I

TEXT AND CANON OF THE NEW TESTAMENT

Revisiting the
Corruption
of THE
New Testament
Manuscript, Patristic, and Apocryphal Evidence

Daniel B. Wallace
EDITOR

►V Kregel
Academic & Professional
Revisiting the Corruption of the New Testament: Manuscript, Patristic, and
Apocryphal Evidence
© 2011 Daniel B. Wallace
Published by Kregel Publications, a division of Kregel, Inc., P.O. Box 2607,
Grand Rapids, MI 49501.

All rights reserved. No part of this book may be reproduced, stored in a re­
trieval system, or transmitted in any form or by any means—electronic, me­
chanical, photocopy, recording, or otherwise—without permission of the
publisher, except for brief quotations in printed reviews. All rights reserved.
Library of Congress Cataloging-in-Publication Data

Revisiting the corruption of the New Testament: manuscript, patristic, and

apocryphal evidence / [edited by] Daniel B. Wallace.
p. cm.—(The text and canon of the New Testament)
1. Bible. N.T.—Evidences, authority, etc. 2. Bible. N.T.—Canon. I. Wallace,
Daniel B.
BS2332.R48 2010
225.1—dc22
2010029717

Printed in the United States of America

11 12 13 14 15 / 5 4 3 2
 Contents

Preface / 9
Series Preface / 13
Abbreviations / 15

Chapter 1—LOST IN TRANSMISSION: How Badly Did the Scribes

Corrupt the New Testament Text? / 19
Daniel B. Wallace

Chapter 2—THE LEAST ORTHODOX READING IS TO BE

PREFERRED: A New Canon for New Testament Textual
Criticism? / 57
Philip M. Miller

Chapter 3—THE LEGACY OF A LETTER: Sabellianism or Scribal

Blunder in John 1.1c? / 91
Matthew P. Morgan

Chapter 4—PATRISTIC THEOLOGY AND RECENSION IN

MATTHEW 24.36: An Evaluation of Ehrman’s Text-
Critical Methodology / 127
Adam G. Messer

Chapter 5—TRACKING THOMAS: A Text-Critical Look at die

Transmission of the Gospel of Thomas / 189
Tim Ricchuiti

Chapter 6—JESUS AS ΘΕΟΣ: A Textual Examination / 229

Brian J. Wright

Scripture Index / 267

Ancient Sources Index / 273
Person and Subject Index / 279

I
 Preface

his inaugural volume of the Text and Canon of the New Testament

T series includes essays by six authors. All of the chapters focus on

issues in textual criticism—in particular, how badly the scribes, who
copied their exemplars by hand, corrupted the text. All but one of the
chapters deals specifically with NT textual criticism; one addresses tex­
tual issues related to an early apocryphal text, the Gospel of Thomas.
The writing of each essay tells its own story. Chapter 1 is an ex­
pansion of a paper I delivered at the fourth annual Greer-Heard Point-
Counterpoint Forum, held at New Orleans Baptist Seminary in April
2008. Dr. Bart Ehrman and I dialogued there over the issue of corrup­
tion of the NT manuscripts. Four other scholars joined the debate:
David Parker (Birmingham University), Dale Martin (Yale University),
Michael Holmes (Bethel University and Seminary), and William Warren
(New Orleans Baptist Seminary). The text of all die presentations, along
with a few others, was published by Fortress Press as The Reliability of
the New Testament: Bart D. Ehrman and DanielB. Wallace in Dialogue,
edited by Robert Stewart. I express my gratitude to Fortress Press for
allowing me to publish an expansion of my paper in this book. This
first chapter, because it was delivered to a nonspecialist and, to a large
degree, lay audience, is in some ways out of sync with the rest of the
chapters in this book. But because it frames the discussion that the rest
of the book addresses, it is appropriate as the entree to the topic of this
volume.
The other five chapters were written by former interns of mine

grateful to DTS for having an academic intern program, which allows

faculty to help develop scholars in the making. Since the early 1990s, I
have had multiple interns every year. To date, I have enjoyed the privi­
lege of mentoring more than fifty students through this δοΰλο? pro­
gram—an apt nickname because each student is required to put in 400
hours of work without pay. Each professor shapes the internship how
he or she wants. My own approach has been to spend between sixty and
one hundred hours with the interns during the school year: we usually
meet in my home for several hours each week. Each intern is required
to write a paper that he or she will present at a regional scholarly con­
ference. That paper is read to the intern group in rough draft and, later,
final draft before it is read at the conference. The greatest joy of my

9
 Preface

academic career has been to work closely with these future professors,
pastors, priests, missionaries, and leaders of the church and to see them
mature through the rigors of our year together. Most have gone on for
doctoral work, and they collectively have published dozens of articles
in theological journals, as well as an increasing number of scholarly
monographs. Many, if not most, of these publications were expansions
on these students’ intern papers.
The chapters in this volume went through multiple layers of peer
review. These five men all read their papers first at the Southwestern
Regional Evangelical Theological Society conference. I was impressed
with each of these papers so sufficiently that I requested that these
ThM students be allowed to read them at the national Evangelical
Theological Society conference in November 2008, held in Providence,
Rhode Island. Normally, master’s students are not permitted to read
papers at the annual ETS conference, but these essays were so good
that permission was granted. Several textual critics interacted with the
papers—both before the conference and at it. Valuable feedback thus
came from the intern group initially, then from the regional conference
and individual scholars, and finally from the national conference of the
ETS.
When I proposed to Dr. Bruce Ware, the program chair of the 2008
ETS conference, that these papers be considered for the conference, I
soon realized that they should get even wider exposure. Jim Weaver,
director of Academic & Ministry Resources at Kregel Publications, was
enthusiastic about the possibility of these essays comprising the bulk
of the inaugural volume of the Text and Canon of the New Testament
series.
All five chapters address, directly or indirectly, issues raised in
Bart Ehrman’s The Orthodox Corruption of Scripture, a monumental
work that has raised numerous questions about intentional corrup­
tion of the NT by proto-orthodox scribes in the early centuries. Two
of the essays address a single verse (Matt 24.36 and John 1.1), one
a methodological issue (whether the least orthodox reading is to be
preferred), one an analogous matter (the textual transmission of the
Gospel of Thomas), and one a foundational theological issue (whether
the autographic text ever spoke of Jesus as θεός). Thus three of the
essays deal with global issues (i.e., not focused on a single verse or
textual problem), while two deal with an individual textual problem;
one of these latter two focuses on patristic interactions over Matthew
24.36, while another examines the texts of two manuscripts in John
1.1. The chapter on the text of Thomas may seem out of place in this
volume, but it is both groundbreaking in its treatment and highly rel­
evant for tire transmission of the NT text: the text of Thomas has un­
dergone significant changes that made it more compatible with the
rest of the Nag Hammadi codices.

10
 Preface

As this manuscript was going to the publisher, I received an email

from Dr. Ehrman requesting critical input on his Orthodox Corruption
ofScripture. Oxford University Press had requested that he update that
eighteen-year-old work. In response, I sent Dr. Ehrman the drafts of
most of the chapters of this manuscript. This volume might, then, have
some impact on the second edition of Orthodox Corruption ofScripture.
Regrettably, because of where our respective rnanuscripts were in the
publishing pipeline, the reverse could not be the case.
My fellow contributors and I offer up these essays in the spirit of
collegial dialogue and bring them as offerings of the mind to our Lord
Jesus Christ. May he be pleased and magnified by our efforts.
—Daniel B. Wallace, editor

11
I
Text and Canon of the New Testament

ew would question that Jesus Christ is the most important figure in

F human history, let alone in Western civilization. Our primary sources

for him are the twenty-seven books that we call the New Testament.
There are numerous challenges that face historians, Christians, theo­
logians, and skeptics alike: Do the Gospels tell the truth about Jesus?
Are they historically reliable? Were they later documents, written by
noneyewitnesses in every case? Was Paul's view of Jesus contrary to the
view of the early apostles? Do the books of the New Testament contra­
dict each other theologically or historically? Were the letters purported
to be by Paul, Peter, James, and Jude really written by these men? Can
we recover the autographic text of the New Testament with any reason­
able assurance? What about the other NT documents—is the author­
ship traditionally assigned to them accurate? How should we interpret
these books? Are they normative for Christians today—that is, do they
speak to believers authoritatively with reference to faith and practice?
AH these and many more questions are exceedingly important. But
two areas are foundational: First, can we recover the autographic text;
that is, can we determine, through rigorous analysis of surviving manu­
scripts and scribal methods, what that text, in all its essentials, looked
like? Do we dare go further, even hoping to recover the autographic
wording in all its particulars? If so, what methods and materials would
be of critical importance in this endeavor? Second, should these twenty­
seven books be treated with more authority than the myriad of books
that were written by Christians in the early centuries after the death and
resurrection of Jesus? If so, why? Did the early church get it right con­
cerning these books and their authors, or do we need to modify their
conclusions or even revamp them? This series explores these questions
of textual and canonical criticism. Its focus is whether these books bear
an authority above all others and, if so, what these books actually say.
The hope for the series is that fresh dialogue, collegial interaction, and
new evidence and arguments will enable us to better grasp the text of
the New Testament and its place in the life of the church today.
—Daniel B. Wallace, editor

13
 Abbreviations
t

APOCRYPHA AND PSEUDEPIGRAPHA

Gos Heb Gospel of the Hebrews
Sir Sirach
Wisd Sol Wisdom of Solomon

APOSTOLIC FATHERS
1 Clem. 1 Clement
Herm. Sim. Shepherd ofHermas, Similitudes
Ign., Eph. Ignatius, To the Ephesians
Ign., Trail. Ignatius, To the Trallians
Ign., Phld. Ignatius, To the Philadelphians
Ign., Pol. Ignatius, To Polycarp
Ign., Rom. Ignatius, To the Romans
Ign., Smym. Ignatius, To the Smyrnaeans
Mart. Pol. Martyrdom ofPolycarp
Pol., Phil. Polycarp, To the Philippians

Old Testament
Exod Exodus
Deut Deuteronomy
IChr 1 Chronicles
Job Job
Ps/Pss Psalm(s)
Prov Proverbs
Isa Isaiah
Mic Mic

New Testament
Matt Matthew
Mark Mark
Luke Luke
John John
Rom Romans
1 Cor 1 Corinthians
2 Cor 2 Corinthians
Gal Galatians
Eph Ephesians

15
 Abbreviations

Phil Philippians
Col Colossians
llhess 1 Thessalonians
2Thess 2 Thessalonians
lTim 1 Timothy
2 Tim 2 Timothy
Titus Titus
Phlm Philemon
Heb Hebrews
Jas James
IPet 1 Peter
2 Pet 2 Pet
1 John 1 John
2 John 2 John
3 John 3 John
Jude Jude
Rev Revelation

Greek Texts
NA27 Novum Testamentum Graece, 27th ed.
Tischendorf8 Novum Testamentum Graece, Sth ed.
UBS4 Greek New Testament, 4th revised ed.

Modern English Versions

CSB The Christian Standard Bible
ESV The Holy Bible: English Standard Version
KJV King James Version
NASB New American Standard Bible
NET The New English Translation
NTV New International Version
NJB The New Jerusalem Bible
NLT New Living Translation
NRSV New Revised Standard Version
RSV Revised Standard Version

CLASSICAL AND ANCIENT CHRISTIAN SOURCES

Eusebius, Eccl. theol. Eusebius, De ecclesiastica theologia
{Ecclesiastical Theology)
Eusebius, Hist eccL Eusebius, Historia ecclesiastica
{Ecclesiastical History)
Hippolytus, Philos. Hippolytus, Refutatio omnium haeresium
(Philosophoumena) {Refutation ofAU
Heresies)
Josephus, Ap. Josephus, Contra Apionem {Against Apion)

16
 Abbreviations

Origen, Comm. Jo. Origen, Commentarii in evangelium

Joannis (Commentary on John)
Origen, Matt. Comm. Origen, Commentarium in evangelium
Matthaei (Commentary on Matthew)
Philo, Somn. Philo, De somniis (On Dreams)
Tertullian, Ad uxor. Tertullian, Ad uxorem (To His Wife)

GENERAL
ca. circa
CE Common Era
LXX Septuagint
MS, MSS manuscript(s)
MT Majority Text, Masoretic Text
NT New Testament
OT Old Testament
TR Textus Receptus
TSKS Artide-substantive-Kai-substantive

MAJOR REFERENCE WORKS AND SERIES

ANF Ante-Nicene Fathers
BDAG A Greek-English Lexicon ofthe New
Testament and Other Early Christian
Literature
BP Biblia Patristica: Index des Citations et
Allusions Bibliques dans la Litterature
Patristique
IGNT/IGNTP International Greek New Testament
Project
L&N The Greek/English Lexicon ofthe NT Based
on Semantic Domains
MPG/PG Patrologia Graecae
NPF/NPFS The Nicene and Post-Nicene Fathers: First
Series
J ^5 . ^z ♦ ς'ζ z>
Patrologia Latina
TCE The Catholic Encyclopedia
TLG Thesaurus Linguae Graecae

PAPYRI
NHC Nag Hammadi Codices
P. Oxy. Oxyrhynchus Papyri

PERIODICALS
AJT American Journal of Theology
Bib Biblica
BSac Bibliotheca Sacra

17
 Abbreviations

CBQ Catholic Biblical Quarterly

CRBR Critical Review ofBooks in Religion
CT Cuneiform Textsfrom Babylonian Tablets
in the British Museum
EFN Estudios defilologia neotestamentaria
EvT Evangelische Theologie
ExpTim Expository Times
HibJ Hibbert Journal
HTR Harvard Theological Review
HTS Harvard Theological Studies
JBL Journal ofBiblical Literature
JETS Journal of the Evangelical Theological
Society
JSNT Journalfor the Study of the New Testament
JTS Journal of Theological Studies
Mus Museon: Revue deludes orientales
NovT Novum Testamentum
NTS New Testament Studies
PSB Princeton Seminary Bulletin
RelSRev Religious Studies Review
ResQ Restoration Quarterly
SJT Scottish Journal of Theology
ThTo Theology Today
TJ Trinity Journal
TRu Theologische Rundschau
TS Theological Studies
TSK Theologische Studien und Kritiken
TynBul Tyndale Bulletin
VC Vigilae Christianas
ZNW Zeitschriftfur die neutestamentliche
Wissenschaft und die Kunde der dlteren
Klrche

I
• ♦ ·%· ► 18
I
 1

LOST IN TRANSMISSION
How Badly Did the Scribes Corrupt
the New Testament Text?

Daniel B. Wallace1

or well over a century, the principal attacks on the earliest MSS of

F the NT have come from the far right, theologically speaking. Most
of the attackers are fanatics, but there are some scholars among them.
Occasionally, vitriolic language comes from even those who work with
the Greek text and thus, in some measure, stand apart from the Bible­
thumping preachers in America’s backwaters. They speak of the early
manuscripts as “bastard Bibles” and as coming from a “sewer pipe.”
Such verbal abuse might be expected from the ultraconservative
wing, because their method is apparently dictated, from beginning to
end, by their theological presuppositions. But in recent years, the attack
on the earliest witnesses has also come from the left side of the theo­
logical aisle. Liberals, too, have their fanatics, some of whom may be
bona fide scholars but with expertise far removed from textual matters.
They nevertheless freely opine that the state of the text is in such bad
repair that we must abandon all hope of recovering anything remotely
close to themriginal wording.
Things have changed in the last few years, however: some respect­
able textual critics have joined the ranks of the scornful. Bart Ehrman—a
scholar with impeccable credentials in textual criticism—-has arguably
led the charge. He has been the most prolific among these scholars and
the most intentional in bypassing peer review, appealing directly to the

1. This chapter is an expanded version of a lecture delivered at the fourth, annual

Greer-Heard Point-Counterpoint Forum, held at New Orleans Baptist Seminary
on April 4-5, 2008. A more succinct version appeared in Robert Stewart's The
Reliability of the New Testament: Bart Ehrman and Daniel Wallace in Dialogue
(Minneapolis: Fortress, 2010). Thanks are due to Fortress Press for permission to
publish tiiis expanded essay here.

19 / Chapter 1
 LOST IN TRANSMISSION

general reader. His wildly popular book Misquoting Jesus, published '
in 2005 and based on his 1993 scholarly tome Orthodox Corruption of
Scripture, has had a huge impact among nonscholars.
Normally, Ehrman writes in a clear, forceful style and punctuates his
writing with provocative one-liners and a good measure of wit. I must
confess, however, that his Misquoting Jesus left me more perplexed than
ever. I was not sure exactly what he was saying. Read one way, the book
contradicted what he had written elsewhere; read another way, it was
hardly controversial—and certainly not the sort of book that would
warrant being a blockbuster on the New York Times Bestseller List. So
I acknowledge that I am not sure what all the points of disagreement
between us are. But I do know some.
Whatever our disagreements, there are issues on which we agree.
There is often a gulf between those “inside” a particular scholarly dis­
cipline and those on the outside. When outsiders hear what insiders
are talking about, sometimes they can get quite alarmed. Ehrman says
in the appendix to Misquoting Jesus, “The facts that I explain about
the New Testament in Misquoting Jesus are not at all 'news’ to biblical
scholars. They are what scholars have known, and said, for many, many
years."2 He is right. We do walk on common ground.
There are basically five things that we agree on:

1. The handwritten copies of the NT contain a lot of differences.

We are not sure exactly what the number is, but the best estimate is
somewhere between 300,000 and 400,000 variants. This means, as
Ehrman is fond of saying, that there are more variants in the MSS than
there are words in the NT.

2. The vast bulk of these differences affect virtually nothing.

3. We concur on the wording of the original text almost all the

time.3

2. Bart D. Ehrman, Misquoting Jesus: The Story behind Who Changed the Bible and
Why, 1st paperback ed. (New York; HarperOne, 2007), 253. All quotations are from
the paperback edition.
3. In the appendix to Misquoting Jesus, Ehrman says about his disagreements with
Bruce Metzger, whom Ehrman described as his doctor-father and to whom he had
dedicated Misquoting Jesus: “[EJven though we may disagree on important reli­
gious questions—he is a firmly committed Christian and I am not—we are in com­
plete agreement on a number of very important historical and textual questions.
If he and I were put in a room and asked to hammer out a consensus statement
on what we think the original text of the New Testament probably looked like,
there would be very few points of disagreement—maybe one or two dozen places
out of many thousands” (252). My views of the original text of the NT would tend
to be closer to Metzger’s but would certainly not line up there all the time (e.g,, I
agree with Ehrman that Jesus was angry in Mark 1.41). My point is that those who

2.0 / Chapter 1 Daniel B, Wallace

 How Badly Did ths Scribes Corrupt the New Testament Text?

4. We are even in harmony over several well-known or controver­

sial passages:

Mark 16.9-20— Here Jesus tells his disciples that they can drink
poison and handle snakes and not get hurt. I agree with Ehrman
that this passage is not part of the original text of Mark.

John 7.53-8.11—We both agree that the story of the woman caught
in adultery was not part of the original text of John. It is my favorite
passage that is not in tire Bible.

1 John 5.7—The King James Bible says, “For there are three that bear
record in heaven, the Father, the Word, and the Holy Ghost: and these
three are one.” This would be the most explicit statement about the
Trinity in the Bible, but it is definitely not part of the original text. This
feet has been known by scholars for more than half a millennium.

Mark 1.41—Although most MSS here say that Jesus was moved with
compassion when he healed a leper, we both agree that the original
text probably said that he was angry when he did so. One of Ehrman’s
finest pieces is his provocatively titled article “A Leper in the Hands
of an Angry Jesus.”4 He there gave very strong evidence that Mark
1.41 spoke of Jesus’ anger rather than his compassion. I agree.

5. We both are of the same mind that the orthodox scribes oc­
casionally changed the NT text to bring it more into conformity with
their views.

All these agreements raise a fundamental point: even though we are

looking at the same textual problems and arriving at the same answers
most of the time, conservatives are still conservative, and liberals are
still liberal.
What is the issue then? The text is not the basic area of our dis­
agreement; the interpretation of the text is. Even here, it is not so much
the interpretation of the text5 as the interpretation of how the textual
variants arose and how significant those variants are—that is where

are “reasoned eclectics" (the dominant school of thought in NT textual criticism)

would agree with each other On the vast majority of textual problems.
4. Bart D. Ehrman, "A Leper in the Hands of an Angry Jesus," in New Testament Greek
and Exegesis: Essays in Honor of Gerald F. Hawthorne (Grand Rapids: Eerdmans,
2003), 77-98.
5. To be sure, Ehrman would regard the NT books to be in contradiction with each
other over many issues, while I would regard the basic beliefs of the authors to be
in general agreement. The larger issue is whether what the NT essentially teaches
is true. .Here, the bedrock of all three branches of Christendom is that Jesus Christ,

Daniel B. Wallace Chapter 1/21

 LOST IN TRANSMISSION

our differences lie. Ehrman puts a certain spin on the data.6 If you have
read Misquoting Jesus, you may have come away with an impression of
the book that is far more cynical than what Ehrman is explicitly saying.
Whether that impression accurately reflects Ehrman’s views is more dif­
ficult to assess, but one thing is clear: Ehrman sees in the textual vari­
ants something more pernicious, more sinister, more conspiratorial,
and therefore more controlled than I do. This chapter—indeed, this
book—offers a different narrative.

TWO ATTITUDES, THREE QUESTIONS

To begin with, there are two attitudes that we should try to avoid:
absolute certainty and total despair. On tire one side are King James Only
advocates: they are absolutely certain that the I<JV, in every place, exactly
represents the original text. To be frank, the quest for certainty often
overshadows the quest for truth in conservative theological circles and is
a temptation that we need to resist. It is fundamentally the temptation of
modernism. To our shame, evangelicals have too often been more con­
cerned to protect our presuppositions than to pursue truth at all costs.
On the other side are a few radical scholars who are so skeptical that
no piece of data, no hard fact, is safe in their hands. It all turns to putty
because all views are created equal. If everything is equally possible,
then no view is more probable than any other view. In Starbucks and
on the street, in college classrooms and on the airwaves, you can hear
the line “We really don’t know what the NT originally said since we no
longer possess the originals and since there could have been tremen­
dous tampering with the text before our existing copies were produced.”
But are any biblical scholars this skeptical? Robert Funk, the head of
the Jesus Seminar, seemed to be. In The Five Gospels he said,

Even careful copyists make mistakes, as every proofreader knows. So

we will never be able to claim certain knowledge of exactly what the
original text of any biblical writing was.7

the theanthropic person, died for humanity’s sin on a Roman cross outside of Jeru­
salem and later rose from the dead.
6. Although Ehrman speaks frequently about the hopelessness of the task of trying
to recover the wording of the original text or about the irrelevance of such a task
(two views that seem mutually exclusive), he also speaks about what he and Bruce
Metzger “think the original text of the New Testament probably looked like" (Mis­
quoting Jesus, 252; italics added), as though the task were neither hopeless nor
irrelevant.
7. Robert W. Funk, Roy W. Hoover; and the Jesus Seminar; The Five Gospels: The
Search for the Authentic Words ofJesus (New York: Macmillan, 1993), 6 (italics
added).

22 / Chapter 1 Daniel B, Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

The temporal gap that separates Jesus from the first surviving copies
of the gospels—about one hundred and seventy-five years—corre­
sponds to the lapse in time from 1776—the writing of the Declaration
of Independence—to 1950. What if the oldest copies of the founding
document dated only from 1950?8

Funk’s attitude is easy to see: rampant skepticism over recovering

the original wording of any part of the NT. This is the temptation of
postmodernism.9 The only certainty is uncertainty itself It is the one
absolute that denies all the others. Concomitant with this is an intel­
lectual pride—pride that one "knows” enough to be skeptical about all
positions.
Where does Ehrman stand on this spectrum? I do not know. On the
one hand, he has said such things as the following:

If the primary purpose of this discipline is to get back to the original

text, we may as well admit either defeat or victory, depending on how
one chooses to look at it, because we’re not going to get much closer
to the original text than we already are.
... [A]t this stage, our work on the original amounts to little more
than tinkering. There's something about historical scholarship that re­
fuses to concede that a major task has been accomplished, but there
it is.10

8. Ibid.
9. Those whose writings are very influential in the marketplace of ideas but who are
not biblical scholars make even more unguarded statements. For example, Earl
Doherty declared in Challenging the Verdict (Ottawa: Age of Reason, 2001), “Even
if we had more extensive copies of the Gospels from within a couple of genera­
tions of their writing, this would not establish the state of the originals, nor how
much evolution they had undergone within those first two or three generations. It
is precisely at the earliest phase of a sect's development that the greatest mutation
of ideas takes place, and with it the state of the writings which reflect the mutation”
(39)._......
10. B art D. Ehrman, “Novum Testamentum Graecum Editio Critica Maior·. An Evalu­
ation," TC: A Journal ofBiblical Textual Criticism (1998), revision of a paper pre­
sented at the Textual Criticism Section of the 1997 Society of Biblical Literature
conference in San Francisco. He goes on to argue (in point 20 of his review), “We
can still make small adjustments in the text in places—change the position of an
adverb here, add an article there—we can still dispute the well known textual
problems on which we're never going to be agreed, piling up the evidence as we
will. But the reality is that we are unlikely to discover radically new problems or
devise radically new solutions; at this stage, our work on the original amounts
to little more than tinkering. There’s something about historical scholarship
that refuses to concede that a major task has been accomplished, but there it
is.” This sounds, for the most part, as though he thinks the primary task of tex­
tual criticism—that of recovering the wording of the autographic text—has been
accomplished.

Daniel B. Wallace Chapter 1/23

 LOST IN TRANSMISSION

In spite of these remarkable [textual] differences, scholars are con­

vinced that we can reconstruct the original words of the New
Testament with reasonable (although probably not 100 percent)
accuracy.11

The first statements were made at the Society of Biblical Literature

in an address to text-critical scholars. The last is in a college textbook.
All of this sounds as if Ehrman would align himself more with those
who are fairly sure about what the wording of the autographic text is.
But here is what Ehrman wrote in his immensely popular book
Misquoting Jesus:

Not only do we not have the originals, we don’t have the first copies of
the originals. We don’t even have copies of the copies of the originals,
or copies of the copies of the copies of the originals. What we have are
copies made later—much later.... And these copies all differ from one
another, in many thousands of places.... [TJhese copies differ from
one another in so many places that we don’t even known how many
differences there are.12

We could go on nearly forever talking about specific places in which

the texts of the New Testament came to be changed, either acciden­
tally or intentionally.... [T]he examples are not just in the hundreds
but in the thousands.13

And here is what he wrote in another popular book, Lost Christianities:

The fact that we have thousands of New Testament manuscripts does

not in itself mean that we can rest assured that we know what the
original text said. If we have very few early copies—in fact, scarcely

11. Bart Ehrman, The New Testament: A Historical Introduction to the Early Christian
Writings, 3rd ed. (New York: Oxford University Press, 2003), 481. All quotations
are from this edition.
12. Misquoting Jesus, 10.
13. Ibid., 98. Elsewhere Ehrman says, “Given the problems, how can we hope to get
back to anything like the original text, the text that an author actually wrote? It is
an enormous problem. In fact, it is such an enormous problem that a number of
textual critics have started to claim that we may as well suspend any discussion
of the ‘original’ text, because it is inaccessible to us. That may be going too far”
(58); “In short, it is a very complicated business talking about the ‘original’ text of
Galatians. We don’t have it. The best we can do is get back to an early stage of its
transmission, and simply hope that what we reconstruct about the copies made at
that stage—based on the copies that happen to survive (in increasing numbers as
we move into the Middle Ages)—reasonably reflects what Paul himself actually
wrote, or at least what he intended to write when he dictated the letter” (58; italics
added).

24 / Chapter 1 Darnel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

any—how can we know that the text was not changed significantly
before the New Testament began to be reproduced in such large
quantities?14

The cumulative effect of these latter statements seems to be not only

that we have no certainty about the wording of the original but that,
even where we are sure of the wording, the core theology is not nearly
as “orthodox” as we had thought. According to this line of thinking,
the message of whole books has been corrupted in the hands of the
scribes; and the church, in later centuries, adopted the doctrine of the
winners—those who corrupted the text and conformed it to their own
notion of orthodoxy.
So you can see my dilemma. I am not sure what Ehrman believes. Is
the task done? Have we essentially recovered the wording of the original
text? Or should we be hyperskeptical about the whole enterprise? It seems
that Ehrman puts a far more skeptical spin on things when speaking in
the public square than he does when speaking to professional colleagues.15

14. Bart Ehrman, Lost Christianities: The Battlesfor Scripture and the Faiths We Never
Knew (New York: Oxford University Press, 2005), 219.
15. Part of the evidence for this is what he says in interviews. In one posted on Sep­
tember 25,2006, on the Evangelical Textual Criticism, website (http://evangelical­
textualcriticism.blogspot.com/), he was asked by host P. J. Williams, “Do you think
that anyone might ever come away from reading MisquotingJesus with the impres­
sion that the state of the New Testament text is worse than it really is?” Ehrman
responded, “Yes I think this is a real danger, and it is the aspect of the book that has
apparently upset our modern day apologists who are concerned to make sure that
no one thinks anything negative about the holy Bible. On the other hand, if people
misread my book—I can’t really control that very well.” The cynicism and implicit
condemnation of apologists, coupled with a denial of his own radical skepticism
about the original text, clearly suggests that Ehrman feels that he has not contrib­
uted to this fidse impression. Further, in his final chapter of Misquoting Jesus, Eh-
rinan declares, “The reality, I came to see, is that meaning is not inherent and texts
do not speak for themselves” (216). But if he really believed this, would he have the
right to complain about how people are reading his books?
The reality seems to be that Ehrman has had the opportunity to alter such a
.. .... felscompression in his many radio, TV, and newspaper interviews. But instead of
tempering the misimpression, he usually feeds it. For example, in an interview in
the Charlotte Observer (Dec. 17, 2005)—nine months before his interview by P. J.
Williams·—he said, “When I talk about the hundreds and thousands of differences,
it’s true that a lot are insignificant. But it’s also true that a lot are highly significant
for interpreting the Bible. Depending on which manuscript you read, the meaning
is changed significantly.” No quantitative distinction is made between insignificant
variants and significant variants; both are said to be “a lot” But a qualitative distinc­
tion is made: “a lot are insignificant,” while “a lot are highly significant." Further, in
many of his interviews, he leads off with what appears to have a calculated shock
value, viz., denial of the authenticity of the pericope adulterae.
One other comparison can be made: Both Ehrman and KJVers have a major
point in agreement. They both view the early scribes as having almost a conspirato­
rial motive behind them. (Webster's defines the word conspire in three ways: “1 a: to

Daniel B. Wallace Chapter 1/25

 LOST IN TRANSMISSION

These two attitudes—total despair and absolute certainty—are the

Scylla and Charybdis that we must steer between. There are also three
questions that we need to answer:

1. What is the number of variants—how many scribal changes are

there?

2. What is the nature of variants—what kinds of textual variations

are there?

3. What theological issues are at stake?

THE NUMBER OF VARIANTS

A textual variant is any place among the MSS in which there is varia­
tion in wording, including word order, omission or addition of words,
even spelling differences. The most trivial changes count, and even when
all the manuscripts except one say one thing, that lone MS’s reading
counts as a textual variant.16 The best estimate is that there are between
300,000 and 400,000 textual variants among the manuscripts. Yet there
are only about 140,000 words in the NT. That means that there is an
average of between two and three variants for every word in the Greek
NT. If this were the only piece of data, it would discourage anyone from
attempting to recover the wording of the original. But there is more to
this story.

join in a secret agreement to do an unlawful or wrongful act or an act which be­

comes unlawful as a result of the secret agreement «accused of conspiring to over­
throw the governmental;] 1 b: scheme!;] 2 to act in harmony toward a common
end.” Ehrman does not necessarily see what the proto-orthodox scribes did as a
"secret agreement!’ but he certainly sees them as doing more than acting in har­
mony toward a common end. And if what became the orthodox view started out
in a minority camp struggling for survival, then the fact of the changes the scribes
made could certainly not be made public.) The basic difference is that KJ Vers think
that heretics corrupted the text, while Ehrman thinks that orthodox scribes did. (Of
course, Ehrman is not adamantly against the early Alexandrian manuscripts. But it
does seem that his overriding criterion for determining the wording of the original
[as seen in Orthodox Corruption] is that if a reading even gives off a feint scent of
perhaps being an orthodox corruption, that trumps all other considerations, both
external and internal. In addition to my discussion later in this chapter, see Philip
Miller's “The Least Orthodox Reading Is to be Preferred: A New Canon for New
Testament Textual Criticism,’’ also in this book.)
16. On a popular level among evangelicals, a textual variant is erroneously defined in
terms of the number of MSS that support a reading rather than the number oftextual
differences regardless of MS support For a discussion of this erroneous definition,
see D. B. Wallace, “The Number of Textual Variants: An Evangelical Miscalculation,’'
at http://bible.org/article/number-textual-variants-evangelicaJ-miscalculatiOD, ac­
cessed September 3,2011..

26 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

There are two points to ponder: First, the reason we have a lot of
variants is that we have a lot ofmanuscripts. It is simple, really. No clas­
sical Greek or Latin text has nearly as many variants, because they do
not have nearly as many manuscripts. With virtually every new manu­
script discovery, new variants are found?7 If there were only one copy
of the NT in existence, it would have zero variants?® Several ancient au­
thors have only one copy of their writings in existence, and sometimes
that lone copy is not produced for a millennium. A Ione, late manu­
script would hardly give us confidence that that single manuscript du­
plicated the wording of die original in every respect?9 To speak about
the number of variants without also speaking about the number of
manuscripts is simply an appeal to sensationalism.17 19
18
20
Second, as Samuel Clemens said, "There are lies, damn lies, and sta­
tistics.” A little probing into these 400,000 variants puts these statistics
in a context.
In Greek alone, there are more than 5,600 manuscripts today.21
Many of these are fragmentary, especially die older ones, but the av­

17. For example, the recently cataloged Codex 2882, a MS of Luke from the tenth or
eleventh century, has 29 singular readings not found in any other manuscripts. Yet
the manuscript is, for the most part, an ordinary Byzantine manuscript, and none
of the singular readings are even remotely compelling.
18. Of course, if a Ione MS also had corrections, those readings would count as
variants.
19. This was recognized three hundred years ago by the brilliant Richard Bentley in
his work Remarks upon a Discourse of Free Thinking (London: J. Morphew and
E. Curll, 1713; 8th ed., London; Knaptons, Manby, and Beecroft, 1743), 349 (the
eighth edition, quoted here, was published a year after Bentley’s death, with addi­
tions from his manuscript):

If there had been but one manuscript of the Greek Testament at the restoration of
learning about two centuries ago, then we had had no various readings at all....
And would the text be in a better condition then, than now [that] we have 30,000
[variant readings]?
It is good, therefore... to have more anchors than one; and another MS. to
join the first would give more authority, as well as security.

Bentley’s discussion that the number of variants among NT MSS gives us more as­
surance, not less, concerning the wording of the original occupies a major section
in this book (347-64).
20. Although Ehrman does both in Misquoting Jesus, he seems to emphasize the
former far more than the latter. As NT professor Craig Blomberg observes, “What
most distinguishes the work are the spins Ehrman puts on some of the data at
numerous junctures and his propensity for focusing on the most drastic of all the
changes in the history of the text, leaving the uninitiated likely to think there are
numerous additional examples of various phenomena he discusses when there are
not” {Denver Journal 9 [2006]; accessed online).
21. Eldon Jay Epp, "Are Early New Testament Manuscripts Truly Abundant?,” in Israel’s
God and Rebecca’s Children; Christo logy and Community in Early Judaism and
Christianity; Essays in Honor ofLarry W. Hurtado and Alan F. Segal, ed. David B.

Daniel B. Wallace Chapter 1 / 27

 LOST IN TRANSMISSION

erage Greek NT MS is over 450 pages long. Altogether, there are more
than. 2.6 million pages of texts, leaving hundreds of witnesses for every
book of the NT.
It is not just the Greek MSS that count, either. Early on, the NT was
translated into a variety of languages—Latin, Coptic, Syriac, Georgian,
Gothic, Ethiopic, Armenian. There are more than 10,000 Latin MSS
alone. No one really knows the total number of all these ancient ver­
sions, but the best estimates are close to 5,000—plus the 10,000 in
Latin.22 It would be safe to say that we have altogether about 20,000
handwritten manuscripts of die NT in various languages, including
Greek.
If someone were to destroy all those manuscripts, we would not be
left without a witness, because the church fathers wrote commentaries
on the NT. To date, more than one million quotations of the NT by the
fathers have been recorded. "[I]f all other sources for our knowledge of
the text of the New Testament were destroyed, [the patristic quotations]
would be sufficient alone for the reconstruction of practically the entire
New Testament,” wrote Bruce Metzger and Bart Ehrman.23
These numbers are breathtaking But if left by themselves, they also
resemble Samuel Clemens’s quip about statistics. Far more important
than the numbers is the date of the MSS. How many manuscripts do we
have in the first century after the completion of the NT, in the second,
and in the third? Although the numbers are significantly lower for the
early centuries, they are still rather impressive. Today we have as many
as 12 MSS from the second century, 64 from the third, and 48 from the

Capes, April D. DeConick, Helen K. Bond, and Troy A. Miller (Waco, TX: Baylor
University Press, 2007), 77-107, notes on 395-99. Epp notes that the official tally is
5,752 manuscripts but that this "count of manuscripts... cannot simply be looked
up, because duplications and items misplaced in the lists must be eliminated” (78).
This brings his numbers to 5,494. However, in die summer of 2007, the Center
for the Study of New Testament Manuscripts (see www.csntm.org) sent a team
to Albania to photograph Greek NT manuscripts housed in the National Archive
in Tirana. In die process, at least 24 manuscripts and as many as 30 were photo­
graphed for the first time—MSS that had not been cataloged by the Institutfiir
neutestamentliche Textforschung in Munster, Germany (the differences being that
some of these MSS may have been among the 17 that had been presumed lost for
decades. In the summer of2008, before these MSS had been tabulated by the Mun­
ster institute, Ulrich Schmid of the INTF informed me that the number of extant
MSS now stood at 5,555. As of May 2011, CSNTM had discovered more than 75
Greek NT MSS. Thus the number of extant MSS will soon officially stand at over
5,600.
22. Curiously, Epp claimed as recently as 2007 that there were about “10,000 versional
manuscripts” (Epp, “Are Early New Testament Manuscripts Truly Abundant?,” 77)
although the number of Latin MSS alone reaches that number.
23. Bruce M. Metzger and Bart D. Ehrman, The Text ofthe New Testament: Its Trans­
mission, Corruption, and Restoration, 4th ed. (New York: Oxford University Press,
2005), 126.

28 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt die New Testament Text?

fourth—a total of 124 MSS within 300 years of the composition of the
NT. Most of these are fragmentary, but the whole NT text is found in
this collection multiple times.2425
How does the average Greek or Latin author stack up? If we are
comparing the same time period—300 years after composition—the
average classical author has no literary remains. But if we compare all
the MSS of a particular classical author, regardless of when they were
written, the total would still average at least less than 20 and probably
less than a dozen—and they would all be coming much more than three
centuries later. In terms of extant MSS, the NT textual critic is con­
fronted with an embarrassment of riches. If we have doubts about what
the autographic NT said, those doubts would have to be multiplied a
hundredfold? for the average classical author.26 When we compare the

24. Epp's numbers are slightly different: 11 MSS from the second and third centuries,
52 from die third and fourth, and 48 from die fourth and fifth, for a total of 111
MSS within 300 years of the completion of the NT (“Are Early New Testament
Manuscripts Truly Abundant?," table 6.2, p. 80). The different counts are due to (1)
new discoveries since 2006 (when he tabulated his data) and (2) some MSS that he
did not consider or perhaps overlooked. (Bagnall has recently offered a different
view on the date of our earliest NT MSS, considering them much later than almost
all other textual critics and paleographers have [Roger S. Bagnall, Early Christian
Books in Egypt [Princeton, NJ: Princeton University Press, 2009].) What is most
telling in his essay is that lie hardly speaks of the value of these numbers in rela­
tion to other ancient literature. It is remarkable that in an article intended to show
how sparse the early data for the NT are (e.g., "If the early manuscripts are most
valuable, what value and how much abundance do we have in the mere eleven
manuscripts that have survived from the period up to and around 200 C.E.? At that
point, Christianity had been in existence for two hundred years!" [88]), the author
makes no comparison with other ancient Greco-Roman literature. Without such
a comparison, there is no way to tell how relatively abundant the early NT MSS
are. Further, nowhere does he discuss how much text of the NT was found in these
early MSS.
25. Technically, our doubts should be a thousandfold if there are 20,000 NT MSS (in
Greek and die ancient versions) and less than 20 MSS of an average classical author.
However, if one were only to look at die more ancient copies of such documents,
the ratios could be closer. A hundredfold is an amount that is hard to dispute.
26. Note that I did not say. that we have no doubts about the autographs of these other
ancient writers. But far greater skepticism toward the NT is shown than its MS
testimony would warrant. Further, it is curious that Ehrman can sound so skep­
tical of the wording of the original NT in MisquotingJesus when a part of his basis
for such skepticism is certainty about what some ancient writers said. Part of his
argument against the reliability of NT MSS is his assumption of accuracy of what
certain ancient writers’ texts read, even though we have to do textual criticism on
their extant MSS to try to reconstruct what they wrote. He enlists Seneca (46),
Martial (47), Hennas (48), Irenaeus (53), Dionysius (53), and Ruffnus (54). Most
significantly, he discusses Origen's quotations of Celsus, an antagonist to die Chris­
tian faith who wrote about seventy years before Origen did, with the tacit assump­
tion that die copies of Origen that we have accurately reflect what Origen wrote
and that Origen accurately recorded what Celsus wrote, even though seventy years
separated the two men. If we had the original text of Origen, we would still be

Daniel B. Wallace Chapter 1 / 29

 LOST IN TRANSMISSION

NT MSS to the very best that the classical world has to offer,27 the NT
MSS still stand high above the rest. The NT is by far the best-attested
work of Greek or Latin literature from the ancient world.
There is another way to look at this. If all of the NT MSS that are
definitely or possibly dated to the second century are fragmentary—and
they are—how fragmentary are they? We can measure this in different
ways. First, three out of the four Gospels are attested in the MSS, as
well as nine of Paul's letters, Acts, Hebrews, and Revelation—in other
words, most of the NT books. Another way to look at this is that over
40 percent of all the verses in the NT are already found in MSS within
150 years of the completion of the NT.28
Ehrman seems to say in a couple places that we do not have any
second-century MSS.29 He declared in an interview in the Charlotte
Observer,30 “If we don’t have the original texts of the New Testament—
or even copies of the copies of the copies of the originals—what do we
have?” His response is illuminating: “We have copies that were made
hundreds of years later—in most cases, many hundreds of years later.
And these copies are all different from one another?31 He is saying that
we do not have any manuscripts of the NT until hundreds of years
after the NT was completed. He repeated this statement at the 2008
Greer-Heard Forum debate and, more recently, on the TC-List, an

dealing with a seventy-year gap after Ceisus. But when a similar gap occurs for the
NT MSS, Ehrman says, "We don’t even have copies of the copies of the originals,
or copies of the copies of the copies of the originals. What we have are copies made
later—much later" (Misquoting Jesus, 10); "If we have very few early copies—in
fact, scarcely any—how can we know that the text was not changed significantly
before the New Testament began to be reproduced in such large quantities?" (Lost
Christianities, 219).
27. Homer's Iliad has just over 2,200 extant MSS, while his Odyssey has 141 (Martin
L. West, Homeri Illas, vol 1, Rhapsodias Ι-ΧΠ Continent, Bibliotheca Scriptorum
Graecorum et Romanorum Teubneriana [Stuttgart In Aedibus B. G. Teubneri,
1998], xxxviii-liv; Victor Berard, LOdyssee: Poesie Homerique, vol. 1, Chants
l-VII, Collection des Universites de France [Paris: Societe D’Edition Les Belles
Lettres, 1924], xxxvi-xxxix. The data on the Odyssey, however need to be up­
dated). Nothing in the ancient Greco-Roman world comes dose to this—except, of
course, for the NT and some patristic writers such as Chrysostom. Not counting
patristic citations of the NT, there are still almost ten times as many copies of the
NT as there are Homeric MSS extant today.
28. The specific number is 43 percent. But this does not necessarily mean that every
portion of each of these verses is in these MSS. Thanks are due to Brett Williams
for doing the painstaking work of tabulating the number of verses that are found in
the definite and possible second-century manuscripts.
29. He does not say this in any of his books, as far as I am aware, but he does discount
the number of early MSS by noting their fragmentary character. Cf. Lost Christi­
anities, 219.
30. Dec. 17, 2005.
31. Italics added.

