ETC Series — Investigative Methodology¶

This file defines the methodology for ALL studies in the etc-XX series (The Final Fate of the Wicked). Every analysis agent MUST follow this methodology.

Investigative Methodology (include verbatim in every agent prompt)¶

INVESTIGATIVE METHODOLOGY:
- You are an investigator, not an advocate. Your job is to report what the evidence says.
- Gather evidence from ALL sides. If a passage is cited by ECT proponents, examine it honestly. If a passage is cited by conditionalists, examine it honestly.
- Do NOT assume your conclusion before examining the evidence.
- Do NOT state opinions. State what the text says. Do not use editorial characterizations like "genuine tension," "strongest argument," "most significant challenge," "honestly acknowledge," or "non-intuitive reading." Simply state what each passage says and what each side infers from it.
- When presenting findings, state: "The text says X" (explicit). Then state: "From this, Y interpretation infers Z" and "W interpretation infers V" (inferred).
- Never use language like "irrefutable," "obviously," or "clearly proves." Use "the text states," "this is consistent with."
- The conclusion should emerge FROM the evidence, not be imposed ON it.

Evidence Classification (REQUIRED in every CONCLUSION.md)¶

Every CONCLUSION.md MUST include a multi-tier evidence classification section. The tiers are Explicit, Necessary Implication, and Inference (with four inference subtypes: I-A, I-B, I-C, I-D).

The classification is about what THE BIBLE says, not what individual verses say in isolation: - Explicit: "The Bible says X" — you can point to a verse that says X. - Necessary Implication: "The Bible implies X" — you can point to verses that, when combined, force X with no alternative. - Inferred: "Someone claims the Bible teaches X" — but no verse says X and no combination of verses forces X. Something must be added that the text does not contain. Inferences are further classified into four types (I-A, I-B, I-C, I-D) based on source and direction.

Evidence Hierarchy: E > N > I-A > I-B (resolved by SIS) > I-C > I-D

CRITICAL RULE: Inferences cannot block explicit statements or necessary implications. If explicit texts and necessary implications establish X, the existence of passages that could be inferred to teach not-X does not prevent X from being a necessary implication. The passages cited against X must be evaluated on their own terms — if reading them as contradicting X requires adding a concept, choosing between readings, applying an external framework, or systematizing, then that reading is itself an inference and cannot override what the explicit texts state.

1. Explicit Statements Table¶

#	Explicit Statement	Reference
E1	[What the text directly says — a quote or close paraphrase]	[Book Chapter:Verse]

Rules for explicit statements: - The text must directly say this. Quote or closely paraphrase the actual words of Scripture. - One explicit statement per verse or closely related verse cluster. - Include statements from ALL sides of the debate. - A paraphrase of a single verse in different words is still explicit. "The dead know not any thing" = "The dead lack knowledge" → explicit (same verse, same meaning). - THE MEANING OF WORDS IS EXPLICIT. If a single verse uses a word, what that word means is part of the explicit statement. "Became a living soul" directly means "is a living soul." "Shall die" means the subject dies. These are not separate conclusions — they are what the verse says.

CRITICAL: What the text SAYS vs. what a position INFERS: - E tier = what the text directly states (what both sides must acknowledge as textual fact) - I tier = what a position interprets/infers from that text

Examples (using passages outside this series): - "Jesus wept" (John 11:35) → E (what the text says) - "Therefore Jesus had genuine human emotions" → I-A (systematizes, but derived from E) - "The same word (aionios) modifies 'life' in John 3:16 and 'fire' in Matt 18:8" → E (grammatical fact) - "Therefore both involve the same duration" → I-B (requires interpreting what the adjective means) - "Text says X happened to A, B, and C" → E (what the text states about those subjects) - "Therefore X also happened to D" → I-B (extension beyond stated subjects — requires adding D)

The test: Could both ECT and conditionalist readers accept this as a factual observation about what the text says, even while disagreeing about what it means or implies? If yes → E or N. If no (one side must deny it) → only then classify by position.

2. Necessary Implications Table¶

#	Necessary Implication	Based on	Why it is unavoidable
N1	[What unavoidably follows from explicit statements]	[Which E# statement(s)]	[Why no reader could deny this conclusion from those statements]

Rules for necessary implications: - A necessary implication follows unavoidably from one or more explicit statements. - No additional concept, framework, or interpretation is required — only understanding what the words mean. - Every reader, regardless of theological position, must agree this follows. If a reasonable reader could disagree, it is an inference, not a necessary implication. - Include necessary implications from ALL sides of the debate.

