Erasmus, Vaticanus, and the Comma: What the Familiar Story Leaves Out

There is a version of the history of 1 John 5:7 that can be told before the coffee cools. Erasmus could not find the heavenly witnesses in Greek. He rashly promised to print them if anyone produced one Greek copy. A manuscript was then made to order, Erasmus kept his promise, and an interpolation entered the Textus Receptus and the Authorized Version. Codex Vaticanus, centuries older than Erasmus’s materials, had been sitting in Rome without the words all along. Once these facts are known, the case is said to be over.

The story is effective because several of its pieces are true. Erasmus omitted the Comma from his first two editions. A British manuscript containing it was brought to his attention. He included a form of the passage in 1522 while doubting the value of that witness. Vaticanus does omit the heavenly witnesses. The surviving independent Greek manuscript evidence for the passage is late and sparse. A defense that denies those facts will not survive contact with the documents.

Yet the familiar story also compresses several different questions into one. What did Erasmus know? Why did he alter his third edition? What is the age of the reading as distinct from the age of its surviving Greek copies? What authority did the early modern church claim when it received the passage? Above all, does the biography of an editor determine the words of an apostle?

Wesley Huff’s popular account of modern translations does not offer a sustained treatment of the Comma. It does, however, employ the larger contrast into which the familiar story is ordinarily fitted: Erasmus worked with few and late resources, modern scholars possess much older and more numerous witnesses, and comparison of those witnesses gives a clearer view of what the authors wrote.¹ The Comma is an exacting test of that account because the external Greek evidence strongly favors the shorter form while the history of Latin use, ecclesiastical reception, and doctrinal employment is far older and more complex than a sixteenth-century accident.

The proposition of this essay is deliberately limited. Erasmus did not create the Comma, Vaticanus did not reveal it to be absent from the entire earlier church, and the alleged wager should not be repeated as documented fact. None of those corrections, standing alone, proves the passage genuine. They do prove that editorial history must be stated accurately before it is allowed to govern a canonical judgment.

The editor did not work inside the legend

Erasmus printed his first Greek-Latin New Testament in 1516 and a revised edition in 1519. In both, the Greek text of 1 John 5 moved directly from the earthly witnesses to the following testimony. His accompanying Latin also departed from the familiar Vulgate form. The omission drew criticism from Edward Lee and Jacobus Stunica, who accused him of endangering Trinitarian doctrine and preferring defective Greek copies to the church’s Latin Bible.

Erasmus answered that the Greek manuscripts available to him lacked the words. He did not argue that the Trinity was false, nor did he present the shorter text as a discovery that overthrew Christian doctrine. He appealed to other passages against Arianism and questioned whether the Comma performed the theological work its defenders assigned to it. His position was textually skeptical and theologically orthodox by his own profession, though his exegesis and his account of ecclesiastical authority were contested from several directions.

The famous promise or wager is not found in Erasmus’s extant explanation of the affair. What the documents show is controversy, a demand for Greek evidence, the announcement that a British codex had been found, and a later editorial concession. Those events can easily grow into a wager when retold, but the dramatic form adds an intention and obligation that the surviving sources do not establish. A responsible criticism of Erasmus has enough material without inventing a bet.²

The British codex, now Trinity College Dublin MS 30 and commonly called Codex Montfortianus, is a late manuscript. Its paper, textual relationships, and early ownership place it near Erasmus’s own period. Its form of the Comma is Latinized, and Erasmus suspected that its text had been adjusted to the Vulgate. He did not simply copy it without thought. He adapted its wording when he inserted the heavenly witnesses into the third edition.

That decision can be described in more than one way. It was a concession under theological and ecclesiastical pressure. It was also an editorial recognition that the passage had a received Latin history and now had at least one Greek witness before him. Erasmus retained doubts and disclosed them. He neither believed that the late codex had become ancient nor treated his printed main text as a transcript of one manuscript.

The Complutensian Polyglot further complicates the fable that one hurriedly manufactured codex created the received reading. Its New Testament volume had been printed before Erasmus’s first edition, though papal authorization delayed publication of the completed Polyglot. It contained the heavenly witnesses in Greek. The precise manuscript basis of that form remains disputed, and its wording appears indebted to Latin theological and textual judgments. It is not independent early Greek proof. It is, however, evidence that the question was already larger than Erasmus and Montfortianus.