30 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?
Λ

international Internet discussion group of biblical textual critics.32 But

this is demonstrably not true. The impression Ehrman sometimes gives
throughout Misquoting Jesus-—box. especially repeats in interviews—is
that of wholesale uncertainty about the original wording, a view that is
probably far more radical than the one he actually embraces.
In light of comments such as these, the impression that many readers
get from Misquoting Jesus is that the transmission of the NT resembles
the “telephone game.”33 This is a game every child knows. It involves a
line of people, with the first one whispering some story into the ear of the
second person. That person then whispers the story to the next person in
line, and that person whispers it to the next, and so on down the line. As
the tale goes from person to person, it gets terribly garbled. The whole
point of the telephone game, in fact, is to see how garbled it can get. There
is no motivation to “get it right? By the time it gets to the last person, who
repeats it aloud for the whole group, everyone has a good laugh.
But the copying of NT manuscripts is hardly like this parlor game,
for many reasons:

1. The message is passed on in writing, not orally. Passing a mes­

sage in writing would make for a pretty boring telephone game!

2. Rather than having one line, there are multiple lines or streams
of transmission.34

32. “[W]e don’t know how much the texts got changed in all those decades/centuries
before our earliest manuscripts, and we have no way of knowing” (TC-List, Nov. 1,
2008; italics added to “centuries” and “earliest”).
33. Ute impression that the transmission of NT manuscripts is like the telephone
game is reinforced by Ehrman's almost complete lack of discussion of early ver­
sions, patristic comments, or the relation between manuscripts that surely brings
us back much earlier than our extant witnesses can do. He knows better than this,
as is revealed by a perusal of Metzger and Ehrman’s Text ofthe New Testament.
34. Elsewhere Ehrman seems to argue that Christians simply destroyed the original
documents “for some unknown reason.” In his discussion of manuscript produc­
tion in Lost Christianities, he says, "In this process of recopying the document
. - ~ .b,y_hand.1jvhathappened to the original of 1 Thessalonians? For some unknown
reason, it was eventually thrown away or burned or otherwise destroyed. Possibly
it was read so much that it simply wore out. The early Christians saw no need to
preserve it as the ‘original’ text. They had copies of the letter. Why keep the orig­
inal?” (217). He is here presupposing two unlikely attitudes: first, that early Chris­
tians just did not care about original documents at all; second, that early Christians
made only one copy of the original text of a NT book.
Kirsopp Lake, Robert P. Blake, and Silva New ("The Caesarean Text of the
Gospel of Mark,” HTR 21 [1928J: 348-49) argued that the scribes in the monas­
teries at Sinai, Patmos, and Jerusalem probably destroyed their exemplars. This
opinion was just one way to make sense of die fact dial most of the manuscripts
in tiiese monasteries were “orphan children without brothers or sisters.” There is,
however, no evidence that early Christians destroyed their own revered manu­
scripts. The only evidence we have of MS destruction by Christians, as far as 1 am

Daniel B. Wallace Chapter 1/31

 LOST IN TRANSMISSION

3. Textual critics do not rely on just the last person in each line but
can interrogate several folks who are closer to the original source.

aware, is of destruction of documents that were viewed as heretical. We have two

statements, both from the fifth century, to this effect. Theodoret of Cyrrhus seems
to imply that he destroyed two hundred copies of Tatian’s Diatessaron {Compen­
dium haereticarum fabularum 1.20 [MPG 83, cols. 369, 372]), though this is not
exactly stated. Rabbula of Edessa, writing in the early fifth century, explicitly says
that die writings of the heretics needed to be destroyed (“Commands and Admoni­
tions... to the Priests and the Benai Qeiama” in Syriac and Arabic Documents re­
garding Legislation Relative to Syrian Asceticism, ed. Arthur Voobus [Stockholm:
Etse, 1960], 48). Though there is no reason to think that his orders were not carried
out, there is again no statement of fact about such destruction. But the fact that
both of these references are from the fifth century (well after heresy and orthodoxy
had been clearly defined) and that copies of the NT per se were not among the
documents destroyed leaves us guessing as to whether earlier groups of Christians
actually did this sort of thing.
Further, there is some evidence that the NT autographa were revered in the
second century. Tertullian, a church father living up through the first quarter of
the third century, chastised his theological opponents about their doubts over
what the original text said. The exact meaning of his statement is somewhat con­
troversial: “Come now, you who would indulge a better curiosity, if you would
apply it to the business of your salvation, run over to the apostolic churches, in

which their own authentic writings are read, uttering the voice and representing
the face of each of them severally’1 (Tertullian, The Prescription against Heretics,
chap. 36; italics added). What is at issue here is the meaning of “authentic" writ­
ings. If this refers to the original documents, as the word in Latin (authenticae)
normally does, Tertullian is saying that several of the original NT books still ex­
isted in his day, well over a century after the time of their writing. He specifically
refers to Paul’s letters sent to Corinth, Philippi, Thessalonica, Ephesus, and Rome,
urging his reader to visit these sites to check out these authentic writings. But if
authenticae does not mean original documents, it would at least mean, in this
context, carefully produced copies. Most likely, the term is here referring to the
originals, but Tertullian’s testimony may not be credible. However, by Tertullian’s
day, carefully done copies of the originals both were considered important for
verifying what the NT authors wrote and may have still been available for con­
sultation. Even taking the worst-case scenario, Tertullian’s statement is an early
documented concern about having the original text or at least accurate copies in
circulation, rather than quietly put on the shelf never to be consulted again. (See
George Houston, “Papyrological Evidence for Book Collections and Libraries in
the Roman Empire,” in Ancient Literacies: The Culture of Reading in Greece and
Rome, ed. William A. Johnson and Holt N. Parker (Oxford: Oxford University
Press, 2009), 233-67, for evidence of the longevity of well-used papyrus MSS in
the ancient world. Houston points out that literary documents were in use an
average of two hundred or more years (248-51).
Regarding the second assumption—that the original text was copied only
once (Ehrman says that it may have worn out from being read, but not from being
copied)—surely it was copied often if it was read often. To suppose that the early
Christians just somehow forgot about the originals or, worse, destroyed them is
contrary to human nature, to at least one early patristic writer’s testimony, and to
all the evidence we have from the first several centuries of the Christian era.

32 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

4. Patristic writers are commenting on the text as it is going

through its transmissional history. And when there are chronological
gaps among the manuscripts, these writers often fill in those gaps by
telling us what the text said in that place in their day.

5. In the telephone game, once the story is told by one person,

that individual has nothing else to do with the story. It is out of his or
her hands. But tire original NT books were most likely copied more
than once and may have been consulted even after a few generations of
copies had already been produced.35

6. There was at least one very carefully produced stream of trans­

mission for the NT MSS, and there is sufficient evidence to show that
even a particular fourth-century MS in this line is usually more accurate
than any second-century MS.

We can illustrate this with two manuscripts that Ehrman and I

would both agree are two of the most accurate MSS of tire NT, if not
the two most accurate: ?>7S and Codex Vaticanus (or B). These two MSS
have an incredibly strong agreement. Their agreement is higher than
the agreement of any other two early MSS. ϊ>75 is 100 to 150 years older
than B, yet it is not an ancestor ofB. Instead, B copied from an earlier
common ancestor that both B and S375 were related to.36 The combina­
tion of both of these manuscripts in a particular reading goes back to
early in the second century.37

35. Ehrman has suggested that as the copying was done, discrepancies would be noted.
At the Greer-Heard Forum, in fact, he painted a scenario in which early copies of
Mark would have been compared to earlier copies to clear up these discrepancies.
If these discrepancies were large enough from the earliest period of copying, would
it not seem lilcely that the autographs would have been consulted?
36. See C. L. Porter, “Papyrus Bodmer XV (ip75) and the Text of Codex Vaticanus,” JBL
81 (1962): 363-76; Porter, ‘An Evaluation of the Textual Variation between Pap75
■·- and-Codex- Vaticanus in the Text of John,” in Studies in the History and Text of the
New Testament in Honor ofKenneth Willis Clark, ed. Boyd L. Daniels and M. Jack
Suggs, Studies and Documents 29 (Salt Lake Qty: University of Utah Press, 1967),
71-80, Gordon D. Fee, “φ7ί, and Origen: The Myth of Early Textual Recen­
sion in Alexandria," in Eldon J. Epp and Gordon D. Fee, Studies in the Theory and
Method ofNew Testament Textual Criticism, ed. Irving Alan Sparks, Studies and
Documents 45 (Grand Rapids: Eerdmans, 1993), 247-73.
37. Fee argues that their agreement goes back to the second century, but he adds that
since ip75 was not a recension but a relatively pure copy in a relatively pure stream
of transmission, Hort had good instincts when he thought of Codex B as repre­
senting “a ‘relatively pure’ line of descent from the original text” (Fee, “Myth of
Early Textual Recension in Alexandria,” 272). Ehrman acknowledges the signifi­
cance of Codex Vaticanus and 1175 in the appendix to MisquotingJesus. Regarding
$75, he says, “This is arguably the most valuable ancient papyrus manuscript of

Daniel B. Wallace Chapter 1/33

 LOST IN TRANSMISSION

Ehrman has asserted, “If we have very few early copies—in fact,
scarcely any—how can we know that the text was not changed signifi­
cantly before the New Testament began to be reproduced in such large
quantities?”* 381 am not sure what large quantities he is speaking about,
since there are more MSS from the third century than there are from
the fourth or fifth century.39
But how can we know? It is a legitimate question. There is a way
to be relatively confident that the text of the fourth century looked re­
markably like the earliest form of the text. g>75 has large portions of Luke
and John in it—and nothing else. Codex B has most of the NT in it. If
B and p75 are very close to each other yet B often has the more primi­
tive reading, we can extrapolate that the text of B is pretty decent for
the rest of the NT. When it agrees with a MS such as Codex Sinaiticus,
which it usually does, that combined reading almost surely goes back to
a common archetype from deep in the second century.40
Nevertheless, Ehrman has carefully and ably described the trans­
mission of the text. He has detailed how the winners succeeded in
conquering all with their views and emerged as the group we might
call “orthodox.
•I< ” What he has said is fairly accurate overall. The only
problem is that his is the right analysis but for the wrong religion.
Ehrman’s basic argument about theological motives describes Islam far
more than Christianity. Recent work on the transmissional history of
both the NT and the Qur’an shows this clearly. Consider the following
points:

the Gospels” (263). Concerning Vaticanus, he says, "This is probably the highest

38. Lost Christianities, 219.

39. The next line, however, suggests that he is speaking about medieval MSS: “Most
surviving copies were made during the Middle Ages, many of them a thousand
years after Paul and his companions had died” (ibid.). The juxtaposition of this sen·
tence with the one questioning whether we can know how significant the changes
were prior to this time is, at best, misleading. Ehrman would acknowledge, as
would most textual critics, that the MSS produced in the Middle Ages are hardly
our most reliable witnesses to the NT text and that we have several sufficient wit­
nesses prior to that time on which to reconstruct the wording of the earliest form
of the text.
40. CL, e.g., Metzger and Ehrman, Text of the New Testament, 277-78,312. Hort be-
lieved that when K and B agreed, their reading went back to a very ancient common
ancestor. That it was not a near ancestor was demonstrated by the thousands of
disagreements between these two manuscripts, suggesting that there were several
intermediaries between the common ancestor and tirese two uncial documents
(B. E Westcott and E J. A. Hort, The New Testament in the Original Greeks vol.
2, Introduction (and) Appendix [Cambridge: Macmillan, 1882], 212-50). Cf. also
Metzger and Ehrman, The Text ofthe New Testament, 312: “With the discovery...
of £66 and £75, both dating from about the end of the second or the beginning of
die third century, proof became now available that Hort's Neutral text goes back to
an archetype that must be put early in die second century?

34 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

1. Within just a few decades of the writing of the Qur'an.» it under­

went a strongly controlled, heavy-handed editing, geared toward “or­
thodoxy,” that weeded out variants that did not conform.
But the NT, as even Ehrman argues, did not suffer this sort of con­
trol early on. instead, Ehrman has often suggested that the earliest de­
cades were marked by free, even wild copying.41

2. Calif Uthman was in charge of the earliest segment of this

heavy-handed editing of the Qur’an. He systematically gathered up any
nonconforming MSS and destroyed them. The originals were destroyed
as well.42 Uthman then claimed that his “canonical" text was the exact
equivalent of the autographs.
There is no real evidence that inexact copies of tire NT were de­
stroyed by ecclesiastical authorities.43 Indeed, there is evidence that
just the opposite took place: defective or deteriorating copies might be
placed in a jar or storage room but not destroyed.44

41. C£, e.g., Ehrman, Misquoting Jesus, 124. I will discuss the nature of the early
copying soon enough, but for now I simply point out that according to Ehrman,
there was extensive uncontrolled copying of the NT in the earliest period.
42. Ehrman opines that perhaps the NT autographs were destroyed. Not only is there
no evidence that this was the case, but there is second-century evidence that the
autographs would have been revered.
43. See nn. 34 and 44 for discussion.
44. Colin H. Roberts (Manuscript, Society, and Belief in Early Christian Egypt
(London: Oxford University Press, 1979], 6-8) gives ample evidence that early
Christians took over the practice of Jews to "dispose of defective, worn-out, or
heretical scriptures by burying them near a cemetery, not to preserve them but be­
cause anything that might contain the name of God might not be destroyed” (ibid.,
7). He was dealing with die earliest period of Christian copying but noted that the
Nag Hammadi MSS (“outside our period”) seem to fit this pattern as well. In ad­
dition, he cited the Dead Sea Scrolls, the Chester Beatty papyri, as well as several
other examples. In more modern times, it is noteworthy to mention the New Finds
manuscripts at St. Catherine’s Monastery of Mt. Sinai. Discovered in 1975, quite
by accident, was a geniza that housed about 1,200 manuscripts and 50,000 frag­
ments of manuscripts. The latest date of any of the MSS was from the eighteenth
----- century;Jheearliest. was the fourth century (about two dozen leaves or fragments
from Codex Sinaiticus). Among the less orthodox MSS were the Protoevangelism
of James and the Assumption of the Virgin. When I visited the monastery in Sep­
tember 2002, Archbishop Damianos expressed surprise to me that the Protoevan­
gelism was among the New Finds manuscripts. I discovered the Assumption of the
Virgin inside the Proto evangelism, occupying a new quire.
What the New Finds illustrate is that the practice of burying MSS at Mt. Sinai
was taking place after the eighteenth century and sufficiently prior to modern times
to have been forgotten by the monks. After Tischendorfs last visit in 1859, the
monastery became increasingly flooded with visitors. This suggests that the geniza
was filled prior to this time. And the fact that leaves from Sinaiticus were buried
there—both from the Pentateuch and from the Apostolic Fathers (i.e., the outer
leaves of the codex, which would be most prone to be loosed from the book)—may
imply that Tischendorf was mistaken when he said that the monks were burning

Daniel B. Wallace Chapter 1/35

 LOST IN TRANSMISSION

3. The closest we come to heavy-handed control for NT MSS

did not occur until at least the ninth century, long after the major
Christological disputes had ended.45 Even then, we do not see defective
MSS getting destroyed.

4. One cannot have it both ways; there cannot be wild copying

by untrained scribes and a proto-orthodox conspiracy simultaneously
producing the same variants. Conspiracy implies control, and wild
copying is anything but controlled.
On the one hand, there was uncontrolled copying of MSS in the
earliest period, but this was largely restricted to the Western text form.46
On the other hand, there was a strand of early copying that may appear
to be controlled. This is the Alexandrian family of MSS. Yet the reason
that MSS of this text form look so much like each other is largely be­
cause they were in a relatively pure line of transmission.47 There was no
conspiracy, just good practices.

5. The reason why Islam has Qur’an MSS that so closely resemble
each other is precisely because tills was official dogma, there was over-
zealous control in the copying of the MSS, and there were severe reper­
cussions to any who erred significantly in their scribal duties. All MSS
ultimately derived from a single copy—a copy that was not identical to
the original text.48

leaves of this codex. For our purposes, it is enough to note that the normative
practice of ancient Christians, even perhaps to modern times, was to bury or hide
sacred texts rather than destroy them.
45. T. J. Ralston (“The Majority Text and Byzantine Texttype Development: The Sig­
nificance of a Non-Parametric Method of Data Analysis for the Exploration of
Manuscript Traditions" [PhD diss., Dallas Seminary, 1994]) notes (in agreement
with von Soden’s assessment) that there was a large editorial push by at least one
scriptorium in the ninth and eleventh centuries, resulting in carefully produced
copies that were very close to each other.
46. It is not entirely insignificant that Ehrman’s preferred reading in several places
that seem to impact Christology is found in the Western text (e.g., Luke 3.22; John
20.28). The burden of proof certainly rests with tire one who would argue that such
a textual tradition has the original wording when the carefully copied tradition
of Alexandria does not. He admits that the Western text is less lilcely to preserve
the best reading when it Jacks support of the Alexandrian witnesses (Misquoting
Jesus, 131). I do agree with Ehrman in at least one Western reading, however. But
opyioSeis in Mark 1.41 has compelling internal evidence in its favor.
47. Ehrman, Misquoting Jesus, 131.
48. This is not at all what the NT transmission was like. See the following discussion
of the work of Uthman in canonizing the Qur’an by starting with his own MS as
the progenitor of all that would follow. Ehrman speculates, without a shred of evi­
dence, that this same phenomenon occurred for NT books: "[Wjhatif only owe of
the copies served as the copy from which all subsequent copies were made... ?”
(Misquoting Jesus, 59).

36 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

Contrast this with the NT: from the earliest times, the NT was
translated into a multitude of languages.49 The transmission of the text
was a growing, living thing, not constrained by ecclesiastical controls
until long after Christianity became legalized. Even then, we know of
nothing like what we see in Islam: scribes not only made plenty of mis­
takes, but they even complained in the margins of their manuscripts
about the weather, the length of the MS they were copying, the clogging
of the ink, and so on.50 This sort of living, hands-on, messy relationship
of the scribes to their holy scriptures is unheard of in Islam. In short,
the Qur’an copying practices were more related to apologetics, while the
NT practices were more related to life.

6. Further, ever since canon was a term meaningfully applied to the

NT, there was never a sense that only the Greek MSS were Scripture. To
be sure, the Reformation sparked a return to the original languages of
the Bible, but the reason was not only purity of the text but clarity in the
proclamation of the message. It is no accident that the Reformers were
the catalyst for the great European translations of the Bible—translations
into the language of the people that could be considered the very Word
of God by the average layman. By way of contrast, the only true Qur’an
is the Arabic Qur’an. All translations are officially suspect. Thus what
Ehrman is describing is right on target but for the wrong religion. He is
describing what has occurred in Islam, not in Christianity.

7. What Westcott said over a century ago is relevant to this

discussion:

When the Caliph Othman fixed a text of the Koran and destroyed all
the old copies which differed from his standard, he provided for the
uniformity of subsequent manuscripts at the cost of their historical

49. Keith Small, a scholar in the United Kingdom who has recently completed his doc­
toral thesis on a comparison of the NT textual transmission and the Qur’an textual
transmission (“Mapping a New Country: Textual Criticism and Qur’an Manu­
scripts” [London School of Theology, 2008]), noted in an email on March 25,2008,
“There was not a program of translation to spread Islam through having people
read tine Qur’an, like there was with the Christian Scriptures. Though one early
jurist, Abu Hanifah (d. AD 767), did rule that a person could recite a vernacular
translation in their prayers, he also is said to have retracted that ruling. The earliest
extant translation I know of is one done into Persian about AD 956 (Ekmeleddin
Ihsanoglu, World Bibliography of Translations of the Meanings of Ute Holy Qur’an,
Printed Translations 1S1S-1980 [Istanbul: Research Centre for Islamic History, Art
and Culture], xxiii)."
50. See Metzger and Ehrman, Text ofthe New Testament, 29, for illustrations. Having
this sort of marginal note in the Qur’an is unheard of. But some of the marginal
notes in the NT MSS are rather impious, showing that the copying was meant
more for die masses than for apologetic reasons.

Daniel B. Wallace Chapter 1/37

 LOST IN TRANSMISSION

foundation. A classical text which rests finally on a single archetype is

that which is open to the most serious suspicions.51

What we see in the NT copies is absolutely nothing like this. Ehrman

tries to make a case for significant theological alterations to the text of
the NT by a group that did not have control over the textfrom the begin­
ning, but the historical ingredients for his hypothesis are missing. It is
like trying to bake a cake with romaine lettuce and ranch dressing.
As Small points out,

The original NT text (the autographic text-form in Epp’s categories)

has been kept remarkably well, and one form of the Qur’an text, a
strongly edited one (a canonical text-form in Epp’s words), has
been preserved remarkably well. This Qur’anic text form (the one
attributed to Uthman though probably a little later—ca. 700 AD)
preserves authentic material, but not in the forms in which it was
originally used or in the complete collection assembled in writing
or orally during Muhammad’s lifetime. Instead, it is a very selective,
heavily edited text. In contrast, the NT is not really the product of
an official process of intentional editing and so preserves more of the
original text within the extant manuscripts. This can be said just on
the basis of Islamic tradition concerning the collections attributed
to have been made by the companions of Muhammad. In the twenty
years after Muhammad’s death until Uthman’s project to standardize
the text, these versions were used extensively in other parts of the
growing Islamic empire, apparently as authoritative scripture. Some
of these are reported to have been in use into the 900’s AD until they
were finally suppressed around 934. My research in the manuscripts
also demonstrates that the majority of the earliest manuscripts
contain this edited text, with the handful of palimpsests pointing
to other textual traditions that were successfully suppressed. These
palimpsests contain the same variety of textual variants that one
can see between the Western and Alexandrian text-types in the
NT tradition—showing that there was a period when the Qur’an
text was more fluid than the majority of manuscripts and Islamic
dogma would lead one to believe. Muslims assume and state that
this Uthmanic text was the original text, though even their traditions
go against the view. It contains original material, but the original
form of that material cannot be reconstructed because Uthman de­
stroyed the autographs and had his authoritative version written in
a defective script which allowed the growth of competing written

51. Brooke Foss Westcott, Some Lessons of the Revised Version of the New Testament,
2nd ed. (London: Hodder and Stoughton, 1897), 8-9 (italics added). Credit is due
to Keith Small for pointing this reference out to me.

38 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

versions and oral recitation systems. Their theological view of me­

chanical inspiration keeps them from adequately engaging with their
own historical sources. What they have done instead is selectively
choose reports that they can use to construct a straight line of “per­
fect” transmission while ignoring the facts which disagree with the
theological construct they want to hold of an eternal book perfectly
transmitted. I think Uthman’s version does probably represent the
main lines of Muhammad’s teaching, though for political reasons
certain parts may have been left out. But we can’t tell for sure be­
cause the autographs were destroyed, not that they wore out in use.
And the main point to get to in all of this is still that the NT and the
Qur’an teach very different things. Also, for whatever integrity one
wants to grant to the transmission of the Qur’an, the NT needs to be
regarded as having more integrity in its transmission process since
there was not such an official editing process after the books were
written. In light of all of this, I think Bart Ehrman’s arguments are
much more appropriate for the Qur’an because for it there can be
demonstrated an official program of textual standardization which
was maintained over three centuries, and in some respects to this
day.52

Concerning conforming the text to the Medieval standard, though

there is a general parallel to this situation to the Qur’an’s, I see it
having a fundamental difference, that while the changes to the NT
were gradual, relatively late in the history of transmission, and pri­
marily for liturgical reasons and to improve the style, the Qur’an’s
form of the consonantal text was determined and maintained from
very early on (within 30-70 years after Muhammad's death) for rea­
sons which had a large ideological/dogmatic component at the outset,
and then that form was further shaped and developed with diacritics
and vowels to maintain and serve various agendas during the next 200
years until the Sunnis came out on top politically in the 900’s and were
able to canonize their version of the text.53
*:x^s? .· .*.**.**^*t*:*^
v:* > v*** +x
In another respect, when Ehrman discusses whether God has
preserved the text of the NT, he places on the NT transmissional
process some rather unrealistic demands—demands that Islam tradi­
tionally claims for itself with respect to the Qur’an but that no bona
fide Christian scholar would ever claim was true of the NT MSS. As is
well known, most Muslims claim that the Qur’an has been transmitted

52. Email from Keith Small, Mar. 11,2008,

53. Email from Keith Small, Dec. 30, 2007. For an excellent survey on tire transmission
of the Qur’an, see now Keith E. Small, Textual Criticism and Qur’an Manuscripts
(Idaho Falls, Idaho: Lexingon, 2011).

Daniel B. Wallace Chapter 1 / 39

 LOST IN TRANSMISSION

perfectly, that all copies are exactly alike. This is what Ehrman demands
of the NT text if God has inspired it. Methodologically, he did not
abandon the evangelical faith; he abandoned a faith that in its biblio-
logical constructs is what most Muslims claim for their sacred text. Or
as C. S. Lewis put it,

The moment [the miracle] enters [nature’s] realm, it obeys all her laws.
Miraculous wine will intoxicate, miraculous conception will lead to
pregnancy, inspired books will suffer all the ordinary processes of tex­
tual corruption, miraculous bread will be digested.54

To sum up the evidence on the number of variants, there are a lot

of variants because there are a lot of manuscripts. Even in the early
centuries, tire text of the NT is found in a sufficient number of MSS,
versions, and writings of the church fathers to give us the essentials
of the original text.

THE NATURE OF THE VARIANTS

How many differences affect the meaning of the text? How many
of them are "plausible” or “viable”—that is, found in manuscripts with
a sufficient pedigree that they have some likelihood of reflecting the
original wording? The variants can be broken down into the following
four categories:

♦ spelling differences and nonsense errors

♦ minor differences that do not affect translation or that involve

synonyms

♦ differences that affect the meaning of the text but are not viable

♦ differences that both affect the meaning of the text and are viable
✓
Of the hundreds of thousands of textual variants in NT MSS, the
great majority are spelling differences that have no bearing on the
meaning of the text.55 The most common textual variant involves what

54. C. S. Lewis, Miracles: A Preliminary Study, 1st Touchstone ed. (New York: Touch­
stone, 1996) 95 (italics added).
55. Even in the earliest form of the text, we see spelling variations by the same author.
Perhaps die most notable example is to be found in John 9.14-21: in the space of
eight verses, the evangelist manages to spell the third-person singular aorist ac­
tive indicative of ανοίγω three different ways (άνέωξεν in v 14, ήυέψξεν in v 17,
ήνοιξεν inv21).

40 / Chapter 1
 How Badly Did die Scribes Corrupt the New Testament Text?

is called a movable nu. But whether the nu appears in these words or

not, there is absolutely no difference in meaning.
Several of the spelling differences are nonsense readings. These
occur when a scribe is fatigued, inattentive, or perhaps does not
know Greek very well. For example, in 1 Thessalonians 2.7, the man­
uscripts are divided over a very difficult textual problem. Paul is de­
scribing how he and Silas acted among the new converts in their visit
to Thessalonica. Some manuscripts have “we were gentle among you,”
while others have “we were little children among you.” The difference
between the two variants is a single letter in Greek (νήπιοι vs. ήπιοι). A
lone medieval scribe changed the text to “we were horses among you"!
The Greek word for horses (ίπποι) is similar to the other two words. Or
consider a reading that Matthew Morgan noted in his chapter for this
book: in John 1.30, the scribe of Codex L writes “after me comes air”
instead of “after me comes a man” (όπίσω μου έρχεται αήρ for όπίσω
μου έρχεται άνήρ).
After spelling differences, the next largest category of variants
are those that involve synonyms or do not affect translation. They are
wordings other than mere spelling changes, but they do not alter the
way the text is translated or, at least, understood. A very common
variant involves the use of the article with proper names. Greek can
say "the Mary” or “the Joseph” (as in Luke 2.16), while English usage
requires the dropping of the article. So, whether the Greek text has
“the Mary” or simply “Mary,” English will always translate this as
“Mary."
Another common variant is when words in Greek are transposed.
Unlike English, Greek word order is used more for emphasis than for
basic meaning. That is because Greek is a highly inflected language,
with a myriad of suffixes on nouns and verbs, as well as prefixes and
even infixes on verbs. One can tell where the subject is by its ending,
regardless of where it stands in the sentence.
Take, for example, the sentence, “Jesus loves John.” In Greek, that
statement can be expressed in a minimum of sixteen different ways,
- though- the-translation would be the same in English every time. Once
we factor in different verbs for “love” in Greek, the presence or absence
of little particles that often go untranslated, and spelling differences, the
possibilities run into the hundreds. Yet all of them would be translated
simply as "Jesus loves John.” There may be a slight difference in em­
phasis, but the basic meaning is not disturbed.
If a three-word sentence like this could potentially be expressed
by hundreds of Greek constructions, how should we view the number
of actual textual variants in the NT manuscripts? That there are only
three variants for every word in the NT, when the potential is almost
infinitely greater, seems trivial—especially when we consider how many
thousands of manuscripts there are.

Daniel B. Wallace Chapter 1/41

 LOST IN TRANSMISSION

The third largest category involves wording that is meaningful but

not viable. These are variants found in a single MS or group of MSS
that, by themselves, have little likelihood of reflecting the wording
of the autographic text. In 1 Thessalonians 2.9, one late medieval
MS speaks of “the gospel of Christ” instead of “the gospel of God,”
while almost all the other MSS have the latter. Here, “the gospel of
Christ” is a meaningful variant, but it is not viable because there is
little chance that one medieval scribe somehow retained the wording
of the original text while all other scribes for centuries before him
missed it.
The final, and by far the smallest, category of textual variants in­
volves those that are both meaningful and viable. Less than 1 percent
of all textual variants belong to this group, but even saying this may
be misleading. By “meaningful,” we mean that the variant changes the
meaning of the text to some degree. It may not be terribly significant,
but if the reading impacts our understanding of the passage, then it is
meaningful.
For example, consider a textual problem in Revelation 13.18:

Let the one who has insight calculate the beast’s number, for it is the
number of a man, and his number is 666.

• A few years ago, a scrap of papyrus was found at Oxford

University’s Ashmolean Museum. It gave the Beast’s number
as 616. And it just happens to be the oldest manuscript of
Revelation 13 now extant.

• This was just the second MS to do so. (The other MS, not
quite so early, is a very important witness to the text of the
Apocalypse, known as Codex Ephraimi Rescriptus.) Most
scholars think that 666 is the number of the Beast and that
616 is the neighbor of the Beast. But it is possible that the
Beast’s number is really 616.
♦ .X
• But what is the significance of this, really?561 know of no
church, no Bible college, no theological seminary that has
a doctrinal statement that says, “We believe in the deity of
Christ, we believe in the virgin birth of Christ, we believe
in the bodily resurrection of Christ, and we believe that the
number of the Beast is 666.”

56. That this variant is not entirely trivial can be confirmed, by Ehrman’s view of things:
in discussing Johann Albrecht Bengel's quirky views of eschatology, Ehrman de­
clares, *If the number of die Antichrist were not 666 but, say, 616, that would have
a profound effect" {MisquotingJesus, 111).

42 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

• This textual variant does not change any cardinal belief of

Christians, but, if original, it would send about seven tons
of dispensational literature to the flames!

Although the quantity of textual variants among the NT MSS num­

bers in the hundreds of thousands, those that change the meaning pale
in comparison. Less than 1 percent of the differences are both mean­
ingful and viable. There are still hundreds of texts that are in dispute.
I do not want to give the impression that textual criticism is merely a
mop-up job nowadays, that all but a handful of problems have been
resolved. That is not the case. There are hundreds of passages whose in­
terpretation depends, to some degree, on which reading is followed, but
the nature of tire remaining problems and their interpretive significance
are probably far less monumental than many readers of Misquoting
Jesus have come to believe.

THE THEOLOGICAL ISSUES INVOLVED

Finally, what theological issues are involved in these textual vari­
ants? Ehrman argues that the major changes that have been made to the
text of the NT have been produced by "orthodox” scribes, that they have
tampered with the text in hundreds of places, with the result that the
basic teachings of the NT have been drastically altered. Ehrman’s basic
thesis that orthodox scribes have altered the NT text for their own pur­
poses is certainly true, and this occurs in hundreds of places. Ehrman
has done the academic community a great service by systematically
highlighting so many of these alterations in his Orthodox Corruption
of Scripture. However, the extent to which these scribes altered these
various passages and whether such alterations have buried forever the
original wording of the NT are a different matter. Indeed, the very fact
that Ehrman and other textual critics can place these textual variants in
history and can determine the original text that was corrupted presup­
poses that the authentic wording has hardly been lost.57
In the concluding chapter of MisquotingJesus, Ehrman summarizes
his findings as follows:

57. In response to Eldon Epp’s adoption of a new direction for the goal of NT textual
criticism, Moisds Silva noted, "[F] or us to retreat from the traditional task of tex­
tual criticism is equivalent to shooting ourselves in the foot. And my exhibit A
is Bart Ehrman’s brilliant monograph The Orthodox Corruption of Scripture.. ..
Although this book is appealed to in support of blurring the notion of an original
text, there is hardly a page in that book that does not in fact mention such a text or
assume its accessibility.... Indeed, Ehrman’s book is unimaginable unless he can
identify an initial form of the text that can be differentiated from a later alteration”
(M. Silva, "Response,” in Rethinking New Testament Textual Criticism, ed. David
Alan Black [Grand Rapids: Baker, 2002], 149).

Daniel B. Wallace Chapter 1/43

 LOST IN TRANSMISSION

It would be wrong... to say—as people sometimes do—that the changes

in our text have no real bearing on what the texts mean or on the theo­
logical conclusions that one draws from them.... In some instances,
the very meaning of die text is at stake, depending on how one resolves
a textual problem: Was Jesus an angry man [Mark 1.41]? Was he com­
pletely distraught in the face of death [Heb 2.9]? Did he tell his disciples
that they could drink poison without being harmed [Mark 16.9-20] ? Did
he let an adulteress off the hook with nothing but a mild warning [John
7.53-8.11]? Is the doctrine of the Trinity explicitly taught in the New
Testament [1 John 5.7-8] ? Is Jesus actually called die “unique God” there
[John 1.18]? Does the New Testament indicate that even the Son of God
himself does not know when the end will come [Matt 24.36]? The ques­
tions go on and on, and all of them are related to how one resolves diffi­
culties in the manuscript tradition as it has come down to us.58

I have dealt with the referenced passages in detail in my essay "The

Gospel according to Bart,” published in the Journal of the Evangelical
Theological Society.59 What 1 present here is much briefer and more
selective.
Ehrman’s summary paragraph gives us seven passages to consider:

Mark 16.9-20
John 7.53-8.11
IJohn 5.7 (in the KJV)
Mark 1.41
Hebrews 2.9
John 1.18
Matthew 24.36

The first three passages have been, considered inauthentic by most NT

scholars—including most evangelical NT scholars—for well over a
century.60 The presence or absence of these passages changes no fun­

58. Ehrman, MisquotingJesus, 207-8.

59. "The Gospel according to Barb A Review Article of MisquotingJesus Bart Eh-
rman,” JETS49 (2006): 327-49.
60. Ehrman writes as though the excision of such texts could shake up orthodr»:
convictions. Such is hardly the case. I am aware of no confessional statements at
seminaries, Christian colleges, or major denominations that were retooled in the
slightest because of the excision of these verses. Yet it should be noted that these
two passages are the largest textual problems in the NT by far. As one scholar
complains, “[Ehrman’s] first extended examples of textual problems in the New
Testament are the woman caught in adultery and the longer ending of Mark. After
demonstrating how neither of these is likely to be part of the originals of either
Gospel, Ehrman concedes that ‘most of the changes are not of this magnitude’ (p.
69). But this sounds as if there are at least a few others that are of similar size, when
in feet there are no other textual variants anywhere that are even one-fourth as

44 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

damental doctrine, no core belief—despite the fact that there is much

emotional baggage and certainly historical significance attached to
them.
In the next three passages, Ehrman adopts readings that most tex­
tual critics would consider spurious. I think his assessment is correct
in one of them (Mark 1.41), possibly correct in one (Heb 2.9),61 and al­
most surely incorrect in the last (John 1.18).62 Nevertheless, even if his
text-critical decisions are correct in all three passages, the theological
reasons he gives for the changes are probably overdone, as, for example,
in the case of Matthew 24.36.6364
In Matthew’s version of the Olivet Discourse, we read, “But about
that day and hour no one knows, neither the angels of heaven, nor the
Son, but only the Father” (NRSV). Ihe words “nor the Son” are not
found in all tire MSS, which raises a significant issue: Did some scribes
omit these words from the text of Matthew, or did other scribes add
these words? Ehrman is firmly convinced that the words were expunged
by proto-orthodox scribes who bristled at the idea of the Son of God’s
ignorance.
As Adam Messer notes, this passage “is Ehrmaris example par
excellence!^ Ehrman discusses it explicitly at least half a dozen times in
MisquotingJesus.65 And in an academic publication, he calls it “the most
famous instance” of doctrinal alteration.66

In Misquoting Jesus, he argues, “The reason [for the omission] is not

hard to postulate; if Jesus does not know the future, tire Christian
claim that he is a divine being is more than a little compromised.”67
Ehrman does not qualify his words here; he does not say that some
Christians would have a problem with Jesus’ ignorance. No, he says
that the Christian claim would have a problem with it.68 If he do es not

long as these thirteen- [sic] and twelve-verse additions” (Blomberg, DenverJournal

9 [2006]; accessed online).
61. But see Krista M. Miller, "Evaluating the Reading Χωρι? Θεού in Hebrews 2:9 in
Light of Patristic Evidence” (Th.M. thesis, Dallas Theological Seminary, 2010).
62. See Brian Wright,""Jesus as Θεός: A Textual Examination,” in this book, for a dis­
cussion of John 1.18.
63. See Adam G. Messer’s extended treatment of this verse, especially in relation to pa­
tristic writings, in “Patristic Theology and Recension in Matthew 24.36: An Evalua­
tion of Ehrman’s Text-Critical Methodology)' in. this book.
64. Messer, “Patristic Theology and Recension in Matthew 24.36,” 130.
65. Misquoting Jesus, 95,110,204,209,223 n. 19,224 n. 16.
66. Bart D. Ehrman, Studies in the Textual Criticism of the New Testament, vol. 33 in
NTTS, ed. Bruce M. Metzger and Bart D. Ehrman (Leiden: Brill, 2006), 333.
67. Misquoting Jesus, 20i.
68. This sort of absolute statement seems to be either intentionally provocative or un­
intentionally ambiguous. Either way, it leaves a distinct impression on the reader
that Jesus’ deity is foreign to the original NT. This impression was reinforced by
Ehrman himself in an interview with Diane Rehm on National Public Radio (Dec.

Daniel B. Wallace Chapter 1/45

 LOST IN TRANSMISSION

mean this, he is writing more provocatively than is necessary and is

misleading his readers. If he does mean it, he has overstated his case.

Ehrman suggests that the omission would have arisen in the late
second century, as a proto-orthodox response to the Adoptionist heresy.
This is possible, but there are three problems with this hypothesis.

