“Textual Criticism Has No Explicit Biblical Guidelines”: The Admission That Controls the System

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.

Lane Keister does not deny inspiration or preservation. He openly brings theological considerations into textual criticism. Yet his own account places the decisive work in probabilistic canons that may conflict, have no revealed order of priority, and must finally be weighed by the critic. The question is therefore not whether judgment is unavoidable. It is what office that judgment occupies.

Lane Keister’s 2017 essay on the canons of New Testament textual criticism contains one sentence more important than any of its eight canons. He says the discipline lacks express biblical directives. He therefore presents the rules by which he works, acknowledges that other textual critics will arrange the evidence differently, and does not claim that every Christian ought to adopt his formulation.

Keister then adds three qualifications. Every canon has “probabilistic force only.” The canons can point in different directions and must be weighed against one another. And because the discipline is both science and art, the judgment of the critic cannot be eliminated: “It is unavoidable.”

This is commendably candid. Keister does not conceal the role of critical judgment behind an air of mechanical certainty. He knows that no algorithm simply receives a list of manuscript data and prints the apostolic text. Dates must be estimated, genealogical relations inferred, patristic citations interpreted, transcriptional probabilities imagined, and competing considerations assigned relative weight. Anyone who has examined an apparatus knows that this is true.

But Keister’s admission also exposes the controlling problem in his preservation model. If Scripture does not regulate textual criticism explicitly, if the proposed canons bind no one, if every canon is merely probabilistic, if the canons conflict, and if the critic must finally weigh them, then what governs the passage from preserved manuscripts to the text of Scripture?

The immediate answer is: the trained judgment of the critic.

That does not prove that the judgment is careless, unbelieving, or insincere. It does mean that the method’s final discriminating act is not supplied by manuscript facts alone. Nor, on Keister’s account, is it supplied by an explicit biblical hierarchy of textual criteria. The text printed above the apparatus at a disputed place is the conclusion of an expert judgment made among probabilities.

The problem is not that Keister admits judgment. The problem is what office that judgment occupies.

Keister Must Be Stated at Full Strength

A fair critique must begin with what Keister actually believes.

He believes that God inspired definite words. He believes that God preserved the Old and New Testaments by special providence. In his 2024 answer to Chris Thomas, he emphatically located that preservation “through the manuscripts.” He does not believe that the Bible was lost and resurrected by modern scholarship. He does not believe that a critic confers divine authority upon a reading. He regards both the King James Version and the English Standard Version as the Word of God, and he rejects alarmist claims that textual variants have destroyed Christian doctrine.

Nor does Keister profess a religiously neutral method. One of his canons explicitly limits the preference for a more difficult reading by the character of God: because God cannot lie, a reading that truly contradicts Scripture cannot be original. His appeal to providence is also thoroughly theological. He describes textual criticism as “reading in the book of God’s providence.”

These facts prohibit several easy but irresponsible accusations. Keister is not an unbelieving rationalist. He is not trying to dethrone Scripture. He does not deny preservation. He does not pretend that the critic is neutral. And nothing in his essay gives us warrant to diagnose his motives.

The criticism here concerns the architecture of his position. A scholar may sincerely intend to serve Scripture while employing a method that gives scholarly judgment a constitutive function at disputed places. Humility of intention does not settle hierarchy of function.

Keister’s account may be summarized as a sequence:

No explicit biblical guidelines → nonbinding critical canons → probabilistic and sometimes conflicting results → necessary critical weighing → a reading selected as most likely original.

The theological question is whether this sequence is adequate to Westminster’s confession of an inspired text “kept pure in all ages” and therefore presently authentic.

Scripture Need Not Contain a Textbook of Paleography

The weakest response to Keister would insist that every technical rule of textual criticism must be stated verbatim in a Bible verse. Scripture does not tell us how to date handwriting, identify the same scribe across several leaves, detect a correction under ultraviolet light, or distinguish a citation from an allusion in a church father. It does not provide a manual of codicology, typography, lexicography, or historical research.

