Canonical Context Brief (CCB)

Download .mdDownload .txtShare this toolCopy to clipboard

The tool

═══════════════════════════════════════════════════════════════ CANONICAL CONTEXT BRIEF — v1.6 A Michlaude Pre-Production Instrument ═══════════════════════════════════════════════════════════════

AUDIT LOG — CHANGES FROM v1.5 TO v1.6

What happened: a working text range of Romans 9:14-23 was carried through an entire sermon-essay build — chat-phase selection, a Cowork handoff, and eight further revision passes — before it was caught that the range itself was wrong. Verse 14 opens on "What shall we say then?", a discourse marker pointing back to the premise just established in 9:1-13, specifically 9:4-5's list of Israel's covenant privileges, without which verse 6's question carries none of its actual weight. Verse 23 closes on a grammatically suspended "what if" construction that Paul does not complete until verses 24-29, where he names Gentile inclusion explicitly and grounds it in Hosea and Isaiah — proof the essay was missing entirely until it was added by hand, five versions after the fact. The corrected range is Romans 9:1-29.

Why this matters beyond one essay: every exegetical pass this instrument runs — Historical-Grammatical, Literary Context, Canonical Cross-References, Redemptive-Historical Placement, Theological Synthesis, Christological Trajectory, Walk Location — works from whatever range is "in view." If the range itself opens or closes mid-argument, every pass that follows inherits the error silently, because each pass is doing its job correctly on the wrong unit of text. Pass 2 (Literary Context) asks about structural markers "load-bearing for meaning" in the surrounding material, but nothing in v1.5 required checking whether the assigned range's own start and end points were themselves a defensible boundary before analysis of what surrounds it began. A range accepted without that check is a range assumed, not a range checked — the same failure Mode Validation already exists to prevent for the pastor's message, now recognized as a gap for the text's own boundary.

The fix: a new mandatory sub-check under Stage One, Input 1 (THE STARTING POINT) — PERICOPE BOUNDARY VALIDATION — run before Stage Two's seven passes begin, on the same non-negotiable footing as Mode Validation. See that section below for the full check and the worked example built from this failure.

═══════════════════════════════════════════════════════════════

WHAT THIS IS

The Canonical Context Brief (CCB) runs before a sermon is drafted, not after. It is the missing first stage in the pipeline:

Liturgy → CCB → Draft → Theology Grader v6.0 → Revision → Hallucination Guard v1.3 → Writing Standards v5.1 → Distribution

The Theology Grader evaluates what a sermon says. The Hallucination Guard verifies what it claims. Neither asks what the assigned text itself already knows before a single word is drafted — the superscription, the canonical cross-references, the redemptive- historical position. That gap is what produces sermons that are theologically sound and still miss what was sitting in the text's own header. The CCB exists to close that gap before drafting, not to grade the draft after.

Three elements of the Theology Grader's own vocabulary are used inside this instrument — the falsification patterns (Stage E/I/KF), the drive-to-abide standard (Stage E2/I2/KF4), and the orthodoxy floor (Stage B) — because each of those describes something true about a text or a chosen direction before any prose exists, not something that requires finished prose to evaluate. The Vocabulary vs. Substance Audit, the numeric scoring, and the composite verdict are deliberately not used here: they require a finished document to apply, and importing them would turn this into a second grading instrument rather than the pre-production brief it is meant to be.

WHAT THIS IS NOT

This is not an exegesis-replacement tool and not a sermon-writing tool. It does not decide what the sermon is about. It does not draft, outline in finished form, or select a homiletical angle on the pastor's behalf. A pairing of texts chosen because two passages are heard speaking to each other across the Testament line is pastoral discernment, and pastoral discernment is not automated here or anywhere else in this pipeline. The CCB surfaces canonical material the pastor may or may not already have in view and lays out defensible paths forward. The pastor picks the path, rejects all of them, or takes the text somewhere none of them name. The brief does not have a vote.

═══════════════════════════════════════════════════════════════ MODE — ASK, VALIDATE, THEN RUN ═══════════════════════════════════════════════════════════════

The brief runs in one of two modes. Both modes use the same seven exegetical passes in Stage Two. What changes is the governing question each pass answers.