What belongs here: - The logical entailment of explicit words. If the text says God "only" has immortality (E), then "no one else has immortality" is a necessary implication — it is what "only" means stated as a conclusion. - Combinations of explicit statements that yield an unavoidable conclusion. If E says "breath departs, thoughts perish" and E says "the dead know nothing," then "the spirit returns without consciousness" follows necessarily. - Observable patterns stated as conclusions. If term X is applied to humans and animals (per E statements) and term Y is applied only to humans (per E statements), then "Y, not X, distinguishes humans from animals" is a necessary implication. - Direct entailments. If two explicit facts together produce a conclusion without any additional reasoning, that conclusion is a necessary implication.

Test: Could a plain reader reach this conclusion from the cited explicit statements without importing any theological framework? Would EVERY reader, from ANY theological tradition, agree this follows? If yes → necessary implication. If no → inference.

STRICTER N-TIER TEST (NEW in etc):

Ask these three questions for EACH N item:

Universal agreement test: Would a scholar from the OPPOSITE theological position (ECT if you're conditionalist, conditionalist if you're ECT) necessarily agree this follows from the cited E statements?
If NO → it's an inference, not a necessary implication
No interpretation required test: Does this require choosing between possible meanings, or is it the only possible meaning?
If it requires choosing → it's an inference (I-B)
Zero added concepts test: Does this add ANY concept, framework, or connection not present in the explicit statements themselves?
If YES → it's an inference (requires criterion #1, #3, or #4b)

Common N-tier misclassifications to avoid:

❌ "Since the same adjective modifies both outcomes, both have the same duration" → WRONG (requires interpreting what the adjective means — that's criterion #2, making it I-B) ✓ "The same adjective is used for both outcomes" → CORRECT (observable grammatical fact, both sides agree)

❌ "Since text says X to subjects A/B/C, and D shares their destination, D also experiences X" → WRONG (requires inferring same destination = same experience — that's I-B) ✓ "The text says X to A, B, and C. The text says D goes to the same destination." → CORRECT (two separate observable statements)

❌ "Since the text uses identical language for two groups, they experience the same outcome" → WRONG (requires choosing between 'same language = same experience' and 'same language = same category of result' — that's I-B) ✓ "The text uses identical language for both groups" → CORRECT (observable fact)

If an N item requires defending it or explaining why someone should accept it, it's probably an inference.

3. Inferences Table (4-Type Taxonomy)¶

#	Claim	Type	What the Bible actually says	Why this is an inference	Criteria
I1	[The claim about what the Bible teaches]	[I-A/I-B/I-C/I-D]	[What the relevant verses actually say — cite E# and N# items AND include actual verse references and what they say]	[What must be added beyond what the text contains]	[Which criterion/criteria apply]

The 4-Type Inference Taxonomy¶

Two dimensions create a 2×2 matrix:

	Aligns with E/N	Conflicts with E/N
Derived from E/N	I-A (Evidence-Extending)	I-B (Competing-Evidence)
Not derived from E/N	I-C (Compatible External)	I-D (Counter-Evidence External)

I-A (Evidence-Extending): Uses ONLY vocabulary and concepts found in E/N statements. Only an inference because it systematizes multiple E/N items into a broader claim. Strongest inference type.

I-B (Competing-Evidence): Some E/N statements support it, but other E/N statements appear to contradict it. Genuine textual tension where both sides can cite Scripture. Resolved by the Scripture-Interprets-Scripture (SIS) protocol.

I-C (Compatible External): Reasoning from outside the text (theological tradition, philosophical framework, historical context) that does not contradict any E/N statements. Supplemental only — adds information the text does not contain but does not override anything the text says.

I-D (Counter-Evidence External): External concepts that require overriding, redefining, or qualifying E/N statements to be maintained. Weakest inference type — requires the text to mean something other than what it says.

Mechanical Tests for Classification¶

Source test (derived vs. external): Strip away the systematization. Are ALL remaining components found in the E/N tables? YES = text-derived (I-A or I-B). NO = external (I-C or I-D).

Direction test (aligns vs. conflicts): Does the claim require ANY E/N statement to mean something other than its lexical value? YES = conflicts (I-B or I-D). NO = aligns/compatible (I-A or I-C).

Consistency checks: - I-A should only require criterion #5 (systematizing) and optionally #4a (SIS). If it requires #1, #2, or #3, it is misclassified. - Every I-B must have E/N items on BOTH sides. If only one side has E/N support, it is I-A or I-D. - Every I-D must override at least one E/N statement. If it overrides nothing, it is I-C.

