The 99 Percent Argument Is Not a Majority-Text Rule

Part II: Mark 16 and Scrivener · Article 2 of 3

Stephen Steele treats the 99-percent figure at Mark 16 as though the Trinitarian Bible Society had announced a universal algorithm: at every variation unit, count the surviving Greek manuscripts and print the reading with the largest number.

TBS does not profess that method. Its doctrine identifies the Received Text as the Greek text preserved through ecclesiastical transmission and represented in the Reformation-era printed tradition. Numerical prevalence is important to that account, but the Society is not a Majority Text organization in the strict editorial sense.¹

Steele therefore defeats a rule TBS does not claim. His examples still expose statements that require correction. They do not prove that the only remaining rule is “whatever the KJV says.”

Three Different Arguments from Number

An appeal to 99 percent can mean at least three things.

It can be a decisive rule:

The reading with the most surviving manuscripts is original.

---

1. Trinitarian Bible Society, “Statement of Doctrine of Holy Scripture,” Quarterly Record 571 (April–June 2005): 10–13, https://cdn.ymaws.com/www.tbsbibles.org/resource/collection/01C074CC-748F-4C67-86AC-A9926A25241A/Issue%20571%20-%20April%202005%20from%20web.pdf. ↩

It can be a rebuttable presumption:

A reading transmitted in nearly all surviving manuscripts should be retained unless unusually strong contrary evidence overcomes that prevalence.

---

1. Trinitarian Bible Society, “Statement of Doctrine of Holy Scripture,” Quarterly Record 571 (April–June 2005): 10–13, https://cdn.ymaws.com/www.tbsbibles.org/resource/collection/01C074CC-748F-4C67-86AC-A9926A25241A/Issue%20571%20-%20April%202005%20from%20web.pdf. ↩

Or it can be an internal critique:

A method that claims to respect the whole manuscript tradition must explain why two codices outweigh a reading known in the second century and transmitted in almost every later Greek copy.

---

1. Trinitarian Bible Society, “Statement of Doctrine of Holy Scripture,” Quarterly Record 571 (April–June 2005): 10–13, https://cdn.ymaws.com/www.tbsbibles.org/resource/collection/01C074CC-748F-4C67-86AC-A9926A25241A/Issue%20571%20-%20April%202005%20from%20web.pdf. ↩

The first is mechanical majority rule. The second is an evidentiary principle. The third tests whether another method weighs witnesses consistently with its own stated regard for transmissional evidence.

TBS’s Mark 16 argument is strongest in the second and third forms. The percentage does not stand alone. It accompanies Irenaeus, ancient versions, major uncials after the fourth century, ecclesiastical use, and the peculiar layout of Vaticanus. It measures the exceptional character of the shorter form within surviving Greek transmission.

Steele quotes the number, turns it into the first proposition, and then asks why TBS does not count the same way everywhere.² That inference requires proof.

Received Text and Majority Text

The Received Text and the Majority Text agree at most places because both stand chiefly within the Byzantine transmission. They differ in method and at a limited number of readings.

A modern Majority Text edition collates manuscripts, classifies their relationships, and prints the reading judged to represent the majority form. A Received Text position receives a settled Greek textual tradition through the church’s use and the mature Reformation editions. It allows manuscript evidence to confirm, illuminate, and defend that received text, but it does not reopen every reading to a fresh vote whenever another manuscript is catalogued.

That position can be criticized. Steele may argue that ecclesiastical reception is being defined too narrowly, that the Reformation print settlement receives unwarranted finality, or that a closed edition cannot correct known errors. Those are direct objections.

It is not a direct objection to say that a non-majority reading disproves majority rule when the position has already denied majority rule as its constitutive method.

TBS’s own statement makes the distinction visible. It says the received Greek text was acknowledged throughout the Byzantine period and preserved in the overwhelming majority of extant manuscripts. It also acknowledges that a few readings in the printed Received Text came through Latin and lack presently extant Greek support.³ One may think those propositions sit uneasily together. The Society plainly knows it is not asserting numerical unanimity.

