From “Erasmus’s Handful” to the Church’s Text: What the Small-Base Argument Leaves Out

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.

Erasmus had only a small number of Greek manuscripts immediately available when preparing the first published Greek New Testament. That fact is sufficient to refute an infallible-Erasmus theory. It is not sufficient to show that Erasmus created the textual tradition represented by those manuscripts, that the Received Text is identical to his hurried 1516 edition, or that a modern critical reconstruction follows merely because more manuscripts are now catalogued. A small editorial base is not the same thing as a small transmissional base.

The historical objection is familiar because it is rhetorically powerful.

Erasmus arrived in Basel. He possessed only a handful of late Greek manuscripts. His work proceeded rapidly. His copy of Revelation was defective. The first edition contained typographical and editorial errors. Yet this rushed production, we are told, became the foundation of the Textus Receptus and then acquired an authority wholly disproportionate to its evidence. Modern scholars now possess thousands of manuscripts. Why should the church prefer a sixteenth-century text built from a few late copies to an edition informed by the whole surviving tradition?

Put that way, the verdict seems almost automatic. A few manuscripts must lose to thousands. A hurried Renaissance humanist must lose to international committees. A first-generation printed edition must lose to five centuries of discovery. But the framing conceals several questions beneath one numerical comparison. How many manuscripts did Erasmus use immediately? What tradition did they represent? Is the Received Text identical to his first edition? How did subsequent editions revise it? And do thousands of catalogued artifacts constitute thousands of independent witnesses at every disputed place?

Those questions do not have the same answer.

Lane Keister is right to insist that “Access to any manuscripts was fairly limited” during the production of the Received Text. He is also right that no known pre-print manuscript is verbally identical to a later printed TR edition. See his chapter-by-chapter review of Why I Preach from the Received Text. A preservationist who evades these facts weakens his own case.

Yet the small-base argument proves much less than Keister needs it to prove. It establishes the fallibility and evidential limitations of Erasmus’s first edition. It does not establish that the church lacked the New Testament text before Erasmus, that the tradition embodied in his manuscripts began with those manuscripts, or that the modern critical text is the necessary remedy.

The argument wins against an infallible Erasmus. It does not win against providential preservation. The strongest form of Keister’s objection should be stated without caricature. First, a small sample increases editorial risk. If an editor works from only a few manuscripts, a local peculiarity in one copy can enter the printed text without adequate correction. The danger becomes greater when only one usable witness is available for a book or passage.

Second, late manuscripts do not by themselves give direct documentary access to the earliest centuries of transmission. A twelfth-century codex may preserve an ancient reading, but its date cannot simply be ignored. The greater chronological distance allows more time for copying changes to accumulate. Third, Erasmus’s first edition was not a leisurely critical monument. The Greek text was prepared and printed under significant pressure. The surviving 1516 Novum Instrumentum is a magnificent historical achievement, but it is not free from printing mistakes or questionable editorial decisions. No one should canonize a printer’s deadline.

Fourth, Revelation presents special difficulties. Erasmus had a defective and lacunose Greek exemplar, made use of Latin evidence at points, and produced readings whose precise editorial history remains disputed. Those problems may not be waved away by saying providence. Fifth, later manuscript discoveries are genuine evidence. A doctrine of preservation cannot make a newly available ancient witness irrelevant merely because Erasmus did not see it. A late witness may preserve an early reading; an early witness may preserve a secondary reading. Both require evaluation.

Sixth, the claim that Erasmus’s edition was pure simply because it became influential would beg the question. Influence may arise from providence, printing, politics, commerce, scholarship, ecclesiastical adoption, or some combination of them. The theological claim must interpret the history without replacing historical argument. Every one of these points should be granted. None entails the conclusion that the Received Text is merely Erasmus’s few manuscripts inflated into dogma.¹

What the Objection Proves—and What It Does Not

The small-base argument is useful only if its conclusion is kept proportional to its premises.

The historical fact legitimately establishes: Erasmus’s 1516 edition was fallible.

The historical fact does not establish: The Greek textual tradition began in 1516.

The historical fact legitimately establishes: His immediately available manuscript base was limited.

The historical fact does not establish: The readings in those manuscripts had only a narrow prior transmission.

The historical fact legitimately establishes: Some local decisions deserve reexamination.

The historical fact does not establish: Every Received Text reading is an Erasmian conjecture or peculiarity.

The historical fact legitimately establishes: The first edition contained errors later editions could correct.

The historical fact does not establish: The Received Text is identical to the first edition.

The historical fact legitimately establishes: No single manuscript exactly equals a later TR edition.

The historical fact does not establish: A transmitted textual tradition cannot possess recognizable identity across differing witnesses.

The historical fact legitimately establishes: More evidence can expose an editor’s mistake.

