What “In Depth” Means for a Church-Ready Reader

This appendix explains the Coherence-Based Genealogical Method at the level an editor uses it, while remaining clear enough for a diligent churchgoer. The CBGM is not a black box. It is a disciplined workflow that begins with exhaustive collation, proceeds through carefully defined variation units, registers agreements across witnesses, evaluates the direction of change inside each unit, and then aggregates those micro-judgments into reading-level relationships among witnesses. The method’s strength lies in its capacity to expose patterns in a contaminated tradition without pretending that a single, rigid family tree can explain all relationships. Its limit lies in the fact that its decisive step depends on editorial judgments about prior and posterior readings. A documentary-first posture keeps that limit in check by letting early, reliable witnesses carry first weight—especially the Alexandrian papyri and Codex Vaticanus (B, 300–330 C.E.)—while employing internal criteria as disciplined servants, not masters.

The Data the Method Requires and How It Is Prepared

CBGM presupposes full transcriptions of witnesses for a given book. A “witness” is any source that testifies to wording: Greek manuscripts first of all; early versions when their renderings map securely to a Greek reading; and patristic citations when they quote precisely. Before any counting can be done, the transcriptions must be normalized. Normalization is not homogenization. It preserves meaningful differences while smoothing those that are orthographic or otherwise non-distinctive. Without this step, a computer would record false disagreements and inflate noise.

The method also requires a clear catalog of witnesses with secure sigla, reliable dating, and coverage maps that indicate which sections are extant and legible. This is why a book-by-book approach is essential. Each book has its own witness profile, and the set of overlapping variation units changes with lacunae and damage. When a papyrus such as P66 (125–150 C.E.) covers long stretches of John, or when P75 (175–225 C.E.) provides sustained testimony in Luke and John that aligns closely with B, the agreement can be measured directly. Where coverage is patchy, the method automatically limits claims to what the overlaps allow.

Variation-Unit Segmentation: The First Non-Negotiable Decision

Everything CBGM does is downstream from how editors define the boundaries of a variation unit. A variation unit is the smallest span where competing readings are in genuine competition. Deciding where a unit begins and ends is not clerical; it is historical description. If a scribe made a single change that produced both a lexical substitution and a word-order adjustment, the editor should keep those effects inside one unit so the later computations do not treat one historical act as two. If unrelated differences cluster in proximity, they should be separated so that coherence does not blur causation.

Conservative segmentation guards against statistical artifacts. When unit boundaries mirror scribal phenomena on the page, pre-genealogical agreement counts what actually happened and genealogical inference does not overcount a single act. When boundaries are inflated or arbitrary, later tallies become fragile. Responsible use of CBGM therefore begins with restraint at the segmentation stage. This is also where documentary priorities shape the editor’s instincts. Early witnesses that preserve a tighter, less smoothed construction help an editor see the contours of a genuine unit more clearly.

Enumerating Readings and Subreadings with Accurate Attestation

Within a defined unit, the editor lists all distinct readings. If two forms differ only orthographically while preserving the same lexical choice, they are registered as subreadings. If they change sense, structure, or lexical substance, they are distinct readings. Attestation is then assigned to each reading. This stage is purely descriptive. It tells us who reads what; it does not tell us who is right. Accuracy here is essential because every later inference depends on a trustworthy list of who supports each option.

A simple example clarifies the point. If the rivals are “Lord” and “Jesus” in a clause, the editor must be certain that versional and patristic evidence actually points to one of those Greek forms rather than to a translation habit or a paraphrase. A Syriac rendering may be consistent with either option; a Latin witness may collapse two Greek possibilities. Care at this stage prevents the later coherence metrics from resting on ambiguous evidence.

Pre-Genealogical Coherence: Mapping Agreement Before Inference

CBGM begins with counting. Pre-genealogical coherence measures how often two witnesses agree across the set of defined units. This descriptive layer is objective in the simple sense that preferences do not alter the count. If Witness A and Witness B read the same way in a high percentage of overlapping units, they display strong pre-genealogical coherence. If they diverge often, they display weak coherence. This map is valuable because it reveals clusters that deserve attention and exposes mixture where agreement shifts unpredictably.