Rules for inferences:¶

An inference is a claim about what the Bible teaches that no verse explicitly states and no combination of verses necessarily implies. Something must be added beyond what the text contains.
State each inference as a Bible-wide claim ("The Bible teaches X"), not as a verse-specific interpretation ("Verse Y means Z"). Then show why no verse says it and no combination forces it.
Always include the actual verse references and what they say in the "What the Bible actually says" column — not just E#/N# numbers. The reader should see the biblical evidence without cross-referencing.
Include inferences from ALL sides (both ECT and conditionalist, both holistic and dualistic, etc.)
Identify what reasoning step is required that the text itself does not state
Note when an inference requires applying concepts not found in the text

An inference MUST require at least one of these: 1. Adding a concept the text doesn't state — e.g., "self-existent immortality" when the text just says "immortality" 2. Choosing between two possible readings — e.g., interpreting "nephesh departing" as a conscious entity departing vs. life departing 3. Applying an external framework — e.g., "under the sun" limitation on Ecclesiastes statements 4. Cross-referencing (split into 4a and 4b — see SIS section below) 5. Systematizing into a doctrine — e.g., combining multiple texts into a comprehensive theological position

If you cannot identify which of these an inference requires, it is probably a necessary implication and should be moved.

Scripture-Interprets-Scripture (SIS) Principle¶

What Changes from etc2¶

etc2 criterion #4 ("Bridging to a different passage non-obviously") treated ALL cross-referencing as inferential. This is wrong — using clear passages to interpret unclear ones is standard hermeneutics.

Criterion #4 splits into two: - #4a (SIS with verified textual connection) — NOT an inference trigger. When a clear passage interprets an unclear one, and the connection is verified (shared vocabulary, OT quotation, tool-verified parallel score, or the text itself establishes the connection), this is standard hermeneutics, not an inference. Document the connection. - #4b (cross-referencing without verified textual connection) — IS an inference trigger. The reader must supply the connection between passages. The link depends on the interpreter's judgment, not on the text itself.

How to Document SIS Connections¶

When using #4a (SIS with verified connection), document: 1. The clear passage and the unclear passage 2. The nature of the connection (shared vocabulary, OT quotation, parallel score, self-reference) 3. Why the clear passage is clearer (using the clarity criteria below) 4. How the clear passage determines the reading of the unclear one

I-B Resolution Protocol¶

When an inference has competing textual support (I-B), apply this 5-step process:

Step 1: Identify tension. List E/N items FOR and AGAINST the claim.

Step 2: Assess clarity of each E/N item on a 3-level scale: - Plain: Directly addresses the topic using relevant vocabulary; no interpretation needed - Contextually Clear: Addresses the topic but requires genre/audience/context awareness - Ambiguous: Could plausibly be read either way

Step 3: Count and weigh. Plain statements outweigh Ambiguous ones (not a mere vote count). The weight is determined by clarity level, not just quantity.

Step 4: Apply SIS. Plain statements determine the reading of Ambiguous ones. The clear interprets the unclear.

Step 5: State resolution. One of: - Strong — Plain statements on one side with only Ambiguous statements on the other - Moderate — Mix of Plain and Contextually Clear on the dominant side - Unresolved — Substantial Plain/Contextually Clear statements on both sides

Clarity criteria (what makes a passage "clearer"): 1. Directness of vocabulary — actual words vs. figurative language 2. Genre — didactic > apocalyptic > parabolic 3. Scope — universal statement > specific situation 4. Frequency — repeated across authors/testaments > single occurrence 5. Self-interpretation — text explains its own meaning = maximally clear

I-B Resolution Subsection¶

After the Inferences table, include a dedicated I-B Resolution subsection for EACH I-B inference with the full 5-step analysis:

#### I-B Resolution: [Inference #] — [Short description]

**Step 1 — Tension:**
- FOR: [E#, N# items supporting the claim]
- AGAINST: [E#, N# items opposing the claim]

**Step 2 — Clarity Assessment:**
| Item | Level | Rationale |
|------|-------|-----------|
| E# | Plain/Contextually Clear/Ambiguous | [Why] |
| ... | ... | ... |

**Step 3 — Weight:**
[Summary of how items on each side weigh]

**Step 4 — SIS Application:**
[How plain statements determine reading of ambiguous ones]

**Step 5 — Resolution: [Strong/Moderate/Unresolved]**
[Explanation]

Classification Decision Trees¶

Apply these trees mechanically when classifying every evidence item. Every gate is a binary YES/NO answerable from the text itself. Work through each tree in order. Do not skip gates.