Mark 15:3

Steele’s first local test is Mark 15:3. The Authorized Version reads that the chief priests accused Christ of many things, “but he answered nothing.” Most Greek manuscripts and the principal later Received Text editions lack the clause.

This is a legitimate difficulty for any TBS publication that places the verse on a list of modern “omissions” without explanation. The clause is not the stable reading of the mature Greek tradition in the same way Mark 16:9–20 is. Calling its absence a drastic modern alteration conceals its limited support.

Steele overstates the other side when he presents the wording as though it existed only because Scrivener designed a Greek text to match the KJV. The phrase had antecedent printed Greek authority. It appears in the Complutensian Polyglot and in the first two editions of Stephanus, though not in his influential 1550 edition. Beza discussed the Complutensian reading favorably in his annotation even while excluding it from his own text.⁴

Those facts establish a narrower conclusion:

• The AV did not invent the clause.
• Scrivener did not translate it from English into Greek.
• The reading was known in Greek print before 1611.
• Its surviving Greek support and place in later Received Text editions are nevertheless weak.

TBS should not use Mark 15:3 as though it were textually equivalent to Mark 16. Steele should not use its weakness to prove an English origin that the chronology excludes.

Ephesians 3:9

Ephesians 3:9 is harder. The Received Text reads koinōnia, “fellowship,” where the overwhelming Greek evidence reads oikonomia, “dispensation,” “administration,” or “stewardship.”

The words are visually similar, and koinōnia has some late Greek attestation. Yet the external evidence for oikonomia is immense: the early papyrus, the uncials, almost all minuscules, versions, and patristic citations. Even defenses of the Received Text reading acknowledge that it is a tiny minority.⁵

This example refutes the claim that the numerical majority alone constitutes the Received Text. It also requires more than an appeal to the English phrase “fellowship of the mystery.” If the reading is to be defended, its advocates must offer a transmissional, internal, or reception-historical case adequate to the evidence.

What it does not prove by itself is motive. A textual tradition can contain a weak reading without having been created to serve an English translation. Koinōnia stood in Greek editions before the AV. The KJV inherited and translated it. The chronological direction again runs from Greek print to English, even if one concludes that the Greek reading should now be corrected.

Steele’s argument moves from “this reading is poorly attested” to “the KJV is the only standard.” The premise supports criticism of the reading; it does not establish the cause assigned to its retention.

Revelation and the Zero-Percent Rhetoric

Revelation supplies Steele’s most forceful examples because Erasmus worked from a defective exemplar and sometimes supplied Greek from Latin. Some mature Received Text readings have no presently identified extant Greek witness. Revelation 16:5 has become the emblem of that difficulty.

Those cases must be distinguished before percentages are assigned.

At Revelation 16:5, “O Lord” and “who shall be” are separate variation units. The first has direct Greek attestation outside the printed Textus Receptus. The second has no presently identified surviving Greek witness, but Beza expressly said he restored it from an old, trustworthy manuscript. Whether his report was accurate remains disputed.

At Revelation 20:12, “God” is a minority Greek reading, not a zero-manuscript English creation. At Revelation 21:24, the additional clause appears in a pre-Erasmian Greek manuscript, though the reading probably arose when commentary and biblical text were confused. Revelation 21:14, by contrast, presents a genuine error in the received printing: the third “twelve” is overwhelmingly attested and was omitted despite standing in Erasmus’s exemplar.

The details will be examined in Article 8. Here they show why “zero percent” cannot function as a rhetorical basket into which every difficult Revelation reading is thrown.

An Internal Critique of Critical Weighting

The 99-percent argument has its greatest force when directed toward critical weighting.

Modern eclectic criticism does not count manuscripts mechanically. It weighs age, genealogy, geographical spread, scribal behavior, versional testimony, patristic citation, and internal probability. That is why two early witnesses can outweigh hundreds of later ones.

But such weighing can become opaque. If the overwhelming transmission of a passage, explicit second-century attestation, and long ecclesiastical reception are all overcome by a narrow set of witnesses and internal judgments, readers are entitled to ask what evidence could ever overturn the favored stream.