The descriptive map typically highlights the stabilizing significance of early Alexandrian witnesses. In Luke and John, the close alignment of P75 (175–225 C.E.) with B is visible and weighty. Agreement between a late second–early third-century papyrus and a fourth-century codex witnesses to a stable line of text that reaches back very near to the autographs written in the first century C.E. Where such anchors exist, the later steps of the method should respect them; where such anchors are thin, the later steps must be stated with proportioned confidence.

Local Stemmata: The Editorial Hinge Expressed as Arrows

After description comes direction. A local stemma is the miniature diagram an editor draws for a single unit to indicate which reading is prior and how the others arose from it. The editor asks which reading best explains the origin of its rivals by ordinary scribal tendencies and authorial style. If a shorter reading fits context and easily generates a later explanatory expansion, or if a less polished word order can be smoothed by a scribe but would rarely be made harsher, the earlier reading is placed at the head.

This step is where internal criteria belong. Intrinsic probability assesses fit with the author’s usage and immediate context. Transcriptional probability assesses how a scribe would likely copy one form into another, considering habits such as harmonization to parallels, conflation, expansion, smoothing, and omission of short particles. A documentary-first approach requires that internal criteria serve external anchors, not displace them. When early Alexandrian papyri and B agree, that agreement weighs heavily because it anchors a reading near the earliest recoverable stage of the text. Internal criteria then explain why that anchored reading generated later forms.

Genealogical Coherence: Aggregating Local Arrows into Reading-Level Relations

When local stemmata exist for many units, CBGM aggregates them to see whether one witness frequently carries prior readings relative to another in overlapping units. If Witness A regularly holds the reading judged earlier where Witness B holds a posterior reading, A functions as a plausible source for B at the level of readings. The tool represents this with a line from A to B. The thickness of the line indicates how many units support that direction. Thick lines indicate persistent direction; thin lines warn that only a few units drive the relation.

Genealogical coherence does not claim that B was copied from A as a physical exemplar. It claims that the readings in A often stand earlier relative to those in B. This reading-level relation respects contamination. A single witness may have more than one strong potential ancestor because its readings flow from multiple streams. The network that results is not a family tree; it is a realistic picture of influence in a tradition where mixture is common.

Potential Ancestors and Substemmata: What the Labels Mean

A “potential ancestor” is the witness that often supplies earlier readings relative to another witness across overlapping units. For each witness, the tool lists its closest potential ancestors and thereby builds a substemma. The substemma is not a claim about a copying chain; it is a practical way to say which earlier witnesses best account for the readings in a given manuscript. Because the New Testament tradition exhibits contamination, a witness may have different potential ancestors in different sections, and that is not a flaw. It is an honest reflection of how scribes worked.

Understanding this terminology protects students from common misunderstandings. When a graph shows a line from A to B, it is not asserting that a scribe of B had A open on the desk. It is saying that, in many places, the judgments made in local stemmata place A’s readings earlier than B’s. That is a modest but useful claim, and it is entirely testable in the units that generated the line.

The “Initial Text”: Operational Definition and Practical Consequences

In CBGM usage, “initial text” names the reading judged to stand at the head of the tradition in a given unit. Across a book, that set of readings composes the reconstructed text for publication in a research edition. The term is operational, not mystical. It does not automatically equal the autograph; it is the editor’s best reconstruction of the earliest recoverable form for that unit. Because the potential-ancestor network is computed relative to these initial-text decisions, the global portrait inherits the editorial grid. When the grid changes—because a unit is resegmented, or a local arrow is redrawn—the network can change.

This tether does not invalidate the method. It simply requires candor. Editors and teachers should say plainly that graphs visualize many small judgments and that those judgments rest, wherever possible, on early, high-quality witnesses. When a unit’s assigned initial text lacks Greek manuscript support, the fact should be transparent and confidence should be proportioned accordingly.