No confessional bibliology worthy of the name should object to such auxiliary sciences merely because their procedures are not individually expressed in Scripture.

The Westminster Confession of Faith itself prevents that mistake. Chapter 1.6 says that the whole counsel of God is either expressly set down in Scripture or may be deduced from it by “good and necessary consequence.” It also recognizes that some circumstances are ordered by Christian prudence according to the general rules of the Word. Reformed theology has never limited biblical government to propositions stated in precisely the form later controversies require.

The question, therefore, is not whether there is a verse saying, “Prefer geographically diverse attestation,” or, “A reading that explains the rise of its competitors deserves special consideration.” The question is whether the entire critical enterprise—its goal, starting point, permissible inferences, authority structure, and relation to the church’s received text—is governed by Scripture’s doctrine of Scripture.

There is a categorical difference between using noninspired tools to study an acknowledged text and using nonrevealed norms to determine which words constitute the acknowledged text. Grammar helps us interpret Scripture. Photography helps us read a damaged manuscript. Historical inquiry helps us understand a scribe or printer. None of these tools is thereby made the supreme rule of faith.¹

But when a critical method decides whether a clause belongs to Scripture, its conclusion identifies the verbal content of the supreme rule. At that point, “the Bible does not explicitly tell us how” cannot close the theological inquiry. The method need not be dictated in technical detail, but it must remain bounded by what Scripture teaches about its own inspiration, preservation, availability, reception, and authority.

No method is disqualified merely because Scripture does not state its technical rules verbatim. But no method that identifies Scripture’s words may be insulated from Scripture’s doctrine of how those words are given, preserved, received, and appealed to.

Eight Canons—and No Revealed Order of Priority

Keister proposes eight canons. They deserve to be treated as serious historical judgments rather than dismissed as arbitrary preferences.

Keister’s canon: Older witnesses tend to preserve older readings

Legitimate ministerial use: Age can supply relevant historical evidence

The unresolved question: How much weight does estimated age receive when later witnesses are more numerous or continuous?

Keister’s canon: Geographically diverse attestation strengthens a reading

Legitimate ministerial use: Wide circulation may indicate antiquity

The unresolved question: How are geography and genealogical dependence reconstructed and weighted?

Keister’s canon: Genealogically related witnesses should not be counted as independent

Legitimate ministerial use: Raw manuscript totals can mislead

The unresolved question: How certain is the proposed genealogy, and when does family coherence outweigh numerical or ecclesiastical continuity?

Keister’s canon: The more difficult reading is often earlier

Legitimate ministerial use: Scribes sometimes smooth or clarify

The unresolved question: Which difficulty is genuinely transcriptional, and when does this preference reward accidental or corrupt readings?

Keister’s canon: The reading that best explains the others has a stronger claim

Legitimate ministerial use: Competing readings should be related historically

The unresolved question: Explanatory superiority depends upon a reconstruction of scribal motives and sequence

Keister’s canon: Continuous attestation strengthens a reading

Legitimate ministerial use: Historical continuity matters to a doctrine of preservation

The unresolved question: Can an unattested or long-hidden reading overcome continuity, and by what theological rule?

Keister’s canon: The majority reading has a stronger claim, other things equal

Legitimate ministerial use: Numerical prevalence is evidence, not an embarrassment

The unresolved question: What counts as “other things equal,” and who determines when they are?

Keister’s canon: Early patristic testimony may carry great weight

Legitimate ministerial use: Fathers can localize readings early

The unresolved question: Citations are difficult to date, identify, and distinguish from paraphrase or harmonization

None of these considerations is inherently illegitimate. Several are indispensable. A ministerial critic should ask how old a witness is, whether witnesses are independent, whether a scribe could have produced one reading from another, whether a reading was broadly and continuously used, and what the early church cited.

The structural difficulty appears when the canons conflict. Suppose the older extant manuscripts support one reading, the numerical majority another, continuous ecclesiastical use another, and internal difficulty still another. Keister does not offer a revealed order by which the conflict must be resolved. Indeed, he disclaims universal obligation for his rules and emphasizes their probabilistic character.