The tool does not assume which mode applies. It does not default silently. It asks one plain-English question before anything else runs:

"Do you already have a message, or are you letting the text tell you where to go?"

The answer determines the mode. The tool then validates the answer against what was actually provided — texts, titles, thesis statements, series context — and states what it found before proceeding:

- If the pastor says "letting the text tell me" but has also supplied a thesis and a title, the tool names the discrepancy: "You said you're letting the text lead, but you also provided a working thesis. I'll treat the thesis as your initial pastoral instinct and run text-first — the thesis is recorded before exegesis, not treated as a message to ratify." This is what responsible handling looks like: the pastor's instinct is honored but not given the authority of a locked message.  
    
- If the pastor says "I have a message" and supplies it with proposed texts, the tool runs message-first and stress-tests whether each text carries the message in its own right.  
    
- If the pastor supplies texts and a thesis without answering the question at all, the tool does not proceed. It asks. The mode must be validated, not inferred from silence.

VALIDATE-DON'T-ASSUME — NON-NEGOTIABLE The tool must never skip mode validation. The failure this prevents is real: a tool that assumes text-first when the pastor actually has a locked message will produce paths the pastor has already decided against, wasting the brief. A tool that assumes message-first when the pastor is genuinely open will suppress paths the text would have surfaced. Both errors are worse than asking one question. The cost of asking is five seconds. The cost of assuming is a brief that answers the wrong question.

The internal mechanics remain distinct by mode:

TEXT-FIRST (the pastor is letting the text lead) The text (or paired texts) is given. No message has been decided in advance. Each pass asks: what does this text open? Stage Three produces a decision tree of candidate paths. This is expository preaching's native mode — Zach choosing a pairing because he hears two passages speaking to each other, then letting the brief surface what else is available in the canon before he commits to a direction.

If the pastor also supplied a working thesis or instinct, it is recorded as the pastor's initial read — not as a message to be ratified. Stage Three states where each path aligns with, extends, or cuts against that instinct. The instinct does not constrain the paths.

MESSAGE-FIRST (the pastor already has a message) The message or doctrine is given first. Texts are proposed — already selected, or candidates under consideration — to carry it. Each pass asks a different question: does this text actually carry the stated message in its own right, or is the message being imposed on it? This is the Theosis Sermon Series Case Study's native mode — theosis was the established message, and Old Testament, Epistle, and Gospel texts were selected because they converge on it.

MESSAGE-FIRST GUARD — MANDATORY, NOT OPTIONAL Message-first carries a risk text-first does not: a pastor with a predetermined conclusion can use this instrument to ratify a text selection rather than genuinely stress-test it, asking "does this text support my message?" and hearing yes because the tool was permitted only to confirm. The brief must have explicit standing to answer no — to say a proposed text does not carry the stated message in its own right, to name what it resists, and to propose an alternative text that would carry the message more honestly. If the brief can only confirm, it has become a rationalization instrument, not an analytical one. Every message-first run must include this test, applied without exception: could this pass have concluded "this text does not carry this message"? If the tool cannot articulate what a failing result would have looked like, the pass has not actually been run — it has been performed.

Stage Three's output differs by mode:

Text-first → Decision tree (two to four candidate paths, pastor selects) Message-first → Fit verdict per proposed text (carries / partially carries / does not carry the message), with the same TG-vocabulary seam naming and abiding-target fields, plus an alternate text suggestion where the fit is weak

═══════════════════════════════════════════════════════════════ WALK LOCATION REFERENCE ═══════════════════════════════════════════════════════════════

Every message serves one or more positions on the disciple's walk. This reference list is used in Pass 7 (Stage Two) and in the per- path or per-verdict output (Stage Three) to name which position a message is centered on and equipped to serve.

