bibliography retains 12 scholarly references and four grouped project-definition entries. No case placeholder remains in the manuscript.

### 9.2 Draft 0.2 fingerprint

`DARVO_ARTICLE.md` SHA-256: `b67adf7fc64cb94d63e40ddd9313633602bbade17776beaddac39b23ff208c69`.

### 9.3 Completed local validation

- Checked the four extended case quotations against A038, A041, A045, and B052, and resolved every cited case message ID against the 179-entry index. Confirmed the omission comparison and original-author counts in a separate source review.
- Confirmed sections 3–6 and the complete reference list are byte-identical to draft 0.1. Framework sources, the canonical diagram, the original case source/index, and archived 0.1 web downloads retain their pre-edit hashes.
- Counted 7,595 main-text words; no unfilled case or opening/conclusion insertion point remains. The only added display equation is the scoped case derivation; all original formal results remain unchanged.
- `npm run test:reader`: seven checks passed, including source fidelity, current inventory, contextual/citation links, mathematics rendering, and reader identity behavior. `npm run test:study`: four checks passed, including corpus integrity and exclusion of full temporary-study content from production output.
- Web lint and production build passed. Local Chrome checks passed for the full article, four responsive widths, contextual navigation, citations, language/position preservation, completion, and print behavior. The generated 24-page verification PDF contains the case evidence, eight-condition table, and within-case contrast without pending-case placeholders.
- No website deployment, live-provider analysis, corpus-performance evaluation, new scholarly-source verification, or narration recording occurred. The approved excerpts are part of the local article; the complete temporary case reader remains development-only.

## 10. Theoretical revision — draft 0.3 · 15 September 2026

**Editorial scope.** The author selected a theoretical-first manuscript, with the case to be added later. Draft 0.3 removes the case-dependent opening, abstract claims, case analyses, recurrence application, and concluding return. Section 7 develops the general complete-response derivation, concurrent responsibilities, same-incident intersection, and recurrence. Section 8 states research and evaluation requirements without presenting the project's implementation as an empirical result. No empty case placeholders remain.

**Preserved work.** The complete [draft 0.2 archive](DARVO_ARTICLE_DRAFT_0_2.md) retains SHA-256 `b67adf7fc64cb94d63e40ddd9313633602bbade17776beaddac39b23ff208c69`. The original corpus, its analysis, and historical records above remain available for later incorporation. The archive is not a current source for the article reader and is not added to its public source-asset allowlist.

**Formal continuity.** Sections 3–6 and the references retain their draft 0.2 text. The general section 7.1 derivation requires each of the eight predicates to follow from an explicit, consistent premise set under common bindings. Section 7.2 distinguishes shared family operations from both full mechanisms occurring in the same incident. The canonical definition, diagram, and detailed framework specifications remain unchanged.

### 10.1 Draft 0.3 length

The counting convention includes section headings, keywords, captions, table text, and prose, while excluding front matter, mathematical expressions, image alternatives, and references. Word tokens consist of Unicode letters or digits with internal apostrophes or hyphens retained.

| Section | Main-text words |
|---|---:|
| 1. Abstract and keywords | 224 |
| 2. The question that remains | 559 |
| 3. The self-protective family | 940 |
| 4. Scholarly origin and conceptual distinctions | 894 |
| 5. A formal framework for the complete response | 1,513 |
| 6. Five propositions | 1,021 |
| 7. Derivation, scope, and recurrence | 678 |
| 8. Research implications | 384 |
| 9. Conclusion | 240 |
| **Total** | **6,453** |

### 10.2 Draft 0.3 fingerprint

`DARVO_ARTICLE.md` SHA-256: `eb754a45e3008d78709cee56925c15a5fb01c9ecc3e645632b17f66a01b7d8e9`.

