There Are No Brute Manuscripts

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.

“He that is first in his own cause seemeth just; but his neighbour cometh and searcheth him.” — Proverbs 18:17

“Let the manuscripts speak for themselves.”

Few statements sound more reasonable. The Christian who disputes a modern critical text is often told to stop imposing theology upon the evidence. He should set aside confessional commitments, examine the witnesses, and follow the tangible facts wherever they lead. Surely visible readings and datable materials provide neutral ground upon which believers, heretics, and unbelievers may reason together.

The appeal contains an important truth. Manuscripts are real objects. A leaf either contains a word or it does not. A scribe either corrected a line or he did not. Evidence must be reported honestly, and no doctrine of preservation authorizes a Christian to invent a reading, conceal a variant, alter a date, or dismiss an inconvenient witness. A confessional critic ought to be more scrupulous about facts, not less.

But manuscripts do not speak for themselves. Human beings describe, classify, relate, and infer from them. A manuscript can display a reading; it cannot assign that reading historical weight. It can be older than another copy; it cannot announce that its wording is therefore more ancient. It can agree with five witnesses and disagree with five hundred; it cannot determine their dependence or identify which group preserves the apostolic text.

The slogan confuses the existence of objective evidence with the existence of self-interpreting evidence. That confusion is one of the academy game’s most effective concealments. It allows a network of historical assumptions to appear under the innocent name the facts.

This article will argue that there are no brute manuscripts. Textual facts are objective because God created and governs the world in which they exist. Yet their significance is never autonomous, and textual judgments never arise from data alone. Date, number, genealogy, internal probability, authorial style, and conjecture all operate within an account of what Scripture is and what kind of transmission history may be expected. The confessional critic must use evidence fully while refusing to let an unconfessed philosophy masquerade as evidence.

What Is A “Brute” Manuscript?

In this discussion, brute does not mean physical, obvious, or intellectually crude. A brute fact is an allegedly uninterpreted fact: a datum whose meaning and evidential force are thought to be independent of God, revelation, and any governing view of reality. A brute manuscript would be a witness that enters the critic’s study without covenantal identity and yields conclusions without the direction of prior commitments.

No serious critic explicitly describes a manuscript that way. The brute-fact ideal usually appears rhetorically: one side is said to possess evidence while the other possesses theology; one follows manuscripts while the other follows presuppositions.

That contrast is false. The choice is never between evidence and presuppositions. It is between evidence interpreted under one set of presuppositions and evidence interpreted under another.

This claim does not make facts subjective. A faded line does not change because the examiner changes his worldview. Nor does it mean that every interpretation is equally plausible. Some collations are accurate and others are careless. Some inferences explain more of the evidence with fewer difficulties. Some claims can be falsified by looking at the page.

Christian epistemology does not dissolve the external world into the knower’s mind. The world is real because God made it, and human knowledge is possible because man bears God’s image in an ordered creation. Denying brute facts therefore secures objectivity: facts are what they are within God’s comprehensive knowledge and providence.

The narrower point for textual criticism is simple: a visible mark is one thing, and a judgment about its relation to the authentic text is another.

Four Levels Of A Textual Claim

Much confusion can be removed by distinguishing four levels often compressed into the phrase the manuscript evidence.

First, there is observation. Manuscript X reads A at a particular place. A corrector erased a word. The page has lost its outer margin. These claims are closely answerable to the artifact.

Second, there is classification. The hand belongs approximately to a certain century. The correction is contemporary or later. The witness has a particular relation to a group of witnesses. Classification draws upon comparative knowledge and may carry a range of uncertainty.

Third, there is historical inference. Reading A is earlier than reading B. One reading arose by omission, harmonization, expansion, or grammatical correction. A surviving witness may preserve a textual state older than the physical copy that carries it. These claims attempt to explain a history no scholar directly observed.

Fourth, there is textual judgment. Reading A belongs to the authentic text of Scripture and reading B is a corruption. This conclusion uses observations, classifications, and inferences, but it is not identical with any of them.¹

The distinctions are not walls. Observation requires competence; classification affects inference; inference may cause reconsideration. Nevertheless, separating the levels exposes a common move: a disputable historical judgment is described as though it were simply the visible contents of a codex.

The confessional critic should repeatedly ask: What was seen? What was inferred? What principle connected the two? How certain is each claim? That discipline protects both evidence and honesty.

A Date Is Not A Verdict

An early manuscript is valuable evidence. It places a reading in existence no later than the production or correction of that witness. Other things being equal, a copy closer in time to the apostolic writing has passed through fewer copying generations than a much later copy.

But other things are rarely equal, and the date of a manuscript is not the date of every reading within it.