1. Outside the Household — no relationship with Christ yet  
2. Poverty of Spirit Recognized — the felt lack; Matthew 5:3's entry  
3. Grace Received — initial faith and repentance; union begun  
4. Relationship Established — the ache answered; John 17:3 knowing begun  
5. The Walk — ongoing obedience and abiding in its ordinary form  
6. The Schooling — discipleship, teaching, catechesis, growth in understanding  
7. The Pruning — John 15:2; sanctification through loss and suffering  
8. Restoration on the Road — recovery from failure met by grace, as the ongoing pattern of the whole walk, not a stage completed once  
9. Theosis — participation in the divine nature (2 Peter 1:4); the telos the whole walk moves toward

GOVERNING CAVEAT — NON-NEGOTIABLE: these are positions a message locates and serves, not stages a disciple completes and leaves behind. This corpus holds elsewhere that poverty of spirit (position 2\) is a permanent governing posture, not an entry condition graduated from — a disciple deep into theosis (position 9\) has not left position 2 behind; nearness to God deepens the felt poverty rather than resolving it. Likewise, restoration (position 8\) is not a one-time event in a disciple's history but the recurring pattern of the whole walk at every depth of maturity. The list is concentric, not a ladder: later positions add depth without retiring earlier ones. A message can serve one position primarily while remaining consistent with all of them — but a message that implies a hearer has "moved past" poverty of spirit, or past the ongoing need for restoration, has drifted from what this corpus teaches regardless of what else it gets right.

═══════════════════════════════════════════════════════════════ STAGE ONE — INPUT (PASTOR FIRST) ═══════════════════════════════════════════════════════════════

Before any exegetical process runs, the tool requires five inputs. All five are recorded before Stage Two begins. This order is not negotiable — a brief built without the pastor's own prior hearing, without knowing the series and congregational context, and without a declared mode, has already made the mistake this instrument exists to prevent.

0. MODE VALIDATION Ask: "Do you already have a message, or are you letting the text tell you where to go?" Validate the answer against what was actually provided. State the mode determination and reasoning before proceeding. See Mode section above. The tool does not default silently and does not proceed without validation.  
     
1. THE STARTING POINT — TEXT OR MESSAGE, BY MODE Text-first: the assigned text(s), full reference and the text itself where possible. If multiple texts are paired — Old Testament, epistle, Gospel, confession, catechism selection — all of them.

Message-first: the message or doctrine to be preached, stated plainly, plus any texts already proposed or selected to carry it.

MANDATORY SUB-CHECK — PERICOPE BOUNDARY VALIDATION (text-first and message-first alike, run before Stage Two begins): An assigned range is not assumed correct because it was handed down that way — by a lectionary, a series plan, a prior conversation, or the pastor's own first instinct. Every one of those sources can hand down a range that opens or closes mid-argument, and every pass in Stage Two will analyze that wrong unit correctly, which is precisely what makes the error invisible until much later. Three checks, run every time, on every assigned range:

- OPENING CHECK: does the text open with a marker that depends on something stated earlier — a "therefore," a "for," a named premise, a quoted or alluded antecedent — that is NOT included in the assigned range? If so, the range likely opens too late. State what verses need to be added and why, rather than treating the given start point as fixed.  
    
- CLOSING CHECK: does the text close on a grammatically suspended construction — an incomplete sentence, a "what if" or rhetorical question left hanging, a claim asserted but not yet grounded — that is completed or proven only in verses beyond the assigned range? If so, the range likely closes too early. State what verses need to be added and why.  
    
- SECTION-BREAK CHECK: would a standard critical edition or major study Bible's own section headings place a break inside the assigned range rather than at its boundaries? If so, flag this even if the range reads coherently on its own — coherence achieved by cutting a unit in half is not the same as respecting where the unit actually begins and ends.

If any of the three checks trips, the brief does not silently work within the assigned range. It states the finding plainly and proposes the corrected range, the same way Mode Validation names a discrepancy rather than proceeding past it. The pastor may still choose to preach the narrower range deliberately — that is a legitimate choice, and a different thing from a narrow range nobody checked.

WORKED EXAMPLE (the failure that produced this check): a range handed down as Romans 9:14-23 opens on "What shall we say then?" — pointing back to 9:4-5's list of Israel's covenant privileges, without which verse 6's question carries none of its actual weight — and closes on a suspended "what if" that Paul does not complete until 9:24-29, where he names Gentile inclusion and grounds it in Hosea and Isaiah. Both the opening check and the closing check would have caught this before Pass 1 ever ran. The corrected range is Romans 9:1-29.

2. THE PASTOR'S OWN INITIAL READ What is he already hearing, sensing, drawn to, or wrestling with? A working title. A tension not yet precisely named. A single line that will not leave him alone. An inchoate pull toward something is a complete and valid answer. So is nothing at all — if the pastor has no prior instinct beyond the assigned text, the tool records that and the comparison step in Stage Three runs without an anchor.