### 10.3 Review and local reader integration

- A separate read-only review found no material issues in the revised argument. It checked consistency, common bindings, premise establishment versus aggregation, incident intersection, and recurrence; the suggested terminology refinement from independent statements to distinct relationships was applied.
- Reader metadata and all eight contextual definitions are bound to draft 0.3's reviewed hash. The generated inventory is 167 content blocks, 14 display equations, three tables, one canonical diagram, and 16 reference entries. The general derivation replaces the case-specific derivation; the five named propositions retain their proofs.
- Source-preservation checks confirmed the archived manuscript, sections 3–6, references, canon, diagram, framework sources, corpus source/index, historical ledger records, and prior public source assets. Local links resolve in the current manuscript, archive, and ledger.
- `npm run test:reader`: seven checks passed for source fidelity, inventory, contextual/citation links, mathematics, and reader identity behavior. `npm run test:study`: four checks passed after the build, including complete temporary-study exclusion from production output. `npm run lint` passed.
- `npm run build:disabled` passed, including English and Portuguese article metadata for draft 0.3. The initial generic build correctly refused the local emulator-enabled configuration; the inactive production target preserves the public activation boundary.
- Local Chrome verification passed against Patterna's existing server on port 5174: complete article, original diagram, four responsive widths, context and citation navigation, focus restoration, language/position preservation, complete printing, and version-bound reader state. The initial port 5173 attempt reached a different project's server; neither server was changed.
- No new empirical finding, scholarly-source verification, runtime evaluation, or deployment is claimed by this revision.

## 11. Full correctness review — draft 0.4 · 15 September 2026

### 11.1 Findings and repairs

The review covered the complete manuscript, canon, corrected DARVO specification, scientific writing principles, authorial voice guide, scholarly claim map, and reader annotations. Separate formal, editorial, and scholarly reviews were reconciled with the manuscript. The central propositions survived review; no defeating countermodel was found under their stated premises.

| Finding | Repair and scope |
|---|---|
| The conclusion compressed D2 into “Configuration requires substantive support,” although the form is assessable separately. | State that a **supported complete response** requires both configuration and substantive support. The conjunction and D2 are unchanged. |
| The admissible interpretation class was implicit in “stated meanings.” | Fix meanings, actor/incident bindings, and background rules independently of the conclusion. Every interpretation satisfying them and the premises remains admissible, including a defeating countermodel. Nonempty consistency remains required. |
| Dependency discussion emphasized positive derivation. | Explicitly show both L→Q and reviewed contraposition ¬Q→¬L. This clarifies existing strict implication, not a new classifier or permission to promote raw flags. |
| Recurrence's category domain and occurrence identity were implicit. | Define μ as a mechanism or family category; e₁≠e₂ denotes different underlying occurrences. Each witness must be established and complete by the cutoff. Mixed-family recurrence does not establish repetition of every member. |
| The abstract could read as claiming successful protection; the PAST table could endorse the false attribution. | Scope protection to the represented account and state that past conduct **is presented as** the cause of impossibility. Mental states and success retain separate evidential requirements. |
| The conclusion contained an instruction about not softening findings. | State the logical result affirmatively. D2's repeated possibility formulations are replaced by the demonstrated compatibility of its two constructed models. |

The exact canon quotation, diagram, predicates, conjunctions, five propositions, scholarly claims, and bibliography remain unchanged. No empirical case was reintroduced. The earlier case archive and original corpus remain separate.