1. It is somewhat startling that virtually no church father seems to

have any problem with the words “nor the Son” until the fourth century,69
though several comment on this very passage. Irenaeus (late second cen­
tury), Tertullian (late second and early third century), and Origen (early
third century) all embraced the deity of Christ, yet none of them felt
that this passage caused any theological problems.70 Irenaeus goes so far
as to use Christ’s ignorance as a model of humility for Christians.71 If
the scribes were simply following the leads of their theological mentors,
the lack of any tension over this passage by second- and third-century
fathers suggests that the omission of “nor the Son” either was not a reac­
tion to Adoptionism or was not created in the late second century.

2. If the omission was created intentionally by proto-orthodox

scribes in the late second century, it most likely would have been cre­
ated by scribes who followed Irenaeus’s view that the four Gospels were

8, 2005). Ms. Rehm asked a vital question: “Has any central doctrine of Christian
faith been called into question by any of these variations?” Ehrman’s response is
illuminating:

Well, yes. In the eighteenth century one of the first scholars to start studying
these materials was a man in Germany named Wettstein, who ended up losing his
teaching post because he pointed out that a number of the changes in the oldest
manuscripts compromis ed the teaching of the deity of Christ, and they threatened
the doctrine of the Trinity, that some of the oldest manuscripts didn't support the
view ofJesus as divine.

Two things are notable about this response. First, rather than citing any tex­
tual problems in the NT, Ehrman enlists the name of Wettstein, a scholar who,
more than two centuries ago, came to the conclusion that the deity of Christ and
the Trinity had a dubious textual basis. Second, he seems to say that these funda­
mental doctrines are in jeopardy. Essentially, Ehrman appears to be agreeing with
Wettsteiris assessment. It is no wonder that toward the end of the interview, Ms.
Rehm sighs, "Very, very confusing for everyone who hears you, reads the book, and
thinks about their beliefs.”
69. Athanasius is the first father to mention any problem with the Son's omniscience
when discussing this passage (see Messer, "Patristic Theology and Recension in
Matthew 24.36,” 153).
70. Messer, "Patristic Theology and Recension in Matthew 24.36.” Specifically, Messer
challenged Ehrman’s claim that Origen knew of the shorter reading {ibid., 151).
71. See Messer, “Patristic Theology and Recension in Matthew 24.36,” 146.

46 ! Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

the only authoritative books on the life of Jesus.72 But the parallel pas­
sage in. Mark 13.32 definitely has the words “nor the Son.” (We know of
almost no MSS that omit the phrase there.) Even though Mark was not
copied as frequently as Matthew in the early centuries of the Christian
faith, the proto-orthodox would have regarded it as Scripture by the
end of the second century. Why did they not strike the offensive words
from Mark?

3. If the scribes had no qualms about deleting “nor the Son” in

Matthew, why did they not delete the word “alone"? Without “nor the
Son,” the passage still implies that the Son of God does not know the date
of his return: “But as for that day and hour no one knows it—not even
the angels in heaven-except the Father alone” Since the Father is speci­
fied as the only person who intimately knows the eschatological cal­
endar, it is difficult to argue that the Son is included in that knowledge.73
This point is not trivial. It cuts to the heart of Ehrman's entire
method. In Orthodox Corruption, he argues that the reason the same
MS can vacillate in the kinds of theological changes it makes is due
to “the individuality of the scribes, who, under their own unique cir­
cumstances, may have felt inclined to emphasize one component of
Christology over another.”74 But he immediately adds,

It strikes me as equally likely, however..., that the same scribe may

have seen different kinds of problems in different texts and made the
requisite changes depending on his perceptions and moods at the mo­
ment of transcription.75

If this kind of logic is applied to Matthew 24.36, we would have to say

that the scribe had a major mood swing, because justfour words after he
deleted “nor the Son,” he could not bring himself to drop “alone.”

72. Irenaeus speaks of them as though they were as certain as the four winds of the
"" earth (IrenaeliS.'A^aiMsiHeresies 3.11.8). Although his logic may leave something
to be desired, it is not insignificant that he speaks as though this matter had been
settled for some time—at least for the proto-orthodox. The Ebionites, however,
utilized only Matthew’s Gospel, but they would have nothing to gain from omitting
“nor die Son” from their copies of Matthew.
73. It is not until Basil (mid- to late fourth century) that we see a patristic writer af­
firm the omission and argue that “alone” in Matt 24.36 does not exclude the Son.
See Messer, “Patristic Theology and Recension in Matthew 24.36,” 160. Basil could
have easily changed the text or argued that “alone” was not found in all the MSS,
if that were the case. But the fact that he attempts to adjust to the passage with the
“alone” as a lone speed bump shows that, like other fathers, he "tended to clarify
[his] theology rather than change texts" (ibid., 161).
74. Ehrman, Orthodox Corruption, 282 n. 16.
75. Ibid.

Daniel B. Wallace Chapter 1/47

 LOST IN TRANSMISSION

A recent critique of Ehrman’s overarching method at this juncture

did not mince words:

If this view is accurate, then how can we have any possibility of deter­
mining the theological motivations involved in textual changes? With
statements such as these, it becomes nearly impossible to falsify any hy­
pothesis regarding theological tendencies.... Rather than verify his con­
clusions through the rigorous work ofevaluating individual manuscripts,
the major prerequisite in Ehrman’s methodology is the alignment of a
favorable theological heresy with particularly intriguing variants.76

Another reviewer complained about the wax nose on Ehrman’s pro­

nouncements over the theological Tendenz of the orthodox scribes with
these words:

[N]o matter what textual problem relating to the central theme and
soul of the Bible (i.e., the Trinitarian God) may be found in the manu­
script tradition . . . , one can always postulate a motivation for an
orthodox corruption, whether or not it is probable. This disingen­
uous method can be applied because no matter whether an article
is left off or added, a word slightly shifted or removed, due to ortho­
graphic errors or any other unintentional type, it often changes the
meaning just enough that there is bound to be a heresy which would
benefit from the change. If an article is missing, it may seem that the
unity of the Godhead is in danger. If the article is present, it may ap­
pear to threaten their distinct personalities. If a phrase exemplifying
Jesus’ humanity is removed, it was obviously to combat the heresy of
Adoptionism. If it is added, it was obviously to combat the heresy of
Sabellianism.77

My point on Matthew 24.36 is not that Ehrman’s argument about

the omission of “nor the Son” is entirely faulty, just that it is not the
only option and does not tell the whole story. In fact, several as­
pects of the problem have apparently not been considered by him,
yet this is his prime example of orthodox corruption. It strikes me
that Ehrman is often certain in the very places where he needs to be
tentative, and he is tentative where he should have much greater cer­
tainty. He is more certain about what the corruptions are than about
what the original wording is, but his certitude about the corruptions

76. Matthew P. Morgan, “The Legacy of a Letter: Sabellianism or Scribal Blunder in

John 1:1c?” (paper presented at the Southwestern Regional Meeting of the Evan­
gelical Theological Society, Southwestern Baptist Theological Seminary extension
campus, Houston, TX, Mar. 29, 2008), 31.
77. Messer, “Patristic Theology and Recension in Matthew 24.36,” 181.

48 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

presupposes, as Silva has eloquently pointed out, a good grasp of the

original wording.
To sum up, although Ehrman’s reconstructions of the reasons for
certain textual corruptions are possible, they often reveal more about
Ehrman’s ingenuity than about the scribes’ intentions. Or, as Gordon
Fee said, “[ujnfortunately, Ehrman too often turns mere possibility into
probability, and probability into certainty, where other equally viable
reasons for corruption exist.’’78 A very high proportion of Ehrman’s ex­
amples could easily be classified as accidental errors rather than inten­
tional changes. If they are accidental, then, by definition, one cannot
claim theological motives.

OF THE NEW TESTAMENT MANUSCRIPTS

Finally, I offer three arguments for the essential reliability of the NT

1. Ehrman argues repeatedly in Misquoting Jesus that we do not

have the earliest copies of the NT and that the copies we do have are
several generations removed from the autographs. From this, he sug­
gests that we have to be agnostic about what these early copies looked
like, that they could have changed the original text in some very signifi­
cant ways. But he elsewhere contradicts this argument by assuming that
the earliest copies were virtually flawless.
In Misquoting Jesus, he observes that although we do not have first-
century MSS, we get a glimpse of what they would have looked like by
an examination of the Synoptic Gospels. He argues that Mark was the
first Gospel and that Matthew and Luke each independently copied it.
Most students of the NT agree with this assessment.
Ehrman then suggests that Matthew and Luke give us the best clue
about how scribes did their work. He claims that “the authors of the
NT were very much like the scribes”79 and that “they, like the scribes,
-••were changing scripture.”80 A comparison of virtually any triple-tradi­

78. Gordon D. Fee, review of The Orthodox Corruption ofScripture, in Critical Review
cfBooks in Religion 8 (1995): 204.
79. Misquoting Jesus, 211.
80. Ibid., 213. See the whole discussion on pp. 211-15. On p. 135, in discussing the tex­
tual problem in Mark 1.41, Ehrman suggests, “There is even better evidence than
this speculative question of which reading the scribes were more likely to invent.
As it turns out, we don't have any Greek manuscripts of Mark that contain this
passage until the end [sic] of the fourth century, nearly three hundred years after
the book was produced. But we do have two authors who copied this story within
twenty years of its first production.” He goes on to tacitly assume that the copy of
Mark that the other evangelists were using was virtually identical to the original,
when he says, "It is possible ... to examine Matthew and Luke to see how they

Daniel B. Wallace
 LOST IN TRANSMISSION

tion pericope in the Synoptic Gospels reveals that, indeed, Matthew

and Luke exercised some freedom in their use of Mark. But in order
for Ehrman to make his point that Matthew and Luke changed the text
of Mark, he assumes that the copies of Mark that they each used had
not already gone through drastic changes. Ehrman is not alone in this
assumption. Virtually every Gospels scholar begins with the assump­
tion that Matthew and Luke were reading nearly identical copies of
Mark. All of redaction criticism is based on this assumption. But if their
copies of Mark were not significantly changed by the time that they got
to Matthew and Luke, then the evangelists are not at all like the scribes.
One cannot have it both ways.
Indeed, the evangelists were not like the scribes. They each had
something to contribute to the story from their own perspective
and certainly did not envision themselves as mere copyists of Mark’s
Gospel. Matthew has a style of writing, as well as certain motifs, that
are different from Luke’s. Each evangelist displays a clear pattern in
his presentation of the gospel. How can each Gospel be so distinc­
tive if scribes had already drastically changed the text of Mark? Either
Matthew came across a copy of Mark that was prefabricated in exactly
the same direction that he would have taken his Gospel, or he was the
creative source that changed it. The very fact that Gospels scholars as­
sume that the Matthean motifs are really Matthew’s—-rather than some
nameless scribe’s—suggests that any scribal corruptions to Mark before
Matthew came across it were trivial and insignificant. And the fact that
Matthew’s motifs are clearly distinguishable from Luke’s indicates that
the important changes to Mark’s Gospel were made by the evangelists
rather than by a previous scribe.
There are hints here and there that the copies of Mark that Matthew
and Luke used were not identical to Marie’s original. But they almost
always involve minor alterations.81 Since that is the case, what can we
say about the scribes in the very first generation of copies? Although

changed Mark, wherever they tell the same story but in a (more or less) different
way” : '
81. Perhaps die most significant alteration involves μόνος in Matt 24.36, assuming
that ουδέ ό υιός is authentic (though see my previous discussion of this text, as
well as Messer’s chapter): this does not seem to be in line with Matthew’s elevated
Christology, for it makes it doubly explicit that the Son does not know the day
or the hour. What would possess Matthew to add insult to injury by adding the
μόνος? Where else in Matthew is his Christology ever lower than Mark's? Also,
there may be some instances in which Matthew has a historical present that paral­
lels an aorist in Mark; since Mark has twice as many historical presents as Mat­
thew (151 to 78), while Matthew has about 50 percent more material, die ratios
are greater than two to one. The instances of ευθύς are not significant, since this is
always Mark’s word choice (over ευθέως), except in 7,35, which has no parallel in
Matthew. Other instances are surely to be found, but to my knowledge, no one has
done any serious work on this matter.

50 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

they certainly were not professional scribes, they did not significantly
alter the text?2

2. Ehrman repeatedly speaks of the first two hundred years of

copying as uncontrolled, giving the impression that all MSS of this era
were riddled with mistakes, both unintentional and intentional. The
scribes, it seems, were undisciplined and wild, freely adding or sub­
tracting words whenever they wanted to?3 But this is not the whole
story It certainly describes the Western text type, but the Western text
does not comprise the only family of MSS in the early centuries?4 (And
even if it did, the very nature of an uncontrolled tradition is antithetical
to changes that conformed to a singular system of beliefs, because one
does not get order out of chaos.)8284
83
85
The standard introduction to NT textual criticism puts things in
perspective:

It would be a mistake to think that the uncontrolled copying practices

that led to the formation of the Western textual tradition were fol­
lowed everywhere that texts were reproduced in the Roman Empire.
In particular, there is solid evidence that in at least one major see of
early Christendom, the city of Alexandria, there was conscious and

82. Ehrman holds that Matthew and Luke were written within twenty years of Mark
{Misquoting Jesus, 135). His dates for Matthew and Luke would thus probably be
ca. 85 GE. But he also holds that die Alexandrian stream of transmission finds its
roots deep in the second century and that it was a very pure stream. Consequently;
we are dealing with a gap of two or three decades in which we have nothing to go
on for what the scribes were doing. Rather than having hundreds ofyears without
a witness to the NT, then (as Ehrman said), we have a few decades. But these few
years will hardly bear the weight of his thesis of radical changes to the text Further,
on a trajectory, the reconstructions scholars make of the Synoptic Gospels’ texts
really show how little these documents must have changed prior to the Alexandrian
scribes doing their work. Finally, if we can pinpoint the greatest textual upheaval
as between 85 CE and as late as 110 or 120 CE, Ehrman’s main thesis of theological
change by proto-orthodox is shipwrecked on the rocks of theological trajectories,
since the great majority of orthodox convictions, according to Ehrman, arose after
this time.
83. MisquotingJesus, 45-69 and passim.
84. Even here, Ehrman is not playing fair with the data. He assumes early corruption
by Western scribes, yet this is based on patristic testimony, not MS evidence.
There are no second-century Western MSS, but there are second-century Alex­
andrian MSS. Ehrman’s thesis thus implodes on the only basis that he considers
valid.
85. If all of our early MSS were products of uncontrolled copying practices and
sloppy scribes, we would still be in relatively good shape to recover the original
wording; the many differences among the MSS, precisely because they were not
controlled, would significantly help scholars to weed out the variants. If there is, in
addition, one stream of transmission that was relatively pure, our lot is improved
immeasurably.

Daniel B. Wallace Chapter 1/51

 LOST IN TRANSMISSION

conscientious control exercised in the copying of the books of the

New Testament.
Alexandria... had a long history of classical scholarship.... It is no
surprise, then, to find that textual witnesses connected to Alexandria
attest a high quality of textual transmission from the earliest times. It
was there that a very ancient line of text was copied and preserved.86

These words are found in the fourth edition of The Text of the New
Testament, by Bruce Metzger and Bart Ehrman. As I have already
pointed out, the kind of scenario Ehrman needs in order to demonstrate
any kind of subversive and pervasive orthodox corruption involves both
control and conspiracy. The Qur’an fits the bill, but die NT does not.

3. Although it is true that die orthodox occasionally corrupted

the Scriptures (just as the nonorthodox occasionally corrupted the
Scriptures),87 the emerging Christology of the proto-orthodox was not
the major force behind most of die intentional changes to the NT MSS.
The larger impetus was harmonization, especially in the Gospels. As
Ehrman recognizes,

[t]his scribal tendency to “harmonize" passages in the Gospels is

ubiquitous. Whenever the same story is told in different Gospels,
one scribe or another is likely to have made sure that the accounts
are perfectiy in harmony, eliminating differences by strokes of their
pens.88

Although Ehrman overstates the point, it is true that narrative har­

monization was a stronger impetus than a high Christology. To put if
bluntly, to the early Christian scribes, the historicity of Christ was more
important than their doctrine of Christ. Virtually all Gospels MSS har­
monized passages between Matthew, Mark, and Luke.89 They even did

86. Metzger and Ehrman, Text ofthe New Testament, 277-78 (italics added).
87. Ehrman insists that history was written by the winners; in Ehrman’s construct,
this mantra seems to assume that only the winners’ writings will be found. But his
entire edifice is based on the belief that he can locate earlier layers of the NT copies
that were undefiled by the proto-orthodox. He does not deal with the possibility
that some of the MSS may have been tampered with by the nonorthodox. In his
reconstruction, there are thus only two types of readings: pure readings and those
corrupted by the orthodox. But if some readings that have escaped the net of the
proto-orthodox have survived, is it not equally possible that some readings that
were created by the nonorthodox also have survived?
88. Misquoting Jesus, 97.
89. Fee puts it more strongly: "[Cjertain MSS and text-types have a much higher fre­
quency of harmonization than others, although no MS is completely guiltless”
(“Myth of Early Textual Recension in Alexandria,” 269).

52 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

so in such a way that would, at times, turn a high Christology into a

highly suspect Christology.
As an example, consider the story of the rich young man who asks
Jesus how to obtain eternal life:

“Good teacher, what must I do to inherit eternal life?"

Jesus said to him, “Why do you call me good? No one is good except
God alone.” (Mark 10.17-18)

Matthew changes this passage in a couple of key ways:

“Teacher, what good tiling must I do to gain eternal life?”

He said to him, “Why do you ask me about what is good? There is only
one who is good.” (Matt 19.16-17)

In Mark’s story, the reader is confronted with the question of Jesus’

identity in a startling way: Is Jesus suggesting that he is not God? Vincent
Taylor argued that “[Jesus’] use of the question along with [his] state­
ment that God alone is good implies a contrast of some kind between
Jesus and God.”90 But in Matthew’s version, “good” is no longer attached
to “teacher” in the young man’s question. This allows Jesus’ response to
be less of a threat to an orthodox Christology.
As we would expect, many scribes—especially the later orthodox
scribes—harmonized these two passages. But what they changed was
not Mark’s Gospel but Matthew's. In the hands of these scribes, the
young man now' says in both Gospels, “Good teacher.” and Jesus says in
both Gospels, “Why do you call me good?"91
Why would these scribes change Matthew’s wording instead of
Mark's? The most likely reason was because the story in Luke’s Gospel

90. Vincent Tayior, The Gospel according to St. Mark (London: Macmillan, 1952), 426.
Taylor, Swete, Lagrange, and others mention several patristic writers who com­
mented on this tert. It seems that they struggled with this one yet did not, for the
most parL change the text.
91. Mark 10.17 has the rich young man say, διδάσκαλε άγαθέ, τί ποιήσω ινα ζωήν
αΙώνιον κληρονομήσω; Jesus responds, τί με λέγει? αγαθόν; ούδεί? αγαθός· ει
μή εΐ? ο θεό?. ΝΑ27 lists no variants here. Matt 19.16 has the young man say (in
NA27) διδάσκαλε, τί άγαθόν ποιήσω (να σχώ ζωήν αιώνιον; Jesus responds (ν.
17), τί με έρωτα? περί τον αγαθού; εΐ? έστιν ό αγαθό?. But several witnesses in
Matthew, particularly of the Byzantine strain, have the young man say διδάσκαλε
άγαθέ, τί αγαθόν ποιήσω, ινα έχω ζωήν αιώνιον; Jesus responds, τί με λέγει?
άγαθόν; ούδει? αγαθό? ει μή εΐ? δ θεό?. Orthodox Byzantine scribes, along with
a host of others, changed the young man’s address to Jesus in Matthew to “good
teacher,” and Jesus' response to “Why do you call me good? No one is good except
one, God.” At these points, the wording in both Gospels is now identical, but the
scribes have made it so by harmonizing Matthew to Mark, rather than the other
way around.

Daniel B. Wallace Chapter 1/ 53

 LOST IN TRANSMISSION

already conformed to the wording in Mark’s, and it was easier to change

one story rather than two. But this clearly illustrates that the scribal
tendency to harmonize the Gospels could trump their tendency toward
a high Christology. What seems to drive much of Ehrman’s text-critical
method is the belief that the least orthodox reading is to be preferred.92
But the story of the rich young man shows that there are other factors
that need to be weighed and that this criterion clearly is not the most
important.

CONCLUSION
It would have been an impossible task for me to try to address all
the passages that Ehrman puts forth as examples of early orthodox cor­
ruption of the text, but I have tried to raise some questions about his
method, his assumptions, and his conclusions. I do not believe that
the orthodox corruptions are nearly as pervasive or as significant as
Ehrman does. I have tried to show that there is no ground for wholesale
skepticism about the wording of the autographic text and that Ehrman
is far less skeptical than the impression he gives in the public square.93
Even Ehrman does not think that any essential belief of the Christian
faith is jeopardized by the variants. In the appendix to Misquoting Jesus,
added to the paperback version, there is a Q&A section. I do not know
who the questioner is, but it is obviously someone affiliated with the
editors of the book. Consider this question asked of Ehrman:

Bruce Metzger, your mentor in textual criticism to whom this book is

dedicated, has said that there is nothing in these variants of Scripture

92. This has been carefully documented by Philip Miller's chapter in this book, *"Ihe
Least Orthodox Reading Is to be Preferred: A New Canon for New Testament Tex­
tual Criticism?”
93. Ehrman hints here and there that he is not nearly as skeptical as reading his Mis­
quoting Jesus might lead one to believe. For example, in his conclusion to the sec­
tion “Examples of the Problem,” which illustrates "complications in knowing the
'original text,”’ he says (62),

For my part, however, i continue to think that even if we cannot be 100 percent
certain about what we can attain to, we can at least be certain that all the surviving
manuscripts were copied from other manuscripts, which were themselves copied
from other manuscripts, and that it is at least possible to get back to the oldest and
earliest stage of tire manuscript tradition for each of the books of the New Testa­
ment. All our manuscripts of Galatians, for example, evidently go back to some
text that was copied; all our manuscripts of John evidently go back to a version
of John that included the prologue and chapter 21. And so we must rest content
knowing that getting back to the earliest attainable version is the best we can do,
whether or not we have readied back to the “original” text. This oldest form of the
text is no doubt closely (very closely) related to what the author originally wrote,
and so it is the basis for our interpretation of his teaching.

54 / Chapter 1 Daniel B. Wallace

 How Badly Did the Scribes Corrupt the New Testament Text?

that challenges any essential Christian beliefs (e.g„ the bodily resur­
rection of Jesus or the Trinity). Why do you believe these core tenets
of Christian orthodoxy to be in jeopardy based on die scribal errors
you discovered in the biblical manuscripts’91*

Note that the wording of the question is not “Do you believe...” but
“Why do you believe these core tenets of Christian orthodoxy to be in
jeopardy ... ?” This is a question that presumably came from someone
who read the book very carefully. How does Ehrman respond?

The position I argue for in Misquoting Jesus does not actually stand at
odds with Prof. Metzger’s position that the essential Christian beliefs
are not affected by textual variants in the manuscript tradition of the
New Testament.95

Suffice it to say that viable textual variants that disturb cardinal doc­
trines found in the NT have not yet been produced.96

91 Misquoting Jesus, 252.

95: ΙΒΙ3’Ήβ goes on to say, “What he means (I think) is that even if one or two passages
that are used to argue for a beEef have a different textual reading, there are still
other passages that could be used to argue for the same belief For the most part, I
think that’s true.”
96. Yet, surprisingly, Ehrman at times seems to suggest that we must know exactly
what an author wrote—down to the very words—before we can understand any
of his message ("[T]he only way to understand what an author wants to say is to
know what his words—all his words—actually were” [Misquoting Jesus, 56; italics
added]). The fact that he can argue for the authenticity of certain variants because
of how they fit in with the biblical author’s argument presupposes that he has suf­
ficient data from the rest of the Gospel or epistle to determine what that author’s
viewpoint is. And if this is the case, would this not undercut his entire skeptical
position, since otherwise we could not understand any of the NT because we are
not sure of all the particular words in any book?

Daniel B. Wallace Chapter 1/55

 2

THE LEAST ORTHODOX

READING IS TO BE PREFERRED:
A New Canon for New Testament
Textual Criticism?

Philip M. Miller1*

rior to the publication of Novum Instrumentum Omne by Desiderius

P Erasmus on March 1, 1516, the text of the Greek NT was hand

copied for over fourteen centuries by ancient scribes who introduced
variant readings as they transcribed the text. Textual critics usually
group sources of these variant readings into two categories: uninten­
tional alterations (e.g., dittography, haplography) and intentional altera­
tions (e.g., harmonization, conflation).
Within the category of intentional alterations, those alterations
made for theological and doctrinal considerations have garnered
particular interest. While nearly all modern textual critics would ac­
knowledge that scribes did alter the text for such reasons at times, Bart
Ehrman has argued that theologically motivated alterations in the NT
are both more significant and more pervasive than was previously tin
--derstoodr-Ehrman has concluded that the NT text was corrupted at

1. Ulis chapter Is dedicated to my grandparents, Jesse and Elizabeth Hensarling, who

have shown me the priceless value cf abandoning one’s life for Christ, and to Eu­
gene and LaVerne Miller, who have taught me the value of encouragement, leader­
ship, and service to one’s family, church, and country. They are my heroes. I want
also to express my gratitude to my wife and dear friend, Krista, whose love and
encouragement have been the fuel of this project from tire very beginning. Thank
you.
2. See especially Bart D. Ehrman, The Orthodox Corruption afScripture; The Effect of
Early Christological Controversies on the Text of the New Testament (Oxford: Ox­
ford University Press, 1993); Bart D. Ehrman, Misquoting Jesus: The Story behind
Who Changed the Bible and Why (San Francisco: HarperSanFrancisco, 2005).

57 / Chapter 2
 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

the hands of orthodox scribes who “occasionally altered their texts to

make them say what they were already believed to mean.”3 Ehrman’s
increasing commitment to this judgment appears to significantly shape
his approach to textual criticism.
As this chapter will show, a new, implicit canon is emerging as an

orthodox reading is to be preferred. While Ehrman denies the influ­

ence of such a canon, his methodology and resulting textual decisions
indicate otherwise.345 This chapter explores this undeclared canon of
unorthodoxy—-(1) surveying its historical backdrop; (2) examining its
influence on Ehrman’s methodology in The Orthodox Corruption of
Scripture and, secondarily, in his popular work Misquoting Jesusf and
(3) critiquing its role within Ehrman’s methodology and considering its
value to the discipline of NT textual criticism.
To show that a new, implicit canon of unorthodoxy is emerging as
an undeclared tenet in Ehrman’s text-critical methodology, the second
section of this chapter is devoted to examining his methodology both
deeply and broadly. Detailed examination of three textual problems
reveals the readings can be explained without resorting to orthodox
corruption. Having examined these three problems in depth, attention
shifts to a synthetic approach to surveying the major textual problems
discussed by Ehrman in The Orthodox Corruption of Scripture. Such
a survey reveals that Ehrman has consistently argued for the least or­
thodox readings, which are often at odds with the NA27/UBS4 text,
even in “A”- and “B”-rated decisions.
In addition to describing the line of argument, it seems prudent
to mention the limitations of this chapter. First, it is not my inten­
tion to draw any textual conclusions or even to add to the discussion
of the textual issues under consideration. Rather, my aim is to highlight

3. Ehrman, MisquotingJesus, 175.

4. In an online posting on March 27, 2009, on die Yahoo! group textualcriticism,
Ehrman wrote in response to Tommy Wasserman’s question about using orthodox
corruption as a canon,, "I have never argued, and never plan to argue, that the theo­
logical orthodoxy of a reading should be used as a criterion for deciding the text”
(“RE: Greenlee’s strawman (??) in The Text of the NTf http://groups.yahoo.com/
group/textualcriticism/message/4608 [accessed Mar. 27, 2009]). In a subsequent
reply to the same thread on the same day, he added, “I may have overstated the
case when I said that I never would consider theological orthodoxy as a ‘criterion’
Maybe I would eventually.... I thought then, and still think now, to use orthodoxy
as a criterion would involve presupposing my conclusions. If it is widely accepted
that there were such forces at work, however, I suppose I could see the possibility
of using the question of orthodoxy as a criterion. But I would be very very careful
in doing so, as there were often lots of other factors at work” (http://groups.yahoo.
com/group/textualcriticism/message/4611 [accessed Mar. 27,2009]).
5. Ehrman’s Misquoting Jesus reflects his more recent thinking on these issues. How­
ever, due to its popular nature, our interaction with Misquoting Jesus will be lim-

58 / Chapter 2 Philip M. Miller

J A New Canon for New Testament Textual Criticism?

the evidence that reveals the influence of the canon of unorthodoxy in

Ehrman’s methodology. Second, my reliance on secondary literature
is intentional, since my focus here is chiefly text-critical methodology.
Third, for the sake of space, only a selection of the textual problems in
The Orthodox Corruption ofScripture and MisquotingJesus are explored;
the texts selected are representative of how Ehrman employs his canon
of unorthodoxy. Fourth, the critique of Ehrman’s canon of unorthodoxy
offered in this chapter is an initial one; this is not the final assessment. As
more data become available, greater clarity will surely result.

THE CANON OF UNORTHODOXY

AND ITS HISTORICAL BACKDROP
.................. When I assert that a new, implicit canon of unorthodoxy has be­
come an influence in Ehrman’s text-critical methodology, I do not mean
that textual critics have never argued along these lines before. In fact,
the reality of theologically motivated alterations is rooted deep in the
history of textual criticism.6

The Early Church

Perhaps the earliest reference to theologically motivated alterations
may be in Revelation 22.18-19, possibly a condemnation of scribes who
might be tempted to add to or subtract from the words of prophecy
contained therein.7 Clearer evidence is the almost unanimous opinion
of the early church fathers that the NT texts were sometimes changed
in light of Christological concerns.8 This assessment was normally ex­
pressed in the context of accusing heretics of changing the text for their
own interests.

$
6. For farther support and detail for this chapter, see Peter M. Head’s excellent article
** "Christology and Textual Transmission: Reverential Alterations in the Synoptic
Gospels,” NovT 35, no. 2 (1993): 105-29, to which this chapter is greatly indebted.
- JZ. Jewish^and early Christian, literary works sometimes concluded with an admoni­
tion to later scribes to copy carefully (so Letter ofAristeas 310-11; Eusebius, Hist.
eccL 5.20.2, citing Irenaeus; cf. 1 Enoch 104.10-13; 2 Enoch 48.6-9; b. Megilloth
14a). Due to the use of these lands of warnings in apocalyptic literature as well
as possible allusions to the Deuteronomic warning passages (Deut 4.1-2; 12,32;
29.19-20; cf. Josephus, Ap. 1.42-43), the case can be made that careless scribal al­
terations are not the primary aim of this warning. Rather, it is most likely intended
for more meddlesome parties bent on distorting the contents of the book for their
own purposes (cf. Rev 2.14,20-23). See G. K. Beale, The Book ofRevelation, New
International Greek Testament Commentary (Grand Rapids: Eerdmans, 1999),
1150-54.
8. Several of these opinions will be discussed here. For a faller treatment, see Head,
"Christology and Textual Transmission,” 106; E. S. Buchanan, "Ancient Testimony
to the Early Corruption of the Gospels,” BSac 73 (1916): 177-91.

IV Philip M. Miller Chapter 2 / 59

*
έ
I
 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

Origen (185-254 CE) at times indicated that the source of tex­

tual variations was in either the negligence of some scribes or the per­
verse audacity of others.9 When refuting the accusation of Celsus—that
Christians altered the text of the Gospels repeatedly in order to avoid
criticism—Origen argued that it was the heretics, not the orthodox, that
were responsible for variation in the textual tradition.10
The writings of Eusebius (263-339 CE) preserve portions of
Dionysius’s writings that mention some heretics who went so far as to
falsify even tire Scriptures.11 Eusebius also recorded an anonymous ac­
count of some disciples of Theodotus the Cobbler who intentionally
corrupted their copies of the Scriptures.12 The reality of theologically
motivated alterations to the text is therefore acknowledged early on by
the church fathers.
Regarding specific textual problems, the fathers argued that certain
readings arose precisely because of theological meddling. For example,
Origen argued that heretics were responsible for adding “Jesus” to the
name of Barabbas in Matthew 27.16-17, because an evildoer could not
bear the name “Jesus?13 Again, Origen claimed that in Luke 23.45 the
enemies of the church refashioned the text as an attack on the Gospels,
since an eclipse was impossible at the time of a full moon.14 Ambrose
(339-97 CE) suggested that the Arians inserted “nor the Son” in Matthew
24.36 to strengthen their cause.1516Of particular interest is the argument
by Epiphanius (315-403 CE) that the orthodox removed Luke 22.43-44
out of fear, because they did not understand tire strength and perfection
of Jesus’ moment of vulnerability and need in Gethsemane.10 Numerous
other examples could be given.17 Certainly many of the fathers’ text-critical
conclusions would be considered unpersuasive in light of current scholar­

9. Origen, Matt. Comm.'ser. 15.14; see Bruce M. Metzger, "Explicit References in the
Words of Origen to Variant Readings in New Testament Manuscripts,” in Histor­
ical and Literary Studies: Pagan, Jewish, and Christian, New Testament Tools and
Studies 8 (Leiden: Brill, 1968), 78-79.
10. Origen, c. Celsus 2.27.
11. Eusebius, Hist. eccl. 4.23.12.
12. Eusebius, Hist. eccl. 5.28.16. <
13. Origen, Matt Comm. ser. 121; see Metzger, "Explicit References in the Words of
Origen,” 94.
14. Origen, Matt. Comm. ser. 134; see Metzger, "Explicit References in the Words of
Origen,” 96.
15. Ambrose, Defide 5.16.
16. Epiphanius, Ancoratus 31. This perspective is echoed by Anastasius Sinaita (Ho­
degos 148) and Photius (Epistle 138 to Theodore). However, Ehrman and Plunkett
have persuasively argued that these verses were added by the orthodox to combat
against docetic theology. See Bart D. Ehrman and Mark A. Plunkett, "The Angel
and the Agony: The Textual Problem of Luke 22:43-44,” CBQ 45 (1983): 401-16;
Ehrman, Orthodox Corruption, 187-94.
17. See Bruce M. Metzger, ‘The Practice of Textual Criticism among the Church
Fathers ” in New Testament Studies: Philological, Versional, and Patristic, New

60 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

ship, yet it is significant that they believed theologically motivated altera­

tions to be a reality.

Modern Textual Criticism

As modernism was dawning, Richard Simon was at the front in

suggesting that such variants in the text were the result of scribal ac­
tivity rather than intentional distortion by heretics, as the church fa­
thers usually indicated.18 Simon postulated that difficult texts prompted
marginal explanations that eventually were copied into the text by later
scribes.
Johann Wettstein was the first to discuss a principle for textual criti-
........ cism in passages that have variants with theological impact. In 1730,
Wettstein argued that when there are two variant readings, the one that
seems more orthodox is not immediately to be preferred.1920 Wettstein un­
derstood “more orthodox” readings as those favoring the orthodox side
of a doctrinal
neutral, favoring neither theological position. He asserted that the het­
erodox could not have corrupted the text, because the manuscripts were
in the possession of the orthodox. He also contended that since scribes
altered the writings of the church fathers in the direction of orthodoxy,
. similar changes should be expected in the NT text. Thus, for Wettstein,

Johann J. Griesbach extended Wettstein’s position, claiming that the

reading that favored orthodoxy was an object of suspicion.20 Griesbach

Testament Tools and Studies 10 (Leiden: Brill, 1980), 189-98; Head, “Christology
and Textual Transmission? 105-29.
18. Richard Simon, Histotre critique du texte du Nouveau Testament, Ou I'on etdblit
la Verite des Ades sur lesquels la Religion Chretienne estfondee (Rotterdam: Chez
Reinier Leers, 1689), 355; Simon, A Critical History of the Text of the New Testa­
ment, 3 vols. (London: R. Taylor, 1689), 2:123.
19. Johann J. Wettstein, Novum Testamentum Graecum editionis receptae, cum Lec-
__ . tionibus .Vadantibus Codicum MSS., Editionum aliarum, Versionum et Patrum,
necnon Commentario pleniore ex Scriptoribus veteribus, Hebraeis, Graecis, et
Latinis, historiam et vim verborum illustrante, 2 vote. (Amsterdam: Ex officina
Dommeriana, 1751-52), 2:864. Hie masterful prologue to this work, Animad-
versiones et cautiones (2:851-74), is a republication of his earlier Prolegomena ad
Testamenti Graeci editionem accuratissimam, e vetustissimis codicibus denuopro-
curandam: in quibus agitur de codicibus manuscriptis Novi Testamenti, Scripto­
ribus qui Novo Testamento usi sunt, versionibus veteribus, editionibus prioribus,
et Claris interpretibus; et proponuntur animadversiones et cautiones, ad examen
variarum lectionum Novi Testamenti, 4 vote. (Amsterdam: Rengeriana, 1730),
1:165-201.
20. Johann J. Griesbach, Novum Testamentum Graece, Textum adfidem Codicum Ver-
sionem et Patrum recensuit et Lectionis Variatatem, 2nd ed., 2 vote. (London: P.
Elmsly, 1796), 1:62.

Philip M. Miller Chapter 2 / 61

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

viewed these readings with suspicion because a spurious reading that

beautifully confirmed a dogma of the church or forcefully destroyed a
heresy would most likely work its way into the text due to the scribes’
general tendency toward inclusiveness.
In terms of dissent, Frederick Scrivener did not regard Griesbach’s rule
as generally acceptable, although he conceded its utility in navigating a few
variants.21 John Burgon more adamantly criticized the principle, calling it a
"monstrous canon.”22 However, lais critique must be tempered by his argu­
ments elsewhere that both heretics and the orthodox were responsible for
corrupting the NT text.23 The most significant and decisive voice of dissent
came from Hort24 in 1882, when he wrote that "even among the numerous
unquestionably spurious readings of the New Testament there are no signs
of deliberate falsification of the text for dogmatic purposes.”25 According
to Hort, dogmatic positions may have influenced the choice between rival
readings, but they were not the cause for the readings in the first place.26
Instead, Hort suggested that passages that appear to be doctrinally moti­
vated are the result of carelessness or laxity, not malice. In concert with
this assertion was Bludau’s argument that the manuscripts of the NT were
not easily susceptible to deliberate falsification, given the vigilance exer­
cised over their production by all concerned, including, at times, opposing
parties.27
Many scholars have subsequently challenged Hort’s statement, in­
sisting that theological considerations played a role in the rise of variant
readings. Significant textual critics of the twentieth century viewed Hort’s

21. Frederick H. A. Scrivener, A Plain Introduction to the Criticism of the New Testa­
ment, for the Use ofBiblical Students, ed. Edward Miller, 4th ed., 2 vols. (London:
George Bell and Sons, 1894), 2:251-52.
22. John W. Burgon, The Traditional Text of the Holy Gospels Vindicated and Estab-
lished, ed. Edward Miller, 2 vols. (London: George Bell and Sons, 1896), 1:66.
23. John W. Burgon, The Causes of the Corruption of the Traditional Text of the Holy
Gospels (London: George Bell and Sons, 1896), 191-231.
24. While both Brooke Foss Westcott and Fenton John Anthony Hort were responsible
for their foundational text Introduction [and] Conclusion (New York: Harper and
Brothers, 1882), Hort is the one usually credited with the major advancements in
the text-critical methodology found therein.
25. Westcott and Hort, Introduction [and] Conclusion, 282; Hort notes that Marcion is
the one exception to this rule.
26. Westcott and Hort, Introduction [and] Conclusion, 282-83; see also J. M. Bebb,
“The Evidence of the Early Versions and Patristic Quotations on the Text of the
Books of the New Testament,” in Studia Biblica et Ecclesiastica, vol. 2 (Oxford:
Clarendon, 1890), 224-26.
27. August Bludau, Die Schriftfalschungen der Hdretiker: Ein Beitrag zur Textkritik
der Bibel (Munster: Aschendorffschen Verlagsbuchhandlung, 1925); cf. also Bruce
M. Metzger and Bart D. Ehrman, The Text of the Mew Testament: Its Transmis­
sion, Corruption, and Restoration, 4th ed. (Oxford: Oxford University Press, 2005),
282-83.