Tree 1 — Tier Classification¶

Start: You have an observation or claim about what the Bible says.

Q1: Does this directly quote or closely paraphrase the actual words
    of a specific verse or verse cluster?
    NO  → go to N-CHECK
    YES → go to E-CHECK

E-CHECK:
  E1: Is this the plain lexical meaning of those words —
      no concept, framework, or interpretation added beyond what
      the words themselves require?
      YES → TIER: E (Explicit). Stop. Go to Tree 3 (E-Positional).
      NO  → go to N-CHECK

N-CHECK:
  N1: Does this follow unavoidably from one or more E-items?
      NO  → TIER: I (Inference). Stop. Go to Tree 2 (I-Type).
      YES → go to N2

  N2: Would a scholar from the OPPOSITE theological position
      necessarily agree this follows from the cited E-items,
      without any additional reasoning?
      NO  → TIER: I. Stop. Go to Tree 2.
      YES → go to N3

  N3: Does reaching this conclusion require choosing between
      two possible meanings of a word or phrase?
      YES → TIER: I. Stop. Go to Tree 2.
      NO  → go to N4

  N4: Does this add ANY concept, framework, or connection not
      already present in the cited E-items themselves?
      YES → TIER: I. Stop. Go to Tree 2.
      NO  → TIER: N (Necessary Implication). Stop. Go to Tree 4 (N-Positional).

Tree 2 — I-Type Classification¶

Start: You have an item classified as I (Inference). Which subtype?

SOURCE TEST:
  S1: Strip away any systematization. Are ALL remaining components
      of this claim found verbatim or directly in the E/N tables?
      YES → text-derived → go to DIRECTION TEST (text-derived)
      NO  → external → go to DIRECTION TEST (external)

DIRECTION TEST (text-derived):
  D1: Does this claim require any E/N statement to mean something
      other than its plain lexical value?
      NO  → TYPE: I-A (Evidence-Extending). Stop. Go to Tree 5 (I-Positional).
      YES → TYPE: I-B (Competing-Evidence). Stop. Go to Tree 5.
            [I-B requires a full SIS Resolution subsection in the analysis.]

DIRECTION TEST (external):
  D2: Does this claim override, redefine, or qualify any E/N statement?
      NO  → TYPE: I-C (Compatible External). Stop. Go to Tree 5.
      YES → TYPE: I-D (Counter-Evidence External). Stop. Go to Tree 5.

CONSISTENCY CHECKS (run after typing — reclassify if any check fails):
  I-A check: Does it require ONLY criterion #5 (systematizing)?
             If it also requires #1, #2, or #3 → reclassify.
  I-B check: Does it have E/N items on BOTH sides?
             If only one side has E/N support → reclassify as I-A or I-D.
  I-D check: Does it override at least one E/N statement?
             If it overrides nothing → reclassify as I-C.

Tree 3 — E-Item Positional Classification¶

Start: You have a verified E-item. Does it support Conditionalist, ECT, or Neutral?

STEP 1 — VOCABULARY SCAN:

  V1: Does the verse use destruction/cessation vocabulary
      applied to the wicked?
      Keywords: destroy, perish, die, death, consume, devour,
      burn up, second death, cut off, blotted out, ashes, be no more
      YES → Candidate: CONDITIONALIST → go to Step 2

  V2: Does the verse use conscious-ongoing-existence vocabulary
      applied to the wicked?
      Keywords: torment, tormented, weeping, gnashing, in flames,
      in agony, thirst, remember, no rest, suffering, cannot die
      YES → Candidate: ECT → go to Step 2

  Both V1 and V2 YES → note both; Step 2 determines which survives.
  Both V1 and V2 NO  → NEUTRAL. Stop.


STEP 2 — FOUR VALIDATION GATES:
(Must pass ALL four. Failure at any gate → go to Step 3.)

  GATE 1 — SUBJECT GATE:
    Is the grammatical subject a literal human being?
    Automatic FAIL if subject is any of:
      • A symbolic entity (a figure in an apocalyptic vision or heavenly scene)
      • An allegorical character in a parable (a narrative figure used to illustrate a point)
      • A nation, city, or people used as a prophetic type or figure
      • A non-human spirit being
    PASS → continue to Gate 2
    FAIL → record "Subject is [X], not a literal human." → go to Step 3.

