The Strongest Critical-Text Case: Rules against Straw Men

Revision note: Substantially revised July 20, 2026. The argument has been reordered according to the classical rhetorical movement, compressed where necessary, and given one consolidated Notes and Further Reading section.

A Case Worth Testing Must Be the Best Case

The thesis of this series is deliberately demanding. It proposes that the modern critical-text ecosystem cannot independently reconstruct the whole Westminster system with the same directness, precision, stability, and canonical coherence as the received-text tradition from which that system historically arose. A conclusion that large cannot be established by defeating small arguments.

It would be easy to assemble careless popular statements about textual criticism. One could answer the person who believes that editors simply follow the two oldest manuscripts, the reader who assumes that every bracketed verse is false, or the preacher who says that a doctrine remains unaffected because he can still find its vocabulary somewhere else. Those claims deserve correction, but refuting them would not test the strongest critical-text case.

Nor would it be enough to place a received reading beside an ESV rendering and declare the theological matter settled. A modern critical-text advocate may offer external evidence for the shorter reading, internal reasons why a scribe would expand the wording, a local interpretation that preserves a legitimate distinction, parallel passages that supply cumulative warrant, and a systematic argument that does not depend upon the disputed sentence. Unless those resources are admitted and tested together, the alleged failure belongs to a caricature rather than to the position under review.

The governing rule is therefore simple:

No negative verdict counts unless it survives the strongest responsible, textually consistent, and canonically complete reconstruction available to the critical-text position.

That rule is not a ceremonial disclaimer. It determines which evidence must be gathered, which objections must be answered, how opposing claims may be attributed, and when the series is permitted to assign a grade below successful derivation. A conclusion reached without that work will be withdrawn or revised.

Fair Representation Is a Confessional Duty

The duty to represent an opponent accurately does not arise from academic etiquette alone. It follows from the moral theology this series professes.

Westminster Larger Catechism 144 places under the ninth commandment the duties of “preserving and promoting of truth,” maintaining the good name of one’s neighbor, receiving a good report willingly, and giving charitable esteem to others. It also requires “from the heart, sincerely, freely, clearly, and fully, speaking the truth.” Question 145 forbids “rash, harsh, and partial censuring” and “misconstructing intentions, words, and actions.”¹

Those requirements govern theological controversy. A critic may not preserve the truth of his conclusion by misdescribing the reasoning of the person he opposes. He may not infer a corrupt intention merely from a harmful consequence. He may not quote a scholar’s qualification away, move from one individual’s statement to an entire class, or describe an edition as doing what only one user of that edition has done.

The same catechism also forbids flattering sinners, hiding faults, and excusing what ought to be reproved. Charity is therefore not indulgence. A fair account may conclude that an argument is circular, that an editorial method is unstable, that a reading weakens a confessional proof, or that an attempted harmonization fails. It may reach those conclusions firmly. What it may not do is acquire rhetorical certainty by suppressing evidence that would make the conclusion harder to prove.

The ninth commandment thus cuts in both directions. It forbids caricature, and it forbids false exoneration. The right standard is neither hostility nor flattery, but truth stated in its proper proportions.

Steelmanning Is Not Neutrality

The modern term steelmanning describes the practice of presenting an argument in its strongest responsible form before criticizing it. That practice does not require pretending to possess no theological commitments.

This series begins from confessional commitments concerning revelation, preservation, the authority of the original tongues, the unity of Scripture, and the legitimacy of good and necessary consequence. Critical-text advocates likewise reason from convictions about manuscripts, transmission, historical evidence, editorial responsibility, and the relation between inspiration and textual reconstruction. No party enters the discussion as a view from nowhere.

Fair representation asks a different question from neutrality. It asks whether the critic has described the position being tested so accurately that a well-informed adherent could recognize it, and so strongly that defeating it would matter. A presuppositional critique can grant another system’s proposed premises for the sake of the test and ask what follows from them. That temporary grant is neither an endorsement of the premises nor a surrender of the critic’s own standard. It is an internal test of the claim on its strongest available terms.

Three acts must therefore be distinguished:

1. Understanding: stating what the opposing position actually affirms and why.
2. Provisional admission: allowing its operative text, evidence, and strongest exegesis to function within a controlled reconstruction.
3. Final judgment: evaluating whether the resulting construction is textually warranted, canonically coherent, and sufficient for the confessional proposition.

One may perform the first two without prejudging the third. Indeed, a negative conclusion has little value unless the first two have been done.

The Strongest Responsible Critical-Text Case

There is no single person called the critical-text advocate, and no critical edition contains a doctrinal manifesto. The following statement is therefore a controlled synthesis. It gathers claims that a responsible contemporary advocate could hold together; it is not a quotation attributed to every editor, translator, or Reformed user of a modern edition.