62
I A New Canon for New Testament Textual Criticism?

position to be untenable, including Harris,28 Howard,29 Conybeare,30 and

Lake.31 Of note was Howard's suggestion that there was a "tendency on
the part of the scribes to insert in the text of the Gospels what they knew
to be established in the belief and practice of the Church,”32 According to
Conybeare, the church ensured that only orthodox readings were pre­
served in the manuscript tradition.33 For him, the most likely place to
find the earliest reading was in the early church fathers, since he believed
the manuscript tradition to be dominated by readings approved by the
church34 Building on the foundation Conybeare laid, Lake added further
that only the slightest manuscript evidence was needed to determine the
earliest reading.35 These four scholars stood in sharp contrast to Horts
position and argued that theological corruption significantly shaped the
transmission of the text and that the earliest reading may be without sig­
nificant manuscript support.
By the middle of the twentieth century, most of the standard works
acknowledged that scribes did sometimes alter the text of the NT for
theological and doctrinal considerations.36 Prominent among these

28. J. Rendell Harris, Side-Lights on New Testament Research, Angus Lectures for 1908
(London: Kingsgate, nd.), 29-35. Harris discusses examples of dogmatic alterations
in the writings of Marcion and Tatian as well as Luke 4.16; 22.43-44; and 23.34.
29. Wilbert E Howard, “The Influence of Doctrine upon the Text of the New Testa­
ment,” London Quarterly and Holborn Review 6, no. 10 (1941): 1-16.
30. Frederick C. Conybeare, “Three Early Doctrinal Modifications of the Text of the
Gospels,” W1 (1902): 96-113.
31. Kirsopp Lake, The Influence of Textual Criticism on the Exegesis ofthe New Testa­
ment (Oxford: Parker and Sons, 1904).
32. Howard, “Influence of Doctrine upon the Text of the New Testament,” 12.
33. Conybeare, "Three Early Doctrinal Modifications of the Text of the Gospels,”
96-113.
34. Frederick C. Conybeare, “The Eusebian Form of the Text Matth. 28, 19,” ZNW 2
(1901): 275—88. Conybeare misread die Eusebian citation and argued that "in my
name” represents the earliest text of Matt 28.19, despite die fact that this reading
has no MS support. He proposed that the earliest reading was altered by orthodox
scribes who sought to align it with orthodox Trinitarianism.
35. Lake, Influence of Textual Criticism.
. . 36.. .Maurice.Gogueh ie iexie.et les editions du Nouveau Testament grec (Paris: E. Le-
roux, 1920), 64-67; Daniel Pio oij, Tendentieuse Varianten in den Text der Evan­
gelic (Leiden: Brill, 1926); Kirsopp Lake, The Text of the New Testament, rev. Silva
New, 6th ed. (London: Rivingtons, 1928), 6; Ldon Vaganay, Initiation ά la critique
textuelle neotestamentaire (Paris: Bloud et Gay, 1934), 53-54; C. S. C. Williams,
Alterations to the Text of the Synoptic Gospels and Acts (Oxford: Blackwell, 1951),
5; Leon E. Wright, Alterations of the Words ofJesus as Quoted in the Literature
of the Second Century (Cambridge, MA: Harvard University Press, 1952), 58-68;
Alexander Souter, The Text and Canon of the New Testament, rev. Charles Stephen
Conway Williams, 2nd ed. (London: Duckworth, 1954), 106; Heinrich J. Vogels,
Handbuch der Textkritik des Neuen Testaments, 2nd ed. (Bonn: P. Hanstem, 1955),
178-82; Jacob H. Greenlee, Introduction to New Testament Textual Criticism
(Grand Rapids: Eerdmans, 1964), 68; M. R. Pelt, "Textual Variation in Relation to
Theological Interpretataion in the New Testament” (PhD diss., Duke University,

Philip M. Miller Chapter 2 ! 63

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

was Bruce Metzger, who identified two groups of alterations: (1) “those
which involve the elimination or alteration of what was regarded as doc­
trinally unacceptable or inconvenient”37 and (2) “those which introduce
into the Scriptures ‘proof for a favorite theological tenet or practice.”38
C. S. C. Williams described these “tendencious, reverential and
doctrinal alterations" as highlighting some of “the most fascinating
problems in the whole field of the textual criticism of the New
Testament.”39 What is most notable about Williams's description is his
commitment to recognize the goodwill of the scribes. He held that the
changes these scribes introduced into the textual tradition were not
sourced in malice, deceit, or agenda but, rather, derived from rever­
ence for Christ. Metzger, while wrestling with the difficulties altera­
tions raised in his Textual Commentary on the Greek New Testament,
likewise noted the suppression of certain readings “for reverential
considerations”40 the omission of doctrinally difficult words,41 and
secondary improvements “introduced from a sense of reverence for
the person of Jesus.”42 Peter M. Head deemed these “reverential altera­
tions” as evidence of “the scribe’s involvement in his work understood
as an act of devotion to the divine Christ.”43

Orthodox Corruption ofScripture

As Ehrman has pointed out, the last forty years have seen signifi­
cant developments in postulating the role of theology in connection
with the transmissional history of the NT.44 The first major develop­
ment came in 1966, in Eldon Epp’s masterful study on the theological
tendency of Codex Bezae in Acts, in which he determined that 40
percent of the variant readings were anti-Judaic.45*This finding was

1966); Eberhard Nestle, Einfuhruxg in das griechische Neue Testament, 2nd ed.
(Gbttingen: Vandenhoeck und Ruprecht, 1899), 161-62; Bruce M. Metzger, The
Text ofthe New Testament, 2nd ed. (Oxford: Oxford University Press, 1968)» 201.
37. Metzger, Text of the New Testament, 2nd ed., 201. Prior to Bart Ehrman’s involve­
ment in The Text of the New Testament, Metzger acknowledged the reality of or­
thodox corruption. These comments are repeated in Metzger and Ehrman, Texiof
the New Testament, 4th ed., 266.
38. Metzger, Text ofthe New Testament, 2nd ed., 201; Metzger and Ehrman, Text of the
New Testament, 266.
39. Alterations to the Text of the Synoptic Gospels and Acts, 5.
40. Bruce M. Metzger, A Textual Commentary on the Greek New Testament, 2nd ed.
(Stuttgart: Deutsche Bibelgesellschaft, 1994), 56, note on Matt 27.16-17.
41. Ibid., 51-52, note on Matt 24.36; cf. 164-66, “Note on Western Non-Interpolations."
42. Ibid., 200, note on John 11.33.
43. Head, “Christology and Textual Transmission,” 129.
44. Metzger and Ehrman, T&tf ofthe New Testament, 283.
45. Eldon J. Epp, The Theological Tendency of Codex Bezae Cantabrigiensis in Acts,
Society for New Testament Studies Monograph Series 3 (Cambridge: Cambridge
University Press, 1966), 171.

64 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

significant because it seemed to indicate that the scribe of Codex

Bezae or his tradition was anti-Judaic and that bias intentionally or
unintentionally found its way into the text as it was copied.46 This con­
clusion called into question a scribe’s ability to remain objective in
transcription.
The second major development occurred in 1993, with the publica­
tion of Ehrman’s The Orthodox Corruption of Scripture,47 in which he
argued that "scribes occasionally altered the words of their sacred texts
to make them more patently orthodox and to prevent their misuse by
Christians who espoused aberrant views ”48 His masterpiece is the most
thorough treatment to date of theologically motivated alterations in the
text of the NT. However, significant critiques of Ehrman’s method and
conclusions remain.49
In The Orthodox Corruption of Scripture, Ehrman began his ar­
gument by summarizing Bauer’s view of early Christianity: that early
Christianity was diverse, full of competing views, especially as it re­
lates to Christology.50 Ehrman reasoned that the scribes, who were
aware of and possibly embroiled in the debated Christological issues,
occasionally altered the text as they copied it, making it more pa­
tently orthodox and less subject to abuse by their opponents.51 When
orthodoxy “won," he argued, the scribes naturally gave preference
to the most orthodox readings,52 and thus the most orthodox read­
ings came to dominate the textual tradition. The implicit conclusion

4-6. Metzger and Ehrman» Text ofthe New Testament, 284.

47. Bart D. Ehrman, Orthodox Corruption of Scripture: The Effect of Early Christo-
logical Controversies on the Text ofthe New Testament (Oxford: Oxford University
Press, 1993).
48. Ehrman, Orthodox Corruption, xt Again on p. 4, Ehrman claimed that "theological
disputes, specifically disputes over Christology, prompted Christian scribee to alter
the words of Scripture in order to make them more serviceable for die polemical
bTA· modified their manuscripts to make them more patently 'orthodox'
task. Scribes
and less susceptible to ‘abuse4 by the opponents of orthodoxy?
49. For reviews of Ehrman’s Zfce Orthodox Corruption ofScripture, see J. Neville Bird·
sail, Theology 97, no. 780 (1994): 460-62; J. K. Elliott, NovT 36, no. 4 (1994): 405-6;
JSordonJXFee, CRBR 8 (1995): 203-6; Michael W. Holmes, RelSRev20t no. 3 (1994):
237; Bruce M. Metzger, PSB 15, no. 2 (1994): 210-11 David C. Parker, JTS 45, no.
2 (1994): 704-8. See also Stratton Ladewig, "An Examination of the Orthodoxy of
the Variants in Light of Bart Ehrman's The Orthodox Corruption ofScripture? (ThM
thesis, Dallas Theological Seminary, 2000);Ivo Tamm “Iheologisdi-christologische
Varianten in der frilhen Uberlieferung des Neuen Testaments?” (Magisterschrlft,
Westfelische Wilhelms-Universitit Munster, 2001); Daniel B. Wallace, “The Gospel
according to Bart: A Review Article of Misquoting Jesus by Bart Ehrman,” JETS 49,
no. 2 (2006): 327-49.
50. Ehrman, Orthodox Corruption, 4, 7—11; c£ Walter Bauer, Rechtgldubigkeit und
Ketzerei im dltesten Christentum, Beitrage zur historischen Theologie 10 (TUbingen:
J. G B. Mohr [Paul Siebeck], 1934).
51. Ehrman, Orthodox Corruption, 3-4.
52. Ibid., 8,27.
I
I
k

Philip M. Miller Chapter 2 / 65

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

from Ehrman’s assertions is that when multiple readings exist for a

Christologically significant passage, the reading that is least orthodox
is most likely to be the earliest reading.
There were two primary reasons for this conclusion. First, only the
least orthodox reading could have given rise to the widely divergent
views of Christology found in the first three centuries, while the or­
thodox reading would have strongly supported the orthodox position
only. Second, the orthodox scribes would be more likely to alter un­
orthodox and suborthodox readings so that they reflected an orthodox
position, rather than the other way around.58 Ehrman attempted to dem­
onstrate his case by giving examples where he believed the orthodox
scribes altered Scripture to clearly affirm orthodoxy over and against the
Adoptionists, Separationists, Docetists, and Patripassianists.*54 Thus, he
maintained, the orthodox scribes created variants that "establish the or­
thodox character ofthe text, either by promoting more fully an orthodox
understanding of Christ or by circumventing the heretical use of a text
in support of an aberrant teaching.”55
In a more recent statement about his textual-critical approach,
Ehrman advises in a 2009 web posting, “I think it is important to be very
clear about methodology. At any point where there is textual variation,
one has to decide what the earliest form of the text is and what the latter
changes (corruptions) were. Only once that decision is made can you go
on-to consider whether theological debates were involved in making the
corruption.”56 This is the two-step method Ehrman sought to employ in
The Orthodox Corruption of Scripture·. (1) establish the oldest reading
using accepted text-critical methods and (2) only then inquire about

53. Ibid., 27.

54. In Orthodox Corruption, Ehrman addressed anti-adoptionistic corruptions of
Scripture in chapter 2 (47-118), antl-separatlonist corruptions in chapter 3 (119-
80), anti-docetic corruptions in chapter 4 (181-261), and anti-patripassianist cor­
ruptions in chapter 5 (262-73).
55. Ehrman, Orthodox Corruption, 28 (italics mine).
56. Ehrman, “RE: Greenlee’s strawman (??) in The Text of the NTT In a subsequent
reply to the same thread (see n. 4), Ehrman continues, to identify some of the rea­
sons for his two-step method:

When I wrote The Orthodox Corruption of Scripture, I wanted to be very very

careful not to argue in a circle, that orthodoxy affected scribes, that's why they
changed the text, as we can see in places where orthodoxy affected the readings
they created. Instead I wanted to argue on other grounds what the oldest form
of the text was, and then to see, once that was established, whether the reason
for the change may have been theological debates affecting the scribes. I thought
then, and still think now, to use orthodoxy as a criterion would involve presup­
posing my conclusions. If it is widely accepted that there were such forces at work,
however, 1 suppose 1 could see the possibility of using the question of orthodoxy
as a criterion. But I would be very very careful In doing so, as there were often Lots
of other factors at work.

66 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

theological influence. Maintaining the order of this method is essential

to Ehrman’s line of argument. At issue is how faithfully this two-step
method has been applied.
As is shown within the history of textual criticism, Ehrman is not
alone in his thesis that theologically motivated alterations to the text
of Scripture occurred; the “canon of unorthodoxy" has been discussed
for centuries. Nevertheless, Ehrman seems to have taken this idea to
a new level, arguing that orthodox corruption is both more extensive
and more significant than previously understood. He also implied that
significant doctrines of orthodoxy were at stake when he wrote that
textual variation raised “such questions as whether the Gospels could
have been used to support either an ‘adoptionisic’ Christology... or
one that was 'anti-docetic'...» whether Luke has a doctrine of the
atonement.... whether members of the Johannine community em­
braced a gnostic Christianity..., and whether any of the authors of the
New Testament characterizes Jesus as God.”57 Ehrman has proposed
that significant issues of orthodoxy are on the table for discussion in
light of these variants, including whether Jesus is ever characterized
as God by any of the NT authors. Ehrman apparently has entertained
these possibilities because of his increasing commitment to the canon
of unorthodoxy.
While Ehrman is not alone in viewing doctrinal considerations
as playing a role in the rise and transmission of variant readings, he
distinguishes himself in at least two ways. First, he appears to give
strong preference to the least orthodox reading. When there are mul­
tiple readings of varying levels of orthodoxy, Ehrman seems to choose
the least orthodox reading almost by default. He gives the impres­
sion that orthodox readings are guilty until proven innocent. Second,
Ehrman allows the canon of unorthodoxy high priority in his text-
critical methodology, as the next section will demonstrate.

THE CANON OF UNORTHODOXY DEMONSTRATED

IN EHRMAN’S METHODOLOGY
Ehrman’s acceptance of the canon of unorthodoxy {the least or­
thodox reading is to be preferred) is observed by a carefill examination
of his text-critical methodology. Though the select textual problems
discussed in this section are intended not to represent the totality of
Ehrman’s methodology, they seem to reflect instances where the canon
of unorthodoxy has dear influence. The examination of each textual
problem will consist of three steps: (1) a summary of Ehrman’s thesis on
the problem, (2) a discussion of both the external and internal evidence

57. Metzger and Ehrman, Text of the New Testament, 284-85 n. 32; cf. Ehrman, Or­
thodox Corruption, 276-77.

Philip M. Miller Chapter 2 / 67

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

regarding the problem, and (3) an analysis of Ehrman’s methodology in

light of the evidence.

Matthew 24.36

In the Olivet Discourse, when commenting on the timing of his re­

turn, Jesus stated, “But as for that day and hour, no one knows it, not even
the angels in heaven, nor the Son, but only the Father.”58 Ehrman's argu­
ment here is that orthodox scribes omitted the words ουδέ ό υΙός, “nor
the Son,” from Matthew 24.3659 because “it suggests that the Son of God is
not all-knowing and could therefore be used by Adoptionists to argue that
Jesus was himself not divine.”60 Adoptionists believed that Jesus was born
a human and that at some point, usually identified as his baptism, he was
adopted as the Son of God. Ehrman argues that this “phrase in Matthew
was seen as problematic by Christian scribes”61 who “appear to have
taken umbrage at the notion that Jesus was not all-knowing or spiritu­
ally perfect."62 Thus, he maintains, scribes dropped the words ουδέ ό υιός
from the text because “if Jesus does not know the future, the Christian
claim that he is a divine being is more than a little compromised.”63
There are two relevant readings present in Matthew 24.36, and
these readings may be translated as follows:

But as for that day and hour, no one knows it, not even the angels in
heaven, nor the Son [ουδέ ό υιός], but only the Father.

But as for that day and hour, no one knows it, not even the angels in
heaven, but only the Father.

There are two possibilities for this variant: either (1) ουδέ ό υιός
was original to the text and was later omitted, or (2) ουδέ ό υιός
was not original to the text but was added later. Ehrman argues that
ούδέ ό υιός was the earliest form of the text and was later omitted
by orthodox scribes who were troubled by the implication that Jesus
did not know the timing of his own return. Others have argued that
ουδέ ό υιός was added to the text out of a desire to harmonize the
Matthean account with Mark 13.32, which includes the phrase ουδέ
t C/
0 U109.

58. All Scripture quotations are my own translations unless noted otherwise.
59. Ehrman discussed Matt 24.36 in Orthodox Corruption (91-94) and Misquoting
Jesus (95, 203-4).
60. Ehrman, Orthodox Corruption, 92.
61. Ibid.
62. Ibid., 94.
63. Ehrman, Misquoting Jesus, 204.

68 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

FIGURE 2.1: SUPPORT FOR TWO READINGS IN MATTHEW 24.36

ούδέ ό υιός K*·™ B D Θ/13 28 1505154Τα

Vgn>ss gyjpa ann geo>. Diatessaron3™ Irenaeus13'
b

Origen111 Chrysostom Cyril (Hesychius); Hilary

Ambrose Latin mssac!!',aietome Augustine Varimadum
OMIT K'LWAi1 33 157 180 205 565 579 597 700
892 1006 1010 1071 1241 1243 1292 1342 1424
Byz [E F G Η Σ] Leet it®1,1 vg syrs'pJl cop“,meg'b0 geoA
Didymus**’; Phoebadius Greek mssa“-loAmSr0!i Jerome
Greek mssse&lDi'!ra“

The external evidence largely supports Ehrman’s contention for

the inclusion of ουδέ ό υΙός. Some important witnesses support this
reading, including such early Alexandrian and Western manuscripts as
2«id
b dQf13 pc it vg™** Ir!". However, the exclusion of ουδέ δ υιός also
bears support from a1, as well as L W f1 33 Byz vg sy cop. On the basis
of external evidence, the inclusion of ούδέ ό υιός certainly possesses
better support, although the evidence is divided.
The solution to this textual problem must primarily rely on the
strength of the internal evidence. Ehrman argues, “[N]ot only is the
phrase ουδέ ό υιός found in our earliest and best manuscripts of
Matthew, it is also necessary on internal grounds.”64 This necessity is
derived from Ehrman’s conviction that this is a clear example of or­
thodox corruption. He reasons that scribes were uncomfortable with
Jesus’ ignorance of his own return and therefore omitted this embar­
rassing text This is a persuasive explanation, especially in light of the
fact that the interpretation of this passage continues to be debated.65
However, it is notable that ουδέ ό υιός is solidly preserved in the par­
allel text in Mark 13.32:

But as for that day and hour, no one knows it, not even the angels in
heaven, nor the Son, but only the Father.

Only X pc vg“ omit the words ουδέ ό υιός in Mark 13.32. If the idea of
Jesus' ignorance about the timing of his return was so embarrassing to
the orthodox scribes, one begins to wonder how Mark 13.32 escaped

64. Ehrman, Orthodox Corruption, 92 (italics mine).

65. Metzger proposed another line of reasoning in support of the longer reading. He
proposed that the two ούδέ were best understood as correlative within the sen­
tence structure and therefore that the internal evidence supported the longer
reading {Textual Commentary, 2nd ecL, 51-52). However, this does not completely
close the case, since ούδέ may be used in the simple sense of “not even," in which
case the shorter reading is tenable.

Philip M. Miller Chapter 2/69

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

with such a clean pedigree.66 It seems inconsistent to construct a histor­

ical setting in which orthodox scribes intentionally modified Matthew
out of embarrassment yet left the identical embarrassing material in
Mark intact. Although the scribes were hardly consistent, for such a
theologically significant issue we would expect at least some consis­
tency on the part of a few scribes. Furthermore, the variant reading in
Matthew 24.36 can be accounted for differently.
The other possibility is that ουδέ δ υιός was added to the text out
of a desire to harmonize Matthew 24.36 with the parallel statement in
Mark 13.32.
I· t
l· FIGURE 2. 2: HARMONIZING MATTHEW 24.36 WITH MARK 13.32

Mark ουδέ οί άγγίλοι έν ούρανφ ουδέ ό υίός el μή δ πατήρ

13.32
Matthew ουδέ οι αγγ€λοι των ουρανών si μή ό πατήρ
24.36 μόνος·

The inclusion of ουδέ ό υΙός in Matthew 24.36 can be easily explained

by harmonization; the scribe could have harmonized these passages ei­
ther through unintentional or intentional alteration. If the alteration was
unintentional, the scribe’s familiarity with the text of Mark did him a
disservice when he copied Matthew, for he added to the verse without
realizing it. If it was intentional, the scribe may have sought to elimi­
nate an apparent inconsistency in the recorded words of Jesus by har­
monizing the statements. However, intentional change seems less lilcely,
since the word μόνος is found in Matthew but not in Mark. If this were
intentional harmonization to eradicate an apparent inconsistency in the
recorded words of Jesus, these statements would likely be changed in
both Gospels. Since they are not, it seems reasonable that unintentional
harmonization may have given rise to the addition of ουδέ ό υ'ιός in
Matthew 24.36.67 In addition, because Matthew generally softens Mark’s
harsh statements as he incorporates material into his own Gospel, it is
likely that Matthew himself dropped the phrase ουδέ ό υιός and added
μόνος. In doing this, Matthew would have kept true to the ipsissima vox

66. Matthew was copied and used more frequently than Mark, but this hardly accounts
for the great disparity in the external evidence if Ehrman is correct that orthodox
scribal corruption gave rise to this reading. Since both Gospels were viewed as
authoritative by the early church, the Adoptionists would have been able to prove
their case in Mark as well as in Matthew.
67. Harmonization of Matthew to Mark seems to be the better explanation, despite the
tendency in the MSS for Mark to conform to Matthew, rather than the other way
around.

70 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

of Jesus, while softening Mark’s harsh statement regarding Jesus’ igno­

rance of his return.68
In discussing the textual problem of Matthew 24.36, Ehrman be­
gins to reveal the role of the canon of unorthodoxy in his text-critical
methodology in a number of ways (though it should be acknowledged
that Ehrman is in good company in regard to this textual problem, since
many textual critics believe ουδέ ό υώ$ is the earliest reading):69

1. When Ehrman says that ούδέ ό υιός is “necessary on internal

grounds,”70 he seems to be overstating his case. The internal evidence
does not necessitate the originality of ουδέ ό υιό?. This is a debated text,
a fact that Ehrman's word choice does not reflect.

2. Ehrman overstates the importance of this passage to his­

toric orthodoxy. Ehrman claims that “if Jesus does not know the fu­
ture, the Christian claim that he is a divine being is more than a little
compromised.”71 Yet orthodoxy flourished despite the identical state­
ment’s solid presence in Mark 13.32.

3. Because viable, alternate explanations exist for the textual evi­

dence, Ehrman’s thesis is unconvincing in this passage.

4. Ehrman has indicated his desire to follow a two-step meth­

odology: (1) establish the oldest reading using accepted text-critical
methods and (2) only then inquire after theological influence. He has
stayed true to this approach in that he has identified tire strong external
evidence for the inclusion of ούδέ ό υιός1. However, based on external
evidence alone, it is difficult to establish this reading as the oldest
reading. In some sense, Ehrman has strayed from his two-step method
by placing the bulk of his argument on the notion of scribal embar­
rassment over Jesus’ ignorance of his own return. At the very least, this
reveals his methodological tendencies.

68. An alteration of this kind by an evangelist is distinct from scribal alterations, be­
cause of the authorial role. An author’s role is to create new material and to redact
existing materials, while a scribe's role is transmission of material. Thus, if Mat­
thew softened die statement previously recorded by Mark, he would be exercising
his role as an author and evangelist. Gospel redaction is not analogous to scribal
activity.
69. UBS1 and NA27 both include ούδέ ό ulos in their respective texts. UBS1 gives this
reading a “B” rating (The Greek New Testament, 4th ed. [Stuttgart: Deutsche Bibel-
gesellschaft, 1993], 95; c£ Metzger, Textual Commentary, 2nd ed., 51-52).
70. Ehrman, Orthodox Corruption, 92.
71. Ehrman^ Misquoting Jesus, 204.

Philip M. Miller Chapter 2/71

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

In the end, Ehrman’s confidence in the necessity of the reading of

ουδέ ό υιός in Matthew 24.36 is derived from the fact that it aligns well
with the canon of unorthodoxy, not because of overwhelming textual
evidence. In light of the fact that this textual problem is debated, it is
significant that Ehrman chooses to speak of the necessity of ούδέ ό υιός.
It appears that Ehrman finds this reading “necessary” precisely because
he holds that the least orthodox reading is to be preferred.

John 1.18
The textual problem in John 1.1872 is notoriously difficult and has
been the subject of sharp disagreement. John 1.18 can be translated,

No one has ever seen God, but the unique one, himself God, who re­
sides in the Father’s bosom, has made him known.

Scholars have long debated the textual evidence for whether the reading
should be μονογενής θεός, which can be translated as "the unique one,
himself God” or “the unique God,”73 or ό μονογενής υιός, “the unique
Son.” However, most textual critics are swayed by the strength of the
external evidence in favor of μονογενής θεός, as Ehrman concedes.7*
Nevertheless, Ehrman states that “the variant reading of the
Alexandrian tradition, which substitutes God for Son, represents an or­
thodox corruption of the text in which the complete deity of Christ is
affirmed.”75 Ehrman claims that orthodox and gnostic scribes corrupted
this text to prevent its use by the Adoptionists, who taught that Jesus
was not God but merely a man adopted by God.76 Ehrman affirms, "For
the scribe who created this variant, Christ is not merely portrayed as the
‘unique Son.' He himself is God, the ‘unique God,’ who is to be differenti­
ated from God the Father, in whose bosom he resides, but who nonethe­
less is his co-equal.”77 He writes that “now [after the alteration] Christ is
not merely God’s unique Son, he is the unique God himself!”78 Thus, he
concludes, “this Alexandrian reading derives from an anti-adoptionistic
context, and therefore represents an orthodox corruption.”79

72. Ehrman discussed John 1.18 in Orthodox Corruption (78-82) and MisquotingJesus
(161-62).
73. Ehrman translated μονογενής θεός as “the unique God.” It is better to translate
μονογενής as a substantival adjective. See Wallace, “Gospel according to Bart,”
344-46.
74. Ehrman, Orthodox Corruption, 79.
75. Ibid., 78.
76. Ibid., 82.
77. Ibid.
78. £ku:maxi, Misquoting Jesus, 162.
79. Ehrman, Orthodox Corruption, 82.

72 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

There are five readings for this textual problem in John 1.18. For this
discussion, the first three readings are most relevant. The first two read­
ings differ in that one has the article while the other does not. Both read­
ings support the translation "No one has ever seen God, but the unique
one, himself God [μονογενής θεός // ό μονογενής θεός], who resides in
the Father’s bosom, has made him known.” The third reading supports the
translation “No one has ever seen God, but the unique Son [c μονογενής
υιός], who resides in the Father’s bosom, has made him known.” The fourth
reading is a conflation of the first two readings with the third, resulting in
"the unique Son of God” (μονογενής υιός θεοί) which gives mild support
to the third reading. The fifth reading omits “Son” and “God" entirely and
is translated substantively as “the unique one” (ό μονογενής).

FIGURE 2.3: FIVE READINGS IN JOHN 1.18

μονογενής θεός 3J66 Κ* Β C* L syrp'geo2 Origen^ w Didymus

Cyril174
ό μονογενής θεός g>75 K2 33 copto Clement573 Clementfrora TheodQDiS1β Origen61274
Eusebius377 Basil172 Gregory-Nyssa Epiphanius Serapion172
Cyril2·'4
ό μονογ€νής υιός A C3 W’w ΔΘΨ 0141 f f3 28 157 180 205 565 579
597 700 892 1006 10101071 1241 1243 1292 1342
1424 1505 Byz [EF G H] Led it»· ■=·«.«vg syri·h·
arm eth geo1 Slav Irenaeus1111/3 Clementfroia'n,e‘>dwus w
| Clement175 Hippolytus Origen1*172 Letter of Hymenaeus
Alexander Eustathius Eusebius477 Serapion172 Athanasius
Basil172 Gregory-Nazianzus Chrysostom Theodore
Cyril174 Proclus Theodoret John-Damascus; Tertullian
Hegemonius Victorinus-Rome Ambrosiaster Hilary577
Ps-Priscillian Ambrose10711 Faustinus Gregory-Elvira
Phoebadius Jerome Augustine Varimadum
V 4 * V ·♦ 9 * Λ ♦ > Κ' . : . '· X · V>
μονογενής υιός itq Irenaeus’1'175; Ambrose1711 vid
θεοΰ
ό μονογενής vgms Ps-Vigilius172

The external evidence for this textual variant is inconclusive, but

it strongly favors the first or second readings of μονογενής θεός // ό
μονογενής θεός. Noteworthy support includes the earliest and best
witnesses, such as X* B C* // ίβ75 K' 33. Ehrman concedes, “[I]t must
be acknowledged that the first reading [μονογενής θεός] is the one
found in the manuscripts that are the oldest and generally considered

Philip M. Miller Chapter 2 / 73

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

to be the best—those of the Alexandrian textual family.”80 Again he

writes, ’“[Tjhe Alexandrian reading is more commonly preferred by
textual critics, in no small measure because of its external support.”81
However, the reading ό μονογενής υιό? does have support from Codex
Alexandrinus, which reflects the Byzantine text type in the Gospels,
and from the majority of the witnesses, especially the later Byzantine
manuscripts C3 Θ Ψ f1113 Byz lat.
Like most of the textual problems that Ehrman discusses, the in­
ternal evidence is crucial to his thesis, for it is in this evidence that he be­
lieves “the real superiority of ό μονογενής υιός shines forth.”82 Ehrman
argues that the earliest form of the text read “the unique Son” but that
orthodox and gnostic scribes altered it to read “the unique God,” thus
inscripturating a high Christology in John 1.18 and preventing the
Adoptionists from having a textual foothold.
The alternate argument is that μονογενής θεός was original83 and
that a scribe changed it to ό μονογενής υιός because it fits well with
Johannine style. The title ό μονογενής υίός is a Christological title
unique to John (John 3.16,18; 1 John 4.9). A scribe familiar with John's
language could have been transcribing on "mental autopilot” and
simply supplied the Johannine Christological title. Given the scribes’
tendency to move in the direction of the author's style rather than away
from it, μονογενής θεός is the harder reading, which is generally to be
preferred. In contrast, it is difficult to see why a scribe would change
ό μονογενής υίός to μονογενής θεός, since the phrase μονογενής
θεός occurs nowhere else in the Bible. In contrast, μονογενής υίός
appears elsewhere in Johannine material (John 3.16, 18; 1 John 4.9).
Therefore, since μονογενής θεός is the harder reading, it would seem
to be preferred.
Ehrman, however, argues that this reading is much too hard, for
two main reasons. First, he contends that although some transla­
tions understand μονογενής as a substantival adjective in apposition
to θεός, an adjective is never used substantially when it immediately
precedes a noun of the same inflection.84 On this point and counter to
Ehrman, Daniel B. Wallace has presented seven clear examples where
an adjective is used substantivally when preceding a noun of the same
inflection.85
Ehrman’s second argument is that even though μονογενής θεός
possesses the stronger support, it “is virtually impossible to understand

80. Ehrman, Misquoting Jesus, 161-62.

81. Ehrman, Orthodox Corruption, 79.
82. Ibid.
83. The articular θεό? is almost certainly a scribal emendation to the anarthrous θεό?;
θεό? without the article is a much more difficult reading.
84. Ehrman, Orthodox Corruption, 81.
85. Wallace, “Gospel according to Bart," 344—46.

74 I Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

within a Johannine context.”86 He asserts that the reading μονογενή?

θεό? must be translated “the unique God,” resulting in a modalistic
equation of the Son with the Father. Ehrman writes, “The problem, of
course, is that Jesus can be the unique God only if there is no other God;
but for the Fourth Gospel, the Father is God as well. Indeed, even in this
passage the μονογενή? is said to reside in the bosom of the Father. How
can the μονογενή? θεό?, the unique God, stand in such a relationship
to (another) God?”87
Interestingly, despite Ehman’s objections, the translators that
prefer the reading μονογενή? θεός have not found it impossible to
understand within a Johannine context. They translate μονογενής as
a substantival adjective in apposition to θεός,88 hence as “the unique
one, himself God.”89 Even Ehrman concedes elsewhere that the con­
text of the first chapter of John demands that a distinction between
the person of the Father and the person of the Son be maintained,
both of whom John indicates to be fully God. John 1.1 states, “In
the beginning was the Word, and the Word was with God, and the
Word was God’,’ demonstrating that the Word was with God, thus
distinct from God, and also fully God.90 This distinction is seen again
in 1.14: “And the Word became flesh and took up residence among
us. We saw his glory—the glory of the one and only who came from
the Father, full of grace and truth.” The Word is from the Father and
therefore not the same as the Father. Again, as Ehrman has noted,
1.18 itself does not allow for a confusion of the members of the
Godhead, for Jesus is declared to be distinct from the Father since
he is at the Father’s side: "the unique one, himself God, who resides
in the Father’s bosom, has made him known.” Thus, given the strong
contextual distinction between the Father and the Son in the first
chapter of John, the reading μονογενή? θεό? does not necessarily
lead to a modalistic understanding of the Godhead. This reading
is therefore the harder reading, but not an impossible reading as
Ehrman has indicated.

86. Ehrman, Orthodox Corruption, 79.

87. Ibid., 80.
88. For a defense of understanding μουογβνήε as a substantival adjective, see Wallace,
“Gospel according to Bart,” 344-46.
89. E.g., the NET reads, “No one has ever seen God. The only one, himself God, who is
in closest fellowship with the Father, has made God known" (italics mine); the NIV
reads, “No one has ever seen God, but God the One and Only, who is at the Father's
side, has made him known” (italics mine); with slight variation, the ESV reads, "No
one has ever seen God; the only God, who is at the Father's side, he has made him
known” (italics mine), and it offers in the footnotes, “the only One, who is God.”
90. It is preferable to understand as qualitative in John Lie. See Daniel B. Wal­
lace, Greek Grammar beyond the Basics: An Exegetical Syntax of the New Testa­
ment (Grand Rapids: Zondervan, 1996), 266-69.

Philip M. Miller Chapter 2 / 75

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

Ehrman’s discussion of John 1.18 demonstrates more of how the

canon of unorthodoxy functions in his methodology:

1. It is worth noting that Ehrman relies heavily on internal evi­

dence, which brings with it a greater degree of subjectivity.

2. Ehrman paints the textual decision in John 1.18 as the

choice between two unorthodox variants. On the one hand, Ehrman
suggests that the earliest reading, ό μονογενή? υιό?, was under­
stood as adoptionistic. On the other hand, Ehrman indicates that
ό μονογενή? θεό? would result in a modalistic reading incompat­
ible with Johannine theology. According to Ehrman’s perspective,
no matter how the text is reconstructed, the text is less than or­
thodox or at least lends itself to supporting heterodox Christology.
If Ehrman is correct that an orthodox scribe was responsible for the
changes in the text of John 1.18, then regardless of which reading
is original, an orthodox scribe was responsible for producing a less

3. Ehrman confuses a reading that contains a strong orthodox

statement of Christ’s deity in John 1.18 with modalistic theology. At
best, this seems to indicate a lack of evenhandedness in his exegetical
method given the wider context of John’s Gospel.

4. As quoted earlier, Ehrman speaks of the hypothesized orthodox

corruption of this passage as factual, when it is, in fact, debated. This
presentation of the data overstates the conclusiveness of his case.

5. Given the chance to argue for either an anti-adoptionistic or

anti-modalistic corruption, Ehrman has chosen to argue for the former.
This is no doubt partly due to the early dating of?)66 and £>75, manuscripts
that cannot possibly represent fully developed Modalism. However, one
can imagine that Ehrman could have easily argued that ό μονογενή?
υιό? was an anti-gnostic or anti-patripassianistic orthodox corruption?1
What is significant is that here, like elsewhere, Ehrman has argued for
the reading with the lowest Christology.

6. In terms of following his two-step methodology, Ehrman seems

to fall short of the mark. While he offers a number of arguments to sup­
port his favored reading, it is only when he turns to his thesis of or­
thodox corruption that “the real superiority of ό μονογενή? υιό? shines

91. See Ehrman’s discussion of the flexibility of categorizing these corruptions in Or­
thodox Corruption, 282 n. 16.

76 / Chapter 2
 A New Canon for New Testament Textual Criticism?

forth.”92 In making this move—a move more severe than the one made in
Matthew 24.36—Ehrman trades his preferred method for a questionable
one. However, this trade is necessary if he is to offer a compelling case for
orthodox corruption in John 1.18.

Ehrman's confidence that δ μονογενή? υιό? is the reading of choice

seems to stem not from the conclusiveness of the evidence but from its
coherence with the canon of unorthodoxy: the least orthodox reading is
to be preferred,

Hebrews 2.9b

In his discussion of Hebrews 2.9b,9394 Ehrman maintains that the

proto-orthodox and Gnostics were debating during the second cen­
tury over the significance of Jesus’ death.* The Gnostics argued that
the divine element “Christ” had left the man Jesus prior to the cruci­
fixion and therefore that Jesus died without “Christ.”95 In contrast, the
orthodox affirmed that Jesus and the Christ were one and the same,
both in life and in death.96 With this backdrop, Ehrman argues that
Hebrews 2.9b originally said that Jesus would experience death "apart
from God” (χωρίς· θεού) on behalf of everyone:97 “The Gnostics could
readily take die original text to mean that the divine element within
Jesus had already left him prior to his death, so that he died ‘apart’
from God, that is, abandoned by that divine being who had sustained
him during his ministry.”98 Ehrman concludes, “[I]t appears that the
scribes of the second century who recognized the heretical potential
of the phrase χωρΐ? θεού changed it by making the simple substitu­
tion of... χάριτι θεοΰ [“by the grace of God”], thereby effecting an
orthodox corruption that came to dominate the textual tradition of the
New Testament.”99
There are two readings offered for Hebrews 2.9b:

...by God’s grace [χάριτι θεοΰ] he would experience death on behalf

ofeveryore. —

92. Ehrman, Orthodox Corruption, 79.

93. Ehrman discusses Hebrews 2.9b in Orthodox Corruption (146-50) and Misquoting
Jesus (144-48).
94. Ehrman, Orthodox Corruption, 150.
95. Ibid. This understanding of the crucified Jesus being separated from the immor­
tality of die divine is represented well in the Apocalypse of Peter (James M. Rob­
inson, The NagHammadi Library, rev. ed. [San Francisco: HarperCollins, 1990]).
96. Ehrman, Orthodox Corruption.
97. Ibid., 149.
98. Ibid., 150.
99. Ibid.

Philip M. Miller
 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

... he would experience death apartfrom God [χωρίς θεού] on behalf

of everyone.

FIGURE 2.4: TWO READINGS IN HEBREWS 2.9b

χάριτι θεού p46 X A Β C D Ψ 075 0150 6 33 81 104 256 263 365 424*
436 459 1175 1241 1319 1573 1739v r?!d 1852 1881 1912
1962 2127 2200 2464 Byz [KL P] Leet it*·b· ά·v vg
C0R1p·

gyjp.h.pa! COpsa,bo,fay ggQ g|ay acc.ioOngcn Qngen^

Athanasius Didymus Chrysostom mss KC-ro'n>eodore Cyril
Theodoret1/2 Ps-Oecumenius; Faustinus Jerome

χωρίς θεού 0243 (0121b) 424cvid 1739“· vgms syr’”nss Origen*v r’kt
mssscc·10 Theodore Nestoriansacc·
TheodoretW:,sra; Ambrose mssacc'ofa°me Vigilius Fulgentius

The external evidence for the first reading is nearly conclusively in

favor of χάριτι θεού as the preferred reading. This is one of the rea­
sons the UBS4 committee gave χάριτι θεού an “A” rating, “indicating
that the text is certain.”100 Not only are the earliest and best witnesses,
such as Sp46 a A B, united in their support of this reading, but nearly all
manuscripts support this reading. The reading Ehrman proposes, how­
ever, has external support from only three Greek witnesses from the
tenth century and later.101 Of these three, the only significant witness
is Codex 1739, which is a copy of an early and important manuscript.
This reading is also discussed by several church fathers and given sup­
port by one Vulgate manuscript and some copies of the Syriac Peshitta.
In discussing this problem, Daniel B. Wallace notes, “[WJith external
evidence this weal; many textual critics would dismiss this reading
entirely.”102 Ehrman acknowledges the difficulty in entertaining this
poorly attested variant when he writes that in “the case of Hebrews 2:9
there is a direct clash between these two lands of evidence [external and
internal]. Although the surviving documents are virtually uniform in
stating tirat Jesus died for all people ‘by the grace of God’ (χάριτι θεού),
the force of the internal evidence compels us to accept as original the
poorly attested variant reading, which states that Jesus died ‘apart from

100. For an explanation of the letter rating system used by the UBS4 committee, see n.
115.
101. Ehrman identified only MSS 0121b and 1739 in support of his favored reading (Or­
thodox Corruption, 146), although he acknowledged the presence of MS 424cv«i in
an endnote (176 n. 132).
102. Wallace, "Gospel according to Bart,” 337.

V ♦ ...... / Chapter 2
V “ 78 Philip M. Miller
 A New Canon for New Testament Textual Criticism?

God’ (χωρίς θεού).”103104Compellingly strong internal evidence can out­

weigh strong external evidence. In this case, Ehrman finds the internal
evidence to be compelling and decisive.
Concerning the internal evidence, Ehrman is wise to probe for
some explanation for the reading χωρίς θεού, since it is the harder
reading and is fairly early, as evidenced by the patristic quotations and
Codex 1739.1M There are three proposed explanations for how these
readings arose. The first explanation proposed by Ehrman offers that
the original reading, χωρίς θεού, was modified by orthodox scribes to
read χάριτι θεοΰ because the original reading seemed to affirm a het­
erodox belief that Jesus was separated from die divine Christ prior to
his death.105 Ehrman is confident tiiat “the external evidence notwith­
standing, Hebrews 2:9 must have originally said that Jesus died 'apart
from God?”106
However, there are two other viable solutions to this textual problem
that call into question Ehrman’s certainty. A second explanation is that
χάριτι was copied as χωρίς due to scribal lapse or carelessness.107 This
view is easier to understand when the text is seen in majuscule script:
χ^ριτι·θ·γ//χιι>ριθ·θ·γ. Thus the change between χάριτι and χωρίς may
be accidental.
A third and more persuasive explanation is that the phrase χωρίς
θεού was a marginal note on Hebrews 2.8b that was confused by a later
scribe as a correction for χάριτι θεού in 2.9b.108 Hebrews 2.8b quotes
from Psalm 8.6 when it says, “when he subjected all things to him, he
left nothing that is not subject to him.” A similar passage is found at 1
Corinthians 15.27-28, which also quotes from Psalm 8.6: “For ‘he has
put everything in subjection under his feet.’ But when it says ‘everything’
has been put in subjection, it is evident that this does not include the one
who put everything into subjection to him. And when all things are sub­
jected to him, then the Son himself will be subjected to the one who sub­
jected everything to him, so that God may be all in all.” In other words,
Paul argues that all things will be subject to Jesus except for the Father.
The theologian reading Hebrews 2.8b—“when he subjected all things to
himrhe leftmothingthat is not subject to him”—may have remembered 1

103. Ehrman, Orthodox Corruption, 146 (italics mine).

104. Krista Miller has demonstrated that the patristic evidence, which is often cited as
a basis for entertaining χωρίς 9eou as a viable variant, actually offers stronger sup­
port for the reading χάριτι 9c ου upon close examination (Krista M. Miller, “Evalu­
ating the Reading Χωρΐ$ Gteov in Hebrews 2:9 in Light of Patristic Evidence” [ThM
thesis, Dallas Theological Seminary, 2010]).
105. Ehrman, Orthodox Corruption, 150.
106. Ibid., 149 (italics mine).
107. Metzger, Textual Commentary, 2nd ed.T 594.
108. Ibid.; F. E Bruce, The Epistle to the Hebrews, rev. ed., New International Commen­
tary on the New Testament (Grand Rapids: Eerdmans, 1990), 32,70-71.