  GATE 2 — GRAMMAR GATE:
    Does the original-language grammar unambiguously support
    the proposed positional reading with no alternative parsing?
    Automatic FAIL if any of the following apply:
      • Reading depends on punctuation absent from the original
        (Greek has no commas — comma placement is editorial, not textual)
      • A connective particle that means "and" is treated as "as/even as"
        (simultaneity or equivalence assumed from a word that expresses sequence)
      • A modifier (adverb, adjective, prepositional phrase) attaches
        grammatically to a different noun than the reading requires
      • A temporal or durational word could grammatically attach to
        a different clause or noun than the proposed reading assumes
    PASS → continue to Gate 3
    FAIL → record which specific grammar issue. → go to Step 3.

  GATE 3 — GENRE GATE:
    Is the passage didactic prose?
    Didactic = direct teaching, epistle, law, narrative report,
               direct-speech prophecy ("Thus says the Lord...")
    Automatic FAIL if passage is:
      • A parable (fictional narrative with characters illustrating a point)
      • An apocalyptic vision (symbolic imagery in Revelation, Daniel)
      • A typological narrative (historical event used as shadow/type)
    PASS → continue to Gate 4
    FAIL → record "Genre is [parabolic/apocalyptic/typological]." → go to Step 3.

  GATE 4 — HARMONY GATE:
    Is the proposed positional classification consistent with
    all other E-items in the master evidence file?
    Check:
      • Same-author E-items on the same topic
      • Cross-testament E-items sharing key vocabulary or subject
      • E-items already classified in the opposite direction on this topic
    If a conflict is found, apply SIS resolution:
      • Use clarity criteria (didactic > apocalyptic > parabolic;
        universal > specific; repeated > single occurrence)
      • Does the clearer passage govern the reading of the less clear?
        YES → apply that reading, re-run Gate 4 with corrected reading
        NO  → unresolvable conflict → FAIL → record conflicting E-item → go to Step 3.
    No conflict found → PASS → CLASSIFICATION STANDS. Stop.


STEP 3 — RECLASSIFICATION CHECK:
(Reached when any gate fails. Failure means the proposed classification
does not hold. It does NOT automatically mean Neutral.)

  RC1: State what the gate failure revealed:
       Gate 1: Subject is [symbolic/figurative/non-human]
       Gate 2: Grammar allows alternative parsing: [describe it]
       Gate 3: Passage is [parabolic/apocalyptic]
       Gate 4: Conflicting E-item [ID] says [what it says]

  RC2: Form the CORRECTED textual observation by applying
       the gate's correction:
       Gate 1: Restate observation with the actual (symbolic) subject
       Gate 2: Restate using the grammatically valid alternative parsing
       Gate 3: Restate as genre-appropriate reading
       Gate 4: Restate incorporating the SIS resolution from conflicting E-item

  RC3: Re-enter STEP 1 with the corrected observation.
       Does the corrected observation pass V1 (destruction vocabulary)?
         YES → CONDITIONALIST. Stop.
       Does the corrected observation pass V2 (conscious-torment vocabulary)?
         YES → ECT. Stop.
       Neither applies → NEUTRAL. Stop.

  NOTE: The corrected observation is still an E-item — the text says
  what it says. Only the positional direction changes. The item remains
  in the Explicit table; only its position column is updated.

Generic example of reclassification flow:

Scenario	Gate that fails	RC2: Corrected observation	RC3: Result
Verse uses torment vocabulary; subject is a symbolic vision-figure	Gate 1: subject not a literal human	Restate with actual subject: "A symbolic figure experiences [X]." Re-enter V1/V2: neither destruction nor human-torment vocab applies to a symbolic entity as evidence about human fate.	Neutral
Verse uses conscious-state vocabulary; grammar allows two parsings, one of which removes the ECT implication	Gate 2: alternative parsing valid	Restate using the grammatically valid alternative parsing. Re-enter V1/V2: does the corrected reading contain destruction vocab? If yes → Conditionalist. Neither vocab? → Neutral.	Conditionalist or Neutral depending on corrected vocabulary
Verse uses torment vocabulary; passage is a parable	Gate 3: parabolic genre	Restate as genre-appropriate observation: "A parable depicts a character experiencing [X]." Re-enter V1/V2: parabolic depiction does not constitute didactic teaching about human fate.	Neutral
Verse appears to support one position; same author elsewhere explicitly states the opposite	Gate 4: harmony conflict with clearer E-item	Apply SIS: restate using the reading the clearer passage governs. Re-enter V1/V2: does the SIS-governed corrected reading contain positional vocabulary?	Depends on corrected vocabulary — may be Conditionalist, ECT, or Neutral

Tree 4 — N-Item Positional Classification¶

Start: You have a verified N-item. Same as Tree 3 with one additional gate first.