At that point, the canons do not determine the reading. The critic determines how the canons apply, which evidence is reliable, and how much each consideration counts. Two critics may use the same eight canons in good faith and reach different results because their judgments about genealogy, scribal behavior, continuity, and relative weight differ.

The canons are not raw facts. They are weighted interpretations of facts.

A manuscript’s date is a historical conclusion. Its relation to another manuscript is a genealogical conclusion. The likelihood that a scribe simplified a sentence is a psychological and transcriptional conclusion. The significance of geographic spread is a historical conclusion. Even the classification of a reading as “difficult” is interpretive. The physical witness is evidence; the critical value assigned to it is an inference.

This does not make textual criticism impossible. It makes its theological subordination indispensable.

Probability Can Serve Certainty; It Cannot Ground Authority

Probability belongs to responsible human reasoning. A damaged leaf may leave us less certain about one letter than another. A patristic citation may probably support a reading without reproducing it exactly. An editor may make a fallible judgment while the text judged remains objectively definite. Confessional bibliology does not require omniscience in the critic or equal psychological certainty about every disputed letter.

The decisive distinction is between the fallibility of our recognition and the formal status of the object recognized.

A court may possess an enacted statute even while lawyers disagree about one damaged archival copy. A family may possess an authentic letter even while two readers dispute a faded word. The dispute does not transform the statute or letter into the cumulative product of their reconstructions. Their judgment remains answerable to an antecedently existing and possessed object.

In the ecclesiastical case, Westminster 1.8 does more than affirm that the autographic words probably survive somewhere in the evidence. It confesses that the Hebrew and Greek Scriptures were kept pure and are “therefore authentic,” so that the church’s final appeal is to them. Preservation issues in present judicial availability.

Probability may assist the church in correcting a defective copy or assessing an isolated claim. It may serve certainty by removing a local obscurity. But probability cannot become the ground of Scripture’s authority. God’s Word is not authoritative because a critic has assigned a reading the highest current likelihood. The critic’s conclusion is authoritative only insofar as it faithfully recognizes the Word already given and preserved.

The modern eclectic model often affirms that proposition in theory. Its practice, however, makes the current identification of the Word at each contested locus depend upon the result of an ongoing probabilistic adjudication. The text is definite in the absent autograph; the evidence is available in the manuscripts; but the church’s identification of the autograph is locally mediated by revisable expert judgment.

That is sufficient for a doctrine of recoverability. It is not yet the same thing as a doctrine of a formal, identifiable, publicly possessed apographic text.

Judgment Is Unavoidable—but It Need Not Be Sovereign

Keister is correct that human judgment is unavoidable. The Received Text position does not escape judgment. Editors compared manuscripts. Printers corrected proofs. Translators chose English equivalents. Ministers interpret syntax. Readers decide between competing claims. Appeals to ecclesiastical reception also require historical and theological reasoning.

The alternative to critical judgment is not no judgment. The alternative is ministerial judgment under an antecedently received rule.

That distinction concerns office rather than psychology.

Ministerial textual criticism receives the inspired and preserved text as the church’s present standard. It compares individual witnesses, recognizes copying faults, corrects printing errors, and investigates difficult places. Its judgments are real and fallible, but they are made within a prior commitment to the identity, continuity, and public availability of the text.²

Magisterial textual criticism need not announce, “The critic is above God.” It becomes functionally magisterial whenever the identity of Scripture at a disputed place is constituted for the church by an open-ended critical decision among all extant alternatives, without an antecedently identifiable apographic standard capable of judging that decision.

The difference can be compressed into one sentence:

Ministerial criticism judges copies by the text; magisterial criticism makes the text the cumulative output of judgments about copies.

Keister intends the first relation. Yet his methodological sequence tends toward the second. His preserved object is the manuscript aggregate, which contains original readings and secondary competitors. His canons discriminate among them. Because the canons are probabilistic and may conflict, the critic’s judgment supplies the final discrimination. At the contested point, the church receives as text the reading that the critical process presently ranks first.