Philip M. Miller Chapter 2 I 79

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

Corinthians 15.27-28 and written χωρ ις θεοΰ, "except for God,” as an in­
terpretive comment in the margin. When a scribe copied the manuscript
with the marginal notation, he may have understood it to be a correction
for χάριτι θεοΰ in the following verse. A simple error in substitution can
therefore explain the rise of χωρίς· θεοί as a secondary reading in this
passage.109
In his analysis of Hebrews 2.9b, Ehrman reveals how determinative
the canon of unorthodoxy is for his methodology.

1. As Ehrman himself acknowledges, he is placing the burden of

his decision on the internal evidence, over and against solid external
evidence to the contrary. In the end, this continued reliance on internal
evidence makes Ehrman’s methodology increasingly subjective.

2. While Ehrman cites this passage as a clear example of orthodox

corruption, there are two other plausible explanations for the textual ev­
idence. Hebrews 2.9b is therefore not as clearly in support of Ehrman’s
thesis as he seems to portray.

3. Ehrman’s confidence level seems improper given the textual

merits supporting the other reading. As quoted earlier, Ehrman uses
words like “compels” and “must” to affirm a reading that many textual
critics deem untenable in light of the overwhelming external and in­
ternal evidence to the contrary.110

4. Ehrman overestimates the theological importance of this

variant to historic orthodoxy. Elsewhere, when arguing for χωρίς θεού
as the best reading, Ehrman notes that it fits well with the picture of
Jesus in Hebrews that “repeatedly emphasizes that Jesus died a fully
human, shameful death, totally removed from the realm whence he
came, the realm of God. His sacrifice, as a result, was accepted as the
perfect expiation for sin. Moreover, God did not intervene in his pas­
sion and did nothing to minimize his pain. Jesus died “apart from God.”’111
This description of Christ offered by Ehrman is in complete accord with·
historic orthodoxy.

5. With regard to his two-step method, Ehrman simply falls flat

in Hebrews 2.9b. Ehrman seems to find evidence for orthodox cor­
ruption there precisely because he is looking for it. In studying this

109. Metzger, Textual Commentary, 2nd ed., 594; Bruce, Epistle to the Hebrews, 32,
70-71; Hans-Friedrich Weiss, Der Brief an die Hebrder, Kritisch-exegetischer
Kommentar uber das Neue Testament 13 (Gottingen: Vandenhoeck und Ruprecht,
1991), 200’202.
110. Metzger, Textual Commentary, 2nd ed., 594.
111. Ehrman, Orthodox Corruption, 149.

80 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

textual variant, it is unlikely that someone would conclude that the

orthodox scribes systematically purged the original reading from all
but three tenth-century Greek manuscripts. It seems that in his desire
to find orthodox corruption, Ehrman ferrets out passages that illus­
trate his point?12 This may be a case of “begging the question,” a logical
fallacy in which the proposition to be proven is assumed implicitly
or explicitly in one of the premises. In this case, Ehrman’s premise
is that orthodox corruption occurred in this passage. This premise is
what allows him to see merit for the reading χωρίς θεού despite its
paltry support. Once this reading is given credence, he can argue for
its originality. His argument for its originality then becomes the basis
for his conclusion that orthodox corruption has occurred in this pas­
sage. His argument is circular. He has presupposed his conclusion. In
proving his case for orthodox corruption in Hebrews 2.9b, Ehrman
must presume that it did take place. Otherwise, a variant like χωρίς
θεοΰ would never have the internal evidence to muster consideration.
In this example, it appears that Ehrman has deviated from Iris method
of choice.

Ehrman’s confident dismissal of the almost conclusive evidence

against his favored reading indicates that he has embraced the canon
of unorthodoxy. Affirming that the least orthodox reading is to be pre­
ferred, Ehrman is able to dismiss the strength of the external evidence
in favor of the canon that is more dominant for him. The canon of un­
orthodoxy holds such influential sway over Ehrman’s methodology that
he can dismiss the evidence for χάριτι θεού that the USB4 committee
found to be persuasive to the point of certainty.

Other Texts with Extended Treatment

The limits of this chapter do not permit an exploration of all the tex­
tual variants Ehrman discusses in tire Orthodox Corruption ofScripture.
However, the methodological patterns mentioned here are present in
many of Ehrman’s discussions. The following table shows the passages
to which Ehrman gives specific and extended treatment in Orthodox112

112. Wallace calls this "Ehrman's overall agenda of exploiting the apparatus for or­
thodox corruptions, regardless of the evidence for alternative readings” (“Gospel
according to Bart,” 338 n. 40). However, Wallace has softened his stance on He­
brews 2,9b, in part because the father of reasoned eclecticism, GOnther Zuntz,
prefers this reading (Zuntz, The Text ofthe Epistles: A Disquisition upon the Corpus
Paulinum [Oxford: Oxford University Press, 1953], 34-35). Zuntz bases some of
his preference off the patristic evidence, which has been recently examined and
found to better support χάριτι θβοϋ (Miller, “Evaluating the Reading Χωρ'ι? Θεοϋ
in Hebrews 2:9”).

Philip M. MiUer Chapter 2 / 81

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

Corruption113 While dozens of additional passages receive ample atten­

tion, these seem to be prominent in Ehrman’s argumentation.

FIGURE 2.5: PASSAGES WITH TEXTUAL VARIANTS

DISCUSSED IN ORTHODOX CORRUPTION

Least Orthodox
NA27/UBS4
Textual Reading
Agreement UBS4 Rating115
Problem114 Preferred by
with Ehrman?
Ehrman?

Luke 3.22 Yes No B

Mark 1.1 Yes No . c
1 Timothy 3.16 Yes Yes A
John 1.18 Yes No B
1 John 4.3 Yes No Unrated
Mark 15.34 Yes Yes B
Hebrews 2.9b Yes No A
Luke 22.43-44 Yes Yes A
Luke 22.19-20 Yes No B
Luke 24.12 Yes No B
Luke 24.51-52 Yes No B

In light of the tabulated results of Ehrman's text-critical method­

ology, some key observations are apparent.

113. This table contains all textual discussions that bear their own textual heading, de­
noting the prominent place they hold in Ehrman’s discussion.
114. These textual problems are arranged according to the order in which Ehrman dis­
cusses them in Orthodox Corruption on the following pages, respectively: 62^67/
72-75,77-78,78-82,125-35,143-75,146-50,187-94,198-209,212-17,227-32.
115. The UBS4 committee explains the letter grades used to indicate their level of con­
fidence in the reconstructed text as follows: “The letter A indicates that the text is
certain. The letter B indicates that the text is almost certain. The letter C, however,
indicates that the Committee had difficulty in deciding which variant to place in
the text The letter D, which occurs only rarely, indicates that die Committee had
great difficulty in arriving at a decision" [The Greek New Testament, 4th ed. [Stutt­
gart: Deutsche Bibelgesellschaft, 1993], 3*). Thus the letter assignment in column
4 of fig. 2.5 corresponds to the reading accepted by the UBS committee in column
3; a “No” in the third column indicates that die UBS committee's preferred reading
is different from Ehrman’s, and the letter assignment signifies the level of certainty
the committee had in their decision. The higher dieir certainty is, die more signifi­
cant Ehrman’s dissent is.

82 / Chapter 2 Philip M. Miller

*
 A New Canon for New Testament Textual Criticism?

1. Ehrman uniformly argues for what he believes is the least or­

thodox reading. This is to be expected, since Ehrman is arguing for the
orthodox corruption of Scripture. However, this shows the consistency
with which Ehrman is applying the canon of unorthodoxy.

2. In over 72 percent of these examples, Ehrman’s textual decisions

are in disagreement with the NA27/UBS4 text. While this text form is far
from definitive, it does represent mainstream NT text-critical scholar­
ship. Of note is the fact that the late Bruce Metzger, whom Ehrman
regarded as “the world’s leading expert in the field,”116 was influential in
the shaping of the NA27/UBS4 text. It appears that Ehrman’s increasing
commitment to the canon of unorthodoxy may be disassociating him
from the NA27/UBS4 text and from the views of his mentor.

3. Not only does Ehrman frequently disagree with the NA^/UBS4

text, but he dissents from one “Unrated,” one “A,” five “B”s, and one
“C” rated decisions. This means that in 63 percent of these examples,
Ehrman disagrees with what the UBS4 committee considered to be read­
ings that were “certain” or “almost certain.”117 If Ehrman’s methods are
regularly resulting in textual decisions that are at odds with the NA27/
UBS4 text, especially "A”- and “B”-rated readings, it calls into question
the value and legitimacy of that methodology.

4. In every case that Ehrman disagrees with the NA27/UBS4

text, there are compelling reasons for choosing the more orthodox
reading, which happens to be the reading found in the NA27/UBS4 text.
Furthermore, in each case where Ehrman is listed in agreement with
the NA27/UBS4 text, there are alternate explanations for the textual evi­
dence aside from orthodox corruption.118

Ehrman’s application of the canon of unorthodoxy not only es­

tranges him from the widely accepted NA27/UBS4 text but also fails to
prove his thesis. To be sure, Ehrman discusses passages where theologi­
cally-motivated alterations played a role in the transmission of Scripture.
However, these passages are not nearly as significant as one might ex­
pect. The confidence of Ehrman’s conclusions is derived not primarily

116. Ehrman, Misquoting Jesus, 7.

117. For calculation purposes, the “Unrated" reading is considered as an “Aerated
reading, since the UBS4 committee did not consider the problem worthy of discus­
sion because of its certainty.
118. 1 Tim 3.16 could be understood as an unintentional error; rather than an or­
thodox corruption, although I consider this explanation dubious. Mark 15.34 does
not appear to be an issue of orthodoxy at all. The text of Luke 22.43-44 could be
noncanonical historical material that found its way into the text independent of
theologically motivated alterations.

Philip M. Miller Chapter 2 / 83

 THE LEAST ORTHODOX REAPING IS TO BE PREFERRED:

from the force of the evidence but from the presupposed canon, of un­
orthodoxy. For Ehrman, it appears that the least orthodox reading is to
be preferred and that this presupposed canon often results in textual
decisions that are at odds with the mainstream reconstruction of the
text.

THE CANON OF UNORTHODOXY:

INITIAL CRITIQUE, VALUE, AND ROLE

In light of the evidence that Ehrman utilizes the canon of unortho­

doxy in his assessment of textual data, this third section offers an initial
critique of his canon and an assessment of its value and role within NT
textual criticism. This critique will focus, first, on Ehrman’s thesis and,
second, on the canon of unorthodoxy itself.

A Critique ofEhrman’s Thesis

To argue that the canon of unorthodoxy should be accepted, several

tenets must be demonstrated to be valid.

1. A compelling sociohistorical setting in which scribes were per­

suaded to alter the NT texts must be established historically.

2. The textual problems in question must prove themselves to be

significant, even central, to the Christological debates.

3. These variant readings must be demonstrated to be the result

of intentional scribal activity with a theological agenda, rather than of
unintentional or alternate sources of variation.

4. Theologically motivated alterations must account for a signifi­

cant number of variants in order for the canon of unorthodoxy to carry
substantial weight in navigating text-critical problems.

If the canon of unorthodoxy is to be accepted, all four of these tenets

must be shown to be probable. This initial critique centers on these four
tenets and is followed by a critique of the canon itself.

1. The first area of critique relates to the proposed sociohistor­

ical setting in which orthodox corruption is said to have taken place.
Because orthodox Christology involves truths held in tension (e.g., the
hypostatic union), there are many ways to assemble the NT teachings in­
correctly, resulting in unorthodox doctrine. The first three centuries are
replete with discussion of these doctrinal tensions as well as the errors

84 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

associated with the overemphasis of one truth to the neglect of another.

This time frame of conflicting theologies provides Ehrman with a rich
supply of options for postulating a sociohistorical setting to explain the
textual data. If the variant affirms the deity of Christ, Ehrman can call
it anti-adoptionistic. If it affirms his humanity, Ehrman can call it anti-
docetic. If it affirms the Son’s oneness with the Father, Ehrman can call
it anti-separationist. If it affirms the distinction between the Father and
the Son, Ehrman can call it anti-patripassianistic. In the end, there is
always a hypothetical, sociohistorical setting to which Ehrman can ap­
peal. No matter to which side of the scale a variant seems to weigh,
Ehrman can tip the scales in favor of orthodox corruption.
However, the convenience of this historical setting can cut the
other way,119 as Ehrman acknowledges.120 If the presumed opposing
party changes, the result is an opposite reconstruction of the data121; a
good example is John 1.18. There are always opportunities for finding
orthodox corruptions in a sociohistorical setting where orthodoxy is
in the middle, holding the truth in tension between two opposing par­
ties. Ehrman has offered a compelling case that a sociohistorical setting
existed in which there is a possibility of intentional scribal alterations
in the direction of orthodox theology; whether he has shown its wide­
spread probability is another matter.

2. The second area of critique relates to Ehrman’s ability to show

that these passages are significant, even central, to the Christological
debates. The most obvious way in which this can be proven is to find
these passages or variants quoted in defense of a particular view by
either the church fathers or their opponents. At times, Ehrman does
just this; yet where many of his examples assume that a given passage
was the subject of debate, evidence sufficient to affirm that it was or
was not is simply unavailable. What further complicates this issue is
that it is unnecessary to assume that these various groups appealed to
the NT Scriptures as the primary basis for their theological positions.
Differing groups may have primarily found their theological perspec­
tive in philosophy, the Old Testament, mystery religions, or sociolog­
ical movements. They may not have appealed to the NT as the primary
basis for their theological position. If this is the case, it is less likely that
the orthodox would be compelled to modify the Scriptures they held
in such esteem. Ehrman has argued persuasively that some passages

119. Adam G. Messer, "Patristic Theology and Recension in Matthew 24.36: An Evalu­
ation of Ehrman’s Text-Critical Methodology" (ThM thesis, Dallas Theological
Seminary, 2009), 34,45-46.
120. Ehrman, Orthodox Corruption, 2S2 n. 16.
121. Ibid.

Philip M. Miller Chapter 2/85

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

were central to the Christological debates, and he simply assumes the

centrality of other passages.

3. Concerning the third tenet, there are many cases in which

Ehrman has not presented enough compelling evidence to conclude that
intentional scribal alteration is the best possible explanation for the tex­
tual data. It is hardly valid to deduce that a scribe must have intention­
ally altered the text for theological reasons when the textual evidence
is reasonably explained by alternate hypotheses. Ehrman’s conclusions
are inconclusive in many cases. The best evidence for this is seen in his
frequent dissent from the NA27/UBS4 text. As we have seen, Ehrman ex­
presses a surprising degree of certainty in dissenting from the choices
made by the UBS4 committee. Gordon Fee poignantly states, “[Ujnfor-
tunately, Ehrman too often turns mere possibility into probability, and
probability into certainty, where other equally viable reasons for corrup­
tion exist.”122

4. The fourth postulate is the most difficult to demonstrate, since

it rests on the validity of the preceding tenets. Since Ehrman has not
demonstrated the first three tenets adequately, the fourth is also in
question. At this point, the examples proffered by Ehrman are insuf­
ficient to conclude that the least orthodox reading is to be preferred. It
does not appear that theologically motivated alterations plague the NT
textual tradition to nearly the same degree as other sources of textual
variants. Ehrman acknowledges that theologically motivated variants
most likely number only in the dozens.123 The small number of con­
clusively demonstrated instances of orthodox corruption provides an
inadequate basis for asserting that the canon of unorthodoxy ought to
be accepted.

A Critique of the Canon of Unorthodoxy

Concerning the merits of the canon of unorthodoxy itself, several
critiques can be offered.

1. The canon that the least orthodox reading is to be preferred is

too universal and needs to be nuanced carefully, for the least orthodox
reading is not to be preferred in most cases, in light of the textual evi­
dence. As demonstrated earlier, Ehrman’s application of this canon
seems to have taken him beyond the evidence at times.

122. Fee, review of The Orthodox Corruption ofScripture, 204 (italics his).
123. Ehrman, Orthodox Corruption, 46 n. 124·; however, Ehrman entertained the idea
that they may number into the hundreds.

86 / Chapter 2 Philip M, Miller

 A New Canon for New Testament Textual Criticism?

2. The canon of unorthodoxy seems properly to be understood not

as an independent canon but as a specific extension of Bengel’s canon that
the harder reading is to be preferred. This is interesting, because Ehrman
at times argues for the least orthodox reading over and against the harder
reading (e.g., John 1.18). Since the canon of unorthodoxy seems to be a
derivation of the canon of the harder reading, they should not conflict.

3. A strong emphasis on the canon of unorthodoxy fits best within

rigorous eclecticism. Ehrman’s employment of the canon of unortho­
doxy frequently overrides compelling external evidence. He appears
to be granting increasing weight to the internal canon of unorthodoxy
while decreasing the weight of external evidence, especially the date and
character of witnesses. As Ehrman begins to increasingly emphasize in­
ternal evidence, his approach appears to be moving toward a more rig­
orous eclectic approach to textual criticism.

4. The canon of unorthodoxy rests on the premise of widespread

orthodox corruption. Ehrman has demonstrated that extensive or­
thodox corruption is possible, but its probability seems unlikely.

5. Placing a high value on the canon of unorthodoxy as a premise

seems to bias the textual decision prior to a complete examination of the
evidence. Ehrman frequently argues for the originality of variants that
fit his sociohistorical model of transcription well but suffer from poor
attestation and little likelihood of going back to the original. If Ehrman
did not place a high value on the canon of unorthodoxy, it is unlikely that
many of the variants he chooses would be considered as viable.

In the end, the canon of unorthodoxy appears to be presupposi-

tionally driven. The fact that the UBS4 committee and Ehrman come
to widely different conclusions at so many points is an indication that
different methods are being used to analyze die textual data. It has been
the goal of this chapter to show that the difference is in the applica­
tion- of-the -canon of unorthodoxy, which Ehrman implicitly holds to
while the UBS4 committee does not. As noted earlier, the UBS4 com­
mittee acknowledges the reality of theologically motivated corrup­
tions.124 However, the committee does not appear to be guided by the
canon of unorthodoxy in the same way that Ehrman is. It seems that
the difference in textual conclusions derives less from the weighing of
textual data and more from Ehrman’s commitment to the canon of un­
orthodoxy. The result is that Ehrman finds orthodox corruption in a
far greater number of passages than does the UBS4 committee. That
Ehrman does not see himself as allowing presuppositions to enter into

124. Metzger, Textual Commentary, 2nd ed., 52,56,164-66,185,200.

Philip M. Miller Chapter 2/87

 THE LEAST ORTHODOX READING IS TO BE PREFERRED:

his methodology is made clear when he writes, “[T]o use orthodoxy

as a criterion would involve presupposing my conclusions.”125 However,
it seems that “orthodoxy as a criterion" is exactly what Ehrman is em­
ploying in at least some of the cases discussed in this chapter. Martin
Hengel’s warning about the danger of allowing presuppositions to drive
methodology comes to mind. Hengel warned of the danger of both “an
uncritical, sterile apologetic fundamentalism” and the “no less sterile
'critical ignorance’” of radical liberalism.126 In both cases, the approaches
are the same; it is the presuppositions that set them apart. This critique
does not fully apply to Ehrman, but his textual decisions seem to indi­
cate a methodology shaped by presupposition.

One of the perennial dangers that confront scholars in every discipline

is the tendency to become one-sided and to oversimplify the analysis
and resolution of quite disparate questions. In textual criticism, this
tendency can be observed when a scholar, becoming enamored of a
single method or criterion of textual analysis, applies it more or less
indiscriminately to a wide variety of problems.127

CONCLUSION
Ehrman’s thesis concerning orthodox corruption of Scripture, while
innovative and thought provoking, still remains hypothesis. Where he
has persuasively argued for it, the texts in question are not as central to
the message of the NT or to orthodox theology as Ehrman indicates.
As Wallace concludes, “Regarding the evidence, suffice it to say that
significant textual variants that alter core doctrines of the NT have not
yet been produced.”128129In the end, it seems that Ehrman’s conclusions
extend beyond the evidence and may have distorted his perspective of
the textual transmission. In addition, the canon of unorthodoxy itself
appears to be inadequate due to the lack of evidence supporting it and
the distorting force it appears to apply to textual analysis.
In conclusion, two seemingly divergent positions must be affirmed.
On the one hand, theologically motivated alterations are a reality and
are the best explanation for some textual variants?29 These theologically

125. Ehrman, "RE: Greenlee’s strawman (??) in The Text of the NT”
126. Martin Hengel, Studies in Early Christology (Edinburgh: T&T Clark, 1995), 57-58.
127. Metzger and Ehrman, Text of the New Testament, 304.
128. Wallace, “Gospel according to Bart,” 346-47 (italics his). While Ehrman has pro­
duced a fascinating survey of theologically motivated alterations, when it comes
to the key evidence to cement his argument, Ehrman still has not found what he is
looking for.
129. Possible examples include Matt 24.36; 27.16—17; Mark 1.41; Luke 2.33; 2.38; 8.3;
11.4; 22.43f; 24.51; 24.53; John 3.13; 11.33; Acts 1.14; 2.41; 4.24; 5.32; 9.22; 15.29;
Rom 9.4; Gal 1.3; 1 Peter 1.22.

88 / Chapter 2 Philip M. Miller

 A New Canon for New Testament Textual Criticism?

motivated alterations are best accounted for in Bengel’s maxim that the
more difficult reading is to be preferred)™ The harder reading for or­
thodox scribes would indeed have been unorthodox or suborthodox
131 Although the importance Ehrman ascribes to these altera­
readings.130
tions has been overstated, he is correct that theologically motivated
alterations occurred on occasion. On the other hand, the canon of un­
orthodoxy seems to move beyond the evidence.
How, then, can these alterations be understood? It seems best to
affirm that orthodox scribes were inclined toward reverence for Christ
and were committed to their faith. At times, that reverence and com­
mitment influenced their scribal activity, introducing alterations into
the textual tradition. For the most part, these alterations do not appear
to be sourced in malice, deceit, or agenda. These corruptions do not
“establish the orthodox character of the text”132 but, rather, clarify and
confirm that character. While important questions remain, it is signifi­
cant to note that no cardinal doctrine of orthodoxy is at stake in light of
the variants presented by Ehrman. He concedes that most cases of tex­
tual variation do not affect the orthodox understanding of the physical
resurrection and deity of Jesus Christ.133
In the end, we can conclude that the least orthodox reading, by it­
self, is not a viable canon for determining the preferred reading.

130. Johann Albrecht Bengel, Η ΚΑΙΝΗ ΔΙΑΘΗΚΗ: Novum Testamentum Graecum

ita adomatum ut Textus probatarum editionem medullam, Margo variantium
lectionum in suas classes distributarum locorumque parallelorum delectum, ap­
paratus subjunctus criseos sacrae Millianae praesertim compendium limam
suppiementum ac fractum exhibeat, ed. Johann Albrecht Bengel (Tubingen: I. G.
Cottae, 1734), 433.
131. Head, “Christology and Textual Transmission,” 109.
132. Centra Ehrman, Orthodox Corruption, 28.
133. In the final Question & Answer session at the fourth annual Greer-Heard Point-
Counterpoint Forum (*The Textual Reliability of the New Testament: A Dialogue
between Bart Ehrman and Daniel Wallace” April 4-5, 2008), I had the following
recorded exchange with Ehrman:
Miller: “Dr. Ehrman, at this point in scholarship does the... in your opinion,
”~dees“the -earliest reconstructible form of the text portray an orthodox under­
standing of, say, the resurrection and the deity of Christ?”
Ehrman: "I’m not sure what the orthodox understanding of... You mean that
Jesus was physically raised from the dead and that he was God?"
Miller: ‘"That Jesus was bodily raised from the dead and then that he’s both
God and man.”
Ehrman: “I don’t think that the... I don’t think that the text... the texts af­
fect that one way or the other. I think... My own view is that the biblical authors
thought that Jesus was physically raised from the dead. My own view is that most
of the biblical authors did not think that Jesus was God. The Gospel of John does. I
think Matthew, Mark, and Luke do not think that Jesus was God. It’s hard to know
what Paul’s view about Jesus’ divinity is, in my opinion. So I think different authors
had different points of view, but I don't think... in most cases I don’t think that it’s
affected by textual variation.”

Philip M. Miller Chapter 2 / 89

 3

THE LEGACY OF A LETTER

Sabellianism or Scribal Blunder in John 1.1c?

Matthew P. Morgan

egarding the study of NT transmission, it has been observed that

R “manuscripts have many more stories to tell if you listen closely to

the sounds of the details they preserve for today’s world”1 Although
the study of manuscripts has a rich heritage, the smaller, often more
obscure stories, contained in every manuscript, remain untold; deserv­
edly, the majority of attention has focused on the character, date, and
quality of text preserved in a particular manuscript.2 Thus the majority
of these investigations seldom deal with the historical perspective con­
tained in the transmission of a given document; instead, their greater

1. Thomas J. Kraus and Tobias Nicklas, “The World of New Testament Manuscripts:
‘Every Manuscript Tells a Story/” in New Testament Manuscripts: Their Texts and
Their World, ed. Thomas J. Kraus and Tobias Nicklas, Texts and Editions for New
Testament Study 2 (Leiden: Brill, 2006), 4.
2. See, eg., Ernest Cadman Colwell, Studies in Methodology in Textual Criticism
of the New Testament, New Testament Tools and Studies 9 (Leiden: Brill, 1969);
Colwell, “Genealogical Method: Its Achievements and Its Limitations’' JBL 66
(1947): 109-33; Colwell, “The Significance of Grouping of New Testament Manu­
scripts,” NTS 4 (1957—58): 73—92; Eldon Jay Epp and Gordon D. Fee, eds., New
Testament Textual Criticism: Its Significancefor Exegesis; Essays in Honor ofBruce
MetzgerfNew Ybrlc Oxford University Press, 1981); Eldon Jay Epp, Perspectives on
New Testament Textual Criticism: Collected Essays, 1962-2004, Supplements to
Novum Testamentum 116 (Leiden: Brill, 2005); Gordon D. Fee, “Textual Criticism
of the New Testament,” in Studies in the Theory and Method of New Testament
Textual Criticism, ed. Eldon Jay Epp and Gordon D. Fee (Grand Rapids: Eerdmans,
1993), 3—44; Bart D. Ehrman, “Methodological Developments in the Analysis and
Classification of New Testament Documentary Evidence,” NovT 29, no. 1 (1987):
22-45; Kurt Aland and Barbara Aland, The Text of the New Testament: An Intro­
duction to the Critical Editions and to the Theory and Practice ofModem Textual
Criticism, trans. Erroll F. Rhodes, 2nd ed. (Grand Rapids: Eerdmans, 1989); Bruce
M. Metzger and Bart D. Ehrman, The Text of the New Testament: Its Transmission,
Corruption, and Restoration, 4th ed. (New York: Oxford University Press, 2005);
Brook Foss Westcott and Fenton John Anthony Hort, The New Testament in the
Original Greek: Introduction [and] Appendix (London: Macmillan, 1882).

91 / Chapter 3
 THE LEGACY OF A LETTER

concern is its contribution or place in the broader pursuit of the "orig­

inal” text.
Although the study of scribal habits provides a personalized
look at individual manuscripts, it is often limited to significant early
manuscripts or families of manuscripts.3 In addition, work on scribal
habits typically pursues the historical nature of the transmission pro­
cess with the aim of forming helpful canons of transcriptional prob­
ability. As a result, little attention is given to the role and testimony
of individual manuscripts and even some of the unique readings they
record.
Within the transmissional history of John 1.1, a single variant
emerges from the vast array of manuscripts containing this sacred text.4
More specifically, this lone variant does not stem from a deviation con­
sisting of a phrase, a word, or even a collection of letters. Instead, its
origin is the addition of a single letter, the article, so that the phrase και
θεό? ήν ό λόγο? reads και ό θεό? ήν ό λόγο? in two eighth-century wit­
nesses, Codex Regius (L [019]) and the first quire of Codex Freerianus
(Ws [032-S]).5
If considered original, this reading with the article has serious
implications for Christology. As the grammarian A. T. Robertson
rightly pointed out, “It is true also that και ό θεό? ήν ό λόγο? (con­
vertible terms) would have been Sabellianism.”6 If correct, this reading
threatens the assertion that evangelical doctrine is unaffected by any
variant.7 At the same time, if the evidence shows that it preserves the

3. Ernest Cadman Colwell, “Method in Evaluating Scribal Habits: A Study of §34S, $546,
ip75” and "Hort Redivivus: A Plea and a Program,” in Studies in Methodology in Tex­
tual Criticism of the New Testament (Leiden: Brill» 1969); James R. Royse, Scribal
Habits in Early Greek New Testament Papyri, New Testament Tools, Studies, and
Documents 36 (Leiden: Brill, 2003).
4. At first glance, it would appear that no variant exists in this verse, since the stan­
dard Greek texts, NA27 and UBS4, do not list a single variant until John 1.3. An­
other critical text (Constantinus Tischendorf, Novum Testamentum Graece, 8th
ed. [1869; repr., Graz: Alcademische Druck-u. Verlagsanstalt, 1965], loc. cit) gives
several citations from the church fathers but lists only one variant for the entire
verse. In a recent publication (U. B. Schmid» W.). Elliott, and D. C. Parker» eds.,
New Testament in Greek IV: The Gospel according to St John, vol. 2» The Ma­
juscules, New Testament Tools, Studies, and Documents 37 [Leiden: Brill, 2007],
189), this same variant is found along with the absence of the nomen sacrum in
1.1b, where the abbreviation θν has been replaced by fteov in the ninth-century
manuscript Codex M (021).
5. The date for Ws is debated. For a more detailed discussion, see “The Place of Co­
dices Regius and Freerianus in Transmission History” below.
6. A. T. Robertson, A Grammar ofthe Greek New Testament in the Light ofHistorical
Research, 4th ed. (Nashville: Broadman, 1934), 767-68.
7. Metzger and Ehrman noted that Johann Albrecht Bengel “came to the conclusions
that the -variant readings were fewer in number than might have been expected and
that they did not shake any article of evangelical doctrine” (Metzger and Ehrman,
Text ofthe New Testament, 158; italics mine). Mor6 recently, this view is echoed in

92 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John Lie?

remnant of an early Sabellian ancestor, it also poses a clear challenge

to the more recent claim that evangelical doctrine is immune to any
viable variant.*8 As a result, to determine the legacy of this loaded
letter in these two manuscripts, it is necessary to listen carefully to
their stories.

THE APPROACH OF THIS STUDY

Three different perspectives inform the legacy and heritage of these
two eighth-century manuscript witnesses and the text of John 1.1c:

1. The historical use of John's Gospel will be examined in the

polemical environment of Sabellianism and its corollaries.9 From this
survey, the date and significance of this particular passage in the de­
bates will be determined. As the dust settles on the issue of chronology,
it is then possible to evaluate the potential threat of scribal influence
during the period of greatest textual turbulence.

2. Scribal behavior and tendencies will reveal the nature and

quality of transmission. If the legacy in this variant lies in a Sabellian
Christology, this theological bias should be manifest throughout
both manuscripts. Thus, as Hort aptly stated, “KNOWLEDGE OF
DOCUMENTS SHOULD PRECEDE FINAL JUDGMENT UPON
READINGS.”10 Subsequently, both Colwell11 and, more recently, Royse12
have continued the rich tradition pioneered by Hort. Royse wrote, ‘As
Hort’s own comments make clear, knowledge of the sorts of errors that
a particular scribe tended to make, and of his overall method and ac­
curacy of copying, is an essential portion of this ‘knowledge of docu­
ments.’”13 That being the case, it is important to examine this loaded
letter through the eyes of the two eighth-century scribes who produced
the two manuscripts. Through a detailed analysis of their unique scribal
proclivities, their legacy will become apparent.

D. A. Carson and. Douglas J. Moo, An Introduction to the New Testament, 2nd ed.
(Grand Rapids: Zondervan, 2005), 31.
8. J. Ed Komoszewski, M. James Sawyer, and Daniel B. Wallace, Reinventing Jesus:
How Contemporary Skeptics Miss the Real Jesus and Mislead Popular Culture
(Grand Rapids: Kregel, 2006), 117.
9. Hurtado wrote, “Particularly in the christological disputes of the early centuries,
it [Gospel of John] was unexcelled as the favorite arsenal of textual ammunition
(often by both sides of the disputes!)" (Larry W. Hurtado, Lord Jesus Christ: Devo­
tion to Jesus in Earliest Christianity [Grand Rapids: Eerdmans, 2003], 349).
10. Westcott and Hort, New Testament in the Original Greek, 31.
11. Colwell, "Redivivus,” 148-71.
12. Royse, Scribal Habits.
13. Ibid., 1.

Matthew P. Morgan Chapter 3/93

 THE LEGACY OF A LETTER

3. Finally, given the grammatical nature of the variant και ό Geos

ην δ λόγος, its story touches the historical use of predicate nominative
constructions in the NT. Looking at normative patterns for the addition
of the article in similar constructions in the NT may show whether its
origin resulted from an intentional theological change.

THE USE OF SCRIBAL HABITS AS A "HISTORICAL LENS"

A popular facet in mainstream textual criticism is the study of read­
ings that show evidence of potential theological motivation. Initially,
Hort and, to a degree, Colwell resisted the notion that textual variants
point to the history of theology.14 However, there is a growing con­
sensus that many variants are helpful barometers for seeing the theo­
logical contours prevalent in the early church. As a result, the notion
of an “original text” is jettisoned in favor of a moving canon contained
within individual manuscripts.15 The primary source of this perspective
is the limited “window of opportunity” presented by the lack of con­

14. Westcott and Hort do not deny theology as a possible influence on textual changes
(Jeff Miller, "(Mis) Understanding Westcott and Hort,0 ResQ 41 [1999]: 155-62).
However, they clearly reject the notion that theological changes were made to
cover up competing doctrines by altering the sacred text. Colwell wrote, “The cur­
rent enthusiasm for manuscript variations as contributions to the history of the­
ology has no solid foundation” (“Redivivus,” 150). Colwell, a pioneer in the field of
textual criticism, denies not the presence of theologically motivated variants but,
rather, their ability to reflect the state of historical orthodoxy. However, Metzger
and Ehrman boldly denied Hort’s claim that “even among the numerous unques­
tionably spurious readings of the New Testament there are no signs of deliberate
falsification of the text for dogmatic purposes” (Metzger and Ehrman, Text of the
New Testament, 282).
15. Some examples will be given here to show the attention this concept is receiving.
Bart Ehrman is perhaps the most outspoken in this regard (Bart D. Ehrman, “The
Text as Window: New Testament Manuscripts and die Social History of Early
Christianity/ in The Text ofthe New Testament in Contemporary Research: Essays
on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes [Grand
Rapids: Eerdmans, 1995], 361-79; Bart D. Ehrman, The Orthodox Corruption of
Scripture: The Effect of Early Christological Controversies on the Text of the New
Testament [New York: Oxford University Press, 1993]). See also Eldon Jay Epp,
“The Multivalence of the Term Original Text’ in New Testament Textual Criticism,”
HTR 92 (1999): 245—81; D. C. Parker, The Living Text of the Gospels (Cambridge:
Cambridge University Press, 1997). For Parker, theological variants are represen­
tatives of a developing canon, so an original text is present in the individual vari­
ants (Living Text, 207-13). Similarly, Epp argued that individual variants should be
viewed as proliferations of canon, so that “the same fluidity that can be observed
in textual variation carries over to canonicity” (“Multivalence,” 278). Therefore, ac­
cording to both Epp and Parker, the NT canon is not a fixed body related to an
“original” text but, rather, a moving target that accommodates the multiplicity of
variants. In a milder manner, Peter Head noted, “The scribe of die New Testament
was a participant in the life and faith of the church, and this life and faith clearly
influenced the process of transmission” (-Peter M. Head, “Christology and Textual

94 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

trolled copying during the second century.1516 Although both Codices

L (019) and WS (032-S) fall outside the time frame of certain theo­
logical debates, it is still possible that they were copied from an early
source. As a result, it is worthwhile to grapple with the relationship of
these two eighth-century manuscripts to much earlier Christological
disputes.

A HISTORICAL SURVEY OF SABELLIANISM

The Roots and Rise ofSabellianism

The variant in John 1.1c may spring from the historical seeds that
later sprouted into the teaching of Sabellius and his closest followers.
Beginning with the offshoot of Gnosticism in Proconsular Asia, the
school of modalistic Monarchianism17 first emerged toward the end
of the second century.18 The central figure for this school of thought
in its infant stages was Noetus of Smyrna, followed later by Praxeas.
Eventually, this view of Christ was given a more systematic and philo­
sophical structure by Sabellius,19 producing Sabellianism in tire East and
a nuanced form called Patripassianism in the West.20

Transmission: Reverential Alterations in the Synoptic Gospels,” NovT 35 [1993]:

128).
16. For a treatment of the transmission^ process prior to the fourth century, see Royse,
Scribal Habits, 19-27; Fee, "Textual Criticism,” 9-10; Head, “Christology and Tex­
tual Transmission,” 106; Metzger and Ehrman, Text ofthe New Testament, 282-87.
Ehrman specifically acknowledged this issue while addressing anti-patripassian-
istic corruptions, stating that "in the early third century, die textual tradition of
the New Testament, as we have seen, had already begun to solidify” (Orthodox
Corruption, 264).
17. For further discussion of the terminology employed here, see William G. Rusch,
ed·» The Trinitarian Controversy (Philadelphia: Fortress, 1980), 8-9; J. N. D. Kelly,
Early Christian Doctrines, 5th rev. ed. (New York: Continuum, 1977), 119—22; Hu­
bertus R. Drobner, The Fathers of the Church, trans. Siegfried S. Schatzmann (Pea­
body, MA: Hendrickson, 2007), 116; Adolf Harnack, History ofDogma, trans. Neil
...... Buchanan, BrdGerman ed., 7 vols. (New York: Dover, 1961), 3:52-53; T. E. Pollard,
Johannine Christology and the Early Church (New York: Cambridge University
Press, 1970), 52; R. P. C. Hanson, The Searchfor the Christian Doctrine of God: The
Arian Controversy, 318-381 (Grand Rapids: Baker Academic, 2005), 310.
18. John Henry Newman, The Arians of the Fourth Century (London: Gilbert and Riv­
ington, 1871), 120.
19. For further discussion of the role and teaching of Sabellius, see Kelly, Early Chris­
tian Doctrines, 121-22; Rusch, Trinitarian Controversy, 9; Harnack, History of
Dogma, 3:52-83; Newman, Arians of the Fourth Century, 211.
20. Some historians make reference to this geographical distinction (cf. Harnack, His­
tory ofDogma, 3:52-53; Hanson, Doctrine of God, 310). Since the origin and study
of text types tends to be geographical in nature, this observation could potentially
have a bearing on the presence of a given reading in the Alexandrian tradition as
opposed to the Western text.

Matthew P. Morgan Chapter 3 / 95

 THE LEGACY OF A LETTER

Sabellianism opposed the “orthodox” position in several ways.21 The

unifying characteristic of modalistic Monarchianism. was the “denial of
the distinctions of Persons in the Godhead.”22 Within this system, adher­
ence to strict monotheism was essential, and the distinction of Christ
from the Father was labeled blasphemy and considered to be the wor­
ship of two Gods (δίθβοι).23 The logical extension of this theology was
Patripassianism, which taught “the idea that it was the Father who suf­
fered and underwent Christ’s other human experiences."24 Although
debatable,25*it seems best to merge these doctrines, given their strong
logical relationship.25
Critical to the idea of a pro-Sabellian influence on the text of John
is the time frame and impact these views had on the church. In the
bigger picture, the conflict was both
H· severe and widespread. As one his­
torian wrote, “This much is certain, however, that in the East the fight
against Monarchianism in the second half of the third century was a
violent one, and that even the development of the Logos Christology (of

21. The term orthodox here refers primarily to the expressions of Origen, Hippolytus,
Eusebius, and Tertullian, whose views were later refined and solidified in the great
creeds of the fourth and fifth centuries. The language of Origen, with its tendency
toward subordinationism, was not sufficient for later scholars and required further
development to be considered orthodox in the sense that we now conceive of it.
Hanson’s comments on this development of doctrine are most helpful here: "The
story is the story of how orthodoxy was reached, found, not of how it was main­
tained. ... ft was only very slowly, as a result of debate and consideration and the
re-thinking of earlier ideas that the doctrine which was later to be promulgated
as orthodoxy arose* (Hanson, Doctrine of God, 70). In a similar manner, Moule
has presented a helpful analogy by contrasting the "evolutionary” view of history
and doctrine, where the end product looks nothing like the initial life-form, with
a ’developmental” view, where “they are not successive additions of something
new, but only the drawing out and articulating of what is there” (C. E D. Moule,
The Origin of Christology [New York: Cambridge University Press, 1977], 3), Only
later, after the dangers and shortcomings of these initial orthodox expressions had
been exposed to rigorous attacks, did more definitive and precise statements begin
to emerge. This definitive understanding of the Father’s relationship to the Son is
evident by die time of Athanasius, as it becomes merely a “presupposition” in his
arguments (Pollard, Johannine Christology, 136-37). .
22. Newman, Arians ofthe Fourth Century, 124.
23. See Kelly, Early Christian Doctrines, 119; Harnack, History of Dogma, 3:55. This
shows that the conflicts of modalistic Monarchianism came about not as the result
of competing Christianities but, rather through the difficulty and struggle of trying
to articulate the reality of Christ’s deity from a. monotheistic framework.
24. Kelly, Early Christian Doctrines, 120.
25. For a slightly different perspective, see M. Slusser, “The Scope of Patripassianism,”
in Studia Patristica, vol. 17, ed. Elizabeth A. Livingstone (Oxford: Pergamon,
1982), 169-80.
26. The reasons for this are twofold: (1) because the logical relationship is so strong
that it is unlikely most people within the church would be able to distinguish
these doctrines and (2) to keep consistency with Ehrman (Orthodox Corruption,
270 n. 6).