A fourth-century copy may reproduce a second-century exemplar. A tenth-century copy may reproduce a much older textual line. An early scribe may introduce an error at once; a late scribe may faithfully transmit an ancient reading. The age of the surviving material establishes a terminus for the witness, not a complete genealogy for its text.

Manuscript survival is historically uneven. Climate, conflict, use, storage, and collection affect what remains. An early copy from a dry region may survive while heavily used copies elsewhere wear out and are replaced.

None of this makes age irrelevant. It makes age one feature within a transmission model. “The oldest manuscripts read B” may be an accurate observation. “Therefore B is original” is an inference requiring additional premises. When those premises are left unstated, a philosophy of transmission has been hidden inside a date.

A Count Is Not A Verdict

The same qualification applies to number.

Numerical preponderance is a real fact. If hundreds of manuscripts contain one reading and two contain another, the distribution demands explanation. Wide ecclesiastical use, geographical extension, and continuity through time can be providentially significant. A confessional doctrine of preservation has good reason to attend to the church’s public possession of the text.

Yet a raw total is not a plebiscite in which every manuscript casts an independent vote. Five hundred copies may descend from a narrow stream or multiply under a standard edition. Conversely, agreement among a few early manuscripts may reflect dependence or local corruption rather than independent antiquity.

“The majority reads A” therefore does not automatically establish A. “The earliest witnesses read B” does not automatically establish B. Both claims become evidentially useful only within judgments about dependence, distribution, historical continuity, and scribal causation.

Confessional Bibliology should not be reduced to counting manuscripts. God promised to preserve his Word, not to make every authentic reading win a simple numerical election in every surviving inventory. At the same time, modern criticism may not treat broad and durable ecclesiastical reception as though it were meaningless debris left by a late and uniformly corrupting church. Number is neither sovereign nor nothing.

Genealogy Does Not Remove Judgment

Modern genealogical methods rightly attempt to move beyond crude appeals to age or number. They compare textual states, measure patterns of agreement, identify possible relationships, and ask which readings are prior and which are posterior. This is an important advance over treating manuscripts as isolated votes.

The strongest form of the method should be stated accurately. The Coherence-Based Genealogical Method does not pretend that a computer mechanically discovers the original text. Gerd Mink’s official description distinguishes observable agreement from genealogical direction and says explicitly that the method does not automatically produce textual decisions. Local stemmata require the full resources of textual criticism; the resulting relationships are then tested and revised iteratively.

That admission proves the present point. Computation can reveal patterns the human eye would miss. It can calculate agreement after readings have been transcribed and variant units defined. It cannot decide by arithmetic alone which reading gave rise to another. Direction must be assigned through textual judgment.²

The procedure is not discredited by this dependence. All historical inquiry joins data and interpretation. The error comes when the resulting graph is presented as though the data generated their own history. A genealogy is a disciplined hypothesis about relationships among surviving textual states. It is not a photograph of the entire providential transmission of Scripture.

Large portions of that history are no longer materially represented. Exemplars perished, scribes used multiple sources, and the same reading could arise independently. A descendant may preserve an older reading at one place and a newer one elsewhere. A model can organize these complexities without becoming an autonomous court over God’s testimony concerning his Word.

Internal Canons Are Compressed Histories

When external evidence does not decide a variant, critics ask which reading best explains the origin of the others. This is necessary. It is also interpretive.

Consider the familiar preference for the shorter reading. Scribes did sometimes expand texts. They incorporated explanatory words, conformed a passage to a parallel, or brought material from a margin into the body. A shorter reading may therefore be prior.

But scribes also omitted through similar endings, repeated letters, difficult syntax, or damaged exemplars. “Prefer the shorter reading” is not a fact seen in the reading; it is a compressed theory about which scribal action is more probable there.

The preference for the harder reading works the same way. A scribe may smooth grammar, remove an apparent contradiction, replace an unusual word, or clarify theology. The more difficult reading may explain the easier.

Yet nonsense is not automatically apostolic. Scribes created difficulties by misreading, transposing, omitting, and joining badly. A modern critic may also misunderstand an idiom. Difficulty must be explained, not merely counted.

The charge of harmonization likewise requires a historical argument. A reading closer to a Synoptic parallel may be secondary because a copyist remembered the parallel. It may also be authentic because the Evangelists sometimes used the same words. A longer christological title may be an expansion, or a shorter title may be an omission. The shape of the variant does not carry a label identifying the scribe’s action.

These canons are not useless. Used as defeasible questions rather than inviolable laws, they help reconstruct plausible histories. The academy game begins when a probability generated by a model of scribal behavior is redescribed as the neutral demand of evidence.

Authorial Style Can Become A Circle

Critics properly compare a disputed reading with an author’s vocabulary, syntax, and habits. Paul does not write exactly like John, and Luke does not write exactly like Mark. A reading foreign to the surrounding author deserves scrutiny.