How CBGM Handles Contamination: The Virtue of Reading-Level Relations

Classical stemmatics drew trees that assumed each copy descended from a single exemplar. That assumption rarely fits the New Testament copying culture. Scribes consulted more than one exemplar, corrected against marginal notes, and imported readings from memory of parallel passages. CBGM addresses this reality by describing relations at the level of readings rather than asserting exemplar chains. The potential-ancestor network shows where a witness often stands earlier relative to another without demanding a pure line of descent. This honesty is one of the method’s genuine contributions. It frees discussion from rigid family rhetoric and focuses attention on unit-level causation.

Where the Documentary Method Must Lead the Way

A documentary-first posture gives the earliest, best witnesses priority. Early Alexandrian papyri such as P52 (125–150 C.E.), P66 (125–150 C.E.), P46 (100–150 C.E.), P75 (175–225 C.E.), and others, when aligned with B, anchor readings that reach very near to the autographs. The close affinity of P75 with B in Luke and John, often summarized around eighty-three percent agreement, displays an early, stable line of text that deserves first attention. This does not grant doctrinal status to any tradition. It simply recognizes chronological and qualitative proximity to the earliest form. CBGM can strengthen this priority by showing how readings likely flowed from these anchors into later mixed witnesses. It should not be used to justify drifting away from these anchors on the strength of internal arguments alone.

A Deep Walk-Through: From Unit Definition to Global Feedback

Consider a unit where the rivals are a compact phrase supported by P75 and B and an expanded phrase supported by a broad array of later witnesses. The segmentation isolates the entire phrase so that the expansion is treated as one act, not several. The attestation is registered: early papyrus plus B for the compact form; later cluster for the expansion. The local stemma places the compact form as prior because expansion is a common scribal tendency and because the compact form fits the author’s known style. Genealogical coherence across the book reveals that the witnesses carrying the compact form often function as potential ancestors relative to the expansion cluster. The initial-text assignment prints the compact form. The global network confirms that, elsewhere in the book, similar arrows run in the same direction. The decision is not reverse-engineered from a graph; the graph visualizes many such unit-level decisions that are themselves anchored by early evidence.

Now consider a second unit where the preferred reading rests chiefly on an internal argument but lacks direct Greek support. The segmentation is careful, the internal pathway plausible, and the global portrait does not contradict the local call. Nevertheless, a documentary-first approach describes the decision as provisional and refrains from building exegesis or doctrine on it. The method has done its work transparently. The editor has done the right thing by stating confidence in proportion to evidence.

Sensitivity to Segmentation: Why Small Boundary Changes Matter

Segmentation is where hidden fragility can enter. If a single scribal act moved a deictic and omitted a connective—and the editor splits them—later computations double-count one act as if it were two independent changes. The opposite error lumps unrelated differences into one unit, blurring causation. Responsible practice compares alternative reasonable segmentations and asks whether the local direction stands under each. If a direction flips only under an artificial segmentation, confidence increases. If a direction depends on an aggressive boundary, confidence should be labeled accordingly. Stating this sensitivity in plain speech strengthens readers’ trust because it shows that conclusions are not brittle.

Thresholds and Line Weights: How Strength of Evidence Is Displayed

The method’s visual lines have thickness because they carry the weight of many units. A thin line is a reminder, not a verdict. It says that the directional relation holds in only a few places and may vanish if those units are re-evaluated. A thick line reflects many units and therefore a relation less likely to evaporate under modest changes. Editors can adjust thresholds for displaying relations so that the graph is neither cluttered with noise nor forced into false simplicity. Readers trained to notice line weight instinctively ask the right follow-up questions: which units carry this relation, and what do the earliest witnesses say there?