96 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

Origen) was directly and lastingly influenced by this opposition.”27 At

the same time, the issue of Christ’s relationship to the Father does not
begin to demand serious attention until the early to mid-third century
at the very earliest.28 This last point is particularly significant since the
manuscript tradition had already gained greater stability prior to this
time period.

The Reaction ofthe Church Fathers

The weight and nature of the Satellian controversies can be seen
from a brief survey of the writings of those men charged with upholding
the apostolic tradition, particularly in their response to Noetus and
others who carried the theological torch after him.

Hippolytus (co. 189-235)29

Perhaps the earliest defense of orthodoxy can be found in the writ­
ings of Hippolytus.30 It is here that the orthodox position and the growing
popularity of Sabellianism within the Roman church became apparent.31
The pivotal work in these modalistic debates was Adversus Noetus, pre­
serving the primary historical record of the initial teachings put forward
by Noetus. Hippolytus defended orthodoxy with an appeal to philos­
ophy rather than Scripture,32 resulting in a crude framework labeled the

27. Harnack, History of Dogma, 3:82 (italics added). See also Kelly, Early Christian
Doctrines, 119; Pollard, Johannine Christology, 52.
28. Hamack indicated that with the presence of Origen in Egypt, the earliest this
schism could have occurred would be 230-40 CE (History ofDogma, 3:83). Simi­
larly, Slusser pointed to the lack of trustworthy evidence after 250 CE ("Scope of
Patripassianism," 169). Even Ehrman acknowledged this problem, stating, “When
these distinctions [between God and the Son] did gain in importance for ortho­
doxy at large, in the early third century, the textual tradition of the New Testament,
as we have seen, had already begun to solidify” (Bart D. Ehrman, The New Testa­
ment: A Historical Introduction to the Early Christian Writings [New York: Oxford
. -, University Press, 1997], 264).
29. Drobner, Fathers ofthe Church, 122-24.
30. Harnack, History ofDogma, 3:62.
31. Other primary figures were the bishops Zephyrinus and CaUistus, whose leader­
ship spanned the late second and early third centuries. From Hippolytus’s polem*
ical language and his record of Callistus’s later excommunication of Sabellius as a
“cover-up” for his own doctrinal problems, there appears to be an internal political
conflict involved in these writings as well (Ehrman, Orthodox Corruption, 262-63).
32. For Hippolytus, the language and structure are concerned more with philosophical
frameworks than with key Scripture citations (Raymond B. Williams, “Origen’s In­
terpretation of the Gospel of John” [PhD diss., University of Chicago, 1966], 247).
His contribution lies more in the connection of a “universal monad,” the ancient
philosophy taught by Heradertus, than with Sabellian views (Kelly, Early Christian
Doctrines, 121).

Matthew P. Morgan Chapter 3 I 97

 THE LEGACY OF A LETTER

"economic trinity?33 where God’s unity was seen in his power, while his
diversity was found in his economy or manifestation.34 This primitive for­
mulation became an initial spark that would later burst into open flames
as the groups worked to express the relationships within the Trinity.

Tertullian (ca. 160—240)35

Little is known about the identity of Tertullian’s opponent Praxeas;36

however, the presence of his Sabellian teaching in Africa is evident in
Tertullian’s strong response in Adversus Praxeam (ca. 213).37 According
to Tertullian’s claim in this writing, Praxeas taught that the Father and
Son were one identical person.38 Tertullian approached the problem
much like Hippolytus, only with greater sophistication. He stated that
God was “three in degree, not condition; in form, not substance; in as­
pect, not power.”39 Arguably his greatest Christological contribution was
the proposal of “the twofold state, which is not confounded but con­
joined in one person—Jesus, God and man.”4041 In Tertullian’s approach,
the dawn of the third century had arrived, and a primitive orthodox ar­
ticulation of tire Trinity was beginning to take shape, but a formal and
concerted appeal to John 1.1 remained absent from the nucleus of the
debates.

Origen (ca. 185-254)^

While earlier church fathers made philosophical arguments,
Origen firmly staked his view of Christ’s relationship to the Father
on the opening verse of John’s GospeL As a serious textual critic, he

33. Kelly pointed out that this formulation of the Trinity was the catalyst that threat­
ened the unity of the Godhead and produced the extreme response found in mo-
dalistic monarchian theology (Early Christian Doctrines, 109).
34. Alexander Roberts and James Donaldson, eds., Hippolytus, Cyprian, Caius, Nova­
tian, Appendix, vol. 5 oi Ante-Nicene Fathers: The Writings of the Fathers Dawn to
A.D. 325 (1886; repr., Peabody, MA: Hendrickson, 2004), 226. See also the com­
ments in Ehrman, Orthodox Corruption, 264.
35. Drobner, Fathers of the Church, 153.
36. Newman, Arians of the Fourth Century, 120-24; Kelly, Early Christian Doctrines,
121; Drobner, Fathers ofthe Church, 163-64.
37. Regarding the date of Adversus Praxeam, both Kelly and Drobner place this work
in the early third century (ca. 213) (Kelly, Early Christian Doctrines, 121; Drobner,
Fathers ofthe Church, 163).
38. The technical language was vox et sonus oris, reflecting the view that the Word had
“no independent subsistence” (Kelly, Early Christian Doctrines, 121).
39. Ehrman, Orthodox Corruption, 264.
40. Drobner, Fathers ofthe Church, 164. The corresponding Latin phrase in Adv. Prax.
27.11 is Videntus duplicem station, non confusum sed coniunctum in una persona,
deum et hominem lesum.
41. Drobner, Fathers ofthe Church, 136-48.

98 / Chapters Matthew P. Morgan

 Sabellianism or Scribal Blunder in Jdhn 1.1c?

also preserved extensive testimony to the exclusivity of the anarthrous

reading of θεός· within the Alexandrian manuscript tradition,42 as the
following segment from his commentary on John shows:

We next notice John’s use of the article in these sentences. He does

not write without care in this respect, nor is he unfamiliar with the
niceties of the Greek tongue. In some cases he uses the article, and in
some he omits it. He adds the article to the Logos, but to the name of
God he adds it sometimes only. He uses the article, when the name of
God refers to the uncreated cause of all things, and omits it when the
Logos is named God.43

In addition, Origen’s understanding of the Trinity, with the Father alone

as αύτόθεος, was dependent on the description of the Son as θεός and
not ό θεός in John 1.1c.44 As a result, Origen, a credible third-century
textual critic, becomes a pivotal historical witness to the exclusion of
the article in John 1.1c.45
Chronologically, if Sabellianism was unable to make inroads while
Origen was in Alexandria, it is doubtful it could wield any influence
on textual transmission prior to ca. 231, when the textual sheriff was
forced to leave for Caesarea.46 The sequence of these events is signifi­
cant, as it makes dubious, at best, the likelihood of a pro-Sabellian
textual tradition in Alexandria during the first quarter of the third
century.

42. For a study of the textual awareness of Origen, see the listing in Bart D. Ehrman,
Gordon D. Fee, and Michael W Holmes, The Text ofthe Fourth Gospel in the Writ­
ings of Origen, vol· 1, Society of Biblical Literature: The New Testament in the
Greek Fathers 3 (Atlanta: Scholars Press, 1992), 38-42.
43. Allan Menzies, “Origen’s Commentary on John,” in The Ante-Nicene Fathers: Trans­
lations ofthe Writings of the Fathers Down to A.D. 325, ed. Allan Menzies, vol. 10
(Grand Rapids: Eerdmans, 1962), 323. Taken from Origen, Comm. Jo. 2.2.13—14,
which reads: Πάνυ δέ παρατετηρημευως και οΰχ ώς ελληνικήν άκριβολογίαν
,οΰκ έπιστάμευος ό Ιωάννης δπου μ,έν τοΐς άρθροις εχρήσατο οπού δε ταΰτα
άπεσιώπησεν,επι μεν τοΰ λόγου προστιθε'ις τό ό,έπιδέ της θεός προσηγορίας
δπου μέν τιθε'ις δπου δε αιρων. Τιθησιν μέν γάρ τό αρθρον, δτε ή θεός όνομασία
επ'ι του άγενητου τάσσεται των δλων αιτίου, σιωπά δέ αυτό, δτε ό λόγος θεός
ονομάζεται. Portions of this Greek text from Origen’s commentary also appear in
Tischendorfs critical apparatus for John 1.1c (Tischendorf, Novum Testamentum
Graece, loc. cit.).
44. Kelly, Early Christian Doctrines, 129—32; Pollard, Johannine Christology, 93-94.
With regard to the concept of αΰτόθεος, my limited investigation found Origen’s
perspective in Origen, Comm. Jo. 2.2.17-20.
45. Although willing to entertain die possibility of the article in John 1.1c, Ehrman
acknowledged the strength of Origen’s comments in Origin, Comm. Jo. 2.2.17-18])
against such a notion (Ehrman, Orthodox Corruption, 179 n. 87).
46. Drobner, Fathers of the Church, 137.

Matthew P. Morgan Chapters / 99

 THE LEGACY OF A LETTER

Eusebius (ca. 260/64-340)S7

Following in Origen’s footsteps, Eusebius provided an even more

pointed discussion of the text of John 1.1c. He clearly asserted what
John’s Gospel does not say:

Not saying, “and the word was the God” [καί ό Seos ήν ό λόγος·],
with the appended article, so that it is not the one that will define
“the One over all” [τον errt πάντων],... therefore it says, “and the
word was God” [και θεό? ήν ό λόγο?], so that we might see God,
“the One over all,” with whom was the Word, and tire Word himself
was God.*48

Like Origen, Eusebius provided additional textual attestation to the lack

of the article with θεό? in John 1.1c, continuing the textual tradition of
his predecessor.49 From tliis point forward, the distinction between the
Father and the Son in the midst of their unity is clearly defined by tine
absence of the article in John 1.1c. Equally important was the continued
presence of the anarthrous θεός·, extending into the fourth century and
beyond tire Council of Nicea (325).

The Historical Viability ofSabellian Influence on the Textual

Transmission of the New Testament
We have already seen the sociohistorical climate that gave birth to
the teachings known as Sabellianism. On one level, both the geograph­
ical spread and the intensity of this conflict certainly had the potential
to influence Christians during the later portion of the third century.
Therefore, it is conceivable that certain orthodox scribes may have had
some theological impetus to alter the text of John 1.1c, given its explicit
use by Origen, Eusebius, and possibly others. To advance the mere pos­
sibility of this influence into a likely explanation, certain variables must
be in place:

1. Sabellianism’s polemical influence must be strong enough to

cause numerous pious scribes to alter the sacred text by universally
eradicating the presence of the article. A fundamental challenge to this
notion was the use of allegorical interpretation by Sabellians to deny the
Logos concept in the prologue of John’s Gospel. As a result, “they did

47. Ibid., 223-24.

48. This is my own translation of the Greek text listed as Eusm“n 121 in Tischendorf’s
critical apparatus for John 1.1c (Tischendorf, Novum Testamentum Graece, loc.
cit.). I was unable to locate this exact statement in Eusebius but found similar ex­
pressions in Eccl. theol. 2.14.3 and 2.17.1-2.
49. See Pollard, Johannine Christology, 282..

100 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

not regard that book [John’s] as justifying the introduction of a Logos,

and the bestowal on him of the title Son of God.”50 Instead, their core ap­
peal was to passages emphasizing Jesus’ unity with the Father (cf. John
10.30; 14.8-10).51 Therefore, Sabellians would be more prone to avoid
John 1.1 than to alter its content, given their preference for other texts
that better articulated their view. It seems doubtful that the Sabellian
approach to John 1.1 would trigger theological sensitivity among even
the most licentious scribe.

2. The strength of the controversy needs to have its greatest impact

during the early second century, when manuscript production was scat­
tered and vulnerable. Since the orthodox arguments were not related to
the text of John 1.1c prior to Origen in the third century, the timing of
Sabellianism does not fit. Likewise, in its earliest forms, the debate ad­
dresses philosophical frameworks and not the prologue to John’s Gospel.
Consequently, it is difficult to posit theological motivations to a text that
the Sabellians preferred to set aside.

3. Finally, it would be reasonable to expect some hesitancy in ref­

erences about tire lack of the article in John 1.1c if competing manu­
scripts existed. However, the voices of both Origen and Eusebius boldly
deny the presence of the article within the early manuscript tradition
while explicitly affirming the anarthrous reading in John 1.1c.

In conclusion, the historical evidence indicates that whatever

theological tensions may have existed, their role in the textual legacy
of John 1.1 was nonexistent. Without earlier roots, the Sabellian con­
flicts would have been unable to produce the fruits of their teach­
ings during the most vulnerable period of NT transmission. So, with

50. Harnack, History ofDogma, 3:63.

51. Multiple scholars affirm John 10.30 and 14.8-10 as key "proof texts" utilized by
the patripassianists: see Kelly, Early Christian Doctrines, 120; Pollard, Johannine
Christology, 52; Harnack, History ofDogma, 3.63; Drobner, Fathers of the Church,
_...1^.3:.Though ^nuch <jf what we know comes through the polemical writings of the
orthodox camp, the available evidence gives at least four reasons to doubt their im­
pact on the textual transmission of John 1.1. First, the debates took place primarily
within the church, where Scripture was held in high regard and was less likely to
be manipulated. Second, as fierce as the debates were, the Sabellians did not turn
to John 1.1 or even the prologue of John’s Gospel as part of their defense. Since the
distinction between the Father and Son is undeniably evident in 1.1b, even if 1.1c
included the articular Seos, the earlier portion of the verse would have presented
even greater challenges. This best explains the use of allegorical interpretation in
John 1.1—18. Third, these writings focus on either monotheistic passages from the
Old Testament (Deut 6.4) or clear statements by Jesus of His unity with the Father
(John 10.30; 14.8-10). Fourth, the time period when these debates were at their
pinnacle happens after die time frame when “uncontrolled” transmission was more
plausible.

Matdiew P. Morgan Chapter 3 / 101

 THE LEGACY OF A LETTER

regard to the sociohistorical context, the likelihood that early scribes

would have seen the transmission of John 1.1c in polemical terms
quickly evaporates. At the same time, the wealth of historical evi­
dence demands a truly exceptional historical reconstruction in order
to propose any Sabellian influence on the textual transmission of
John 1.1c.

AN ANALYSIS OF SCRIBAL HABITS

IN CODICES REGIUS AND FREERIANUS

Terms and Methodology

t
Following, at the macro and micro level, the methodology of E. C.
Colwell for determining scribal behavior,52 “we wish to find a way to
characterize the habits of scribes that will avoid, as far as possible, both
any question-begging assumptions about scribal behavior and any con­
troversial presuppositions about the history of the text.”53 In dlls way,
we can create a scribal rap sheet that can be used to assess the quality
and character of Regius's testimony with regard to John 1.1c.54 In addi­
tion, this analysis will rest on my own collation of Codex Regius and the
first quire of Freerianus against the Robinson-Pierpont Majority Text
(2(X)5)5S of John’s Gospel.56 These variants, individually classified and
categorized, provide a more detailed quantitative analysis that is able
to shed light on the character and background of the manuscript as a
whole.57
Before continuing, it is necessary to establish how the language
of the discipline will be applied throughout the remainder of this
presentation.58 Of particular interest for the determination of scribal

52. Colwell, “Redivivus,” 160.

53. Royse, Scribal Habits, 31.
54. The pivotal nature of studying the scribal habits of entire MSS can be seen in Col­
well "Redivivus,” 160; James R. Royse, “Scribal Tendencies in the Transmission of
the Text of the New Testament,” in Ehrman and Holmes, Text ofthe New Testament
in Contemporary Research, 245; and James D. Yoder, “The Language of the Greek
Variants of Codex Bezae," NovT 3 (1959): 241-42.
55. Maurice A. Robinson and William G. Pierpont, The New Testament in the Original

56. This personal collation has been compared against the extensive collation found in
Schmid, Elliott, and Parker; New Testament in Greek IV, 189—553. The justification
for using the MT rather than the TR as a collating base is found in Daniel B. Wal­
lace, “The Majority Text: A New Collating Base?*’ NTS 35 (1989): 609-18.
57. Royse, Scribal Habits, 42-44.
58. For cogent definitions, see Gordon D. Fee, “On the Types, Classification, and Pre­
sentation of Textual Variation,” in Studies in the Theory and Method of New Tes­
tament Textual Criticism, ed. Eldon Jay Epp and Gordon D. Fee (Grand Rapids:
Eerdmans, 1993), 67.

102 / Chapter 3 Matthew P. Morgan

 Sabeilianism or Scribal Blunder in John L.lc?

habits are the singular readings of a manuscript. For Colwell, a singular

reading was "a reading that had no Greek support in the critical ap­
paratus of Tischendorf's 8th edition.”59 Recently, textual criticism has
pushed for an updated investigation where a singular reading is pre­
served in several major critical texts.60 In keeping with Royse’s expan­
sion of Colwell’s method, singular readings in this study were examined
against NA27, Tischendorf’s eighth edition, and the recently published
IGNT for John’s Gospel.61 Therefore, a singular reading in this study
includes any variant that is preserved in only one Greek manuscript
according to NA27, Tischendorf’s eighth edition, and IGNT.62 Unless
otherwise noted, subsingular readings will conform to the definition
given by Fee: “The meaning of 'sub-singular readings’ is a non-genetic,
accidental agreement in variation between two MSS which are not oth­
erwise closely related.”63 Finally, a nonsense reading will be defined as
“words unknown to grammar or lexicon, words that cannot be con­
strued syntactically, or words that do not make sense in the context.”64
Key evidence in uncovering the legacy of any manuscript is the
nature and quality of its singular and subsingular readings. In these
peculiar strokes are found the slips and secrets that form a personal­
ized portrait of the human instruments used to copy each manuscript.
To a large degree, the story of every manuscript is the story of the
individual whose diligent labor produced it. For our variant in John
l.lc, its testimony hinges on the lives of two scribes, to whose stories
I now turn.

59. Colwell, “Scribal Habits,” 108.

60. Royse, Scribal Habits, 65-67. Royse recommended consulting NA2’, Tischendorf8,
von Soden, Clark, Legg, and the recently published IGNT volumes on Luke and
John. Other works following a similar approach are Peter M. Head, “The Habits
of New Testament Copyists: Singular Readings in the Early Fragmentary Papyri of
John,1’ Bib 85 (2004): 400-401; and Juan Hernandez, Scribal Habits and Theological
Influences in the Apocalypse: The Singular Readings of Sinaiticus, Alexandrinus,
andEphraemi (Tubingen: Mohr Siebeck, 2006), 45.
61. There are a few instances where a reading is listed as singular by Tischendorf but
■ -the-broader-Gollection of evidence in the IGNT proves otherwise. This observation
should present caution in only utilizing a single older work when determining sin­
gular readings. My personal analysis of the singular readings in Codices L and Ws
begins with Tischendorf's reading but is refined by the data available in IGNT. The
text of John l.lc is a good example rtf the added rigor of this method. In almost all
of the critical additions recommended by Royse, the article with Oeds stands as a
singular reading found only in Codex Regius. Howevec IGNT lists this as a subsin-
gular reading with two witnesses, L and Ws (032-S) (Schmid, Elliot, and Parkes
New Testament in Greek IV, 189).
62. An exception to this are certain orthographic variants that are difficult to deter­
mine since spelling was not always standardized, leading most critical texts to ex­
clude these variants.
63. Fee, “Types, Classification, and Presentation,” 67.
64. Colwell, “Scribal Habits,” 111.

Matthew P. Morgan Chapter 3 / 103

 THE LEGACY OF A LETTER

The Place ofCodices Regius and Freerianus in Transmission History

Codex Regius (L [019])

Codex Regius (L [019]), a Gospels manuscript, is universaDy re­

garded as from the eighth century. With regard to text type, it serves as
a primary representative of the secondary Alexandrian65 text.66 The sig­
nificance of this genealogical history is its frequent alignment with the
major and early Alexandrian witnesses Codex Vaticanus (B) and Codex
Sinaiticus (X).67 At the same time, there is also a notable intrusion of
Byzantine readings, producing a polluted form of the Egyptian text.68 In
some sense, the textual history of Regius can be likened to a teenager
whose biological heritage comes from early Alexandria (B X) but whose
most influential friends are primarily Byzantine, resulting in a mixture
of these two forms.
Of significance to the present study is the strong relationship this
manuscript has with the "proto-Alexandrian” text form found in the
combination of Vaticanus (B) and $J75 dated as early as the second cen­
tury CE.69 This past heritage presents the possibility that Regius pre­
serves a more primitive text deliberately removed by previous orthodox
scribes. If this is the case, it is possible to argue that L somehow pre­
served the “original text” Yet there is a tension between the genealogical
quality of this manuscript and the scribe who recorded it. Thus Metzger
and Ehrman noted, “Though badly written by a scribe who committed
many ignorant blunders, its type of text is good, agreeing frequently
with Codex Vaticanus (B).”70*For that reason, a detailed assessment of
scribal proclivity must be compared against the prestigious heritage of
this manuscript to better discern the legacy it leaves behind.

65. For a helpful history behind the transition from the term "late Alexandrian” prior
to Westcott and Hort to the more recent "secondary Alexandrian,” see Carlo M.
Martini, "Is There a Late Alexandrian Text of the Gospels?" NTS 24 (1978): 285-96.
66. This date and type are affirmed in Metzger and Ehrman, Text ofthe New Testament,
7?·, and Aland and Aland, Text ofthe New Testament, 334-35.
67. For a quantitative analysis and subsequent profile of Codex Regius, see Aland and
Aland, Text of the New Testament, 332—37. For more details regarding the signifi­
cance and limitations of this method, see Colwell, "Genealogical Method,” 109-33;
Colwell, “Significance of Grouping,” 73-92.
68. See Aland and Aland, Text of the New Testament, 332-37. Fee noted this in his
analysis of Origen’s text (Gordon D. Fee, “Origen’s Text of the New Testament and
the Text of Egypt," NTS 28 (1982): 350). Elsewhere, he also showed how L, 33, and
Cyril, considered neutral texts, have begun to assimilate the character of standard
Byzantine readings (Gordon D. Fee, "The Text of John in Origen and Cyril of Alex­
andria: A Contribution to Methodology in the Recovery and Analysis of Patristic
Citations," Bib 52 (1971): 370-71).
69. Martini, “Late Alexandrian Text)’291.
70. Metzger and Ehrman, Text of the New Testament, ΤΠ. For further support, see
Parker, Living Text, 83.

Matthew P- Morgan
 Sabelliaiusm or Scribal Blunder in John 1.1c?

Codex Freerianus (Ws [032-S])

Codex W (032), as a whole, is considered to be a late fourth- or

early fifth-century manuscript representing the “Western.” order of
the Gospels (Matthew, John, Luke, Mark).71 According to Metzer and
Ehrman, “The type of text is curiously variegated, as though copied
from several manuscripts of different families of text.”72 Regarding John
1.1, Metzger and Ehrman also noted, “The text of John 1.1-5.11, which
fills a quire that was added about the seventh century, presumably to
replace one that was damaged, is mixed, with some Alexandrian and a
few Western readings.”73 Significant is the recognition that this opening
portion of John’s Gospel was the creation of a different scribe during
a later time period than the remainder of the manuscript. Similarly,
the IGNT clearly distinguishes this portion as supplemental (032-S).
Although Sanders proposed a rather elaborate reconstruction, dating
John 1.1-5.11 before the remainder of the manuscript,74 most scholars

71. Metzger and Ehrman, Text of the New Testament, 80; NA27. The Kurzgefasste Liste
is more specific in dating the MS in the fifth century (Kurt Aland, Kurzgefasste
Liste der griechischen Handschriften des Neuen Testaments [Berlin: Walter De
Gruyter, 1963], 40). Sanders went even further, dating all but John 1-5.11 no later
than the fifth century and probably in the fourth century due to the Diocletian per­
secutions in 303 (Henry A. Sanders and Freer Gallery of Art, The New Testament
Manuscripts of the Freer Collection [New York: Macmillan, .1918], 135-39; Henry
A. Sanders, Facsimile ofthe Washington Manuscript ofthe Four Gospels in the Freer
Collection, with an Introduction by Henry A. Sanders, copy 390 of 435 [Ann Arbor,
MI: University of Michigan, 1912], v).
72. Metzger and Ehrman, Text of the New Testament, 80; cf. also Sanders, New Testa­
ment Manuscripts, 133.
73. Metzger and Ehrman, Text ofthe New Testament, 80.
74. Sanders, New Testament Manuscripts, 134-39. One reason for this chronological
placement is the omission of John 5.12 in the later witnesses Γ (036 [tenth cen­
tury]) and Λ (039 [ninth century]), which he attributed to a relationship with the
parent document that this quire is replacing (128-30, 136). The freshness of the
superscription on the first page compared with the rest of the writing (136), as
well as the low quire number (137), contribute to his view that the first page of the
MS^probably did not include Matthew and may have contained only John” (137).
Later, he pointed to various peculiarities (e.g., enlarged letters, punctuation, and
ornamental dots) from the first quire that are also found throughout the MS (139).
Thus he concluded, *’1he first quire of John is slightly older than the rest of the MS”
(139). However, this explanation is unwarranted for four reasons. First, arguing for
an early date on the basis of a connection with later MSS (Γ, Λ) is inconclusive at
best. Second, it is implausible that a scribe who apparently rewrote the rest of the
Gospels by copying otherfragmentary texts would wrap four Gospels around one
quire and would not rewrite that one quire. If such was his pattern for the rest of
die MS—as Sanders himself articulated—why would he not rewrite the quire at the
beginning of John? (This idea was proposed by Daniel B. Wallace during a review
of the paper on which this chapter is based [February 24,2009].) Third, the various
common traits can also indicate harmonization with the larger portion of the MS,
rather than a core around which the rest of the MS was crafted (this idea was first

Matthew P. Morgan Chapter 3 / 105

 THE LEGACY OF A LETTER

do not find this argument tenable.*75 Instead, a more plausible view con­
siders John 1.1-5.11 to be “a later (probably eighth-century) replace­
ment quire that bears no relation to the rest of the manuscript and made
up for the (presumably) lost original portion.”76 Therefore, until further
evidence can be provided, the assessment of Royse and Goodspeed
should be adopted, dating Ws to sometime in the eighth century.7778
Given an eighth-century date for Ws, its production falls within
the same general time frame as Codex Regius. So how does this close
chronological connection influence the possible relationship between
these two manuscripts? It is at least plausible to suggest that their story
is intertwined in some way. In fact, could their harmony in John 1.1c
serve as the lone remnants of an earlier Sabellian ancestor?

Codex Regius's Scribal Profile

A collation of John’s Gospel in Regius against the Robinson-

Pierpont Majority Text (2005)7i produced a total of 1,321 variants.
From this virtual sea of evidence, a total of 209 (15.8%) singular read­
ings remain, but if we remove insignificant orthographic fluctuations,79
the field narrows to 145 variants (11%) worthy of more detailed con­
sideration.80 Within the confines of these 145 unique recordings lays

presented by J. Bruce Prior in a personal email on February 20,2008). Fourth the

presence of various lengthening devices on the last page of the first quire cause it to
strategically end at 5.11. This points to a later, rather than earlier creation (J. Brace
Prior, email, February 20, 2008).
75. E.g., see Metzger and Ehrman, Text ofthe New Testament, 80-81; James R. Royse,
"The Corrections in the Freer Gospels Codex,” in The Freer Biblical Manuscripts:
Fresh Studies ofan American Treasure Trove, ed. Larry W. Hurtado (Atlanta: So­
ciety of Biblical Literature, 2006), 186; Edgar J. Goodspeed, *The Freer Gospels?
AJTY7 (1913): 599.
76. Royse, "Corrections in the Freer Gospels? 186.
77. Goodspeed, "Freer Gospels? 599.
78. Robinson and Pierpont, New Testament in the Original Greek.
79. The most common orthographic fluctuations involved changing i to η in 16 in­
stances. So μαρτυρία is spelled μαρτυρηα in John 8.13, John 10.31 has λιθασωσιν ’
as λιθασωσην» and φοινίκων is spelled φοινηκων in John 12,13. The change from e
torn occurred 12 times. So με is writtenμαι. in John 1.33, andqpepai? is spelled
ημαιραι^ in Joint 2.19. The inverse of both these alterations were equally popular,
but the replacement of a with o frequently produced singular readings (5). Thus
ουπω reads ουπο in John 7.30, and two changes are found in John 17.2, with καθω?
written as KaSos and εδωκα^ spelled εδοκα$.
80. For an extensive discussion of the issues and complexity of orthographic errors,
see Chrys C. Caragounls, The Development of Greek and the New Testament
Morphology, Syntax, Phonology, and Textual Transmission (Grand Rapids: Baker
Academic, 2006), 339-96; 475-564. Not only is the issue difficult due to our lim­
ited understanding and historical distance from the Koine Greek language, but
the common acceptance of “itacisms” makes determining a unique spelling like
finding a needle in a haystack. Therefore, these variants will not be included in the

106 / Chapter 3 Matthew P, Morgan

 Sabellianlsm or Scribal Blunder in John 1.1c?

the legacy of the scribe who produced Codex Regius. More specifically,
they illuminate any potential theological motivation that might per­
meate the transmission process.*81

Regius's Copying of Units

One component of the scribal profile is the method of copying ob­

served In the singular readings. Given the later date of the manuscript
and the high number of orthographic mistakes, it seems probable that
Regius was the product of a scriptorium.82 In addition, this scribe ap­
pears to have copied one or two letters at a time, in short syllables, rather
than words or phrases. This conclusion is sustained by the pattern of
errors, with over half (88 [60.7%]) involving a single letter, while a sig­
nificantly smaller number (17 [11.7%]) take place at the syllable level.83

general pool of singular readings but will be reserved to support other conclusions
along the way. For a defense of this approach, see Royse, “Scribal Tendencies,” 239.
Alec see T. C. Skeat, "The Use of Dictation in Ancient Book Production,” in The
Collected Biblical Writings ΰ/T.C. Skeat, ed. J. K. Elliott, Supplements to Novum
Testamentum 113 (Leiden: Brill, 2004), 3-32.
81. Royse, Scribal Habits, 51-56.
82. Certainly die date of this manuscript makes its production in a scriptorium more
likely, since that became die standard method when Christianity became the na­
tional religion in the fourth century (Royse, Scribal Habits, 29—30, Metzger and
Ehrman, Text ofthe Hew Testament, 25-30). It also falls prior to the ninth century,
when evidence of stricter controls in the monastic community, including severe
punishment for careless errors, began to occur (Metzger and Ehrman, Text of the
New Testament, 30). With regard to Codex Regius, the volume of nonsense read­
ings involving orthographic confusion lends support to this conclusion, since there
are a large number of orthographic mistakes that produce lexically viable words
that make no sense in context. E.g„ in John 17.14, «ισιν has become ησιν. Like­
wise, there are seven instances where τις is mistakenly written τη$, producing
nonsense in the context (2.25; 33,5; 6-51; 7.20; 1630; 21.12). As these test cases
reveal it appears as though the scribe only hears the letters either in his head or in
a scriptorium and then hurriedly tries to reproduce them in short syllables. Fur­
thermore, if the copy was produced by sight, we should expect fewer cases where
- orffiogfapWc confiisioh produces a lexically viable word. The conditions described
here fit Sheafs argument for dictation (Skeat, "Dictation/ 25). The virtual absence
of unique transpositions can also be explained by a scribe who was copying in units
of letters and/or syllables. Based on the nature of his orthographic mistakes, our
general impression is that this manuscript was copied orally in a scriptorium.
83. A handful of phrases (4) are not listed here. In terms of total variants, the second
largest grouping is words (36). At first glance, this would seem to go against the as­
sertion that this scribe was copying short units by sound. However, in the sorting of
the data, there were a number of instances where a variant was counted as a word
when it could also be classified as a syllable. E.g., there were 8 additions involving
a word. Of these 8 singular readings, 6 consisted of three or less letters. Similarly, 6
of the 7 omissions classified as a word involved variants containing three or fewer
letters. Therefore, it seems that many of the word variants function phonetically
like syllables but are considered words due to their syntactical function.

Matthew P. Morgan Chapter 3 / 107

 THE LEGACY OF A LETTER

This tendency is further reinforced by the large volume and quality of

orthographic variations found throughout this manuscript.8* As the sta­
tistical dust settles, Regius’s scribe emerges as one whose distinctives
are most often shown in either a letter or a single syllable.
<
Corruption Trends and Observations
With a better grasp of the copying methods, it is necessary to stage
the camera over the shoulder of our scribe and observe his tendencies.
Figure 3.1 gives a breakdown of the major categories into which his sin­
gular readings fall.

FIGURE 3.1: BREAKDOWN OF REGIUS’S

SINGULAR READINGS

Substitutions 62 (29.7%)
Transpositions 2 (1.0%)
Additions 23 (11.0%)
Omissions 58 (27.8%)
85
Orthographic84 64 (30.6%)

For a manuscript filled with “many ignorant blunders,”86 the pres­

ence of only two transpositions is striking; yet it further confirms the
probability that this particular scribe possessed a limited grasp of Greek
grammar and was so focused on short sounds that he almost never al­
ters the order. In addition, these adjustments to word order can best be
explained by a skipping of the initial word, which was then added back
later.87
At this point, the scribal story behind Codex Regius is beginning to
emerge. Understanding the scribal proclivities and activities that distin­
guish our main figure, we can now proceed to the story within the story.

84. Significant in Regius are 94 singular readings produced purely by the confusion of
vowel sounds. Furthermore, in approximately 30 instances, this phonetic variation
produces a valid Greek word that makes absolutely no sense in die context.
85. Purely orthographic variants listed here are those where a common vowel substitu­
tion has occurred but the meaning and intention of the word is clearly the same.
86. Metzger and Ehrman, Text ofthe New Testament, 77.
87. The two instances found in Regius’s copy of John appear within a span of four
verses (John 6.23,27), so the close proximity of these variants may suggest that the
scribe was getting rather weary at this point or had his concentration disrupted for
a period of time.

108 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

For this particular scribe, does the meat of his textual transmission indi­
cate a theological legacy or legend?

Quality of Transmission
A critical look at both the integrity and motivations behind the fre­
quency and distribution of errors reveals the potential for theological
variants. Those readings whose nature and character reveal no careless,
transcriptional errors may be our window into the theological world of
our scribe.
The first step in the analysis is to examine the quantity and type
of "nonsense” readings evident in Codex Regius. Metzger and Ehrman
have already speculated that a fair number of singular readings would fit
this category, but the actual data are overwhelming. Of the 145 singular
readings, 87 (60%) are guilty of producing nonsense on the written page.
Figure 3.2 represents a detailed breakdown.

FIGURE 3.2: BREAKDOWN OF REGIUS’S

NONSENSE SINGULAR READINGS

Substitutions 18 (20%)
Transpositions 0
Additions 17(20%)
Omissions — 52 (60%)

Clearly our scribe was prone to erratic behavior with regard to let­
ters, particularly orthographic confusion. This serves as the primary
culprit behind many of these nonsensical variants. For example, in John
6.22, the article τη becomes τι, so that the temporal phrase “on the
next day" (τη επαύριον) becomes the contextually nonsensical question
"What tomorrow?” (τί Επαύριον). Similarly, in John 3.18, the adverb
ηδη-becomes the subjunctive verb ιδη according to Regius. This variant
turns the temporal statement of judgment “condemned already” (ήδη
κέκριταΐ) into “he would see he is condemned” (ιδη κέκριται).88 More
evidence could be brought forward, but it should suffice here to say that
in the copying of short words (e.g., prepositions, relative pronouns, ar­
ticles, etc.), orthographic mistakes that produce nonsense in the word
written on the page abound. A similar wildness is also evident in the
substitution of words. Thus, in John 5.2, “five porticoes” (πέντε στοά?)
becomes "five mouth” (πέντε στόμα).

88. The reading of Regius does not mean “he would see that he is condemned/ which
would require something like ιδη εαυτόν κέκρισθαι.

Matthew P. Morgan Chapter 3 / 109

 THE LEGACY OF A LETTER

In fairness to this scribe, several of his readings do reflect more

deliberation and thought. For example, John 8.20 seems to be a clear
instance of harmonization. In this case, the pluperfect “it had come”
(έληλύθει) is modified to the aorist “it came” (ηλθεν) to match the tense
of the previous verb (επίασεν). However, the most intriguing variant
in John’s Gospel occurs in 12.34, where “this saying” (ό λόγος ούτος)
replaces “the son of man” (ό ύώς του άνθρωπον).89
As we saw earlier, this scribe’s primary characteristic is the produc­
tion of variants when dealing with individual letters. A detailed break­
down of all the singular omissions and additions (in figs. 3.3 and 3.4)
confirms this observation.

FIGURE 3.3: BREAKDOWN OF REGIUS* S

OMISSIONS BY TYPE
| 45 (78.9%)
Letters
Syllables 5 (8.8%)
Words 7 (12.3%)

FIGURE 3.4: BREAKDOWN OF REGIUS’ S

ADDITIONS BY TYPE

Letters 10 (43.5%)
Syllables 5 (21.7%)
Words 8 (34.8%)

When weighed on the basis of these variants’ contribution to the sense

of a passage, virtually all of them make little to no contribution. For instance,
the only omissions that ever produce a significant reading are those that con­
sist of, at least, a syllable or, more typically, a word. When it comes to omit­
ting individual letters, the scribe ofCodex Regius is unable to produce a single
sensical reading in 45 instances. In feet, some of these blunder's are so bad
they are laughable. Thus, in John 130, he wrote “after me comes air" (όπίσω
μου έρχεται αήρ) for “after me comes a man” (όπίσω μου έρχεται άνήρ).
On another occasion, in John 8.45, he changes the conjunction “that” (δτι)
to "the” (ό), which even Tischendorf labels “careless” (ex incuria exddit).90
As for additions, the only case in all 55 ofthis scribe’s singular readings where

89. See “Variant Analysis in Christologically Significant Passages" below.

90. Tischendorf8, Novum Testamentum Graece, loc. cit.

110 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder In John Lie?

λ single letter produces a potentially meaningful variant is in John 1.1c,91 in

the case ofthe article with θεός.
Given the carelessness of the scribe, it seems likely that any textual
variant involving a single letter is an ignorant blunder.9293 The addition of
a single letter, the article with 9eo$, does create a sensical reading, but as
Hort recognized,

Singular readings which make good sense and therefore need imply
no clerical error, but which might also be easily explained as due to a
kind of clerical error already fixed upon the scribe by undoubted ex­
amples, are rendered by the presence of possible clerical error as versa
causa more doubtful than they would otherwise be.93

In the end, the claim that o θεός somehow preserved a Sabellian exemplar
appears to contain more legend than legacy; our scribe has no proclivity
for premeditated theological alterations. Instead, he is just incapable of the
skill and care necessary to make such a fine distinction in a single letter.

Variant Analysis in ChristologicaUy Significant Passages

For the majority of the ChristologicaUy significant variants in John's

Gospel, the testimony is completely unified. At most, only one singular
reading in Codex Regius has any Christological bearing.94 Thus, both in
the classic proof texts for Sabellians, John 10.30 along with 14.9-11, and
in the other Christological bookend to the Gospel,95 John 20.28, no sin­
gular or sub singular readings were created by the scribe of Regius.96

91. Two other instances involving syllables, John 5.3 and 7.8, produce sensical vari­
ants. However, in John 5.3, the addition of the article before πλήθος is arguably
to

nonsense, because It creates awkward grammar in the relationship of this noun to

the following anarthrous adjective (πολύ). In John 7.8, the addition of και at the
beginning of the verse is an insignificant variation.
92. Metzger and Ehrman, Text ofthe New Testament» 77.
93. Westcott and Hort, New Testament in the Original Greek, 233.
“94Γ In 7όΚηΤ2ί34» Codex 'Regius reads ‘this saying* (ό λόγος ουτος) instead of‘this
son of man* (ούτος ό υιός του άνθρώπου). One possible explanation is harmo­
nization with John 7.36, which also contains the phrase τίς ίστιν ό λόγος οΰτος
(much thanks to Dr. Hall Harris III for his input on the use of this expression in
John's Gospel).
95. Several scholars have noted the inclusio formed by John 1.1 and 20.28: see Ray­
mond E. Brown, The Gospel according to John, vol. 1, /-XfZ, Anchor Bible (Garden
City, NY: Doubleday, 1966), 5; William Loader, The Christology of the Fourth
Gospel· Structure and Issues (New York: Peter Lang, 1989), 167: Hurtado, Lord
Jesus Christ, 369.
96. There is a variant in John 20228 involving the omission of die article with the divine
name θεός, which Tischendorf listed in two witnesses (D, 46**). Though Ehrman
called this an "anti-Patripassianist corruption* {Orthodox Corruption, 266), he still
affirmed that John 10.30 and 20.28 were key texts that speak of Jesus deity (Bart D.