GATE 0 — FOUNDATION GATE (N-items only):
  Verify the item genuinely belongs at N-tier before classifying position.
  Apply all three N-tier tests:
    N-Test 1 (Universal agreement): Would a scholar from the opposite
      position necessarily agree this follows from the cited E-items?
    N-Test 2 (No interpretation required): Is this the only possible
      conclusion — not a choice between two readings?
    N-Test 3 (Zero added concepts): Does this add nothing beyond
      what the source E-items themselves contain?
  All three YES → PASS → continue to Tree 3 (Vocabulary Scan onward).
  Any NO → item is misclassified as N → send back to Tree 1 (N-CHECK fails) →
            reclassify as I → apply Tree 2 → apply Tree 5.

Tree 5 — I-Item Positional Classification¶

Start: You have a typed I-item (I-A, I-B, I-C, or I-D). What position does it support?

NOTE: The four-gate validation (Trees 3/4) does NOT apply to I-items.
The inference category already acknowledges that interpretation is required.
Position simply reflects which direction the inference points.

  IP1: Does this inference support the claim that the wicked are destroyed,
       cease to exist, lack immortality, or that death/destruction
       is the penalty for sin?
       YES → CONDITIONALIST

  IP2: Does this inference support the claim that the wicked experience
       ongoing conscious torment or conscious existence after death?
       YES → ECT

  IP3: Does the inference support both, or neither?
       BOTH → verify it is classified I-B (competing evidence from both sides).
               If not I-B, reclassify. Apply SIS resolution to determine
               which direction the weight of evidence favors.
       NEITHER → NEUTRAL.
               (The inference concerns genre, methodology, authorship,
                historical background, or is shared framework for both sides.)

4. Verification Phase (REQUIRED)¶

After completing all tables, run this verification check:

Step A: Verify explicit statements: - Does each E-statement directly quote or closely paraphrase actual verse text? - Is it actually just the plain meaning of the words in the verse? - NEW: Is this what the text SAYS (E) or what a position INFERS from it (I)? If it's an inference, move it.

Step A2: Verify positional classifications of E-items (REQUIRED):

For each E-item classified as ECT or Conditionalist (not Neutral), apply Tree 3 (E-Item Positional Classification) from the Classification Decision Trees section above. This is mandatory — do not assign a positional direction to an E-item without running the full tree.

The tree runs three core checks before a positional classification stands:

Grammar check — Does the original-language grammar actually require this positional reading?
Greek has no punctuation — comma placement is editorial, not a textual fact
A connective particle meaning "and" does not establish simultaneity or equivalence
Verify which noun a modifier grammatically attaches to — the proposed reading must match the actual grammar
Harmony check — Does this classification conflict with any other E-item?
One explicit statement cannot contradict another
Check same-author statements and cross-testament statements on the same topic
Genre check — Is this passage didactic, parabolic, or apocalyptic/symbolic?
Didactic > apocalyptic > parabolic in clarity
A symbolic entity experiencing something in a vision does not establish that literal humans experience the same thing

If any check fails, apply the Reclassification Check (Tree 3, Step 3). A failed gate does NOT automatically mean Neutral — the corrected observation may point to the opposite position. Re-enter the vocabulary scan with the corrected observation to determine the actual classification.

Step B: Verify necessary implications: - Does each N follow unavoidably from the cited E statements? - Could ANY reader deny this conclusion while accepting the explicit statements? If yes → move to Inferences. - Is it actually just a direct quote or close paraphrase of a single verse? If yes → move to Explicit. - NEW: Apply the three N-tier tests (universal agreement, no interpretation required, zero added concepts). If any test fails → move to Inferences.

Step C: Verify inference classifications (source test): For each inference, strip away the systematization. Are ALL remaining components found in the E/N tables? - YES → text-derived (I-A or I-B) - NO → external (I-C or I-D)

Step D: Verify inference classifications (direction test): Does the claim require ANY E/N statement to mean something other than its lexical value? - YES → conflicts (I-B or I-D) - NO → aligns/compatible (I-A or I-C)

Step E: Run consistency checks: - Every I-A: Does it only require criterion #5 (and optionally #4a)? If it requires #1, #2, or #3, reclassify. - Every I-B: Does it have E/N items on BOTH sides? If only one side has E/N support, reclassify as I-A or I-D. - Every I-D: Does it override at least one E/N statement? If it overrides nothing, reclassify as I-C.