The historical fact does not establish: The edition consulting the greatest number of artifacts must have the original reading.

The historical fact legitimately establishes: Exceptional readings require particular defense.

The historical fact does not establish: Exceptional readings define the character of the whole received textual tradition.

The historical fact legitimately establishes: Providence cannot be invoked as a substitute for evidence.

The historical fact does not establish: Providence is irrelevant to the church’s reception and transmission of Scripture.

This distinction is decisive. The historical facts constrain what a preservationist may claim, but they do not supply Keister’s doctrine of preservation. Moving from “Erasmus had few manuscripts at Basel” to “the original is therefore found in the total surviving aggregate by modern critical adjudication” introduces an additional premise. That premise is not a fact about Erasmus. It is a theory about where the preserved text resides and how it becomes identifiable.

Article 3 examined Keister’s answer: the manuscripts collectively are the apographs in which the autographs are found. Competing readings belong to the preserved aggregate, while textual criticism identifies the original among them. The small-base argument makes that model look inevitable by contrasting Erasmus’s desk with the modern catalog.

But the comparison is badly formed. Erasmus’s desk was an editorial location. The modern catalog is a transmissional inventory. One is a small set of physical copies used at a particular moment; the other is the surviving result of many centuries of copying, discovery, registration, and classification. The proper historical comparison is not one desk against one catalog. It is one textual tradition against an alternative account of the whole transmission.

The Edition Is Not the Tradition

The central error is to confuse the manuscripts used to print a text with the historical ancestry of the text they contain. The manuscripts on Erasmus’s desk were not the ancestors of their own readings. They were descendants. Their scribes did not invent the Gospels, Acts, Epistles, or Revelation in the twelfth or fifteenth century. Each codex stood at the end of a prior chain of exemplars. Those exemplars stood at the end of earlier chains. The physical manuscripts immediately available in Basel were few; the transmissional ancestry represented in them was not therefore few.

This does not prove that every reading in a late manuscript is early. It proves only that the date and number of extant artifacts cannot settle the age and breadth of the text they carry. The date of a codex dates the artifact, not automatically every reading in it. A nineteenth-century printing of Augustine does not make Augustine’s sentences nineteenth-century. A medieval copy of an ancient homily may contain both ancient language and later scribal changes. The copy’s date matters, but it is not the date of every element in its exemplar.

The same principle is elementary in textual criticism. Modern critics routinely accept a reading preserved in a later witness when genealogy, versions, patristic citations, or internal considerations indicate that the reading is earlier than the surviving artifact. They also reject readings in early manuscripts when they appear secondary. Keister’s Sturzian commitments themselves acknowledge that readings associated with the Byzantine tradition can be older than the manuscripts that most visibly preserve them.

Therefore, “late manuscripts” cannot function as a complete argument. It identifies the dates of Erasmus’s usable copies. It does not identify the dates of all their readings. Nor does “few manuscripts” establish “narrow tradition.” If several late copies preserve substantially the same form of text, their agreement may reflect a much broader prior history. It may also reflect shared ancestry and therefore less independence than raw numbers suggest. Which explanation applies must be investigated. But that is precisely the point: the number of codices at Basel does not answer the transmissional question by itself. A small editorial base is not the same thing as a small transmissional base.

Erasmus Before and After Basel

The popular story often begins Erasmus’s New Testament work when the printer needed Greek copy. Modern Erasmian scholarship has made that picture more complicated. Erasmus had worked for years on the New Testament, especially on his Latin revision and annotations. He had studied Greek, examined manuscripts in more than one location, recorded variants, and engaged patristic evidence. His first publication was not the sudden product of a man encountering the Greek New Testament for the first time in a Basel library.²

That does not mean every manuscript Erasmus had ever seen directly controlled the continuous Greek text printed in 1516. Immediate copy for the press and the wider evidentiary world of the editor must be distinguished. Recent studies also distinguish the manuscripts physically marked or used as printer’s copy, other manuscripts consulted in Basel, earlier notes from manuscript study, and evidence considered in later editions. Consequently, exact manuscript counts vary according to what is being counted.

The responsible formulation is modest: Erasmus had a small number of Greek manuscripts immediately available for preparing the continuous Greek text at Basel. His wider engagement with New Testament evidence was not limited to the physical copies serving the press. The essays in Patrick Andrist’s edited volume, Basel 1516: Erasmus’ Edition of the New Testament, examine these stages in detail. Earlier studies such as C. C. Tarelli’s “Erasmus’s Manuscripts of the Gospels” likewise asked how the Basel witnesses were used. The history is too technical to rest upon the talismanic repetition of one number.