Matthew P. Morgan Chapter 3 / 111

 THE LEGACY OF A LETTER

An assessment of the theological legacy or legend in Codex L would be

incomplete without examining John 1.18, where the external evidence is
more evenly divided. There, three different textual options appear, which
have been proposed as a chronologically linked progression of theologi­
cally motivated changes.9798 A detailed treatment of this textual problem is
99
not fitting here,9* but figure 3.5 highlights a possible theological progres­
sion and the motivations that generated different readings.

FIGURE 3.5: POSSIBLE PROGRESSION OF

VARIANT READINGS IN JOHN 1:18"

ORIGINAL:100 ό μονογενής υιό?

FIRST CORRUPTION:101 ό μονογενής θεός (anti-adoptionistic)

SECOND CORRUPTION:102 μονογενής θεός (anti-SabeUian/cwnJ

inL)

If this reconstruction is correct, the following alternatives about

Regius's scribe emerge:

1. the scribe holds inconsistent theology,

2. he faithfully copies a conflicted Vorlage, Neor
3. he is completely ignorant of its theological significance.

Ehrman, MisquotingJesus: The Story behind Who Changed the Bible and Why [San
Francisco: HarperSanFrancisco, 2005], 161).
97. Ehrman, Orthodox Corruption, 265-66; Ehrman, Misquoting Jesus, 161-62.
98. For a detailed discussion, see Kenneth Willis Clark, “The Text of the Gospel of John
in Third-Century Egypt," NovT 5 (1962): 19-20; Murray J. Harris, Jesus as God:
The New Testament Use ofTheos in Reference to Jesus (Grand Rapids: Baker, 1992),
74—83; Stratton L. Ladewig, "An Examination of the Orthodoxy of the Variants in
Light of Bart Ehrman’s The Orthodox Corruption ofScripture” (ThM thesis, Dallas
Theological Seminary, 2000), 1-80; Raymond E. Brown, Jesus: God and Man (Mil­
waukee: Bruce, 1967), 12-13; Paul R. McReynolds, "John 1:18 in Textual Variation
and Translation,” in New Testament Textual Criticism: Its Significancefor Exegesis:
Essays in Honour ofBruce M. Metzger, ed. Eldon Jay Epp and Gordon D. Fee (New
York: Oxford University Press, 1981), 105-18; Bart D. Ehrman, Gordon D. Fee, and
Michael W. Holmes, Text of the Fourth Gospel, 60 n. 12; Bart D. Ehrman, “Hera-
cleon and the ‘Western’ Textual Tradition,* NTS 40 (1994): 166.
99. This table does not reflect the view of the author but shows the possible progres­
sion of variants as proposed by Bart Ehrman (Orthodox Corruption, 265-66; Mis­
quoting Jesus, 161-62).
100. Witnesses for this reading are A C3 Ws Δ Θ Ψ SDiy1,13 lat syr^h.
101. Witnesses for this reading are 5p75 K1 33 pc.
102. Witnesses for this reading are X* B C“ syrUro£.

112 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

Assuming intentionality in L’s alterations in John 1.18, the singular

reading ό θ€0ς in John 1.1c potentially represents a _pro~Sabellian
Christology. Yet in the span of only seventeen verses, the same person
produces an equally motivated ζζτζίζ-Sabellian reading in John 1.18.
Given the history of textual changes in 1.18, the potential for theo­
logical influence on the transmissional process is much greater due to
the early manuscript evidence;103 but if 1.18 is our window into the
Vorlage of Regius, we should expect a text that has removed any hint of
Sabellian influence. Consequently, it is unlikely that an internal theo­
logical conflict existed in the exemplar of L, eliminating the preceding
option 2.
Thus the origin of this variant lies squarely on the shoulders of this
eighth-century scribe. This brings us back to two possible explanations:
either his personal motivations were easily conflicted, or he was obliv­
ious to the theological implications of the reading he produced. When
combined with his general ineptitude regarding individual letters, the
best explanation for the addition of the article in John 1.1c once again
lies in a careless mistake on the part of the scribe rather than in any
theological predisposition.
In retelling the story of Codex Regius from the standpoint of scribal
habits, with special attention given to theological motivations, all roads
lead to a common destination; the reading και ό θβός ήυ ό λόγος stands
as the exception to every identifiable pattern of behavior (i.e., singular
readings and theologically significant variants) in this entire eighth-cen­
tury manuscript’s transmission ofJohn’s Gospel.

Codex Freerianus's Scribal Profile

Although somewhat obscure over the last hundred years,104 the ad­
ditional witness preserved in the replacement quire of Codex Freerianus
(only John 1.1-5.11) forms another story within the textual history of
John 1.1c. The limited length and scope of this quire makes the dis­
covery of detailed scribal behavior more difficult; however, given its
chronological· relationship to Codex Regius, it is important to analyze
how these stories intersect. The primary avenue for this investigation
is the subsingular readings shared only by L and Ws, along with other
textual variants involving a few key witnesses (e.g., K [01] B [03]
etc.).105 From this angle, we can explore the degree of dependence

103. Ehrman also stresses this point (Orthodox Corruption, 271 n. 23) to show the cer­
tainty of an early third-century date corresponding to the time period when this
heresy was prominent (see Figure 3.5).
104. See the appendix at the end of this chapter for a survey of the history of Ws in the
critical editions.
105. This investigation does not perform a detailed quantitative analysis but provides
a simple exploration of the 53 instances where both L and Ws agree against the

Matthew P. Morgan Chapter 3 / 113

 THE LEGACY OF A LETTER

between these two documents and the likelihood that their combined
witness has legitimate roots extending back to the early Sabellian con­
troversies. Five key points emerge:

1. A look at the subsingular readings involving only Codex L and

Codex Ws reveals that John 1.1c is the only meaningful variant pro­
duced exclusively in both manuscripts.106 Even though Ws does not
contain a large portion of John’s Gospel, it is striking that there are no
other places where a sensical variant is recorded only in these eighth­
century witnesses. Furthermore, that none of these variants can even be
found in the apparatus of NA27/UBS4 casts further doubt on the quality
of their legacy. This lack of overlap also demonstrates that these read­
ings are isolated and not derivedfrom one another.

2. There are 50 variants where L and Ws join forces against the MT.
A good majority of these (42 [84%]) involve some relationship to the key
witnesses K (01) and B (03).107 The strong presence of K and B (25 variants
[60%])108 in places where these manuscripts intersect is striking. As was
noted earlier, both of these are parental figures in the Alexandrian heri­
tage for Codex L.109 Combined with Metzger and Ehrman’s analysis that
XVs contains some Alexandrian and Western readings,110 these variants
probably represent the common Alexandrian ancestry shared by both
eighth-century manuscripts.

reading of the Robinson-Pierpont Majority Text (2005). This group of variants in

John 1:1-5:11 produced some patterns that provide dues to possible relationships
with other MSS. From these initial observations, a more thorough profile of both L
and W5 in John's Gospel is necessary to gain greater confidence in textual relation­
ships that may exist for these particular MSS.
106. There are a total of 4 subsingular readings, but the other 3 (John 1.19, 39,47) all
involve insignificant orthographic variants: απεστειλαν is spelled απεστιλαν by
both L and Ws in John 1.19, οψεσθε is spelled οψεσθαι in John 1.39, and 1.47 re­
places ιδε withei5e.
107. Though orthographic variants were included in the analysis of subsingular read­
ings shared by L and Ws, they have been thrown out for the remainder of the dis­
cussion, as they do not further our understanding of the relationship between MSS '
(cf. n. 105). Therefore, the total number of variants where L and W5 stand against
the tide of the MT is 50.
108. These statistics do not reflect a pure quantitative analysis or family profile but de­
note how frequently K and B both agree with L and W® against the MT.
109. This observation is all the more striking given Fee's conclusion that K's text type
found in John 1-8 is primarily Western, based on its frequent alignment with D
[06] (Gordon D. Fee, "Codex Sinaiticus in the Gospel of John, “MTS 15 [1968-69]:
23-44). This snapshot seems to show that L and Ws find agreement through a
different gene pool than the predominantly Western character of R. In feet, there
was not a single alignment of L and Ws against the MT where the witness D [06]
was involved. Therefore, it is difficult to see how there is any Western influence on
variants where our two MSS argue against the MT.
110. Metzger and Ehrman, Text ofthe New Testament, 80.

114 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

3. Looking at these 50 variants from another vantage point, there

are 19 instances where both L and W5 find support from a broader range
of early manuscripts in their agreement against the MT.111 Within this
demographic, 12 instances (63%) involve the early Alexandrian witness
B (03). There are also 10 (53%) that are supported by the papyri tandem
of ϊ>δδ and ip75, with 3 others containing at least one of these early papyri.112
From the standpoint of quality, less than half (9) attest to the accepted
reading in the NA27 text.113 Digging deeper into the accepted readings,
5 contain the triad of B jl65 $>75-114, but all include at least B. At the same
time, 5 (28%) of the variants in this category went against the reading
supported by K (01).115

4. There are 23 instances where the reading of Ws appears with

a/ew other witnesses.116 Within this grouping, 6 were included in the
NA27 text, and all were no doubt adopted due to the presence of more
noble ancestry (K B C D ξρδ6 ip75).117 There were 8 readings attested
only in MSS dating from the eighth century or later.118 From this en­
tourage, there was not a single reading warranting inclusion in the
NA27 text. The observation that Ws cannot stand without help from
other early Alexandrian or Western witnesses deals a critical blow to
the possibility that its reading in John 1.1c preserves a unique legacy
from early in the manuscript tradition.

5. A survey of the variants in this replacement quire for W revealed

two nonsense errors involving nomina sacra of the divine names. In John
4.24, the articular nomen sacrum for θεός· was written os rather than
o 9s- Similarly, in the next verse, 4.25, Χριστός was spelled χρ instead
of xs. Then, in 4.48, the article was removed from Ίησοΰς, producing
a singular reading in Ws. Similarly, at the end of the quire, 5.12, the

111. John 1.19,20,37,49frts; 2.15&S, 18; 3.3,4; 4.1, 5, 34,36,37,38,45, 52; 5.10.
112. Of the 12 readings attested by B, 6 of them are supported by both 3)ee and $>75 as
welL

H4.'The three exceptions are John 2.18,4.37, and 4.52.

115. John 1.49bis; 3.3; 4.36, 37. It is interesting to note that John 1.49 plays a significant
role in the articulation and development of Colwell’s rule (E. C. Colwell, "A Definite
Rule for die Use of the Article in die Greek New Testament," JBL 52 [1933]: 12-21,
esp. 13).
116. John 1.44; 2.4,17,24; 3.21, 23, 28; 4.14,17,23,25,27, 30, 31, 35, 42te, 43,46; 5.2,
9, ll&zs.
117. John 3.23; 4.14,25,42,43, 46. All of these readings contain at least 2 of the 6 wit­
nesses previously listed, and 2 of them (John 4.25,43) appear in 5 ofthe 6 witnesses.
118. John 1.44; 2.4,24; 3.21; 4.17,27,35,42. The variant in 1.44 is the most interesting.
It involves the omission of the article o with the proper name Φίλιππο? and is only
supported by the corrector of Codex F (09) and Codex 047, which are dated in
the ninth and eighth centuries, respectively. The substitution in 3.21 changes from
eoTLV toeLOiv only in Ws and«P (044), which are both late witnesses.

Matthew P. Morgan Chapter 3 / 115

 THE LEGACY OF A LETTER

entire verse 12 is missing to make a clean fit with the adjoining section
of the book. Though limited, these tendencies further the notion that
the scribe of this quire was not cautious when copying tire divine name.
Such a trait casts farther doubt on the credibility of his legacy contained
in John 1.1c.

In conclusion, this survey of the quality and type of readings found

in Ws affirms three important things pertaining to Codex Regius and
John 1.1c: *

1. There is no evidence to establish a direct relationship between

these two eighth-century manuscripts. As a result, the occurrences of
the article with θεός found in John 1.1c in both MSS should be consid­
ered isolated readings.

2. Alignment of Codex L and W5 never merits inclusion in the ac­

cepted text of NA27 without support from other key MSS (X B C D m366

3. There are no known instances where a combination of Ws with

a single other witness finds credibility as a potentially “original” reading.

Therefore, the inclusion of W® as a subsingular reading in John 1.1c

does not negate the egregious nature of the scribal behavior in Codex
L, and it farther demonstrates that this combination possesses insuffi­
cient testimony to consider the reading και ό θεός ήν ό λόγος to be a
plausible original.

The Impact ofScribal Behavior on the Variant in John 1.1c

Despite the lengthy and often complex stories of Codices Regius

and Freerianus, the possible conclusions are more straightforward.
Does the tale of these two eighth-century manuscripts and the scribes
that produced them point to a rich, almost extinct theological legacy or
to an innocent blunder mistaken as a theological legend? The answer
lies in the individual scribes and the potential relationship between
these two ancient documents.
In the case of Codex Regius, we entered the world of its scribe to
determine whether this variant faithfully reflects an earlier original or
finds its origin in a mindless slip of the scribe’s pen. Describing the pos­
sibilities, Hort noted, “One MS will transmit a substantially pure text
disfigured by the blunders of a careless scribe, another will reproduce a
deeply adulterated text with smooth faultlessness.”119 For the scribe of

119. Westcott and Hort, New Testament in the Original Greek, 36.

116 / Chapter 3 Matthew P Morgan.

 Sabellianism or Scribal Blunder in John 1.1c?

Codex Regius, the addition of a single letter fits the description of his
most common blunders like a glove. So, although a theological reading
seems legitimate on the surface, placing this variant in the company of
similar variants exposes its careless character. As a result, the addition
of the article in John 1.1c in Regius faithfully preserves the story of a
scribe whose wildness desecrated the purity of his exemplar, rather than
an early Sabellian legacy.
The story of John 1.1c in Codex Regius becomes more complex with
the addition of Ws as a witness. Outside of John 1.1c, there is not a
single meaningful variant found only in L and W8. Also, the textual reli­
ability of these eighth-century manuscripts is weak at best, requiring
reinforcement by multiple witnesses, particularly the pillars of the man­
uscript tradition (8 B C D Φ65 φ75). Although the story of the article’s
origin in W8 remains somewhat obscure, it has no demonstrable rela­
tionship to the transmission of Regius, and its ability to reflect a pristine
early text is irrevocably marred by its company with the equally late and
poorly copied Regius text.

THE GRAMMATICAL VIABILITY

OF THE TEXTUAL VARIANT IN JOHN 1.1C

The History and Significance of Colwell’s Rule in John 1.1c

At times, the heavy clouds of history seem to restrict the clarity
and definitiveness of the story preserved in Codices L and W8, but a
look at the grammatical character of the variant και o Geos ήν ό λόγος
goes a long way in removing the fog. The rationale for this approach can
be found in the opening remarks of Middleton’s magnum opus on the
Greek article:

The student in Theology cannot fail to have remarked, that the exposi­
tion of various passages of the New Testament is by Commentators
made to depend on the presence or the absence of the Article in the
'Greek original.120

When dealing with the text of John 1.1c, the weight of Middleton’s
comments has been shown in the wealth of scholarship that has sought
to understand the meaning of the anarthrous θεός. The foremost concern
of previous grammatical analysis has been in determining whether this
noun should have a qualitative, definite, or indefinite force. As a result,
few have addressed the significance of the preverbal anarthrous predicate

120. T. F. Middleton, The Doctrine ofthe Greek Article, ed. J. Scholefield, 2nd ed. (Cam­
bridge: J. and J. J. Deighton, 1841), xii.

Matthew P. Morgan Chapter 3 / 117

 THE LEGACY OF A LETTER

construction with regard to the textual problem in John 1.1c.121 By adding

the article, this reading makes both the subject and die predicate explicitly
definite. Semantically, it also places the two nouns in a convertible propo­
sition that would explicitly support a Sabellian or modalistic theology.122
The modest goal in this section is twofold: first, to determine how
the grammar and semantics of predicate nominative constructions in
the NT relate to the presence or lack of the article; second, to scrutinize
the grammatical viability of the specific textual variant found in our two
eighth-century manuscripts. As a result, normative grammatical pat­
terns will be used to shed light on the potential legacy of the article in
the two manuscript testimonies.

An Examination ofPredicate Nominative Constructions

Using Είμί in the New Testament

Since it was first published,123 Colwell’s rule has received much

criticism due to some serious methodological flaws in its formula­
tion.124 Appropriately, the approach taken here will first categorize
the use of numerous predicate nominatives in the NT on the basis
of word order and the inclusion or omission of the article. Once the
usage has been arranged according to these factors, it will be further
evaluated in terms of the type of proposition it creates, either subset
or convertible. The final analysis will then seek to compare the vi­
ability of both variants in John 1.1c on the basis of normative usage
in the NT.

The Methods and Limitations of this Study

This study will only examine predicate nominative constructions
that occur with the verb είμί.125 Excluded from the present treat­
ment is the force (e.g., qualitative, indefinite, definite) of the predicate

121. Philip B. Hamer, “Qualitative Anarthrous Predicate Nouns: Mark 15:39 and John
1:1/ JBL 92 (1973): 85. Harner noted this variant in Codex L but quickly dismissed
it on the basis of the theological contradiction it poses with 1.1b.
122. Robertson, Grammar ofthe Greek New Testament, 767-68.
123. Colwell, “Definite Rule/ 12-21.
124. For a detailed history of Colwell’s rule in NT studies, see Matthew P. Morgan, “Ihe
Legacy of a Letter: Sabellianism or Scribal Blunder in John 1:1c?” (IhM thesis,
Dallas Theological Seminary, May 2009), 53-69.
125. Other verbs that utilize predicate nominative constructions are γίνομαι, υπάρχω,
and ευρίσκω, along with the passive form of the verbs οικοδομάω, όυομάζω, καλέω,
and ερμηνεύω. Of these verbs, BDAG only explicitly mentions the use of predicate
nominatives with the passive form of ονομάζω (ad loc.), although the examples
listed with the passive of καλέω (ad loc.) also show its use of the predicate nomina­
tive. However, due to length and scope (John 1.1c), these were not considered at
this point.

118 / Chapters Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

nominative as a primary factor in the presence of the article. In addi­

tion, it will largely ignore cases where the equative verb occurs with
relative pronouns and proper names, since they do not directly apply to
the situation in John 1.1c. Although some consideration will be given to
other semantic influences on the article (e.g., monadic nouns, genitive
modifiers, etc.), word order and function (i.e., subject versus predicate)
will form the structure of this investigation.

Overview ofPredicate Nominative Constructions Using Είμί in the

New Testament

Postverbal Anarthrous Constructions126

For the purposes of the present study, this structural category has
no bearing on either the nature or the semantics of the readings in ques­
tion. Though there are a fair number of instances in the NT (53), the
fact that ο λόγος is uncontested among known manuscripts removes
its voice from the grammatical discussion of John 1.1c. In addition, the
lack of the article with the predicate produces fertile soil for subset,
rather than convertible, propositions.127

Preverbal Arthrous Constructions128

If ό θεός in John 1.1c functions as the predicate nominative,129 it

should conform to normative patterns of other preverbal arthrous predi­
cates in the NT. There are at least two instances (Matt 6.22; Luke 11.4)
that would seem to affirm the possibility of a postverbal articular subject

126. For a more detailed analysis, see Morgan, "Legacy of a Letter? 36-37,70.
127. Daniel B. Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of
the New Testament (Grand Rapids: Zondervan, 1996, 41-46. Wallace stated that
a convertible proposition occurs “when both substantives meet one of the three
_ _ qualifications for S” (Wallace, Exegetical Syntax, 45). Since one of the criteria for
determining the subject is the presence of the article, it is often the case that the
predicate will be anarthrous. In instances where this occurs, the most likely se­
mantic relationship is a subset proposition where the predicate "describes the class
to which the subject belongs” (Wallace, Exegetical Syntax, 41). Within this cat­
egory, just over half of the qualified applicants fit this caricature by forming subset
relationships.
128. For a more detailed analysis, see Morgan, "Legacy of a Letter? 37-39,70.
129. Based on well-documented rules (Lane C. McGaughy, Toward a Descriptive Anal­
ysis of ^EINAJas a Linking Verb in New Testament Greek, SBL Dissertation Series
6 (Missoula, MT: Society of Biblical Literature for the Linguistics Seminar, 1972),
53-54), this condition would not be the case. However, as will be argued later, this
is not a viable option for other reasons. Its treatment here accommodates those
who would label it as predicate in order to avoid even greater problems that arise
when ό θεός1 becomes the subject.

Matthew P. Morgan Chapter 3 / 119

 THE LEGACY OF A LETTER

(ό λόγος). However, this support dwindles to only one unique case con­
taining other influential factors that best explain the arthrous predicate.130
Furthermore, in keeping with less debatable examples,131 the preferred
placement of ό λόγος would be before the verb, with the predicate ό θεός
on the same side.132
Looking at the semantic situation, every instance involving a nomi­
native subject133 forms a convertible proposition indicating an “identical
exchange”134 between them.135 If the reading καϊ ό θεός ήν ό λόγος is
squeezed into this structural category, it would find further semantic
support as a convertible proposition, or, as Robertson points out,
Sabellianism.136
However, the barriers to placing the variant within this camp are
insurmountable. First, this structural category as a whole is extremely
rare, making it unlikely that a scribe would intentionally consider it.
Second, to squeeze into this framework, the variant must swim against
the current of normative word order. Finally, it fails to adopt conven­
tional word order that places both nominatives before the verb. As a re­
sult, it is difficult to embrace a diagnosis that renders the reading και ό
θεός ήν ό λόγος as a viable grammatical option within this classification.

Postverbal Arthrous Constructions137

I now turn to the category that best fits the reading και ό θεός ήν
ό λόγος in John 1.1c. Here the roles will be reversed, with ό θεός being
the subject and ό λόγος the predicate.138 Of all of the potential struc­
tural matches, at least one of the nominatives is impersonal and carries
a distinctively qualitative force. Though the pool of examples is quite

130. The text in Matthew reads ό λύχνο? τοΰ σώματο? έστιν ό οφθαλμό?, and the only
difference in Luke is the genitive modifier attached to the subject: ό λύχνο? τοΰ
σώματο? έστιν ό όφθαλμό? σου. Thus these two texts are identical in terms of how
they arrange the subjects and predicates. It should be pointed out that the lexical
nature of the predicate is impersonal (ό λύχνο?) as well as figurative. Also, both of
these examples contain arthrous genitive modifiers that demand an article for their
nominative counterpart in accordance with Apollonius’s canon.
131. John 6.63te; 15.1; 1 Cor 11.3,25; 2 Cor 3.17.
132. It is interesting that none of the potential constructions Harner proposed for John
1.1c Include this possibility (Harner, “Predicate Nouns,” 84).
133. This statement excludes nominative pronouns, since they do fit the construction in
John 1.1c and are clear in expressing the subject.
134. Wallace, Exegetical Syntax, 41.
135. Matt 6.22; Luke 11.34; John 15.1; 1 Cor 11.3,25; 2 Cor 3.17.
136. Robertson, Grammar of the Greek New Testament, 767-68.
137. For a more detailed analysis, see Morgan, "Legacy of a Letter,” 39-41,70.
138. This distinction is in keeping with the rule that makes word order the determining
factor for the subject when both nominatives have the article (McGaughy, Descrip­
tive Analysis, 53-54).

120 / Chapters Matthew P. Morgan

 Sabeliianism or Scribal Blunder in John 1.1c?

shallow, when an articular subject and predicate are split across the
verb, at least one prefers to be impersonal.
Initially, the variant και ό θεός ήν ό λόγος appears to be in good
semantic and structural company when ό θεός is the subject. On the
surface, then, the semantic “equal sign” formed by the convertible
proposition appears to favor a pro-Sabellian reading; but since the
variant και ό θεός ήν ό λόγος contains two personal nominatives,
Revelation 18.23 presents the only structural parallel within the con­
fines of convertible propositions. Still, the type of equivalence found
in the variant for John 1.1c involves two individuals, not two groups or
classes of people, like those in Revelation 18.23. So, within the scope
of the limited examples in the NT, the reading κα'ι ό θεός ήν ό λόγος
represents the only clear case where two personal singular nominatives
are placed in a convertible proposition. As a result, both the rarity and
the nuances of these constructions place the reading και ό θεός ήν ό
λόγος in a league of its own.

Preverbal Anarthrous Constructions139

So far, the focus has been on the variant reading in John 1.1c com­
pared against other parallel NT constructions, but now 1 turn to focus

the overwhelming favorite. This confirms Colwell’s general conclusion

only forms a textbook structural example but also fits the most popular
form of equative constructions.
The reading και θεός ήν ό λόγος can be shown to fit perfectly
within the grammatical boundaries and tendencies of the NT. From this
analysis, it appears that when the subject occurs opposite an anarthrous
predicate and after the verb, unless it is a pronoun, it prefers the article.
In the case of John 1.1c, the subject (ό λόγος) comports perfectly with
this tendency.
.. ^Although this category, of predicates contains fertile soil for pro­
ducing equative constructions, it yields only a paltry harvest of convert­
ible or subset propositions.140 Based on the limited crop of examples,
the overwhelming majority involve singular nouns in subset proposi­
tions. Similarly, the reading κα'ι θεός ην ό λόγος is a classic example
of a subset proposition where “the word” is classified in the broader
category of God.141

139. For a more detailed analysis, see Morgan, “Legacy of a Letter,’ 41-42,71.
140. Pronouns are not included in the idea of nominatives here because they strip away
any ambiguity regarding the semantic relationship (i.e., convertible or subset).
141. Wallace, Exegetical Syntax, 45.

Matthew P. Morgan Chapter 3 / 121

 THE LEGACY OF A LETTER

Most often, singular nominatives forming a subset proposition

prefer to place both the subject and predicate before the verb. This leaves
only 7 instances where the semantic situation fits the contours of John
1.1c.142 By hiding the subject behind the verb, these contractions lay
additional stress on the class indicated by the predicate, while clearly
specifying the subject with the article.143144
This nuance fits the context of
John 1.1c nicely, balancing the earlier distinction from God (1.1b) with
a strong statement emphasizing the Word’s existence in the category of
Geos. As a result, the reading και θεό? ήν ό λόγο? not only conforms
to its grammatical peers but also provides the best reading within the
context of John 1.1 as a whole.
Although the reading και ό θεό? ήν ό λόγος represents the most
natural way to convey a convertible relationship between two nomina­
tives, if authentic, it would stand as the only clear instance where two
singular personal nouns are interchangeable in the NT.

Grammatical Implicationsfor the Textual Problem in John

As an expert in textual criticism, Colwell formulated his rule to
determine whether the addition or omission of an article should be
“original.” Likewise, the present investigation has sought to deter­
mine whether the story of the variant καί ό θεό? ήν ό λόγο? finds
harmony among its grammatical peers. When the rules and nuances
of nominatives used with είμί are unpacked, the plausibility of this
variant holds very little promise; and in terms of both structure and
semantics (i.e., convertible or subset proposition), its species is virtu­
ally extinct in the NT. More pointedly, it represents the only construc­
tion of its kind, equating two singular personal nominatives without

142. Matt 12.8; Mark 2.28; Luke 6.5 (these first 3 are synoptic parallels); John 1.1c; Acts
28.4; Rom 1.9; 1 Tim 6.10.
143. To illustrate this point, consider the difference between Acts 22.26 and 28.4. The
subject in both is "this man" (ό άνθρωπο? ούτο?), making the only distinction in
the orientation relative to the verb. Though the predicate nouns are different, both
appear without any modifiers or other grammatical influences, lite construction
in Acts 22.26 is typical of the majority (20) of the subset propositions involving
singular nouns that seek to lay stress on the subject. In the context of this passage,
both the soldier and the tribunal (22.27) are stressing their amazement that the
person they are about to flog is a Roman. Meanwhile, in Acts 28.4, Paul has a viper
latched onto his hand, causing the people of Malta to classify him as a murderer.
Therefore, in this instance, the predicate “murderer" (φονεύς) is moved forward
to stress the class that Paul’s recent circumstance reveals. Ln conclusion, these two
examples illustrate the semantic function of different subset propositions and how
they influence the word order and emphasis in the context.
144. For a discussion of the grammatical nuances ofλόγο? and θεό? in the NT and their
influence on the textual problem in John 1.1c, see Morgan, “Legacy of a Letter,"
43-45.

122 / Chapter 3 Matthew P. Morgan

 Sabellianism or Scribal Blunder in John 1.1c?

involving a proper name.145 Meanwhile, the reading καί θεός ήυ ό

λόγος does not face a single grammatical challenge to its integrity.
For these reasons, the tale of an early Sabellian reading in John 1.1c
grows taller under the scrutiny of normative grammatical usage in
the NT.

CONCLUSION

This study began on a quest to listen carefully to the shouts and

whispers of two eighth-century voices with regard to the text of John
1:1c. Although numerous scholars draw attention to the possibility
of theologically motivated corruptions, they rarely move beyond the
realm of speculation to discern the source of these readings (i.e., indi­
vidual MSS, patristic evidence, syntax/internal evidence). During each
step along the winding road of history, this search has sought to discern
whether the reading και 0 θεός ήν ό λόγος speaks of a mostly hidden
Sabellian legacy that has almost vanished from transmissional history.
This task involved muting those voices that represent a mythical legend
rather than a theological landmark. To this end, our pursuit sequestered
three streams of evidence: the roots and rise of Sabellianism, the scribal
character of Codices L and Ws, and the normative grammatical consti­
tution of equative constructions in the NT. Three conclusions result:

1. The most likely scenario for an early Sabellian text and subse­
quent orthodox cover-up would be front-page news about this heresy
during the early second century. This allows for the polemical heat to
rise to the surface of the transmissional process in its most vulnerable
period. Furthermore, this polemical environment would likely include
strong disagreement centered on John 1.1, but if we listen carefully to
the voice of history, the story does not line up.

2. The evidence from the ancient scribes who produced the

.reading και ό θεός.ην ό λόγος is compelling. With Codex Regius, the
addition of a single letter fits the scribe’s most common foibles to a tee,
so the extra article in John 1.1c best demonstrates sloppy habits rather
than an early Sabellian legacy.
Although W5 does not leave sufficient information to form solid
scribal tendencies, its presence deserves some attention. The natures
and dates of Regius and W8 raise the possibility of at least one earlier

145. for an extensive investigation regarding the use of proper names in equative con­
structions, see Mario Cerda, “Subject Determination in Koine Greek Equative
Clauses Involving Proper Nouns and Articular Nouns" (ThM thesis, Dallas Theo­
logical Seminary, May 2005), 1-123.

Matthew P. Morgan Chapter 3 / 123

 THE LEGACY OF A LETTER

parent, but no dependency can be established between these two man­

uscripts. They also appear feeble and frail when left on their own, only
garnering serious consideration when they echo the voice of much
stronger manuscripts (X B C D £>όδ $>75). As a result, it seems very un­
likely that W3 has preserved a Sabellian legacy.

3. Another voice in the halls of history is that of comparable gram­

matical usage in the NT. As with both of the other perspectives, the
reading και ό θεό? ήν ό λόγο? heads into unchartered territory. The
addition of the article changes the semantics from a subset relationship
to a convertible proposition, making Geo? equivalent to λόγο?. Though
structurally sound, it represents the only place in the NT where two
singular personal nouns are made interchangeable. At the same time,
its counterpart, και Geo? ήν ό λόγο? is the epitome of a typical equative
construction. Consequently, I conclude that the notion that the article
with θεό? supports an earlier Sabellian reading is an unsightly myth.

The fascinating story of these two eighth-century manuscripts, with

a single letter that they alone record, has only one ending. To argue that
the reading of John 1.1c originally included the article with θεό? and
preserves a Sabellian Christology is to violate the historical context, the
scribal and transmissional story, and the demonstrable norms of NT
Greek syntax. Such a feat is more than two questionable eighth-century
manuscripts can bear. Rather, their tarnished voice can only join the
chorus of countless others that boldly proclaim that Jesus is God (και
θεό? ήν ό λόγο?).

APPENDIX146
THE HISTORY OF Ws IN THE CRITICAL EDITIONS
At the beginning of this study, a few years ago, there was almost
unanimous confirmation that the reading και ό θεό? ήν ό λόγο? was
found only in the eighth-century manuscript Codex Regius (L [019]).
Looking at standard critical texts such as NA27, Tischendorf’s eighth
edition, Swanson, and UBS4, there was no place where the supplement
to Codex Freerianus (Ws [032-S]) appeared.147 The only critical texts
showing Codex Ws in support of this variant were the editions done
by Hermann von Soden and Augustinus Merk.148 Most recently, the

146. Special thanks go to J. Bruce Prior and T. A. E. Brown, whose kind attention and
interaction strengthened die contents of this appendix.
147. In all publications prior to James R. Royse's Scribal Habits (65), the reading of
Codex L in John 1. lc was more than qualified to stand as a singular reading.
148. Hermann Freiherr von Soden, Die Schriften des Neuen Testaments in ihrer Al~
testen erreichbaren Textgestalt kergestellt aufgrand ihrer Textgeschicte (Gottingen:

124 / Chapter 3 Matthew P. Morgan

 Sabeilianism. or Scribal Blunder in John l.lc?

release of the IGNT presented the first indication that the supplement
of Codex W also supported the articular θεό? in John 1.1c.149
This shift in textual support led to my interaction with Larry Hurtado,
an expert on Codex W, and with two of his colleagues, J. Bruce Prior and
T. A. E. Brown, who have special access to several series of recent digital
images of the manuscript, have had access to the manuscript itself, and
were responsible for the transcription of W5 used in the IGNT. Along with
seeking other expert opinion, my own personal investigation sought to
make sense of the inconsistent record found in this survey of critical Greek
texts. After studying Sanders’ facsimile of the manuscript,160 it seemed pos­
sible that the feint presence of an omicron in John l.lc was due to bleed-
through from the verso side of the leaf or offset from the end of Matthew.151
During this process, Brown was able to draw my attention to some images
that have been made available to the public since 2006 and that conclu­
sively ruled out the possibility of bleed-through or offset as an explanation
for the article.152 As a result, all three scholars that were consulted affirm
the presence of the omicron in John l.lc as “incontrovertible.”153
Important in this discussion is the history of the Freer Gospel man­
uscripts and how their acquisition and public availability fit chrono­
logically with the publication of several Greek texts. The story begins in
1906, when Charles Freer purchased the manuscripts in Cairo, Egypt.
Approximately one year later, they were brought to the attention of
Henry A. Sanders at the University of Michigan, where they were first
analyzed by scholars.154 Along with Sanders, one of the first to publish a
collation of the manuscript was Edgar J. Goodspeed in 1913.155 Until that
publication, textual scholars were largely unaware of this manuscript.

Vandenhoeck und Ruprecht, 1911), 390; Augustinus Merk, Novum Testamentum:

Graece etLatine, 11th ed. (Romae: Sumptibus Pontificii Instituti Biblici, 1992), 307.
149. Schmid, Elliott, and Parker, Afew Testament in Greek lVt 189.
150. Sanders, Facsimile of the Washington Manuscript, 113.
151. The main reasons for this were (1) the clear indication of offset seen in the title of
John’s Gospel onto the last page of Matthew; (2) severe water damage on the top
portion of the page containing John 1; and (3) the location of the p in αρχή above
the omicron in question, which save the distinct impression in the facsimile that
another letter was impressed on top of it.
152. These images addressed two things that could not be adequately seen in the
Sanders facsimile. First, a digital overlay of the recto and verso sides showed that
none of the surrounding omicrons aligned with the letter in John l.lc. Second,
the enhanced digital images of the omicron give strong indication that this was
included in the creation of the page and not part of later water damage,
153. This quotation came from Prior in an email of February 20, 2008, where he wrote,
“As far as 1 am concerned, the reading with the article before ΘΣ in Ws is incontro­
vertible.” Along the same lines, Brown wrote in an email of February 16,2008, that
he was “certain the article was present.” Hurtado also affirmed the presence of an
omicron in an email of February 15,2008.
154. Hurtado, “Introduction,” in The Freer Biblical Manuscripts. 1-3.
155. Goodspeed, “Freer Gospels,” 599-613.

Matthew P. Morgan Chapters / 125

 THE LEGACY OF A LETTER

Considering that Tischendorf died in 1874, it was obviously impossible

for him to know of the additional evidence of W5 supporting the inclu­
sion of the article with θεός in John 1.1c. Similarly, the completion of
Hermann von Soden’s massive work, combined with his death in 1914,
made inclusion of such breaking information rather remarkable.156
More difficult to explain are the recent versions that could have ac­
cessed the collations of Sanders and Goodspeed. For instance, Swanson
seems to have missed the omicron in his survey of the manuscript while
using a “photographic film negative.”157 Similarly, the apparatus of NA27
was produced without resources like the Prior and Brown transcrip­
tion. According to Prior, discussions are underway with Klaus Wachtel
in Munster to include this information in the forthcoming NA28 text.
My own personal examination of the Sanders facsimile confirmed that
the extent of the water damage and resolution of the text around John
1.1c make any Judgment of its reading tenuous at best. However, the nu­
merous digital photographs taken by the Freer and Sadder Galleries and
reviewed by Prior and Brown conclusively affirmed the indusion of the
article in Ws.

156. Credit for the historical backdrop of this section must be given to J. Bruce Prior.
During numerous email interactions, he carefully and extensively answered many
questions regarding die history of scholarship pertaining to this manuscript.
157. J. Bruce Prior, email, February 20,2008·.

126 / Chapter 3 Matthew P. Morgan

 4

Patristic Theology and Recension

in Matthew 24.36
An Evaluation of Ehrman’s
Text-Critical Methodology

Adam G. Messer1

he many differences in the biblical manuscripts seem to dim the

T lit beacon of God’s Word, forcing us to ask, is the light emanating

from the text of Scripture the brilliance of a sacred message, or is it
merely the flames of a dying faith? It has been acknowledged for some
time that alterations in the manuscript tradition were not due merely
to accidental slips of the senses. In fact, the fathers of modern textual

1. This chapter is dedicated to my wife Alicia, who suffered the effects of HG, a debili­
tating and potentially life-threatening pregnancy disease, with our firstborn child.
Every day for four months, she wasted away a little more, seemingly beyond hope;
yet she trusted in the Lord to bring her through it, and he surprised us with his
provision.
It is also dedicated to a woman who offered her aid even while losing her hus­
band and to a man who has been steadily losing his wife to an evil illness yet still
.. .wears the garment .of trust and affection for his Lord and others. The strength of
my wife and friends makes hearts flutter in expectation of what awaits those who
are partakers of the divine nature.
My parents have continually shown me unconditional love and incredible
generosity, and I could not recompense them for their sacrifices of time and re­
sources. Their actions have exemplified for me the overwhelming and nonrecip­
rocal love of God through Christ.
My mentor Dan Wallace has exemplified an exceptional desire to nurture the
next generation of Christian thinkers and researchers. I thank him for his time
and energies, for his endearing quirlanes s that makes our time together lively, and
especially for his concern for my wife.
My appreciation goes to Michael Svigel for his valuable insight into the first
several Christian centuries and for the extra effort he invested in this chapter,
strengthening its argumentation and improving its tone.

127 / Chapter 4
 Patristic Theology and Recension in Matthew 24.36

criticism,2 Westcott and Hort, admitted intentionality, though they pre­

sumably asserted that no alterations were made on account of doctrinal
biases.3 More recently, others have disagreed, finding ample changes
made from theological motivation.
Some purport, for example, that Matthew 24,36 was changed for
theological reasons. In this verse tucked away in the middle of the Olivet
Discourse, Jesus sobers his listeners with a discussion about signs present
at the end of the age but pleads ignorance about the specific time of the
return. Such ignorance is peculiar to readers who come to the text with
presuppositions of his divinity. Why would Jesus, himself a member of
the Godhead, not know the time of his own return? Some feel that this
theological conundrum is precisely the reason the phrase ουδέ δ υιό?
(“nor die Son”) was stricken from a number of manuscripts and that
such an alteration indicates die undoing of a sacred text’s veracity.
Among those whose writings have called into question the textual re­
liability of the NT is Bart Ehrman, whose basic stance is that there are ob­
vious examples of tampering by orthodox scribes and no basis for finding
the original buried beneath an excess of a quarter million other variants.4
Although the predominant interaction with Ehrman throughout this
work assumes his credibility as a scholar—a respect earned through his

2. C£ Jacobus H. Petzer, "The History of the New Testament—Its Reconstruction,

Significance, and Use in New Testament Textual Criticism," in New Testament Tex­
tual Criticism, Exegesis, and Early Church History: A Discussion of Methods, ed.
Barbara Aland and Joel Delobel (Kampen: Pharos, 1994), 12; J. Harold Greenlee,
Introduction to New Testament Textual Criticism, 2nd ed. (Peabody, MA: Hen­
drickson, 1995), 72.
3. F. J. A. Hort, Introduction to the New Testament in the Original Greek (New York:
Harper and Brothers, 1882), 282. But see now Jeff Miller, “(Mis)understanding
Westcott and Hort," RestQ 41, no. 3 (1999): 155-62.
4. Although Ehrman would overtly argue that the great variety of alternative read­
ings in the manuscript record and the wild copying of the early centuries make it
impossible to ever know if our reconstructed copies resemble the original text, this
overt argument would not do justice to his actual claims. While traditional textual
criticism recognizes these factors and acknowledges that absolute certainty over
every word of our reconstructed text is not possible, it still sees no insurmountable
obstacle to a faithful reconstruction, for throughout long centuries, we see reliable
copying in general and can identify and correct most unintentional orthographic
alterations. However, Ehrman argues that since the orthodox emended the text
from theological bias, a faithful reconstruction is implausible. It is no stretch to
enumerate the resulting impressions. First, if the orthodox have made innumer­
able intentional alterations, they were not as concerned with preserving the orig­
inal wording as previously regarded. Second, this makes the long centuries of the
orthodox transmission increasingly enigmatic, because theologically motivated
changes could be introduced without any mechanism for recovery. Third, this
encourages a conspiratorial view of the church’s use of Scripture, which not only
has the undesirable effect of shifting the onus of reliability away from the church’s
faithful preservation of its text but consequently empties modern Christian truth

128 / Chapter 4 Adam G. Messer

 Art Evaluation of Ehrman's Text-Critical Methodology

engaging writing style and valuable contribution to the disciplines of tex­

tual criticism and religious studies-—his recent works have shown a ten­
dency, in my estimation, to overemphasize the tentative nature of certain
conclusions from his seminal work The Orthodox Corruption ofScripture.
For certain conclusions to bear the weight of several of Ehrman’s recently
published views, the supporting historical evidence needs more rigorous
enumeration and demonstration for any single example used to support
Ms primary contention. Here, I wish to highlight with one example what
I feel is an incongruity between tire amount of attention given to the his­
torical evidence and the implications Ehrman draws from the conclu­
sions in Orthodox Corruption. Because no method can remain forever
sufficient nor engender absolute confidence, caution should be exercised
toward sensationalizing results that may unnecessarily be construed to
- run contrary to the NT’s textual reliability—particularly when conclu­
sions draw support from a highly publicized example for which less scru­
tiny has been applied than what is required.
Most concede that there have been theologically motivated changes
in the manuscript tradition, given the thousands of diverse manuscripts
in diverse periods. From among all of these, Ehrman chooses Matthew
24.36 as his case in point. He elaborately introduces it in Orthodox
Corruption of Scripture.5 It riddles his popular work Misquoting Jesus,
where he says, “The reason [for the omission] is not hard to postulate;
if Jesus does not know the future, the Christian claim that he is a di­
vine being is more than a little compromised.”6 It makes the cut in his
Introduction to the New Testament7 as the example of theologically
driven changes. It is called “the most famous example of doctrinal al­
teration” in his Studies in the Textual Criticism of the New Testament,8
and it also appears in the fourth edition of the standard Text ofthe New
Testament,9 which he coauthored with the late Bruce Metzger.