### 11.2 Formal and evidential checks

- D1–D5 and the PAST/FATE anchors were checked against their declared symbols, domains, fixed assumptions, proof constructions, and boundaries. Source-grounded responsibility remains distinct from mere allegation; a countermodel must satisfy the actual premises and background rules.
- All **6,561** raw eight-predicate T/F/U assignments were enumerated. The L→Q review policy withholds **1,458**; the remaining **5,103** produce **1 supported**, **191 unresolved**, and **4,911 false** results. Aggregation agrees with all admissible classical completions of these assignments under L→Q. These are Boolean/evidence-state assignments, not a count of grounded incidents or an exhaustive semantic model of communication.
- **63** finite remedy-domain checks, including the empty domain, verified the PAST quantifier relationship; a distinct outside remedy domain was retained as a scope countermodel. **180** exact-rational utility comparisons verified D5's identity and strict increase, alongside zero-loss, unchanged-probability, and choice-threshold boundaries. The general proofs remain the quantifier argument and algebra in the manuscript; finite checks supplement them.
- The constructed non-cancellation, form/substance, and prior-expectation models satisfy their stipulated premises. Fixed-input audience acceptance leaves classification unchanged. The FATE construction holds forecast and outcome fixed while varying stipulated authorship and correct attribution.
- Recurrence checks cover a single overlapping PAST/FATE incident, duplicate records, two distinct mixed-family witnesses, same-mechanism repetition, another operator, unfinished mechanisms, and completion after the cutoff. Excluding a countermodel merely for disagreeing with the conclusion fails the revised closure rule.
- A second formal review of the repaired §§5.2, 5.4, and 7.3 found no new ambiguity. The supplementary logical-closure and recurrence explanations adopt the same scopes.

### 11.3 Claim-focused scholarly recheck

No material citation or attribution error was found. Fresh source checks confirmed the 2017 abstract's sample and associations; the 2020 and 2023 experimental conditions and reported effect directions; the 2024 total-score associations and correlational design; the 2026 instrument's 18 items and bibliographic year; and Merton's inclusion of apparent vindication. Relevant causal-model, blame-analysis, and responsibility-displacement passages retain their bounded roles.

Sources rechecked: [author publication record and 2017 abstract](https://dynamic.uoregon.edu/jjf/traumapapers.html), [Harsey and Freyd 2020](https://dynamic.uoregon.edu/jjf/articles/hf2020.pdf), [Harsey and Freyd 2023](https://dynamic.uoregon.edu/jjf/articles/hfjiv2023.pdf), [Harsey et al. 2024](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0313642), [Durland et al. 2026](https://dynamic.uoregon.edu/jjf/articles/dhf2026.pdf), [Merton 1948](https://eclass.uowm.gr/modules/document/file.php/NURED263/Merton%20The%20Self-Fulfilling%20Prophecy%201948%20Antioch%20Review.pdf), and [Freyd's account of DARVO](https://www.jjfreyd.com/darvo).

This was a review of the manuscript's assigned claims, not reanalysis of study datasets. The 2017 publisher full text remained inaccessible, so its abstract-limited use is preserved. Previously inspected image-only Freyd, Grice, and Sykes–Matza passages retain their earlier visual-verification records; this pass did not independently reread those scans. The studies are not assigned authority for Patterna's broader family taxonomy.

### 11.4 Current manuscript identity

Draft 0.4 contains **6,584 main-text words** under the existing counting convention: sections 1–9 contain 228, 559, 945, 894, 1,572, 1,030, 730, 384, and 242 words respectively. Reader inventory remains 167 content blocks, 14 display equations, three tables, one diagram, and 16 references.

`DARVO_ARTICLE.md` SHA-256: `44b7801d3566e3dd7caf042c8750da5f19115dec9c0d22b071c7baccfad6c0e8`.

Reader content is versioned as 0.4. Existing 0.1–0.3 source assets and the 0.2 case manuscript are retained. The review introduces no framework, application-service, scoring, corpus, or deployment changes.

### 11.5 Completed source and reader validation

- Display equations and bibliography match draft 0.3 exactly. The canon quotation, framework sources, original diagram, case archive, historical ledger sections, and prior public assets retain their pre-review content. Local manuscript and ledger links resolve.
- The seven reader/source checks and web lint pass. The inactive production build passes, and English/Portuguese metadata identifies draft 0.4 and the unchanged English manuscript language.
- Local Chrome checks pass at `http://127.0.0.1:5174/en/articles/darvo-family`: complete content, mathematics, contextual explanations and focus, citation returns, interface-language/position preservation, four responsive widths, complete printing, and versioned reader behavior. Narration checks use controlled local test media only.
- No real-case addition, new agent-performance claim, live-provider analysis, or public activation was performed.