The method can nevertheless become circular. The critic first constructs an author’s style from the text he regards as original, then uses that style to exclude readings from the original text. If disputed readings helped define the pattern, the conclusion may repeat the initial selection in statistical form.

Large bodies of relatively undisputed material can control stylistic analysis, while genre, quotation, amanuensis, subject, and circumstance may explain variation. The point remains that “this is not Pauline” is an inference from reconstructed Pauline usage, not a sound emitted by the manuscript.

The critic must disclose how the control text was selected, how much variation the model permits, and whether the proposed reading is actually impossible or merely uncommon. Transparency turns a hidden presupposition into an examinable argument.

Conjecture Reveals The Boundary Of A Method

The treatment of conjectural emendation makes the governing worldview especially clear.

For an ordinary ancient work, every surviving witness may preserve corruption at the same place. An editor may conclude that the author’s words have vanished from the entire extant tradition and propose a reading found in no manuscript. The conjecture may be cautious, ingenious, and historically responsible. Nothing in the nature of an ordinary book guarantees that its wording remains attested.³

Many New Testament critics are highly reluctant to print conjectures; modern editions are not casually filled with inventions. The question is one of principle: may the inspired reading have disappeared from the whole preserved Hebrew or Greek inheritance, leaving scholarship to recreate what God no longer supplies?

The confessional answer is no. Westminster confesses that the immediately inspired Hebrew and Greek Scriptures were, by God’s singular care and providence, kept pure in all ages and are therefore authentic for the church’s final appeal. This does not mean that every copy is perfect or that the critic possesses an effortless rule for every variant. It means that the authentic text is preserved in the extant textual inheritance and is to be recognized ministerially, not recreated beyond all witness.

Here theology does constrain method. It excludes universal loss as an unrestricted explanation because God has revealed a providential fact about the object. That is no more illegitimate than excluding a proposed history because it contradicts a securely known date. The real dispute concerns which truths are permitted to count as knowledge.

The Same Facts, Different Weights

Imagine a variant with two readings.

Reading A appears in a large number of later Greek witnesses, several lectionaries, an early version, and repeated ecclesiastical use. Reading B appears in two substantially earlier Greek witnesses, is shorter, and produces a more abrupt sentence. Both early witnesses are related, but the precise degree of dependence is disputed. A patristic citation could support either reading because it is paraphrastic.

Every item in that description may be publicly examined. Yet the list does not calculate its own conclusion.

One critic may give great weight to the age of B, judge its abruptness likely to have been smoothed, and explain A as an expansion that later became dominant. Another may judge the two early witnesses insufficiently independent, explain B as an omission, and regard the version and broad ecclesiastical use of A as evidence of earlier and wider existence. A third may regard the evidence as nearly balanced.

Confessional doctrine does not mechanically select A. A confessional critic may conclude that B is authentic if the complete evidence warrants it. Nor does a restorationist framework mechanically select B. The example shows something more basic: the conclusion emerges from facts plus judgments concerning transmission, scribal behavior, genealogy, providence, and the proper object of recovery.

The honest debate therefore asks not merely, “What evidence do you have?” but also, “By what rule are you assigning it that weight?”

Even The Goal Requires A Doctrine

No method begins until it knows what it is trying to identify.

Older formulations often spoke of recovering the original text or the autographic wording. Some current projects speak more cautiously of the initial text: the form from which the extant tradition can be traced, without necessarily claiming direct identity with a physical autograph in every respect. That terminological shift reflects genuine historical modesty. Surviving data may permit a critic to describe one horizon more confidently than another.

Yet “initial text” is not a neutral object waiting inside the database. It is a goal defined by a theory of textual history. Does the reconstructed starting point represent apostolic wording, an early editorial form, the archetype of surviving witnesses, or merely the earliest recoverable state? Could revelation stand behind that point in a form now lost? The answers determine what counts as success.

The confessional critic seeks the authentic canonical text given by God and kept in the Hebrew and Greek Scriptures. He may distinguish the inspired verbal content from the perished physical autographs. He may acknowledge uneven evidence and fallible local judgments. He may not redefine the object so that the church possesses only a scholarly approximation to a revelation located beyond every extant witness.

The goal of criticism is already bibliology in operational form.

Sin And The Textual Knower

The deepest reason neutrality fails is not that human beings lack intelligence. It is that every knower stands in covenant with God.

Romans 1 does not portray fallen man as empty of knowledge. He knows God through what has been made and suppresses the truth in unrighteousness. Suppression is not always a crude refusal to look. It can take the sophisticated form of rearranging known facts within an interpretation that will not acknowledge God.