More importantly, 1516 was not the final form of Erasmus’s own work. He issued successive editions in 1519, 1522, 1527, and 1535. He corrected errors, altered readings, considered additional evidence, and responded—sometimes well and sometimes poorly—to criticism. The revisions do not make him infallible. They demonstrate that “Erasmus’s first edition” and “the Received Text” are not interchangeable expressions. The 1516 edition is a moment in the history of the Received Text, not the beginning of the New Testament’s textual history and not the end of the printed text’s editorial history.

The Received Text Is Not the 1516 Edition

After Erasmus, other editors and printers continued the process. The Complutensian Greek New Testament, prepared independently in Spain and printed before Erasmus’s edition was published, supplied another early printed witness. Robert Stephanus issued several editions. Theodore Beza produced multiple editions and had access to evidence Erasmus did not possess. The Elzevir editions participated in later standardization. The English translators who produced the King James Bible worked within this printed tradition while comparing editions and readings.

The designation Textus Receptus therefore refers to a succession of closely related printed texts, not to a photographic reproduction of the 1516 volume. Samuel Prideaux Tregelles, although committed to revising the commonly received Greek text on critical principles, narrates this history of editions in his public-domain Account of the Printed Text of the Greek New Testament. His critical stance makes him a useful witness against any romantic fiction that the printed text descended untouched from one Erasmian volume.

This history cuts in two directions.

It prevents the preservationist from making Erasmus the inspired terminus of transmission. The Received Text cannot be defended by treating every 1516 reading as sacrosanct. It also prevents the critic from making every feature of the later Received Text the result of one man’s small Basel library. A later edition may retain an Erasmian reading, correct it, replace it from another source, or stabilize a reading already widespread in the Greek tradition. Each disputed reading has a history. The genealogy of the whole received text cannot be reduced to the genealogy of one printer’s copy. Printing stabilized and propagated a text; it did not create the Greek tradition ex nihilo.

What Public Reception Does—and Does Not Do

The preservationist’s argument becomes strongest when it stops trying to prove too much about Erasmus and asks what happened through the whole process of transmission, printing, correction, translation, and ecclesiastical reception. The early printed editions placed a long-transmitted Greek text into a new medium. Subsequent editions corrected and standardized it. Reformation churches translated from it, preached it, cited it in controversy, used it in theological education, and treated it as Holy Scripture. That history does not make every editor’s decision correct. It does show that the Received Text became something more than Erasmus’s private scholarly proposal.

Reception is evidence of providential recognition, not a sacrament that turns error into inspiration. The church does not create Scripture by accepting it. Neither can ecclesiastical use make a non-apostolic reading apostolic. The causal order matters: a reading is authoritative because God inspired it, and the church is called to recognize and receive what God gave. Yet recognition is not irrelevant. If preservation is public and ecclesiastical rather than merely archival, the stable text actually read, translated, preached, and confessed by the churches bears theological significance.

The alternative is not between an infallible Erasmus and a neutral catalog. The question is whether providence gave the church a recognizable apographic text through ordinary historical means, or preserved all relevant readings in an aggregate whose original wording remains to be constituted by critical method.³

The first position treats reception as ministerial recognition of the inspired text. The second also locates authority in the inspired original but makes its extant identification dependent upon continuing scholarly adjudication. Both affirm providence; they give it different textual and ecclesiastical forms. That is why the small-base objection cannot decide the controversy. It criticizes one early editorial stage. It does not establish which doctrine of preservation best accounts for the full history.

“A Few” Versus “Thousands” Is Not a Textual Apparatus

The numerical contrast becomes even less decisive when the modern side of the comparison is examined. The Institute for New Testament Textual Research’s manuscript catalog is a remarkable scholarly resource. The surviving evidence is vastly greater than what Erasmus could consult. That expansion should be welcomed. It allows earlier print decisions to be tested, manuscript relationships to be studied, and the history of readings to be traced with greater precision.

But a catalog total is not an apparatus total, and an apparatus total is not a count of independent ancestors. Not every registered manuscript contains the entire New Testament. Some contain one corpus, one book, selected lessons, or a small fragment. A papyrus containing several verses of John cannot testify directly to a variant in Romans. A Gospel manuscript cannot vote at Revelation 22. A lectionary’s contents are structured by ecclesiastical readings rather than by continuous coverage of every book. At any particular variation unit, the number of relevant witnesses is smaller than the total catalog.

Nor are all relevant witnesses genealogically independent. Ten manuscripts copied directly or indirectly from a common exemplar do not constitute ten independent origins for a reading. Conversely, one manuscript may preserve a reading derived from an ancient branch no longer represented by numerous surviving descendants. Modern methods increasingly attempt to account for these relationships precisely because raw counting is inadequate.

Keister already understands this. He rejects simplistic majority arguments and insists that the whole manuscript tradition must be evaluated rather than merely counted. But that same principle must govern the rhetoric about Erasmus. If the number of Byzantine witnesses cannot by itself establish a reading, the number of manuscripts unavailable to Erasmus cannot by itself overturn one.