The critic does not create divine authority. But his judgment becomes constitutive for identifying which wording bears that authority.

Calling the process ministerial does not settle the matter. A servant may acknowledge his master while being assigned a task that effectively determines which claimant will be recognized as the master. The issue is not the servant’s declared loyalty. It is whether his office is corrective under a known text or constitutive of the text’s current identity.

Can Providence Supply the Missing Rule?

Keister attempts to place the discipline within theology by describing it as an exercise in reading providence. That instinct is right. Manuscripts did not survive outside God’s government. Their copying, circulation, loss, discovery, collation, and use all occurred within his sovereign decree.

Yet providence cannot function as a substitute textual canon.

Providence explains why the evidence exists. It does not tell us, simply from the fact of survival, which surviving reading is original. God providentially governed faithful copying and scribal error, public use and private neglect, preservation and destruction. A secondary reading survives by providence no less than an original one does. Survival proves divine government, not divine endorsement.

A manuscript’s survival is a providential fact. The originality of its reading is an inference.

Keister himself must agree, because he does not accept every surviving reading. He weighs manuscripts against one another. But once survival is distinguished from authenticity, “reading providence” requires a norm by which the providential evidence is interpreted. Historical events do not interpret themselves.

This is especially important when continuity and hiddenness compete. Keister’s sixth canon gives weight to continuous attestation because he does not believe God allowed his Word simply to vanish for centuries. Yet he also allows that providence may preserve a reading in an obscure or even unique witness. Both propositions are possible at the level of divine power. The theological question is which account accords with God’s revealed promise and the church’s confessed possession of Scripture.

An appeal to hidden providence can always rescue a discontinuous reading from the evidential force of reception: perhaps God preserved it quietly until its recovery. But if that possibility is allowed whenever other canons favor the reading, continuity ceases to regulate the method. It becomes one probability among others, affirmed when supportive and overridden when inconvenient.

Providence must be interpreted by revelation, not revelation reconstructed by an unregulated reading of providence.

Bahnsen and the Myth of the Detached Critic

Greg Bahnsen’s dissertation, The Apologetic Implications of Self-Deception, is not a work on New Testament manuscripts, and it should not be conscripted as though it supplied a textual canon. Its relevance is more modest. Bahnsen’s analysis attacks an idealized picture of human belief as the product of detached rational machinery. Human reasoning is exercised by whole persons: finite, morally situated, capable of inconsistency, and affected by prior commitments.

That observation must not be turned into a charge that Keister is self-deceived. We possess no warrant for such a judgment. Nor does it mean that expertise is worthless. It means that the critic’s imagination and judgment cannot authenticate themselves merely by being scholarly.

Keister already rejects neutrality in one important sense. He excludes readings that contradict the truthfulness of God. But the theological reach of presuppositional analysis extends further. Every critical system assumes answers to prior questions:

• What is the object being sought: an initial text, an autograph, or a received canonical text?

• What has preservation promised: survival of readings, recoverability of wording, or public possession of an authentic text?

• What role does the church’s historical reception play?

• May a reading absent from ordinary ecclesiastical use for many centuries displace one continuously received?

• Is the text the standard by which witnesses are corrected, or is the text the result produced by weighting the witnesses?

• What degree of revision is compatible with the confession that the Scriptures were kept pure in every age?

None of those questions is answered by carbon dating, manuscript counts, or transcriptional imagination alone. They are theological and epistemological. The critic approaches the evidence with a doctrine of what the evidence is evidence of.

This is the proper Bahnsenian point for the present controversy: there is no uninterpreted field of manuscript facts from which a religiously innocent method rises. Evidence is assessed within ultimate commitments. Consequently, the Christian critic must do more than append providence to an otherwise self-governing critical procedure. The procedure itself must be brought under Scripture.