The strongest case may be stated as follows:

The inspired writings were given in determinate words, but the autographic documents are no longer extant. The surviving manuscript tradition contains many copying differences while preserving the overwhelming substance of the text. The church should therefore compare the full available evidence in order to identify the earliest attainable or initial form of the text—and, according to some formulations, the authorial original—rather than canonize any one late printed edition merely because it became traditional.

Readings should be judged through a reasoned combination of external and internal evidence. The age of a witness matters, but neither age nor manuscript count decides the question by itself. A late manuscript can preserve an early reading; many copies can descend from a comparatively narrow source; an early witness can contain individual errors. Genealogical relationships, geographical and versional evidence, scribal tendencies, authorial style, immediate context, and the relative ability of one reading to account for the rise of the others all require consideration.²

Modern tools such as the Coherence-Based Genealogical Method assist that inquiry by comparing states of text and illuminating possible relationships within a heavily contaminated tradition. They do not mechanically choose the biblical wording. Local stemmata require philological judgment, and the method supplies information that editors must interpret rather than an automatic verdict from a computer.³

A reading must not be accepted merely because it is familiar, explicit, orthodox, or useful in controversy. If the best historical and transcriptional evidence indicates that an orthodox expansion is secondary, fidelity to inspiration requires excluding it from the main text. Theology must be derived from the wording judged authentic; a preferred doctrine must not dictate which wording is declared original.

The presence of textual variation does not make the New Testament unrecoverable or Christian doctrine indeterminate. Most variants are inconsequential, the great majority of the text is not seriously disputed, and meaningful uncertainty is ordinarily confined to a limited body of readings. Major doctrines rest on multiple passages and mutually reinforcing canonical evidence. No cardinal Christian doctrine stands or falls solely upon one viable textual variant.⁴

A critical edition does not erase rejected readings from history. Its apparatus records alternatives and permits later scholars to inspect, challenge, and revise editorial decisions. Responsible translations likewise disclose many significant differences in notes. Revision in light of fuller evidence should therefore be understood as accountable scholarship, not as proof that Scripture has vanished.

---

1. Westminster Larger Catechism 144–145, in the Orthodox Presbyterian Church’s Larger Catechism. ↩
2. Bruce M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission, Corruption, and Restoration, 4th ed. (New York: Oxford University Press, 2005), 205–343; Institute for New Testament Textual Research, “Coherence-Based Genealogical Method”, accessed July 17, 2026. ↩
3. Gerd Mink, “The Coherence-Based Genealogical Method—What Is It About?”, Institute for New Testament Textual Research, accessed July 17, 2026; Tommy Wasserman and Peter J. Gurry, A New Approach to Textual Criticism: An Introduction to the Coherence-Based Genealogical Method (Atlanta: SBL Press; Stuttgart: Deutsche Bibelgesellschaft, 2017). ↩
4. Daniel B. Wallace, ‘Challenges in New Testament Textual Criticism for the Twenty-First Century,’ Journal of the Evangelical Theological Society 52, no. 1 (2009): 79–100, especially 95–96. ↩

This is the case the series must test. Its several claims will eventually require separate historical, textual, and theological evaluation. At this stage they define the opponent that may not be replaced with a slogan.

What the Critical-Text Reconstruction Is Allowed to Use

The strongest reconstruction must receive all resources legitimately available within the designated experiment.

It may use the exact main text of NA28 and the current UBS6/GNT6 text, including places where the latter anticipates the forthcoming NA29. It may use the full BHS text and published BHQ fascicles according to their proper editorial character. It may use grammar, syntax, discourse, genre, historical setting, authorial usage, and responsible intertextual relations. It may appeal to every unaffected passage in the canon, not merely the verse in which a variant occurs. It may construct a cumulative proof from several passages and draw conclusions by good and necessary consequence.

It may also use the strongest grammatically viable interpretation of a difficult reading, even where the series ultimately prefers another interpretation. If a canonical distinction resolves an apparent contradiction without violating either context, that distinction must be credited. If a received proof disappears but the exact doctrine remains securely derivable elsewhere, the doctrine receives the positive grade the evidence warrants.

The reconstruction may consult apparatus readings and translation notes as evidence about the textual problem. Their location must remain visible. A reading below the line may help explain the history of the main text, but it does not silently become main-text warrant. An ESV rendering may be defensible, but it does not automatically tell us what NA28 or BHS says. Article 8 will examine those layers in detail; the present rule is simply that every legitimate resource is admitted under its true name.

Finally, the reconstruction may use arguments that this series develops on behalf of the critical-text position, even if no single published author has assembled them in exactly that form. A full edition-exclusive reconstruction of Westminster is not known to have been completed at the scale proposed here. Refusing to consider an available argument merely because an advocate has not yet made it would rig the experiment. But every project-constructed argument must be labeled as such, and it must satisfy the ownership and admissibility rules below.