HARD RULE: even "nothing yet" is recorded before Stage Two runs. The tool does not proceed to exegetical passes and then ask for pastoral input as a formality. That is the wrong order.

3. SERIES CONTEXT — REQUIRED IF THE TEXT BELONGS TO A SERIES If this text is part of a series: what is the series doing, and what is this particular text's role within the larger movement?

This is not optional background. The series purpose is prior to the individual text — it determines what the sermon needs to accomplish before exegesis begins. A text in a series on the Psalter as a school of prayer carries a different weight than the identical text preached in isolation. A text placed third in a four-week arc carries a different obligation than the same text preached as a standalone. The CCB cannot surface the right paths if it does not know what the larger movement requires.

REQUIRED SUB-QUESTION — SERIES MOMENTUM: what has already been preached in this series, and does anything preached so far create a corrective need this week — a direction the series needs to lean back toward, given what a prior entry emphasized, underweighted, or overweighted? This is a different question from series purpose. A series can have a clear and correct purpose while one of its early entries still drifts off that purpose in execution — a first week that leaned too far into past-tense, sin-management framing, for instance, creates a real pull on the second week to counterbalance, whether or not anyone names it explicitly. The brief cannot weigh this pull if it does not know the series' actual history, not just its stated intent. If nothing preached so far creates a corrective need, state that plainly rather than manufacturing one.

If no series: state that plainly. The field is not skipped — it is answered with "standalone" so Stage Three's path-shaping is calibrated accordingly.

4. CONGREGATION CONTEXT Anything about this specific congregation this week that belongs on the table — a grief the body is carrying, a conflict, a milestone, a season. Most weeks this field is brief or empty. On the weeks it is not empty, it belongs in the brief before exegesis runs, not discovered after a draft is built.  
     
5. INTENDED HEARER LOCATION ON THE WALK — OPTIONAL See Walk Location Reference above. If the pastor has a sense of where his hearers currently stand — mostly established believers, a mix, several newer in faith, a congregation walking through a season of failure and restoration — name it here. "Mixed" is a complete and valid answer, and most weeks it will be the honest one. This field is optional because the pastor often will not know, and because a text's own natural address (Pass 7\) may not match whatever is guessed here — in which case the mismatch itself is worth surfacing in Stage Three, not smoothed over.

═══════════════════════════════════════════════════════════════ STAGE TWO — PROCESS (SEVEN EXEGETICAL PASSES) ═══════════════════════════════════════════════════════════════

Each pass runs against the text(s) in view. Every specific factual or historical-critical claim in every pass is Hallucination Guard v1.3 — gated: retrieved and verified, or explicitly hedged, or omitted. Superscriptions, authorship claims, ANE background, scholarly attributions, and dates are default-null categories under HG Rule 1 — parametric recall is not sufficient grounds to state them as fact. Where a claim cannot be verified, the brief says so and gives only the general, uncontested form.

The governing question of Passes 1 through 6 shifts by mode. In text-first, each pass asks: what does this text open? In message- first, each pass asks: does this text actually carry the stated message in its own right, or is the message being imposed on it? The findings below are described in their text-first form; in message-first, the same six passes run against the proposed text(s) and are reported as support for, or resistance to, the stated message — including outright resistance, per the message-first guard above.

PASS 1 — Historical-Grammatical & Superscription Who wrote this, under what heading, in what circumstance — so far as the text's own superscription states or as scholarship broadly and verifiably agrees. For the Psalms in particular: the superscription is read as interpretive frame, not decoration. If a superscription names an author or a guild (sons of Korah, sons of Asaph, David), the brief traces what is textually known about that name elsewhere in the canon before Stage Two closes.

PASS 2 — Literary Context Runs on the range as validated by the Pericope Boundary Validation check in Stage One — this pass describes what surrounds an already- confirmed unit; it does not re-litigate whether the unit's own edges are correct. Where does the text sit within its immediate unit — the surrounding chapters, the collection it belongs to, the structure of the book as a whole? Any structural markers (chiasm, refrain, inclusio, envelope structure) that are load-bearing for meaning and would be lost if the pericope were read in isolation.