Step F: Verify SIS connections: - Is each #4a connection documented with shared vocabulary, OT quotation, or tool-verified parallel? - Is each #4b properly treated as an inference trigger?

Common mistakes to avoid: - Do NOT classify the plain meaning of words as inference. "Only" means only. "Became" means is. "Know not any thing" means lack knowledge. - Do NOT classify observable textual patterns as inference. If Scripture uses term X for animals and humans, that's a fact you can observe. - Do NOT classify unavoidable combinations as inference. If explicit statements A + B together yield C with no alternative, C is a necessary implication. - The COUNTER-CLAIM to an explicit statement or necessary implication is often the real inference. If the text says "the dead know not any thing," the necessary implication is "the dead lack awareness." The inference is "these texts are limited to an earthly perspective" (applies a qualifier the text doesn't state). - NEW: Do NOT classify what a text SAYS as supporting a position. Only classify by position when one side must DENY the textual observation. Otherwise classify as Neutral.

After verification: - Move misclassified items between tables as needed - Update the tally counts - Ensure every inference has a clear Type (I-A, I-B, I-C, or I-D) - Ensure every inference identifies which criterion/criteria apply - Ensure every I-B has a full Resolution subsection - Ensure every necessary implication cites its source E# statements and explains why it is unavoidable

5. Master Evidence File (REQUIRED before writing Tally)¶

Before computing your tally, you MUST read and update etc-master-evidence.md.

Steps: 1. Read etc-master-evidence.md to see all items already registered by prior studies 2. For each E/N/I item in your tables: - If an item with the same verse + same observation (E), same conclusion (N), or same claim + type (I) already exists → note the master ID in your CONCLUSION.md (e.g., "E23 → Master E042") - If no match exists → add it to etc-master-evidence.md with the next available ID 3. Update the "Also In" field for any existing items your study also cites 4. In your CONCLUSION.md, add a note: "Evidence items tracked in etc-master-evidence.md"

Why this matters: etc-18 reads etc-master-evidence.md directly to produce the final deduplicated positional tally. If you skip this step, items from your study will either be uncounted or counted as duplicates. This is the deduplication mechanism — it happens incrementally as each study completes, not retroactively at synthesis time.

6. Tally Summary¶

- Explicit statements: [count]
- Necessary implications: [count]
- Inferences: [count]
  - I-A (Evidence-Extending): [count]
  - I-B (Competing-Evidence): [count] ([N] resolved, [M] unresolved)
  - I-C (Compatible External): [count]
  - I-D (Counter-Evidence External): [count]

7. What CAN Be Said / What CANNOT Be Said¶

What CAN be said (Scripture explicitly states or necessarily implies): - [List — draw from both Explicit and Necessary Implication tables]

What CANNOT be said (not explicitly stated or necessarily implied by Scripture): - [List of things neither side can claim the text directly says or necessarily implies — including things commonly assumed by BOTH sides]

Critical Rules Governing the Hierarchy¶

E > N > I-A > I-B > I-C > I-D. Higher-tier evidence governs the interpretation of lower-tier claims. An I-D claim cannot override an E statement.
Inferences cannot block explicit statements or necessary implications. If explicit texts and necessary implications establish X, the existence of passages that could be inferred to teach not-X does not prevent X from being established.
I-A inferences are the strongest inferences because they use only the text's own vocabulary and concepts. They are inferences only because they systematize.
I-B inferences require the SIS protocol. Both sides have textual support. The resolution must be documented. Plain passages interpret ambiguous ones.
I-D inferences bear the heaviest burden. They require overriding what the text says with concepts the text does not contain. They are valid only if the text itself provides reason to read against its surface meaning.
SIS connections (#4a) are not inference triggers. Using clear passages to interpret unclear ones — when the connection is verified — is standard hermeneutics. Only unverified cross-references (#4b) trigger inference classification.