The Ownership Rule: No Opponent Who Does Not Exist

Steelmanning can itself become unfair if it creates a synthetic opponent stronger than any coherent position.

Suppose one scholar defends reading A because it is the NA28 main text, another prefers apparatus reading B, a translator renders A as though it carried a semantic feature found in B, and a theologian relies on the received reading C. A composite defense that uses A for textual status, B for grammar, the translation for clarity, and C for doctrine may answer every local objection. It does not represent one stable reconstruction.

The ownership rule therefore asks:

1. Could one responsible advocate hold all the selected premises simultaneously?
2. Do the premises arise from the same declared edition policy?
3. Are the proposed interpretations mutually compatible?
4. Would adopting one repair require abandoning a rule used elsewhere?
5. Has any project-created synthesis been identified as the author’s construction rather than another person’s statement?

If the answers are positive, the synthesis belongs in the steelman even if no author has printed it in one place. If not, the alternatives must be tested separately. The series may report that several possible defenses exist; it may not fuse their incompatible advantages into a single invulnerable position.

This rule also prevents a different abuse. The fact that one advocate cannot answer an objection does not prove that the critical corpus cannot answer it. Personal performance and textual capacity are not identical. The project tests the strongest coherent construction the position can own, not the debating skill of a selected representative.

The Admissibility Rule: Responsible, Not Merely Imaginable

The phrase strongest possible case can become an invitation to endless rescue. Almost any wording can be made compatible with almost any doctrine if hypothetical senses, unexpressed qualifications, conjectural sources, and unbounded appeals to mystery are all admitted. A test that must defeat every imaginable explanation cannot reach a conclusion.

The converse limit is equally important. A steelman is not strengthened by allowing its conclusion to exceed its premises. Once the best admissible case is constructed, it must undergo the exact-proposition, premise, closure, and no-borrowing tests. Each indispensable premise is classified as explicitly stated, necessarily entailed, cumulatively probable, merely compatible, not established, or contradicted. A rival model that can affirm every cited passage without affirming the exact confessional conclusion remains a live countermodel. The argument may receive credit for the greatest conclusion it warrants, but not for an unproved exclusion.

This is especially important when repeated examples are mistaken for an exhaustive class. Several passages naming the same three divine persons may establish at least those three; the repetition does not by itself prove that no other divine person exists. Steelmanning gives the critical construction every admissible proof, but closure still has to be demonstrated rather than presumed from Christian tradition.

The correct standard is the strongest responsible case. An argument is admissible when it meets at least five conditions:

• Linguistic viability: the proposed meaning falls within the attested grammar and semantics of the wording.
• Contextual fitness: it makes sense in the immediate discourse, genre, and historical setting.
• Evidentiary support: it rests on relevant manuscript, lexical, exegetical, historical, or canonical evidence rather than mere possibility.
• Methodological consistency: it uses principles the advocate accepts elsewhere and does not change standards solely to escape the present difficulty.
• Doctrinal non-circularity: it does not assume the exact proposition it is supposed to derive unless that proposition has already been independently established.

Publication is not an absolute requirement; new arguments are possible. But an idiosyncratic proposal bears a heavier burden than a standard, well-supported reading. Likewise, popularity does not guarantee responsibility. The point is to exclude both the weakest polemical target and the endlessly adaptable phantom defense.

Evidence before Consequence

Textual controversies often reverse the proper order of argument. A reading appears to weaken a doctrine; therefore, it is said to be corrupt. Or a reading is thought historically prior; therefore, its doctrinal implications are said to be harmless. Neither inference follows.

The analysis must proceed in three stages:

1. Textual judgment: What wording does the edition select, what alternatives exist, and why is one reading argued to be earlier?
2. Exegetical judgment: What does the selected wording mean in its local and canonical contexts?
3. Doctrinal judgment: What proposition can that meaning establish, weaken, leave underdetermined, or exclude?

The stages interact without collapsing. Theology can alert the interpreter to a contradiction, a canonical relation, or a consequence that deserves investigation. It cannot make a poorly attested wording original by doctrinal convenience. Textual evidence can establish which words probably belong in the passage. It cannot decree in advance that every theological consequence of those words must be negligible.

The series will therefore allow a critical reading to win the textual argument and still record a doctrinal loss. It will also allow a received reading to contribute important doctrine while concluding that the reading is textually secondary, if the evidence warrants that result. Confessional bibliology does not gain credibility by making every familiar wording immune to evidence.

What the Steelman Concedes—and What It Leaves Open

The discipline of fair representation gives the critical-text position several substantial concessions.