Thousands of manuscripts do not become thousands of independent votes at every locus. More evidence changes the evidentiary situation; it does not predetermine the verdict. It may confirm the printed reading, expose a peculiarity, divide into complex patterns, or support a competing form. Each case must be examined. “Erasmus had few; we have many” begins an inquiry; it does not conclude one.

The Difficult Readings Must Be Faced Individually

None of this licenses a general appeal to tradition whenever a Received Text reading appears weak. The Book of Revelation must be treated honestly. Erasmus’s exemplar was defective, and Latin influence upon particular readings cannot be dismissed. The precise story of the final verses has often been told with greater confidence than the surviving evidence permits, but uncertainty about the editorial mechanism does not remove the problem. Particular readings require particular evidence.

The same is true of the Comma Johanneum and other exceptional passages. Their authenticity cannot be established by saying only that they occur in a TR edition, nor disproved by reciting the total number of Erasmus’s Basel manuscripts. Each has its own Greek, versional, patristic, liturgical, and reception history. Those histories must be examined rather than compressed into a general origin story.⁴

The proper preservationist response is therefore not to deny anomalies but to limit what they prove. An uncorroborated Erasmian peculiarity should receive less confidence than a reading broadly attested in the Greek tradition and publicly received across later editions. A demonstrable printer’s error may be corrected without reconstructing the New Testament. A difficult exceptional reading may demand a more extensive defense than the stable textual form surrounding it.

Exceptional readings require exceptional defense.

But exceptions should not be made to define the whole. The vast majority of the Received Text did not originate in Erasmus’s mind, his Latin translation, or his printer’s shop. It consists of Greek readings transmitted before him, represented in the manuscript tradition, and carried into later ecclesiastical use. To describe the entire text by its hardest handful of loci is as misleading as describing the modern critical text only by its most controversial bracketed passages. The small-base objection should lower confidence in uncorroborated Erasmian peculiarities. It should not erase the identity of the broader received tradition.

The Right Comparison

Once the categories are separated, two different comparisons appear.

Comparison: Erasmus’s 1516 edition versus a modern critical edition

Proper conclusion: The modern edition possesses a vastly larger evidentiary base, a more developed apparatus, and more sophisticated tools.

Comparison: The received textual tradition versus the modern critical reconstruction

Proper conclusion: The question cannot be decided by manuscript totals; it requires a doctrine of preservation, a history of transmission and reception, and examination of readings.

Confusing these comparisons allows the superiority of a modern apparatus to function as proof of the superiority of a modern textual constitution. It is no such proof. A present-day critical editor knows more about the surviving manuscript record than Erasmus did. That fact is almost trivial. The more difficult question is whether the critical editor’s method rightly relates that knowledge to the text the church has received. Greater information can correct ignorance. It can also operate under a theory that assigns scholarship too constitutive a role.

Likewise, a providential history of reception can ground confidence. It can also be used lazily to protect a weak reading from examination. Neither the age of a method nor the size of a database guarantees ministerial faithfulness. The right question is not simply how many codices sat in Basel. It is what text those codices represented, how the printed text was corrected, how it was received, where its readings stand in the broader transmission, and what authority governs proposed revision.

Providence Through a Process, Not an Editor

The Received Text should not be defended as though God’s preservation depended upon the genius of one humanist. That would place an intolerable burden upon Erasmus and a fragile burden upon the doctrine. Providence works through processes. Scribes copy. Readers compare. Teachers cite. Churches read. Translators render. Printers introduce errors and correct them. Editors consult witnesses. Correspondents report evidence. Later editions repair earlier editions. None of these agents is inspired, yet God is able through fallible means to preserve the words he inspired.

The preservationist claim is not that every stage was pure in isolation. It is that the inspired text continued in faithful apographs, remained available and recognizable to the church, and came to stable public expression through the received textual tradition. That claim permits local variation without turning the text into an unidentified aggregate. It permits ministerial correction without making criticism the constitutive source of the church’s Bible.

Keister asks which manuscript before 1500 exactly equaled the TR. The answer is: no known single manuscript did, and no such identity is required. The preserved text is not identical to every physical witness through which it is transmitted. If that were required, one copying mistake would mean the text ceased to exist. Textual identity resides in the words faithfully transmitted across witnesses, not in the incidental perfection of one artifact.

Keister’s own aggregate model relies upon a similar distinction. He does not claim that one extant manuscript is the autograph. He says the autograph is found across the manuscripts collectively. The dispute, therefore, is not whether the text may be represented across differing apographs. It is whether the relevant collective is the total surviving aggregate awaiting critical discrimination or the publicly received textual tradition within which criticism ministers. That issue cannot be resolved by pointing again to Erasmus’s desk.

Notes and Further Reading