The historical conclusion must therefore be stated without embroidery. Erasmus first printed the shorter Greek form; he faced criticism; a late British Greek witness was produced; he included an adapted form of the Comma in 1522 while recording suspicion about that witness. That is serious evidence against treating his edition as a simple reproduction of early Greek copies. It is not evidence that the passage itself sprang into existence in 1522.

Vaticanus proves an omission, not a complete history

Erasmus knew more about Codex Vaticanus than the popular contrast often permits. His correspondent Paolo Bombace supplied him with readings from the manuscript, including a transcript of 1 John 4:1–3 and 5:7–11. Later, Juan Ginés de Sepúlveda sent or reported hundreds of Vatican readings in an effort to persuade Erasmus that the ancient codex agreed in important respects with the Vulgate. Erasmus therefore did not need a nineteenth-century discovery to learn that Vaticanus omitted the heavenly witnesses.³

This fact cuts in two directions. It refutes the defensive claim that Erasmus would have printed the modern critical text if only he had known Vaticanus. He did know relevant readings from it and did not adopt it as his controlling copy. It also refutes any suggestion that the absence of the Comma from Vaticanus was concealed from the early printed-text tradition. The omission was part of the evidence Erasmus discussed.

Vaticanus remains a formidable witness. It is much earlier than the known Greek copies containing the Comma, and at this point it agrees with the broad surviving Greek tradition. The current critical editions are not choosing the shorter form merely because a celebrated codex sits in the Vatican Library. They are following the convergence of early Greek manuscripts, ancient versions, Greek patristic citation, and a transmissional account in which a Latin Trinitarian interpretation entered the text and later passed into a few Greek copies.

That case must be allowed its full strength. The Comma is not in the continuous text of the great early Greek codices. The known Greek witnesses containing it are few; several place it in a later margin; their wording varies; and some betray translation from Latin. The earliest presently known Greek manuscript with a form in its text is medieval. No honest defense should turn ten late or partly marginal witnesses into an early Greek consensus.⁴

But Vaticanus proves what Vaticanus reads. Together with the other Greek evidence, it proves that the shorter form dominated the surviving Greek transmission. It does not place an observer in the apostolic copy of 1 John. It does not prove that no early Greek exemplar ever contained the passage. It does not erase evidence from another language tradition. Those larger conclusions require a genealogical history, and the history is inferred from surviving effects.

The distinction is the same one this series applied to John 5:4. An artifact is direct evidence for its own text. A proposed ancestor is an inference. In the Comma debate, the inference for secondary origin is powerful because the Greek absence is broad, the Latin forms are unstable, and the passage becomes explicit in Latin sources before it appears in extant Greek copies. Yet powerful inference remains the right category. “Vaticanus omits the words” is observation. “The words began as a Latin gloss and never stood in any Greek archetype” is a historical reconstruction. That reconstruction may be judged superior, but it should not acquire the voice of an eyewitness merely because the alternative has weaker surviving Greek support.

The reading is older than its surviving Greek copies

The Latin and patristic evidence does not rescue the Comma by giving it a fourth-century Greek manuscript. It does something more modest and still important: it demonstrates that the wording and its Trinitarian use long predate Erasmus.

Cyprian, writing in the third century, says that it is written of the Father, Son, and Holy Spirit, “and these three are one.” Whether he quoted a Latin biblical form of the heavenly witnesses or interpreted the earthly witnesses in a Trinitarian manner is disputed. The phrase cannot simply be counted as a continuous-text manuscript, but neither can a writer’s formula “it is written” be dismissed as though it were no evidence at all. Cyprian establishes at minimum that the Johannine words “three” and “one” were being applied to the divine persons in the Latin church well before the surviving medieval Greek witnesses.

The Liber Apologeticus, associated with Priscillian or his circle in the late fourth century, supplies an explicit form naming Father, Word, and Holy Spirit. In the sixth century Fulgentius cites the heavenly witnesses against Arian opponents and appeals to Cyprian’s authority. The passage appears in Old Latin streams, enters a broad portion of the medieval Vulgate tradition, and becomes embedded in Western theological use. The evidence is not uniform: important Vulgate codices omit it, and the Latin wording and order vary. But the reading is not a Renaissance novelty.⁵

This early Latin history presents two rival explanations. On the modern critical account, an allegorical interpretation of spirit, water, and blood developed into a marginal or textual expansion, spread through Latin transmission, and eventually entered a small number of Greek copies and printed editions. On the traditional account, an authentic reading suffered severe loss in Greek transmission, survived or was recognized in the Latin church, and returned to the printed Greek text through providential reception.