It concedes that the recovery of earlier wording is a legitimate scholarly aim; that no responsible method reduces to age or majority alone; that the CBGM is an analytical aid rather than an automated oracle; that apparatuses preserve evidence rather than conceal every rejected reading; that BHS and BHQ must be described as diplomatic Hebrew editions; that the ESV is neither identical with NA28 nor reducible to one editorial choice; that cumulative doctrinal proof is legitimate; that an orthodox longer reading may be secondary; that a difficult shorter reading may be authentic; and that the absence of a familiar phrase does not automatically destroy the doctrine associated with it.

Those concessions leave the governing questions untouched:

• Does the selected critical reading have the best textual claim in each case?
• What does it actually say under the strongest viable exegesis?
• Does the whole critical corpus establish the exact Westminster proposition?
• Are the needed qualifications supplied by the corpus or inherited from the theology being reconstructed?
• Can the same textual and interpretive rules be applied across every dossier?
• Do local solutions remain compatible when assembled into one theological system?
• Does the result preserve the same doctrinal directness and precision, or only a broader orthodoxy?

Steelmanning does not answer those questions for the critical-text position. It removes illegitimate shortcuts from the argument against it.

Objections to the Steelman Requirement

“Critical-text advocates do not extend the same charity”

Sometimes they do not. Received-text and confessional-bibliology positions are frequently reduced to King James Version perfectionism, ignorance of variants, or blind attachment to one late edition. Those misrepresentations should be corrected with the same precision required here. Another person’s violation of the ninth commandment does not authorize ours.

“Granting their premises abandons confessional authority”

It does not. The series provisionally operates with the critical corpus because that is the object of the experiment. It can ask what follows from that corpus without confessing that the editorial method which produced it is sound. An internal critique tests a position under granted premises; a final judgment determines whether those premises and their consequences should be accepted.

“The strongest case will become unfalsifiable”

It would if every imaginable rescue were admitted. The ownership and responsible-admissibility rules prevent that result. A defense must be coherent, linguistically viable, evidenced, non-circular, and methodologically consistent. Alternatives are allowed; improvisation without limits is not.

“Scholars disagree, so no strongest case exists”

Disagreement requires branching analysis, not surrender. Where two major critical interpretations exist, the dossier can test both. If either supplies a successful and consistently ownable reconstruction, the critic has not proved failure. If every responsible branch fails for separately stated reasons, a negative verdict becomes stronger.

“No advocate has attempted the whole Westminster reconstruction”

That is one reason for this project. The absence of a precedent cannot be used as evidence that no reconstruction is possible. The series must build the best case it can from the critical corpus and compatible scholarly resources. It must then label its construction and invite correction from those whose position it seeks to represent.

“The rule gives too much away”

A conclusion is not strengthened by objections it was never forced to survive. If the critical text can rebuild a doctrine after every responsible resource is granted, the article should record success. If the project’s broader thesis is true, it will emerge from repeated, evidence-controlled failures or material weakenings—not from protecting the thesis against its best counterexamples.

Verdict

The strongest critical-text case is not the claim that manuscripts never vary, that editors possess an infallible method, that every printed reading is certain, or that no wording has theological consequences. A responsible case acknowledges variation, editorial judgment, degrees of uncertainty, and the possibility of revision. It argues that the earliest attainable text should be reconstructed from the complete evidence; that neither age nor number decides by itself; that modern genealogical tools assist but do not replace philological judgment; that doctrine must follow the wording judged authentic; and that the broad, cumulative witness of Scripture prevents viable variants from overthrowing cardinal Christian doctrine.

That case must be granted its strongest textually consistent form. It may use the whole canon, good and necessary consequence, the strongest viable local exegesis, every responsible alternative proof, and every compatible argument a well-informed advocate could own. It may not be replaced with “two manuscripts,” “a computer chose it,” “all shorter readings are omissions,” or “editors removed doctrine” without evidence.

The grant has boundaries. A steelman may not alternate opportunistically among incompatible editions and readings. It may not use an ESV rendering as though it were NA28, an apparatus alternative as though it were the main text, or inherited Reformed distinctions as though their textual derivation had already been demonstrated. It may not answer evidence with mere possibility, or fuse mutually inconsistent defenses into an opponent who does not exist.

Matthew 5:22 shows the result. The absence of εἰκῇ does not by itself teach that every anger is sinful or make Christ a sinner. The critical-text advocate has a real textual argument and a serious contextual and canonical reply. The remaining issue is narrower and stronger: whether the shorter reading itself supplies the relevant moral boundary with equal directness, and whether the canonical qualification is derivation or repair. Only the full dossier may decide it.

These anti-straw-man rules do not weaken the proposed critique. They determine whether it deserves to be believed. If the critical-text reconstruction succeeds under them, the series must report the success. If it fails after every responsible resource has been granted, the failure will belong to the strongest case rather than to a polemical substitute.

Notes and Further Reading