Erasmus did not imagine manuscript family genealogies and constructing a text from among them those readings which are most intrinsically and transcriptionally probable.
they didn’t have anything like modern textual criticism.
According to his own words and according to the facts from the varying Greek manuscripts that he used, Erasmus engaged in textual criticism decisions. Erasmus did sometimes speculate concerning which reading he considered most probable. What Erasmus and other textual editors of the 1500's did involved actual textual criticism regardless of whether you may try to deny it. You ignore the fact that Erasmus even introduced some textual conjectures into his text, readings found in no known preserved Greek NT manuscripts.
The 1550 edition of the Greek NT text of Stephanus with its over 2,000 variant readings provides more evidence of the inconsistent, imperfect textual criticism of that day.
Edward F. Hills observed that Stephanus "placed in the margin of his 3rd edition of the Textus Receptus variant readings taken from 15 manuscripts, which he indicated by Greek numbers" (
KJV Defended, p. 117). F. H. A. Scrivener indicated that Stephanus in his preface stated that his sources were sixteen, but that includes the printed Complutensian as one of them (
Plain Introduction, II, p. 189). Tregelles confirmed that “the various readings in the margin are from the Complutensian printed edition and from fifteen MSS” (
Account, p. 30). Brian Walton observed that Stephanus “reckons sixteen Greek copies, which he collated, and out of them noted 2384 various readings, which he though fit to put in the margin of his edition” (Todd,
Memoirs, II, p. 132). Edwin Bissell maintained that
“in the edition of 1550, indeed, the first collection of variations in manuscripts was actually published, numbering two thousand one hundred and ninety-four” (
Historic Origin, p. 128).
Charles E. Hammond claimed: “The influence of prescription already shows itself in the fact that Stephens often follows the text of Erasmus, in defiance of the authority of his manuscripts” (
Outlines of Textual Criticism, p. 11). Herbert Marsh asserted that “in the margin of this [1550] edition there are more than a hundred places, in which he [Stephens] has quoted all his authorities for readings different from his own” (
Course of Lectures, p. 106). F. H. A. Scrivener as edited by Edward Miller maintained that the text of the 1550 edition “is perpetually at variance with the majority” of his fifteen Greek manuscripts and the Complutensian, “and in 119 places with them all” (
Plain Introduction to the Criticism, Vol. II, p. 190). Robert B. Waltz cited “Scrivener’s report that there are 119 places where all of Stephen’s manuscripts read against the TR, but Stephens still chose to print the rendering in previous TR editions” (
Encyclopedia of NT Textual Criticism, p. 855).
The Oxford Encyclopedia of the Reformation noted that "through its [Erasmus's Greek text] being incorporated into the third edition of Robert Estienne's Greek Testament (1550) it influenced strongly the Greek Testament of Theodore de Beza" (Vol. 2, p. 57). Scrivener noted that his “own collation represents Stephen’s first edition as differing from his third in 797 places, of which 372 only are real various readings, the rest relating to accents, or being mere errata” (
Plain Introduction, II, p. 190, footnote 3).
Samuel Tregelles wrote: "Robert Stephens, ten years before, in editing the Latin Vulgate, had made pretty extensive use of MSS.; and in giving the work of Greek collation into the hands of his son Henry, then aged only eighteen, he might have had some thoughts of similarly applying criticism to the Greek text" (
Account, p. 31). Scrivener asserted that “Robert Stephen professed to have collated the whole sixteen for his two previous editions,” but that “this part of his work is now known to be due to his son Henry [1528-1598], who in 1546 was only eighteen years old” (
Introduction, II, p. 190). Edward Miller affirmed: “Robert Stephen did not collate his authorities himself, but employed the services of his son Henry” (
Guide to the Textual Criticism, p. 10). J. Scott Porter also maintained that “the MSS. were collated, and their readings noted, by Henry Stephens, son of Robert, then a youth of eighteen” (
Principles, p. 250). Irena Backus asserted that Robert Stephanus “used Henri’s collations as the sole source of Greek variants for his 1550 edition of the New Testament” (
Reformed Roots, p. 3). John Michaelis as translated by Herbert Marsh pointed out that Robert Stephens “made use of several manuscripts which were collated by his son Henry” (
Introduction to N. T., II, p. 448). Henry Baird quoted Theodore Beza as writing in a preface to his NT about a copy of “our Stephens which had been most carefully collated by his son, Henry Stephens” (
Theodore Beza, p. 236). KJV-only author Laurence Vance acknowledged that the text of Stephanus included the “collations of his son Henry” (
Brief History, p. 13). Jan Krans pointed out that “in a 1565 addition to the preface, Beza informs us that the collations were actually Henri Stephanus’, who was probably asked to do them by his father” (
Beyond What is Written, p. 212). Krans also referred to another source revealing that the collations were done by the son of Robert Stephanus, which is “Henri Stephanus’ own words in the preface to his 1587 New Testament” (p. 212, footnote 6).
Has anyone ever checked and confirmed the accuracy of all of these collations? Scrivener suggested that “the degree of accuracy attained in this collation may be estimated from the single instance of the Complutensian, a book printed in very clear type” (
Plain Introduction, II, p. 190). Scrivener then indicated that “forty-eight, or one in twelve [of Stephen’s citations of the Complutensian] are false” (p. 190, footnote 1). Tregelles maintained that “it may be said, that as the Complutensian text is often
incorrectly cited in Stephen’s margin, we may conclude that the same thing is true of the MSS which were collated; for it would be remarkable if manuscripts were examined with greater accuracy than a printed book” (
Account, p. 31).
Smith’s Dictionary maintained that “while only 598 variants of the Complutensian are given, Mill calculates that 700 are omitted” (III, p. 2131). Marvin Vincent asserted: “Of the Complutensian readings many more were omitted than inserted, and the Complutensian text is often cited incorrectly” (
History of the Textual Criticism, p. 57). In a note, John Eadie commented: “The margin of the New Testament of Robert Stephens, 1550, is not of great value. He did not print all the various readings which his son Henry had gathered, nor did he fully collate all the sixteen MSS” (
English Bible, II, p. 214). Samuel Newth maintained that the manuscripts used by Stephanus were “imperfectly collated” (
Lectures, p. 86). Frederic Gardiner claimed that the collation in this edition “is neither complete nor accurate” (
Principles, p. 5). Marvin Vincent suggested that “the collation, both of the Complutensian and of the manuscripts was partial and slovenly” (
History of the Textual Criticism, p. 57). Vincent wrote: “
The body of manuscript evidence amassed by the Stephens were imperfectly collated in the edition of 1550. Though the authorities stand in the margin, the text is perpetually at variance with the majority of them, and in 119 places, with all of them. No fixed principles regulated the occasional applications of the manuscript readings to the construction of the text” (pp. 63-64). Richard Porson (1759-1808) asserted that “Stephen’s margin is full of mistakes in the readings and numbers of the MSS” (
Gentlemen’s Magazine, May, 1789, p. 386;
Letters, p. 55). Porson maintained that Stephens “has favored us with only a part of the various readings, (probably less than half) and has frequently set down a reading as from one manuscript which belonged to another” (
Letters, pp. 88-89). Charles Hudson reported that the “various readings collated by his son” . . . “are known to be given very inaccurately” (
Greek and English Concordance, p. xiv).