That claim applies to Received Text advocates as well. Ecclesiastical reception cannot become a magical phrase immune from historical examination. Numerical majority cannot become an autonomous rule. Theological confidence cannot excuse factual carelessness. Every side must distinguish fact, inference, and motive; state opponents in their strongest form; and submit its reasoning to the Word it claims to serve.³

Presuppositional critique is not a license for suspicion. It is a demand for disclosed authority.

What Biblical Regulation Would Actually Require

If Scripture does not give a technical handbook of textual criticism, what would it mean for criticism to be biblically governed? At minimum, a confessional method would begin with the following commitments.

First, inspiration produced a determinate verbal text. The object is not a range of equally valid alternatives or a merely doctrinal core. God gave words.

Second, preservation concerns those inspired words. It cannot be reduced to the survival of Christian teaching, the existence of enough evidence for later recovery, or the presence of every original reading somewhere in an undifferentiated aggregate.

Third, preservation issues in availability. Scripture is given to the church as public rule, read in worship, preached, translated, confessed, and appealed to. A text theoretically recoverable from materials the church cannot yet identify is not functioning as that rule.

Fourth, the autographs continue textually in faithful apographs. Particular copies may contain errors without inspiration being transferred to scribes. Yet those errors are recognized as departures from a continuing text, not as coequal candidates waiting indefinitely for critical constitution.

Fifth, ecclesiastical reception is ministerial, not creative. The church does not inspire the canon or manufacture its wording. It receives what God gives. But that reception is a providential means of public preservation and cannot be treated as irrelevant whenever a modern reconstruction prefers another reading.

Sixth, manuscript evidence serves the preserved text. Age, genealogy, geography, transcriptional probability, versions, and fathers all have legitimate uses. They may expose an error in a particular witness or clarify the history of a reading. They do not collectively become an autonomous tribunal over the identity of Scripture.

Seventh, continuity must carry genuine theological weight. It cannot guarantee the perfection of every popular reading, but neither can it be reduced to one defeasible probability among criteria developed without reference to the promise of preservation.

Eighth, critical conclusions remain accountable to a public standard. The church must be able to ask not only whether an editor has used evidence intelligently, but whether his result preserves the identity of the text historically received and confessed as authentic.

These commitments do not remove hard cases. They locate them. They do not make the editor infallible. They define his ministry. They do not abolish probability. They prevent probability from becoming the formal bridge between a lost perfect text and the church’s current Bible.

The Question Keister’s Admission Leaves Unanswered

Keister deserves credit for refusing to overstate his canons. But that restraint sharpens the central question.

If his rules are not expressly revealed, if he does not claim that every Christian must adopt them, if each carries only probability, if they can conflict, and if the critic must supply their relative weight, on what authority may their result displace the reading historically received and publicly used by the church?

It is not enough to answer, “On the authority of the evidence.” Evidence does not weigh itself. Nor is it enough to answer, “By providence.” Providence brought every competing witness into the critic’s hands. Nor can the answer be, “By the authority of Scripture,” unless Scripture’s own doctrine of preservation governs the method by which its wording is identified.

The strongest possible answer is that the critic’s best judgment is a fallible ministry to the infallible autograph. That answer protects the critic from claiming inspiration. But it leaves the church with a formal structure in which the autograph is perfect and absent, the manuscript evidence is abundant and mixed, the canons are nonbinding and probabilistic, and the printed text at disputed places is the revisable result of expert adjudication.

That is a coherent scholarly model. It is not obviously the model confessed by the words kept pure in all ages, therefore authentic.

Keister affirms preservation in name and intention. His methodological admission shows why restoration controls the practice. The manuscript tradition preserves the materials; the canons organize probabilities; the critic supplies the weighing; and the edition presents the result. Scripture’s authority is verbally affirmed throughout, but the church’s access to Scripture’s exact wording at the disputed locus depends upon a judgment for which Keister claims neither explicit biblical prescription nor universal Christian obligation.

The admission therefore controls the system. Once the rules are probabilistic all the way down, the unavoidable question is not whether the critic will judge. It is whether his judgment remains a servant correcting witnesses beneath a publicly possessed text—or becomes the final court that tells the church which wording may count as that text.

Notes and Further Reading