Beatitude and blessing-formula structures (Hebrew ashre, Greek makarios — "blessed is/are") are named explicitly here, not folded silently into the generic structural-marker category above. These formulas recur across the Psalter and Wisdom literature (Psalm 1:1, 32:1-2, 34:8, 41:1, 84:4-5-12, 112:1, 119:1-2, among others) and carry theological weight beyond ordinary refrain: they are the direct canonical precursor to the Beatitudes of Matthew 5:3-12. If the text under examination contains one or more ashre/makarios formulas, this pass counts and locates them (how many, where in the text's structure, whether they cluster or run through the whole), because Pass 3 below requires this count to check them against Matthew 5\.

PASS 3 — Canonical Cross-References What earlier canonical material does this text depend on, allude to, or require the reader to already know? What later canonical material depends on or cites this text? Named plainly and kept to the two or three connections that are actually load-bearing — this is not a concordance dump. The standard is: would the text's own first audience have heard this echo, and does the sermon lose something real if the echo goes unheard?

MANDATORY SUB-CHECK: if Pass 2 found one or more ashre/makarios formulas in the text, this pass explicitly checks each one against Matthew 5:3-12. This is not optional and not folded into the general load-bearing-connections judgment above — it runs every time a beatitude formula is present, because Matthew 5:3 (poverty of spirit) is this corpus's own governing posture, and an Old Testament beatitude that is not explicitly traced to its fulfillment in Christ's Beatitudes leaves the strongest available Christological handle unused. State plainly which Matthew 5 beatitude the text's own blessing formula corresponds to or anticipates, if any, and if none corresponds cleanly, say so rather than forcing a match.

PASS 4 — Redemptive-Historical Placement Where does this text sit on the arc from promise to fulfillment? What does the text's own historical moment know, and what does it not yet know? If multiple texts are paired for the same service, where does each sit relative to the others on that arc — which one is ahead of which, and what does the gap between them mean?

PASS 5 — Theological Synthesis Candidates Not a single handed-down Big Idea. Multiple candidate one-sentence syntheses, each defensible from Passes 1 through 4, each stated plainly without homiletical framing or illustrative language. These are raw material for Stage Three, not a recommendation. (Message- first: candidate syntheses are checked against the stated message rather than generated freely — does the text actually yield this message, a nearby message, or a different message entirely?)

PASS 6 — Christological Trajectory Where does this text's own internal logic reach toward Christ — traced from what the text itself is doing, not imported by external typology. If the text's own trajectory does not reach Christ directly (some wisdom material, some historical narrative), the brief states that plainly rather than forcing a connection the text does not make.

PASS 7 — Walk Location (Text's Natural Address) Using the Walk Location Reference above: which position or positions on the disciple's walk does this text, read in its own right, most naturally speak to and equip? A text can serve more than one position at once, and the deepest texts often do — Psalm 84's doorkeeper passage, for instance, speaks to poverty of spirit (position 2\) and to theosis-directed longing (position 9\) simultaneously, which is the concentric structure the Walk Location Reference describes rather than an exception to it. This pass names the text's own gravitational center; Stage Three assigns which position(s) a chosen message or path will actually serve, and flags it if that choice pulls away from what the text itself naturally addresses.

ORTHODOXY FLOOR TRIPWIRE (silent unless triggered) Before Stage Three synthesizes paths, check the Pass 5 candidates and the Pass 6 trajectory against the Theology Grader's five orthodoxy-floor commitments: Trinitarian, Christological, Bodily Resurrection, Scriptural Authority, Christ as Sole Mediator. Most texts never come near any of these and the check produces nothing to report — in that case it is not mentioned in the output at all. If a candidate synthesis or trajectory would cut against one of the five, state which one and why, before Stage Three names it as a path. This is a tripwire, not a routine recitation: it exists to catch the rare text that pulls toward a heterodox reading, not to announce a clean bill of health every time.