Versions and Fathers: Expanding the Footprint with Discipline

Versions and patristic citations widen the evidentiary base when used with philological care. A version contributes most when its translator’s habits are known for the book and when the rendering points unambiguously to one Greek reading. A patristic citation helps when it is explicit and likely drawn from a manuscript rather than composed from memory. CBGM can register these witnesses in both the descriptive and directional layers. Documentary priorities ensure that versions and fathers confirm and illuminate but do not override Greek anchors. When Greek evidence is thin, their testimony may weigh more heavily, but confidence is still stated in proportion to their reliability.

The Byzantine Cluster in CBGM’s Frame

Inside CBGM outputs, “Byz” typically labels a computed cluster of high agreement relative to the evolving initial text rather than a nineteenth-century essentialized text-type. This reframing is helpful so long as it is understood. Late uniformity can be displayed and measured without assuming doctrinal authority for the cluster and without denying that individual Byzantine witnesses sometimes preserve early readings. The originality question is resolved unit by unit. Where early Alexandrian papyri and B align on a reading that creates the rival forms by ordinary scribal tendencies, that anchored reading deserves priority regardless of the size of a late cluster. Where early anchors are divided and internal direction favors the Byzantine-supported reading, that reading should be considered carefully and, when warranted, adopted.

Advanced Use: Feedback Loops Between Local and Global Levels

One of CBGM’s strengths is the feedback loop between local decisions and global patterns. If the global portrait shows a witness unexpectedly feeding a cluster that departs from early anchors, editors return to the relevant units to test whether the local arrows were drawn under assumptions that need correction. Sometimes the segmentation hid causal links; sometimes an internal argument was overextended; sometimes an overlooked early witness alters the balance. The loop protects the method from premature certainty by making re-examination normal. It also protects readers by ensuring that a visually persuasive global pattern never overrides stubborn unit-level facts.

Case Windows: Jude 5, James 2:3, 1 Peter 5:1, Acts 20:28, and 2 Peter 3:10

A few units illustrate the depth of the method in practice. In Jude 5, the contest between “Jesus” and “Lord” requires precise attestation and a candid local stemma. The internal case for “Jesus” is striking, but the documentary footing must carry the decision. In James 2:3, word order nuances the scene; early Alexandrian support for the sharper arrangement, combined with ordinary smoothing tendencies, yields a stable prior/posterior judgment corroborated by genealogical coherence. In 1 Peter 5:1, the presence of a connective squares with the letter’s discourse and with the common tendency to omit short particles; early witnesses anchor the decision and the global portrait reflects a steady direction from presence to omission. In Acts 20:28, early Alexandrian support for “church of God,” with a clear transcriptional pathway for substitution, carries the day; CBGM describes rather than dictates this result. In 2 Peter 3:10, a reconstructed negative without direct Greek support shows the boundary of the method’s authority; internal arguments can be expressed and coherence computed, but documentary caution remains essential.

Internal Criteria as Servants: Keeping the Hierarchy Clear

Intrinsic and transcriptional probabilities are necessary and useful. They ask, respectively, what the author most likely wrote and what a scribe would likely do when copying. The hierarchy must remain clear. External anchors are first. When papyri such as P75 align with B in Luke and John, or when P66 agrees with B in John, those alignments take precedence because they witness to early states of the text. Internal criteria then explain those alignments, not overturn them. Where early anchors are divided or absent, internal criteria may carry more weight, but the conclusion is labeled accordingly and remains open to refinement when new evidence or better arguments appear.

Practical Replicability: Why CBGM Inferences Are Testable

Because CBGM forces editors to draw local stemmata and to identify the specific units that feed global relations, its inferences are replicable. A critic can review the same units, consider the same alternative segmentations, and decide whether the arrows are justified. Agreement or disagreement becomes concrete rather than rhetorical. This testability is a virtue for the church. It invites mature discussion centered on manuscripts and scribal habits rather than on slogans about families or preferences for novelty.

How CBGM Informs Translation Without Dictating It

Translators need clear, defensible decisions at the clause level. CBGM serves them by making the editorial chain of reasoning explicit. When the initial text is anchored by early witnesses and ordinary transcription, translators can render with confidence. Where a CBGM-era reconstruction rests mainly on internal direction without Greek support, translators retain the freedom to follow the well-attested Greek form and to note the editorial proposal. Either way, the discipline helps translators write precise marginal notes that teach readers why a connective, word order, or lexical choice was adopted.