Positional Tally (REQUIRED in etc-18 Synthesis ONLY)¶

NEW in etc: DEDUPLICATION IS MANDATORY

The etc-18 comprehensive synthesis MUST:

Deduplicate all E/N/I items before creating the positional tally
Two items are duplicates if they cite the SAME verse AND make the SAME observation
Items citing the same verse but making DIFFERENT observations are NOT duplicates
Count each unique evidence item only ONCE
Report both raw and deduplicated counts:
Raw count: Total items from all studies (includes duplicates)
Deduplicated count: Unique items only (duplicates removed)
Classification rules for positional tally:

Classify by position ONLY when one side must DENY the textual observation:

Conditionalist: Item explicitly states the wicked perish, are destroyed, cease to exist, lack immortality, are mortal, die, are consumed, are unconscious in death, or that destruction/death is the penalty for sin
ECT: Item explicitly states the wicked are tormented, suffer consciously forever, that torment language is applied to human judgment subjects, or that ongoing conscious existence continues after death for the wicked
Neutral/Shared: Factual observations BOTH sides must accept even while disagreeing about interpretation:
Grammatical facts (same adjective used, word counts, parsing data)
Statistical observations (vocabulary distributions, occurrence counts)
Genre identifications (apocalyptic, parabolic, didactic)
Subject identifications (who/what is described in a passage)
Textual observations about what is or isn't stated
Any observation both ECT and conditionalist scholars can accept as textual fact

The key question: Can both sides agree this is what the text SAYS/SHOWS, even while disagreeing about what it MEANS? If yes → Neutral.

Examples of Neutral E items (NOT positional): - "Rev 20:10 describes three subjects: devil, beast, false prophet" - "The same adjective (aionios) modifies punishment and life in Matt 25:46" - "Paul expresses preference to be absent from body, present with Lord" - "Basanizo appears in Rev 20:10" - "The lake of fire is the destination for both the devil and humans" - "Rich Man and Lazarus contains dialogue about Abraham's bosom and hades"

Examples of positional E items: - Conditionalist: "The soul that sins shall die" (Ezek 18:4), "God can destroy soul and body" (Matt 10:28), "The dead know nothing" (Ecc 9:5) - ECT: "Tormented with fire and brimstone... smoke of their torment ascends forever" (Rev 14:10-11)

Positional Tally Format¶

## Positional Tally (Deduplicated)

### By Evidence Tier

| Tier | Conditionalist | ECT | Neutral/Shared | Total |
|------|---------------|-----|----------------|-------|
| Explicit (E) | [count] | [count] | [count] | [count] |
| Necessary Implication (N) | [count] | [count] | [count] | [count] |
| I-A (Evidence-Extending) | [count] | [count] | [count] | [count] |
| I-B (Competing-Evidence) | [count] | [count] | [count] | [count] |
| I-C (Compatible External) | [count] | [count] | [count] | [count] |
| I-D (Counter-Evidence External) | [count] | [count] | [count] | [count] |
| **TOTAL** | **[count]** | **[count]** | **[count]** | **[count]** |

**Raw counts (before deduplication):** [total raw] items
**Deduplicated counts:** [total unique] items
**Duplicates removed:** [number] ([percent]% reduction)

Conclusion Tone Rule¶

The conclusion section of every study MUST: - Present the classification results as data - State what the evidence tiers contain - NOT use hedging language like "this doesn't prove X" or "this doesn't disprove Y" - NOT editorialize about what the results mean for either position - Let the numbers and classifications speak for themselves

Example of what TO write: "Across 17 studies, [N] explicit statements and [M] necessary implications use destruction/death vocabulary for the fate of the wicked. [P] ECT claims are classified as I-D (counter-evidence external). [Q] conditionalist claims are classified as I-A (evidence-extending)."

Example of what NOT to write: "This does not prove ECT is wrong" or "This does not disprove conditionalism."

No Editorial Opinion¶

Do NOT characterize passages as being "in tension" with each other — simply state what each passage says
Do NOT call any argument "the strongest" or "the weakest" — present the arguments and the evidence
Do NOT use "genuinely ambiguous" — state the possible readings and note which the text specifies or does not specify
Do NOT say something "requires sustained effort to maintain" — state the reasoning required and let the reader assess
For passages covered by later studies in the series, briefly state what the text says and cross-reference the later study

Cross-References to Other Studies¶

When a passage is examined in depth in another etc-XX study: - Briefly state what the verse says (quote it) - Add: (Examined in depth in etc-XX-slug.) - Do NOT editorialize about the passage — that's for the dedicated study

Prior Study Conclusions¶

Each study should read CONCLUSION.md from prior etc- studies for context on what has been established. Prior findings inform what areas to investigate but not what to conclude. Each study investigates independently.