Greg Bahnsen’s analysis of self-deception is useful here. He describes motivated rationalization as a handling of adverse evidence that can remain active, elaborate, and apparently reasonable. The self-deceiver does not necessarily ignore data; he may manage, reweight, and reinterpret it so that an unwelcome conclusion never receives its due.⁴

This truth must be applied with moral discipline.

It does not follow that every non-confessional decision consciously attacks Scripture or that disagreement proves self-deception. Finitude, inadequate data, training, convention, and ordinary error affect scholarship. A Christian may adopt an autonomous principle inconsistently; an unbeliever may collate with exceptional accuracy. A formal heretic, an unbeliever, and an erring brother should not be collapsed into one personal judgment.

We are ordinarily competent to examine public arguments, not hidden motives. The charge of playing the academy game should therefore identify the published principle and show its incompatibility with Scripture. It should not pretend to read a scholar’s heart.

Nevertheless, charitable restraint must not erase antithesis. A method that excludes God’s testimony from the interpretation of God’s Word is not innocent because its practitioners are courteous or technically gifted. The sin lies objectively in claiming autonomy for human reason. Believers can participate in that sin inconsistently, and unbelievers can display genuine knowledge while suppressing its ultimate foundation.

Does Theology Predetermine The Result?

The standard objection now returns: if preservation excludes certain conclusions in advance, has the confessional critic rigged the inquiry?

Every rational inquiry excludes some conclusions. A historian who knows that an artifact was produced in the tenth century will not call it apostolic. A scientist will not accept a contradiction as a final explanation. A critic who permits universal textual loss has also set a boundary, though a permissive one.

The relevant question is whether the boundary is warranted.

Confessional theology does not dictate the reading at every variant. It does not tell the critic that the longest, most numerous, latest, earliest, or most familiar reading must always win. It does not make collation unnecessary. It supplies revealed truths about the identity, authority, and providential history of the object. Within those truths, extensive ministerial inquiry remains necessary.

Indeed, presuppositions can make inquiry possible without making it pointless. Belief that creation is orderly does not predetermine every scientific measurement. Belief that an author wrote meaningful language does not predetermine the interpretation of every sentence. Belief that God preserved his Word does not predetermine every local textual judgment. It tells the critic what kind of reality his local judgment inhabits.

A Confessional Discipline Of Evidence

A confessional textual method should display at least seven habits.

First, state observations with precision. Name the witnesses, readings, corrections, dates, and uncertainties. Do not substitute “all the evidence” for an actual account of the evidence.

Second, separate observation from inference. Say when a date is approximate, when dependence is hypothesized, and when a proposed scribal cause is only probable.

Third, disclose the governing canon. If the shorter, harder, more widespread, more coherent, or more familiar reading is preferred, explain why that consideration bears weight at this place.

Fourth, use all lawful evidence. Manuscripts, versions, citations, lectionaries, grammar, context, authorial usage, canonical analogy, transmission history, and ecclesiastical reception each have a ministerial place. No single mechanical proxy should replace judgment.

Fifth, reason within revealed preservation. Expect the authentic reading within the preserved Hebrew and Greek inheritance. Do not turn a difficult apparatus into a doctrine of universal loss.

Sixth, distinguish local uncertainty from global skepticism. A critic may remain undecided between readings without concluding that the church lacks the Word of God. The fallibility of one judgment does not transfer to the rule being judged.

Seventh, remain publicly correctable. Confessional commitment is not immunity from evidence. If a collation is wrong, correct it. If a genealogy is misunderstood, revise the argument. If an inference overreaches, narrow it. Submission to Scripture includes repentance from false witness about manuscripts.

These habits do not make textual criticism easy. They make it ministerial.

Conclusion

Manuscripts are witnesses. They must be heard carefully and compared justly, but they do not appoint themselves judge.

A date does not interpret its own significance. A count does not establish independence. A genealogy does not supply its own direction. A shorter or harder reading does not narrate the scribal act that produced its rival. An author’s style does not define itself apart from a text. Even the object called “original” or “initial” is identified within an account of textual history.

There are therefore no brute manuscripts. There are real manuscripts in God’s world, bearing readings through God’s providence, examined by morally responsible knowers, and interpreted according to governing commitments.

The academy game asks the Christian to hide those commitments while accepting the academy’s own as the natural meaning of evidence. He should refuse. He should enter public argument with precision, state facts and principles openly, and expose the point at which observation becomes inference.

The final court is not the oldest codex, the largest count, the most elegant stemma, the most prestigious committee, or the private conviction of the confessionalist. The final court is the Holy Ghost speaking in Scripture. Every manuscript, method, and ministerial judgment has its proper dignity only beneath that authority.

Let the manuscripts testify. Let the critic reason. Let neither be called autonomous.

Notes and Further Reading