The critical account has the simpler path through the extant manuscripts. It explains the late and Latinized Greek forms and the silence of Greek fathers who might have used the verse. The traditional account must explain a catastrophic Greek omission and a predominantly Western survival. It may appeal to early controversy, homoeoteleuton, doctrinal hostility, or the peculiar transmission of the Catholic Epistles, but each mechanism must be proved rather than accumulated as a cloud of possibilities.

The traditional account nevertheless has evidence the caricature cannot absorb. Cyprian’s language, explicit late-antique Latin citation, longstanding liturgical and theological use, the reading’s fit within the witness structure of 1 John 5, its reception in the Reformation churches, and its place in the proof architecture of early Reformed theology are historical facts. They do not equal an early Greek codex. They do mean that “Erasmus added it from one made-to-order manuscript” is not an adequate history of the reading.

David Martin understood the doctrinal and textual issues as connected but not interchangeable. He argued from Latin transmission, fathers, Greek ecclesiastical knowledge, the Dublin manuscript, internal coherence, and reception. Some of his historical claims require correction in light of later manuscript cataloguing. His central warning survives: a Christian should not call a disputed clause dispensable merely because related truths can be proved elsewhere. Before surrendering words received as Scripture, the church must judge both the evidence and the rule by which evidence is made to speak.⁶

An editor cannot become the canon

The critical case against the Comma is among the strongest external cases against a major received reading. That concession is not a retreat from Confessional Bibliology. It is what ministerial textual work is supposed to do: describe the artifacts, expose legends, compare histories, and identify the actual point of difficulty.

The difficulty becomes magisterial when the account of Erasmus is permitted to replace the theological question. If Erasmus included the words for a poor reason, it follows that his reason was poor. It does not follow that the words are not apostolic. If Vaticanus omits them, it follows that Vaticanus is an early witness to omission. It does not follow that Vaticanus is the supreme judge of the church’s text. If late Greek witnesses depend upon Latin, it follows that those witnesses have reduced independent value. It does not follow that Latin evidence has no historical value or that ecclesiastical reception is irrelevant.

The defender of the Comma must also resist a shortcut. Providence is not a label pasted over an unexplained event. To say that the passage was received is to raise questions about the identity, duration, breadth, and authority of that reception. It does not make Erasmus infallible, convert Montfortianus into an ancient codex, or permit weak evidence to be described as strong. A providential argument must be theological and historical at once.

William Whitaker placed the authentic New Testament in Greek, not in a translation and not in the private decree of an ecclesiastical officer. James Ussher’s Body of Divinity likewise directs controversies concerning words to the original text while acknowledging variants caused by copyists and the church’s responsibility to discern the true reading. That Reformed position does not prohibit collation. It prohibits the scholar from becoming what the pope was not allowed to be: an unappealable judge whose sentence creates the authentic text.⁷

The Comma therefore remains a hard case. The Greek documentary evidence favors omission. The pre-Erasmian Latin and patristic evidence proves antiquity of the reading and its theological use, though not by itself its Greek apostolic origin. Early modern reception gives the passage ecclesiastical weight, though not mechanical immunity from examination. Internal and doctrinal arguments matter, though they cannot manufacture an artifact.

That is a less convenient conclusion than the coffee-cup legend. It is also more honest. Erasmus knew Vaticanus, doubted Montfortianus, and printed the passage. The church received words whose history did not begin with him. Modern editors reject those words on a substantial but inferential reconstruction. The dispute is not settled by turning any one of those sentences into the rule of faith.

Part 10 will return to Huff’s reassurance after the textual cases have been heard. The question will no longer be whether Christianity remains recognizable in a modern translation. It will be whether “no major doctrine” is a true verdict on the exact propositions that have changed.

Notes and Further Reading

Series navigation: Previous: “Book or Tree?” · Part 9 of 14 · Next: “No Major Doctrine?”