Calibrating Confidence: Expressing Strengths and Limits in Plain Speech

Editors can and should pair their CBGM outputs with confidence statements that match the evidence. Where early Alexandrian papyri and B agree and the internal pathway is straightforward, the decision can be treated as secure for practical purposes. Where early anchors are divided but the internal case is strong, the decision can be called likely with references to the specific witnesses. Where Greek support is lacking and internal arguments carry the weight, the decision should be described as provisional, commendable for consideration but not a foundation for exposition that requires high certainty. This calibrated language trains readers to live with confident clarity rather than with either bravado or paralysis.

A Brief Chronology to Keep the Big Picture in View

The autographs of the New Testament books were written in the first century C.E., within decades of Jesus’ death in 33 C.E. By the late second and early third centuries, papyri such as P46 (100–150 C.E.), P66 (125–150 C.E.), P75 (175–225 C.E.), and others bear witness to readings that later appear in B (300–330 C.E.). The fourth and fifth centuries produced great majuscule codices—א (330–360 C.E.), A (400–450 C.E.), B (300–330 C.E.), D (400–450 C.E.)—that channel earlier lines of transmission into carefully produced volumes. Modern collation and digital tools allow editors to bring this early, abundant tradition into coherent view. CBGM belongs in this historical arc as a tool for organizing evidence and clarifying reasoning, not as a replacement for documentary priorities.

Common Misunderstandings and Their Correctives

A few misunderstandings often cloud discussion. Some assume CBGM discovers exemplar chains; it does not. It reports reading-level direction. Others assume that coherence proves originality; it does not. Coherence maps agreement and aggregates local judgments; originality is established by weighing those patterns against early anchors and sober internal criteria. Still others assume that a “Byz” label in a graph grants doctrinal privilege or proves lateness; it does neither. It describes high agreement within a computed frame. Clarity about these points prevents needless controversy and keeps attention on the units where real decisions are made.

How to Audit a CBGM-Based Argument as a Serious Reader

A churchgoing reader can audit a CBGM-based claim without specialized tools by asking specific questions. The first question is how the unit is segmented and whether a different reasonable boundary would change the decision. The second is which early Greek witnesses attest the rival readings. The third is what transcriptional pathway is being claimed and whether it matches ordinary scribal habits. The fourth is whether the witnesses that carry the preferred reading routinely stand earlier relative to others in nearby units. The final question is how the editor states confidence. This five-question audit often resolves debates in minutes because it forces the argument into the open and honors the earliest, best testimony.

The Method’s Real Payoff for the Church

The CBGM’s payoff is transparency. It requires editors to show how they got from raw collation to printed line. It provides pastors and teachers with a vocabulary for explaining decisions in clear, non-technical language. It invites readers to think with the manuscripts rather than against them. When used under documentary control, the method strengthens, rather than weakens, confidence in the providentially preserved text. The handful of debated places become classroom opportunities to model disciplined humility, while the vast majority of the text stands secure on early, converging testimony.

A Final Orientation for Deep Users and Beginners Alike

This appendix has unfolded the method from the ground up: data preparation, conservative segmentation, descriptive mapping of agreement, local arrows that express direction, aggregated reading-level relations that respect contamination, and an operational reconstruction labeled “initial text.” The entire workflow gains stability when early papyri and B are allowed their due weight and when internal criteria are kept in their proper role. With these guardrails in place, CBGM is a helpful ally. It makes complex data navigable, compels precise argument, and keeps every claim testable in the actual words of the witnesses.

You May Also Enjoy

Guardians of the New Testament: Literacy, Power, and the Copyists of The New Testament

About the Author

CLICK LINKED IMAGE TO VISIT ONLINE STORE

CLICK TO SCROLL THROUGH OUR BOOKS