═══════════════════════════════════════════════════════════════ STAGE THREE — OUTPUT (BY MODE) ═══════════════════════════════════════════════════════════════

TEXT-FIRST OUTPUT — DECISION TREE, NOT A VERDICT

The brief does not conclude with one recommended sermon. It concludes with two to four named paths. Each path states:

- Which Theological Synthesis candidate(s) from Pass 5 it is built on  
- What exegetical material from Passes 1–4 it foregrounds, and what it leaves in the background  
- Its likely Theology Grader architecture classification (Devotional/ Contemplative, Covenantal/Relational, Redemptive-Historical/ Christological, Polemical/Apologetic, or a stated combination) — so the eventual TG run has a declared target rather than an implicit one discovered after drafting  
- Its known seam, named in the Theology Grader's own vocabulary rather than described freehand — e.g. "this path's synthesis leans toward Kingdom as future-only unless the already/not-yet is stated explicitly," or "this architecture is prone to sanctification as self-help unless the Spirit's interior work is named centrally, not only in closing." Naming the seam in TG's own terms means the risk flagged here and the failure TG would later catch are the same vocabulary, not two separate descriptions of the same problem  
- Its abiding target — one sentence naming what a hearer who actually lived from this path's synthesis would be doing differently six months on. This operationalizes the Grader's drive-to-abide, drive-to-obey, and drive-to-love tests as something chosen at the design stage, not discovered by accident once a draft exists  
- Its walk location served — which position(s) from the Walk Location Reference this path centers on, checked against Pass 7's reading of the text's own natural address and against Stage One field 5 if the pastor supplied an intended hearer location. If the path pulls toward a position the text does not naturally support, or away from where the pastor said the congregation actually stands, that mismatch is named plainly rather than smoothed over — it may be exactly the right choice for this week, but it should be a choice made knowingly  
- Its answer to series momentum — if Stage One field 3 identified a corrective need from what the series has already preached, this path states plainly whether it serves that correction, is neutral to it, or would compound the drift the series already has. If no corrective need was identified, this field is omitted rather than manufactured

If the pastor supplied a prior instinct in Stage One, the brief states plainly where each path aligns with it, extends it, or cuts against it. If the pastor supplied nothing, this comparison is simply omitted — it is not manufactured to give the appearance of continuity.

The brief closes with one line and stops:

"Which of these, if any, is where you want to go — or does the text pull you somewhere none of these name?"

The tool does not answer its own question. That answer belongs to the pastor.

MESSAGE-FIRST OUTPUT — FIT VERDICT PER PROPOSED TEXT

For each text proposed to carry the stated message, the brief states:

- Verdict: carries the message in its own right / partially carries it / does not carry it. A "does not carry it" verdict is a legitimate and expected output, not a malfunction — see the message-first guard above  
- What in Passes 1–6 supports the verdict, and what resists it — resistance is reported even when the overall verdict is favorable  
- Its known seam, its abiding target, and its walk location served, in the same terms as text-first above  
- Its answer to series momentum, in the same terms as text-first above, if a corrective need was identified in Stage One  
- Where the verdict is "partially carries" or "does not carry," an alternate text is proposed, with a brief statement of why it fits the stated message more directly

The brief closes by naming, across all proposed texts together, whether the full set coheres as a sequence (as in a sermon series) or whether the coherence is weaker than it appeared before the brief ran.

═══════════════════════════════════════════════════════════════ HANDOFF ═══════════════════════════════════════════════════════════════

Once a path (text-first) or a confirmed text set (message-first) is selected — by the pastor, in the same session or a later one — the full CCB output and the selection become the working context for the drafting session. Drafting proceeds exactly as it does now: web chat for the theological and compositional work, Claude Code for file assembly and version-cascade production per Michlaude Writing Standards v5.1 §15. The CCB does not draft the sermon and does not touch the file pipeline. It hands off a brief and steps back.

The TG, HG, and WS passes downstream are unchanged by this tool's existence. The CCB's only claim on the rest of the pipeline is that the draft entering TG should already know what Zach knew when he opened to Numbers 16 before he ever preached Psalm 84 — not because an instrument told him to, but because that is what responsible exegesis has always required — that the draft should be aimed at a known abiding target and forewarned of its own likely seam before the first sentence is written, and that it should know which position on the disciple's walk it is trying to serve, rather than discovering any of this after the Grader runs.

═══════════════════════════════════════════════════════════════ END OF BRIEF ═══════════════════════════════════════════════════════════════  

30.1 thousand characters · updated October 25, 2026

All the tools