Mark 16 exposes that question with unusual clarity. The longer ending is early enough to be cited by Irenaeus and broad enough to dominate the surviving Greek tradition. Critical editors nevertheless classify it as secondary. The percentage does not prove originality by itself, but it shows how much contrary evidence the critical judgment must explain.

TBS need not be a Majority Text society to press that criticism. A lawyer may challenge an opponent’s treatment of numerical evidence without making numbers his only rule.

Where TBS Must Be More Careful

The defense of TBS does not require defending every sentence it has published.

Lists of “omissions” too often group together readings with radically different evidence. Mark 16:9–20, Mark 15:3, Ephesians 3:9, and Revelation 16:5 should not be presented under one undifferentiated label. The first has overwhelming Greek support; the second has antecedent printed support but weak manuscript standing; the third is a very small Greek minority; the fourth contains separate units with different evidence.

Nor should theological effect substitute for textual evidence. TBS says that “throne” at Revelation 20:12 obscures divine glory in comparison with “God.” Even if the reading “God” is original, that argument is dangerous. If “throne” were proved original, the Holy Spirit would not thereby have obscured Christ’s glory. Textual judgment must establish the wording before doctrinal rhetoric evaluates the difference.

Steele is right to rebuke such overstatement. His correction would be stronger if he did not answer it with a larger overstatement about the Society’s entire authority structure.

Consistency Properly Defined

Consistency does not require the same evidence to bear the same weight at every passage. It requires the same principles to govern the assessment of whatever evidence exists.

A Received Text method may consider:

• the reading’s place in the continuous Greek church text;
• its presence across mature printed editions;
• Greek manuscript testimony;
• versions and patristic citations;
• the possibility of loss or contamination;
• internal coherence;
• public ecclesiastical reception.

The method becomes ad hoc if those considerations are invoked or dismissed solely according to whether the AV agrees. That is the very point Steele must prove. Non-majority readings create a test case; they do not supply the result in advance.

The proper verdict is therefore twofold:

The 99-percent figure at Mark 16 is not a universal majority-vote rule. It is part of a cumulative defense and an internal criticism of the extraordinary weight assigned to a narrow adverse tradition. Steele’s minority examples show that TBS is not a Majority Text society and that some of its popular rhetoric is careless. They do not prove that the KJV is its final authority.

---

1. Trinitarian Bible Society, “Statement of Doctrine of Holy Scripture,” Quarterly Record 571 (April–June 2005): 10–13, https://cdn.ymaws.com/www.tbsbibles.org/resource/collection/01C074CC-748F-4C67-86AC-A9926A25241A/Issue%20571%20-%20April%202005%20from%20web.pdf. ↩
2. Stephen Steele, “Unpicking a Misleading Meme on Mark 16 (Part 2),” Gentle Reformation, June 24, 2026, https://gentlereformation.com/2026/06/24/unpicking-a-misleading-meme-on-mark-16-part-2/. ↩
3. TBS, “Statement,” 12. ↩
4. Steele acknowledges the Stephanus evidence in his first footnote. See also Michael Marlowe, “Variations within the Received Text Tradition,” cited in Steele, “Part 2.” ↩
5. Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed. (Stuttgart: Deutsche Bibelgesellschaft, 1994), 535; Jeff Riddle, “Text Note: Ephesians 3:9,” Stylos, November 13, 2019, https://www.jeffriddle.net/2019/11/wm-138-text-note-ephesians-39.html. ↩

Mark 16 tests the weighting of witnesses. Scrivener tests the direction of authority. The next article turns to the edition on which Steele places the greatest burden.

Notes

Bibliography

Metzger, Bruce M. A Textual Commentary on the Greek New Testament. 2nd ed. Stuttgart: Deutsche Bibelgesellschaft, 1994.

Riddle, Jeff. “Text Note: Ephesians 3:9.” Stylos. November 13, 2019.

Steele, Stephen. “Unpicking a Misleading Meme on Mark 16 (Part 2).” Gentle Reformation. June 24, 2026.

Trinitarian Bible Society. “Statement of Doctrine of Holy Scripture.” Quarterly Record 571 (April–June 2005): 6–13.