5. BartD. Ehrman, The Orthodox Corruption ofScripture: The Effect ofEarly Christo-
logical Controversies on the Text ofthe New Testament (Oxford: Oxford University
Press, 1993), 91-92,117 nn. 220-21.
6. ”Bart-D;-Ehr-man, Misquoting Jesus: The Story behind Who Changed the Bible and
Why (San Francisco: HarperSanFrancisco, 2005), 204. He mentions Matt 24.36 in
these places; 95,110,204,209,223 n. 19,224n. 16 (Daniel B. Wallace, “The Gospel
according to Bart: A Review Article of Misquoting Jesus by Bart Ehrman," JETS 49
[2006]: 327-49). This list contains two relevant references (95 and 204), two irrel­
evant references (110 and 209), and two footnotes with faulty mappings (223 n. 19
and 224 n. 16).
7. Bart D. Ehrman, The New Testament: A Historical Introduction to the Early Chris­
tian Writings, 4th ed. (New York: Oxford University Press, 2008), 494.
8. Bart D. Ehrman, Studies in the Textual Criticism ofthe New Testament, New Testa­
ment Tools and Studies 33 (Leiden: Brill, 2006), 333.
9. Bruce Manning Metzger and Bart D. Ehrman, The Text of the New Testament: Its
Transmission, Corruption, and Restoration, 4th ed. (New York: Oxford University
Press, 2005), 267.

Adam G. Messer Chapter 4 / 129

 Patristic Theology and Recension in Matthew 24.36

Given that Matthew 24.36 is Ehrman’s example par excellence, it

may be worth seeing if the lens of church history will provide a clearer
view. The patristic evidence may show how the church fathers viewed
this troublesome concept of the Son’s ignorance and could explain both
the time frame and extent of the alterations to the text. Such informa­
tion may help in explaining the specific text-critical problem in Matthew
24.36 and in evaluating Ehrman’s conclusions on the verse. From a
broader perspective, if Ehrman’s analysis of the leading example of or­
thodox corruption is flawed, his whole edifice may require significant
alteration.

PERSPECTIVE ON MATTHEW 24.36

Outline ofthe Problem
Modern Bibles are largely based on the NA27 reading of Matthew
24.36: "Concerning that day and hour no one knows, not the angels in
heaven nor the Son, except the Father alone.” But this was not always
the rendering; past critical editions often excluded the phrase ουδέ δ
υιός (“nor the Son”), because prior to the Alexandrian discoveries of
Sinaiticus and early papyri, a textual stream that lacked ούδέ ό υΙός
blossomed into the Byzantine tradition. This text type shaped the
church for over a millennium, undergirding the King James Version,
which, in turn, dominated the English-speaking world for centuries.
The omission is strongly attested in the majority of manuscripts, mostly
Byzantine, but also with a significant Alexandrian linlc. The presence of
ούδέ δ υιός is attested in important Western manuscripts and our best
Alexandrian exemplars. Further, several versions attest each reading,
and the patristic evidence is split. This textual problem has been diffi­
cult to decide with any ecumenical finality, and critical editors, all aware
of the variant readings, have made different textual decisions.10 Recent
editions have tended to favor the inclusion of ούδέ δ υιός, yet if they
are correct, how do we explain the manuscript and patristic evidence
attesting to the omission?
Bart Ehrman, among others, has suggested that scribes had prob­
lems with Christ’s ignorance and so struck ούδέ δ υΙός from their copies
of Matthew, in response to adoptionistic influences. Although scholars
prior to Ehrman advanced the notion that theological changes affected
the textual tradition, Orthodox Corruption is arguably the most in­
fluential book on the topic. However, whereas Orthodox Corruption
sought primarily to argue the case that it happened, Ehrman’s more re­
cent works have reflected a more radical perspective. He has argued in

10. Scrivener, von Soden, and Merk attest to the omission, while Lachmann, Tischerv·
dorf8, Legg, NA27, and UBS4 contain the phrase.

130 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

these that since orthodox emendation happened, our text, by implica­

tion, cannot be established; he contends that since the originals are for­
ever lost, the hopes for accurate reconstruction lie buried far beneath a
mountain of pious alteration and unchecked fideism.

Methodology

The majority of scribes were not the intellectual giants of their day,
so they would likely derive their understandings of a particular passage
(and the perceived difficulty thereof) from the leading theologians of
the period.1112 If the major teachers of the church had problems with a
text, it is probable that scribes in their region likewise had issues with it.
.... Such a situation would offer the necessary motivation for omitting ουδέ
ό ut6s during the transcription process.
My method for this analysis included perusing every reference
within BP for Matthew 24.36 and its parallel in Mark 13.32 to enu­
merate the fathers’ testimonies?2 The Greek texts come from several

11, The merit of this assertion is also corroborated by Brogan, who takes it a step fur­
ther by arguing that even if their leaders had a theological problem with a passage,
the scribes probably did not introduce the corruptions; rather, their role would
have been limited to the reproduction (perhaps unknowing) of the alterations
made by prominent leaders (John J. Brogan, "Another Look at Codex Sinaiticus,”
in The Bible as Book: The Transmission ofthe Greek Text, ed. Scot McKendrick and
Orlaith O’Sullivan [London: British Library and Oak Knoll Press, 2003], 25). The
Alands implicitly acknowledge an analogous idea (i.e., that alterations were not the
provenance of the scribe) when they speak of the origins of the Western text type:
"Wherever we look in the West, nowhere can we find a theological mind capable
of developing and editing an independent ‘Western text’” (Kurt Aland and Barbara
Aland, The Text of the New Testament: An Introduction to the Critical Editions and
to the Theory and Practice ofModern Textual Criticism, trans. Erroll E Rhodes, 2nd
ed. [Grand Rapids: Eerdmans, 1989], 54).
12. This includes a broad corpus of ancient literature but omits the testimony of certain
fathers (e.g., Jerome, Ambrose, Augustine, etc.) who also referenced Matt 24.36 in
their works. Jerome, in particular, has often been the subject of modern inquiries
.surxoundmg this verse: cf. Emile Bonnard, Saint Jerome» Commentaire sur S. Mat-
thieu, vol. 1, Sources Chretiennes 242 [Paris: Les fiditions du Cerf, 1977]; Dennis
Brown, Vir Trilinguis: A Study in the Biblical Exegesis of Saint Jerome [Kampen:
Peeters, 1992]; J. K. Kitchen, “Variants, Arians, and the Trace of Mark: Jerome and
Ambrose on ‘Neque Filius’ in Matthew 24:36," in The Multiple Meaning of Scrip­
ture: The Role ofExegesis in Early-Christian and Medieval Culture, ed. Ineke van’t
Spijker [Leiden: Brill, 2009]. For Ambrose, cf. A. Bludau, Die Schriftfalschungen der
Hdretiker: ein Beitragzur Texikritik der Bibel [Munster, 1925]; Kitchen, “Variants,
Arians and the Trace of Mark"; Craig Alan Satterlee, Ambrose ofMilan's Method
ofMystagogical Preaching [Collegeville: Liturgical Press, 2002]). While not mini­
mizing the fruit of these and other analyses, distinct evidence from the anterior
(and contemporaneous) period has sometimes been overlooked when drawing
conclusions. The present work has striven to supplement the discussion by fo­
cusing on this other early testimony.

Adam G. Messer Chapter 4 / 131

 1
Patristic Theology and Recension in Matthew 24.36

critical texts (where available), MPG, and TLG. All renderings are per­
sonal translations unless otherwise noted. From this collection of refer­
ences, the most poignant examples of the patristic ways of thinking are
shown for purposes of illustration.
Effort was made to ascertain a particular father’s form of Matthew
based on the four Greek differences with the text of Mark. First, Matthew
uses a TSKS construction for “day” and “hour” (τής ήμέρας και ώρας)
that Mark does not (τής ήμέρας ή τής ώρας). Second, Matthew uses
the conjunction καί where Mark uses ή. Third, the grammatical case of
“heaven” is genitive plural in Matthew (των ουρανών) and dative sin­
gular in Mark (εν ούρανω). Finally, Matthew has the word μόνος at the
end, while Mark lacks it. The fathers seemed to have little difficulty with
small morphological variation surrounding articles and conjunctions in
the first half of the verse and even in the form of “heaven.” So I classify
the verses with the following hierarchical method:

1. If the father identifies his source, we take him at his word.

2. Since the Markan transmission of this verse appears remarkably

unadulterated, texts without ουδέ ό υΙός are generally assumed to be
Matthean.

3. Preference is shown to identifying the source of a reference

containing μόνος as Matthean,13 because the presence or absence of
μόνος has significantly greater rhetorical impact than alterations to the
case and number of “heaven,” “day,” or “hour." Consequently, the fathers
would have been more likely to preserve μόνος in their quotations and
memory, and this fact should weigh as a heavier indicator of the source
than less rhetorically significant differences in form.14

In some cases, a father wrote in Latin, or our only access is through

extant Latin translations. In such cases, it is noteworthy that Latin

13. This assumption seems plausible because no extant manuscripts of Mark from the
first four centuries contain μόνο?. In fact, of the few listed in Tischendorf’s ap­
paratus, none are earlier than the ninth century, and some are much later. This
amount of time (at least five centuries) would create a situation more fertile for
a harmonization from Matthew to Mark, due to a significantly longer time frame
and increased frequency of manuscript duplication.
14. The form of “heaven” will be weighed less than the presence of μόνος but greater
than the other two differences in form. However, enough variation occurs that situ­
ations might be construed that would upset a rigid application of this hierarchical
method. The classification is somewhat of an art and may be interpreted differently
by others more or less qualified. In deference to such endeavors, effort was taken
to footnote the original Greek and Latin texts from which my classifications were
made.

132 ! Chapter 4 Adam G. Messer

 An. Evaluation of Ehrman’s Text-Critical Methodology

versions essentially preserve the differences in the Greek,15 so tentative

judgments may be offered.
Caution must be exercised in how the data of patristic evidence is
applied to the textual reconstruction of Matthew 24.36. The assumption
that a father’s intention was to perfectly quote a verse is demonstrably
false, and each father is subject to a fallible memory, even as their writ­
ings are subject to the same challenges of transmission as the NT.
Despite these caveats, some conclusions are possible. Further, whereas
the challenge to reconstruct the perfect form of a given verse from pa­
tristic evidence may prove insurmountable, the determination of what
ingredients they saw in that verse pays remarkable dividends. Whatever
the exact original words were, the fathers can help us see what concepts
existed in their text, which offers an important angle for textual recon­
struction by highlighting which variants included those concepts and
which did not.16
In addition to researching explicit references to Matthew 24.36 and
Mark 13.32, 1 noted the views of the early church on the Son’s igno­
rance, hopefully illuminating the transmission of this theological doc­
trine when there is no explicit reference made to the verses in question
and in cases where the references do not speak to the issue.

Textual Problem
Although a comprehensive portrait of textual problems in Matthew
24.36 is of secondary importance to the theme of this chapter, manuscript
witnesses, versional testimony, and internal criteria must be added to pa­
tristic evidence in determining the final text. The omission is strongly
attested in the majority of manuscripts,17 predominantly Byzantine,
but not merely so; there is also a smattering of manuscripts from the
Alexandrian and debated "Caesarean" text types, none prior to an early
corrector to Codex Sinaiticus.18 Several versions—namely, the Old Latin,

15. „,.IheLatinequivalents .are (1) ilia et vs. illo vel for καί vs. ή τήϊ, (2) caelorum vs. in
caelo for των ουρανών vs. ev ούρανω, and (3) Latin solus for Greek μόνος.
16. Establishing the exact form of die fathers’ manuscripts can help us in our recon­
struction of the original, because It can help compare the purity of one transmis­
sion stream to another in the conceptual realm (cf. Bart D. Ehrman, “The Use and
Significance of Patristic Evidence for NT Textual Criticism," in Aland and Delobel,
New Testament Textual Criticism, 119-20,134-35).
17. R1 L W y133 Tl et al.
18. The early corrector appears to be contemporaneous with its original production
(H. J. M. Milne, T. C. Skeat, and Douglas Cockerell, Scribes and Correctors of the
Codex Sinaiticus [London: British Museum, 1938], 40). Jongkind affirms Milne and
Skeat’s conclusions regarding the time of these corrections (Dirk Jongkind, Scribal
Habits of Codex Sinaiticus, Text and Studies, 3rd ser., 5 [Piscataway, NJ: Gorgias,
2007], 39).

Adam G. Messer Chapter 4 / 133

 Patristic Theology and Recension in Matthew 24.36

Syriac,19 Coptic,20 and Georgian—attest to the omission. The exclusion

has the patristic support of Didymus, Phoebadius, Ambrose, and Jerome.
The presence of the phrase, however, has a strong pedigree, including
Alexandrian and Western witnesses.21 The Diatessaron, Armenian,
Ethiopic, and Georgian22 versions contain it, as well as several fathers,
including Irenaeus, Origen, and Chrysostom.2324
A few internal considerations enliven the discussion. When the
transcriptional possibilities are evaluated for instances of accidental
orthographic omission in Greek
themselves.25 The most interesting involves a novice scribe overlooking
a faded overbar above the nomen sacrum for "son” (YC) and miscon­
struing the phrase to read “nor the boar.” The resulting verse would
read, “But of that day and hour no one knows, not the angels of heaven,
nor the boar, but the Father alone.” In such a scenario, a pious neophyte
might remove the sacrilegious phrase, under the notion that some tired
or careless scribe had accidentally added it or that a mischievous copyist
had inserted it for sport. Several other scenarios may be invoked by an
appeal to novice, secular, or illiterate scribes, but these are improbable.
Since the chances for accidental errors of sight to occur are small, it
is more likely that if an accidental error occurred, it was an error of
memory.
One such explanation for inclusion involves harmonization.26 The
absence of ουδέ δ υιός· may be original to the Gospel of Matthew. This
possibility is bolstered by a recognizable tendency in Matthew, using
Mark as a literary source, to smooth out the Markan reading in favor
of a higher Christology.27 Matthew may have omitted the phrase ούδέ

19. The Syriac version has its own variety of readings. In support of the omission are
the Sinaitic Old Syriac version (thixd/fourth century), the Peshitta (early fifth cen­
tury), and the Harldean version (616 CE). The presence of “nor the Son” is attested
in the Palestinian Syriac version (sixth century).
20. In support of the omission are the Sahidic, Bohairic, and Middle Egyptian forms of
the Coptic text.
21. K**B D0fiSetaL
22. The Georgian testimony is split. Some manuscripts support die omission, and
others contain ουδέ ό υιό?.
23. A complete list is not enumerated because a more in-depth perusal will be subse­
quently laid out.
24. Options seem to be present in Aramaic and Coptic texts, but given the lack of
scholarly consensus regarding an Aramaic Matthean original and the awlcward
construal of a social scenario allowing a Coptic variant to penetrate the Greek
tradition, these unlikely possibilities are relegated to little more than thoughtful
contemplation.
25. See appendix B.
26. Harmonizations can also be intentional.
27. Plummer mentions examples of Matthew’s higher Christology across parallels:
Mark 3.5 and Matthew 12.13, Mark 6.5-6 and Matthew 13.58, Mark 8.12 and
Matthew 16.4, Mark 10.14 and Matthew 19.14, and statements removed from

134 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

δ υ10£ in his gospel to portray his unique perspective of Jesus.*28 The

parallel passage in Mark 13.32 contains the phrase ουδέ δ υιός·, and this
reading remained remarkably fixed across the centuries.29 If the phrase
is not original to Matthew 24.36, could the early appearance of ουδέ ό
υιό? reflect a harmonization to Mark? How likely was harmonization in
this direction? One attempt to answer this question analyzed over sixty
of “the most significant instances involving harmonization” between
Matthew and Mark and calculated that Mark harmonized to Matthew
around 70 percent of the time in these instances.30 This would seem to
indicate that it was less likely for a scribe to harmonize Matthew 24.36
to Mark 13.32.
If ουδέ δ υιό? is original to Matthew 24.36, could theological moti­
vation be responsible for the disappearance of the phrase from various
manuscripts? Might a scribe with a doctrinal preference for an omni­
scient Christ strike the phrase? In support of this, Metzger and Ehrman
see the originality of the longer reading as a grammatical necessity,31 but
its absence would not create the grammatical inconcinnity they pre­
sume.32 More germane to their argument is that the longer reading is
more difficult. Further, given scribal tendencies to harmonize, had this
phrase not existed in Matthew’s original, Ehrman asserts that omitting
the parallel phrase from Mark (due to Matthew’s influence) would have
been more likely than striking ουδέ δ υιό? from Matthew.33 Metzger

Matthew’s parallels pertaining to Jesus’ perceived ignorance in Mark 5.9, 5.30,

6.38, 8.12, 8.23, 9.16, 9.21, 9.33,14.14, etc. (Alfred Plummer, An Exegetical Com­
mentary on the Gospel according to S. Matthew [Minneapolis: James Family Chris­
tian, 1909], xiv-xvi; cf. also Robert H. Stein, Studying the Synoptic Gospels: Origin
and Interpretation, 2nd ed. [Grand Rapids: Baker Academic, 2001], 91-94). This
prospect assumes the view (Markan priority) currently held by the majority of NT
scholars.
28. He may also have simply recorded what Jesus said on another didactic occasion
pertaining to the same general topic.
29. Only one tenth-century codex (Codex X) and apparently a random Latin manu­
script of the Vulgate attest to the omission. The overwhelming majority of Vulgate
manuscripts (as well as the critical determination) contains the phrase in Mark
——..-13.-32.
30. Powell indicated that the population of harmonizations between Matthew and
Mark did not represent all possible instances but were “probable harmonizations”
that involve “the major manuscripts” (Charles Powell, “The Textual Problem of
Ουδέ Ό TClos in Matthew 24:36,” http://www.bible.org/page.phpTpage_id-2478
[accessed January 25, .2009]).
31. Bruce Manning Metzger, A Textual Commentary on the Greek New Testament, 2nd
ed. (Stuttgart: United Bible Societies, 1994), 52; Ehrman, Orthodox Corruption, 92.
32. Powell, “Textual Problem of Ούδέ Ό Υιός." Central to his argument is the fact
that Matthew has other instances that include just one ouSe (e.g., Matt 6.28,21.32,
25.45, and 27.14), signifying that this word does not require a correlative pair.
33. This follows from two considerations. First, Matthew was more popular and thus
copied more frequently than Mark. Second, Mark apparently harmonized to Mat­
thew significantly more often than die reverse (Powdl, “Textual Problem of Ούδέ

Adam G. Messer Chapter 4 / 135

 Patristic Theology and Recension in Matthew 24.36

simply believed that the chances for a theologically motivated removal

were greater than a harmonization to Mark 13.32.34 Running counter to
this idea are three considerations:

1. Why did the scribe(s) not also remove the phrase from Mark
13.32? Mark remains essentially untouched.

2. Scribal tendency was to conflate rather than remove, and it is

not difficult to imagine that a phrase indicating a Messianic ignorance
would be so vividly associated with this context (“nobody knows the day
or hour”) that the copyist would know this verse by memory. With such
a pregnant phrase, it is easy to suggest a harmonization from Mark.

3. During a textual decision, the reading that differs from its par­
allel is ordinarily preferred, because an explanation is sought for the rise
of the other variants.3536This would make the omission a good candidate
for originality, except that, as mentioned earlier, the presence of ούδέ δ
υιό$ is the harder reading in light of the doctrinal difficulties it may have
presented. These key factors vie for precedence.

From another angle, that Mark seems to have an established

transmission history (no manuscript evidence attests to the omission
until the tenth century) seems difficult to explain apart from a text of
Matthew that matched it for the predominant period in question; oth­
erwise, we might have expected a number of manuscripts attesting to
the omission in Mark. Even so, there are scholars who see in Matthew's
higher Christology the tendency to literarily smooth out potentially
denigrating perceptions of Christ found in Mark’s Gospel and who thus
prefer the originality of the shorter reading.“ Suffice it to say at this
point that most of the text-critical theories that favor weighted external

Ό Ύιό?’; Ehrman, Orthodox Corruption, 92), a tendency, if true, that Ehrman

believes is due to Matthew's comparatively greater notoriety (Ehrman, Orthodox
Corruption, 92).
34. Metzger, Textual Commentary on the Greek New Testament, 52.
35. Metzger and Ehrman, Text of the New Testament, 314.
36. Daniel B. Wallace argued for its original exclusion in his fall 2007 class on textual
criticism (for reasons outlined in the text), as do the notes in the NT translation for
which he was an. editor (NET Bible). John Wu's reinvigorated study of original read*
ings within the Byzantine tradition reflects a preference for the shorter reading» on
the basis that the Byzantine copyists went against their tendency to harmonize
and conflate in Matt 24.36, to explicitly follow its archetype. He posited that the
. interpolation probably happened in the Western and Alexandrian texts early in the
second century (Wei-Ho John Wu, “A Systematic Analysis of the Shorter Read*
ings in the Byzantine Text of the Synoptic Gospels” [PhD diss., Dallas Theological
Seminary, 2002], 60-64).

136 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman's Text-Critical Methodology

evidence3738will prefer the longer reading,88 while the variety of internal

considerations will result in conflicting conclusions.
After surveying the possibilities, two options rise to the fore as
most probable. Either ουδέ ό υιός originally existed in Matthew and
was stricken by a scribe due to theological difficulties presented by the
verse, or the phrase did not exist in Matthew, in which case its presence
in a large number of manuscripts resulted from a harmonization from
Mark 13.32. If we assume that ουδέ ό υιός was part of the original text,
two problems must be explained:

1. What accounts for the theologically motivated omission?

2. How do we explain the predominance of the shorter reading in

the manuscript tradition?

If we assume that the shorter reading is authentic, we must disagree with

our earliest and best witnesses, and we must explain how ουδέ ό υιός
was added, in light of the fact that scribal harmonization of Matthew to
Mark occurred less frequently.
Whereas some combination of external manuscript considerations
and internal considerations tend to rule the day as far as establishing the
original text, the patristic evidence finds its primary contribution in die
realm of establishing the geographic and chronological transmission of
the text and doctrines. Sometimes this plays an important part in the
text-critical process. Whichever variant is original, it appears that both
readings existed by the time the fourth-century Codex Sinaiticus was
composed. Milne and Skeat suggest that before it ever left the scripto­
rium, a corrector had stricken ούδέ ό υιός.39 It was then “readded” by
erasing the diacritical marks indicating its omission. This makes plain
that some deliberation existed for the scribe(s) of Sinaiticus—presum­
ably the result of multiple and contradictory exemplars at Matthew
24.36. In addition, the vitally important fourth-century Codex Vaticanus
bears earmarks of disclosing this variant.40 Because both variants are

37. Powell, “Textual Problem of Ουδέ Ό TfiAs”; Metzger, A Textual Commentary on

the Greek New Testament, 51-52; Ehrman, Orthodox Corruption, 92,117 n* 221.
38. They will do so on the basis of the dominant presence of ουδέ ό υ1ό$ in the Alex­
andrian and Western traditions, generally thought to better reflect the original,
especially when they agree.
39. C£n.l8.
40. Caroli Vercellone Sodalis Barnabitae and losephi Cozza Monachi Basfliani, eds.,
Bibliorum Sacrorum Graecus Codex Vaticanus Auspice Pio DC Pontifice Maximo
(Detroit: Brown and Thomas, 1868; repr., 1982), 35. Next to Matt 24.36 is a dis-
tigma. Miller concludes that distigmai (or what he calls “umlauts1') represent vari­
ants 0. Edward Miller, “Some Observations on the Text-Critical Function of the
Umlauts in Vaticanus, with Special Attention to 1 Cor 1434-35* JSNT 26, no.
2 [2003]: 235). At the SBL conference in November 2009, Peter Head of Tyndale

Adam G. Messer Chapter 4 / 137

 Patristic Theology and Recension in Matthew 24.36

attested in the manuscript tradition by the fourth century, patristic evi­

dence that might play a role in tracking down the origin of this textual
problem is delineated here by this general time frame.

THEOLOGICAL MOTIVATIONS IN MATTHEW 24.36

Background
Some comment is in order regarding who is responsible for the al­
teration in Matthew 24.36. Heretics made copies of the NT, often in
light of concerns they had with the text41 A combination of factors—in­
cluding the sometimes wild copying in early centuries, the multiplicity
of languages into which the church was translating, the lack of central­
ized authority and proliferation, the resulting absence of accountability,
the cost to create a manuscript and their consequent high value, the
competing sub-Christian groups who self-identified as Christians, and
the difficulty in verifying and/or regulating the content of new texts—
all contribute to the very real possibility that manuscripts copied by
heretics could end up being used for communal worship and as exem­
plars by later scribes.42 Ehrman argues that competing Christian groups
struggled for dominance in the first few centuries, each trying to pass
their own traditions down and reflecting a valid form of Christianity.43
While the latter claim deserves additional clarification, it is true that
various groups of various sizes sought to proliferate their truth claims

House, Cambridge, argued that the distigmai were added about a thousand years
after codex B was written. If so, then the fact of such a symbol here may only indi­
cate that the phrase was not in the copies of the Vulgate current at the time these
distigmai were added. However, the fact that no distigma is found at Mark 16.8 in
B is problematic for this hypothesis.
41. Examples include Theodotus, Asclepiodotus, Hermophilus, and Appollonitts (Eu­
sebius, Hist. ecd. 5.28); Marcion (Heikki Raisanen, “Marcion,” in A Companion
to Second-Century Christian "Heretics’,' ed. Antti Marjanen and Petri Luomanen
[Leiden: Brill, 2005], 113-16); Ebionites and their copy of Matthew (Eusebius, Hist,
ecd. 6.17.1); and Valentinians (Irenaeus, Against Heresies 3.11.9).
42. This assertion is ambivalent toward the idea that the majority of copies "were
made by and for individuals on at the very least, for individual churches during the
second and third centuries” (Kim Haines-Eitzen, Guardians of Letters: Literacy,
Power, and the Transmitters of Early Christian Literature [Oxford: Oxford Uni­
versity Press, 2000), 6, 16). If true, this speaks only to the majority practice and
not against the complicated nature of social history that would occasionally have
“interbred’ orthodox and heretical texts. In fact, if the lines between what was
considered “Christian” in antiquity and what was not were truly as variegated as
Ehrman has suggested, particularly in the thesis of Lost Christianities, we might
presume this to have happened much more frequently than we would otherwise
think.
43. Bart D. Ehrman, Lost Christianities: The Battles for Scripture and the Faiths We
Never Knew (New York: Oxford University Press, 2003), ix, 4, 6-7.

138 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

against the ancestors of Nicene theology, whom Ehrman calls the

"proto-orthodox.” As they attempted to pass their traditions and writ­
ings forward, the complicated process known as social life transpired,
and manuscripts shifted hands, were recopied, and intermingled. Such
a view is implicit in C. S. C. Williams’s view of Marcion’s influence on
the textual transmission of Luke and various Pauline epistles: “Strange
as it may seem, the readings adopted by the heresiarch [Marcion] were
more than once reproduced by the copyists of the Church’s ted.”44 He
spoke also to the accuracy of Marcion’s copying: "Marcion did not find
the original text either of tire Epistles or of the Gospel of Luke quite
adequate to support his teaching and he emended the texts of both.’’45
That heretics altered the text has been acknowledged for some time;
in fact, the accepted, if informal, scholarly position was that they were
responsible for the majority of theologically motivated changes—until,
perhaps, the 1993 publication of Orthodox Corruption and its over­
whelming wake. The reason heretical alterations are important to
mention is not that Ehrman or any other scholar would deny them but
because it complicates our inquiry. There were periods in which there
might have been orthodox motivation to change the text, but there were
also times and places of heretical influence.
What reason for alteration -will best account for the patristic evi­
dence and manuscript history?46 Attempting to surmise the time period
during which the alteration occurred, who or what is responsible for it,
and how it happened should give some insight into the reason for the
alteration. Without simply assuming that the group(s) who would profit
from a change to the text would necessarily be those responsible for
them, I hope to offer reasoned conclusions, listing a number of options
for the omission or inclusion in Matthew 24.36.

Who's On Trial?

Orthodox Corruption
—What motivations might exist for the phrase ουδέ ό υιός to be
omitted by the proto-orthodox? To each heresy discussed in this sec­
tion may also be apportioned the motivation to add the phrase ούδέ ό
I
υΙός to support their theology had the phrase originally been lacking in
Matthew.47 Of the plethora of early heresies, a theologically motivated
alteration would occur against heresies that (1) use the phrase ουδέ ά

44. C. S. C. Williams, Alterations to ths Text of the Synoptic Gospels and Acts (Oxford:
Basil Blackwell, 1951), 10-11.
45. Ibid.
46. I use the word alteration to mean either the removal of οΰδέ ό ulos or its addition,
depending on. Matthew’s original state.
47. Such an addition would have been strategically imported from Mark 1332.

Adam G. Messer Chapter 4 / 139

 Patristic Theology and Recension in Matthew 24.36

υιό? or the idea behind it to support their theology, (2) are significant
enough to be taken seriously,48 and (3) overlap the time period in which
the omission appears to have been made.
Since the omission was well attested by Jerome—it took over the
basic Bible of the day by Jerome’s time—and given that it appears in
some versions with roots as far back as the third century CE (Syriac)
and perhaps even the second (Sahidic Coptic),49 this early period will
be the target of investigation. Although many aberrancies litter these
centuries, two stand as good candidates: Adoptionism and Arianism.
The Adoptionists claimed that Jesus was only a man who was adopted
by God at his baptism. The Arians, also called Eunomians, Anomceans,
Anomeans, Heterousians, and Aetians at various points by patristic
writers, believed Jesus was a created being who was not fully God.
Against both perspectives, a phrase demonstrating a self-attested lack
in Jesus’ knowledge could present difficulties for the orthodox view that
he was truly divine. Thus their adherents may have been motivated to
remove ουδέ ό υιός.
Adoptionism did not hit Rome until around 190 CE, and
Epiphanius claims that its originator was Theodotus the Tanner.50 It

48. It is rightly observed that even local heresies were considered of utmost concern
to those leaders and churches they affected. Eplphanius’s Panarion lists scores of
heresies that could be examined in their particulars, yet I here narrow the list by
assuming two things. First, heresies that were widespread and durative would have
greater probabilities of causing manuscript alterations that would survive until the
present times. Second, attention to these many heresies would no doubt further
muddy the waters by greatly expanding the number of possible scenarios. It is, of
course, reflective of good scholarship to mention Adds ripe for fixture study how­
ever, I lean on the following consideration to help choose which rocks to overturn
first: the general consensus of the scholarly world in highlighting certain heresies
no doubt corresponds with the heresies most often attacked by extant patristic
writings. 1 assume that these represent the most fertile initial garden for study, and
I acknowledge that this assumption is only as strong as (since it relies on) the as­
sumption that our extant patristic manuscripts are generally representative of the
total body of ancient ecclesiastical literature.
49. Bruce M. Metzger, The Early Versions of the New Testament: Their Origin, Trans·
mission, and Limitations (Oxford: Clarendon, 1977), 127—32. Although Metzger
mentions Horner's and Thompson's advocacy of a late second-century Sahidic text,
he seems to inconspicuously align himself with Kasser's chronology of Coptic de­
velopment, which suggests that the latter half of the third century is the most likely
period from which to expect a full Sahidic version. Others have asserted that the
earliest Sahidic manuscripts would have appeared within the time frame delin­
eated by 250-350 CE and, perhaps more relevant that the entrance of a standard
version would be even later (350-450 CE) (Frederik Wlsse, “The Coptic Versions of
die New Testament,* in The Text ofthe New Testament in Contemporary Research'.
Essays on the Status Quaestionis, ed. Bart D. Ehrman and Michael W. Holmes
[Grand Rapids: Eerdmans, 1995], 134-35).
50. Epiphanius, Panarion 34. Theodotus the Tanner was active during the late second
century. More precise dates for those considered heretics are sometimes unattain­
able. Approximate dates will be included when available. Epiphanius’s discussion

140 / Chapter 4 Adam Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

soon exerted enough influence to elicit the reactions of concerned,

orthodox apologists.51 Whatever manuscript influence Adoptionists
may have had would have been to stress Jesus’ humanity to the ex*
elusion of his deity. One group of Adoptionists, the Ebionites,52 only
used Matthew,53 so perhaps an orthodox scribe omitted ουδέ ό υΙό?
to undermine their reliance on Matthew 24.36.54 Some evidence ex­
ists that the Ebionites adjusted their own copies of Matthew to suit
their beliefs.55 If this is true, the orthodox adjustment of their manu­
scripts to address a deviant view derived from self-maintained, he­

of Theodotus from Byzantium is available for die English reader in Frank Williams,
The Panarion ofEpiphanius: Bocks II and III (Sects 47-80, De Fide), Nag Hammadi
and Manichaean Studies 36 (Leiden: Brill, 1994), 2,72-77, 91.
51. Irenaeus against the Ebionites and Cerinthus et al. (Against Heresies 1 >26.1-2;
3.11.7; 3.21.1; 4.33.4; 5.1.3), Hippolytus against the Ebionites and Cerinthus et al.
(Refutation ofAU Heresies 7.21-22), Epiphanius against the Ebionites and other
Adoptionists et al. (Panarion 30.1.1; 30.2.1*8; 30.3.7; 30.16.1; 30.26.1; 30.34.6;
65.1.5-10; 65.3.2-4), Eusebius against Cerinthus, Ebionites, and other Adoption­
ists et al. (Hist. eccl. 3.27-28; 5.28; 7.30), etc. Cerinthus la usually considered an
early Gnostic influence, but his teaching about the origin of Jesus was, lilce other
Gnostics (e^., Basilides), essentially the same as the more formal Adoptionists. For
modern summaries of ancient heresies, see Walter Bauer, Orthodoxy and Heresy
in Earliest Christianity, ed. Robert A. Kraft and Gerhard Krodel, trans, the Phila­
delphia Seminar on Christian Origins (Philadelphia: Fortress, 1971); Harold O. J.
Brown, Heresies: Heresy and Orthodoxy in the History ofthe Church (Peabody, MA:
Hendrickson, 1984); Ehrman, Lost Christianities; Ariand J. Hultgren and Steven A.
Haggmark, eds., The Earliest Christian Heretics: Readings from Their Opponents
(Minneapolis: Fortress, 1996); Marjanen and Luo manen, Companion to Second-
Century Christian ‘Heretics?
52. Some scholars believe the Ebio nites existed as early as the composition of the NT,
stemming from an original Jewish following of Jesus (e.g., Robot H. Eisenman,
James, the Brother ofJesus: The Key to Unlocking the Secrets of Early Christianity
and the Dead Sea Scrolls [New York: Viking, 1997], 5-6). Others suggest they
originated early in the second century from Jews disillusioned by repeated failures
to reestablish the kingdom of God (e.g., James D. Tabor, The Jesus Dynasty: The
Hidden History ofJesus, His Royal Family, and the Birth of Christianity [New York:
Simon and Schuster, 2006], 302-3). Since the Ebionite stress on Jesus’ humanity
was similar to the Adoptionists, their influence may have provided die motivation
for· alterations even earlier than 190 CE, perhaps as early as the very beginning.
They also appear to have existed in various places until the fourth century (Tabor,
Jesus Dynasty, 302—3).
53. Irenaeus, Against Heresies 1.26.2.
54. This would make sense of why this alteration only happened in Matthew.
55. Ehrman, Lost Christianities, 102. The Ebionites use of Matthew and their vocal
protest against Jesus’ virgin birth—a problematic denial had they not altered die
first two chapters (or subsections) ofMatthew, which testify to it—make alterations
to Matthew a strong likelihood. Eusebius (Hist, eccl 6.17.1) also hints at this edito­
rialwork. Against this view, Hakkinen argues that the silence of another church fa­
ther (Irenaeus) about Ebionite alteration of Matthew indicates that they apparently
used the same version as the orthodox (Sakari Hakkinen, ‘'Ebiomtes,” in Marjanen
and Luomanen, Companion to Second-Century Christian ‘Heretics? 260-61).
However, he earlier makes the case that Irenaeus was not personally acquainted

AdamG. Messer Chapter 4 / 141

 Patristic Theology and Recension in Matthew 24.36

retical manuscripts might not be as good a fit as it would initially

seem. Other Adoptionists, such as Asclepiodotus and Iheodotus the
Money-Changer,56 were not as isolated as the Ebionites or limited to
sectarian copies of Matthew. Adoptionism, most prominent in the
years 190-275 CE if one includes the overlapping theology of Paul of
Samosata (ca. 200-275 CE),57 undoubtedly had earlier moorings in
Jewish Christianity.
Arianism is the most logical heresy in which to look for an orthodox
advantage, because it became so popular and generated such a strong
reaction. Since Arians saw Christ as a lower, created entity, one can
readily see how an omission of ούδε ό υιός would benefit their orthodox
opponents. Although the exact relationship between Arius (ca. 250-336
CE) and Arianism is in dispute,58 it is certain that Arianism was a full-
fledged controversy by 321 CE, necessitating the Great Council of Nicea
just four years later. Thus strong orthodox motivation for striking ούδε
ό υίός exists from perhaps 280 CE until Arianism’s head was severed at
Constantinople in 381 CE, and motivation may also have existed from
about 190 CE and possibly from the very roots of Christianity, through
Jerome’s inception of the Vulgate. Conversely, had Matthew originally
not contained the phrase, the Arian groups would have been motivated
to add it during this period.

Heretical Corruption

Could heretical movements such as Marcionism, Gnosticism,

Docetism,59 and Sabellianism60 have been responsible for the removal
of ουδέ ό υιός? What would have been their motivation?

with the Ebionites and used inherited information to compose his short paragraph
regarding them in Against Heresies 1.26.2 (ibicL, 250).
56. These disciples of Theodotus the Tanner were active during the rule of Bishop
Zephyrinus (198—217 CE) of Rome (Eusebius, Hist eccL 5.28.3).
57. Paul freely called Jesus 'God.” Whereas this might seem Sabellian, he redefined
godhood to signify mere man infused with the divine. Cf. Brown, Heresies, 96.
58. R. P. C. Hanson, The Search for the Christian Doctrine of God· The Arian Contro­
versy, 318-381 (Grand Rapids: Baker Academic, 2007), 123-28. Hanson notes the
curious affair that many of those dubbed “Arian” did not associate themselves with
Arius. Some even disavowed such a thought. He concludes that Arius “was usually
not thought of as a great man by his followers” (128). However, that his ideas were
somewhat prominent should be inferred from, e.g., Athanasius’s work entitled Dis­
courses against the Arians, the interspersed mention of his name in Hilary’s On the
Trinity (e.g., “Arian fanatics” from 7.7), and the movement called “Arian madness"
in Gregory of Nazianzus’s Orationes 34.8.
59. Docetism shared characteristics with Gnosticism and was modalistic, and it tech­
nically falls under the category of Monarchianism (God as “one" entity).
60. Sabellianism, similarly known by the terms Modalism and Patripassianism, is also
a form of Monarchianism.

142 / Chapter 4 Adam G. Messer

 An Evaluation of Ehnnan’s Text-Critical Methodology

Marcion (ca. 85-160 CE) may be ruled out as a direct candidate, not
because of his theology or manner of readjusting Scripture, but because
he seemed to only accept portions of Luke through Acts and various
Pauline epistles as Scripture. However, his impact on later churches is
worth examination. His Christology bore features of Modalism, in that
the pure God came in the form of (and was) Jesus.61 Phrases assigning
ignorance to Jesus would have been problematic for Marcionism, and
its adherents might have been motivated to change the text. This pos­
sibility would have been bolstered in Matthew 24.36 by an equally
problematic distinction between the Father and Sort Marcion’s church
prospered in the latter half of the second century and was still exerting
influence as late as the fifth century in Syria.62 Because of the complex
interbreeding of ideas in the early centuries, some Marcionite churches
may have begun using orthodox Gospels after the death of their fore­
bear and been motivated to change them. Overall, the relative improb­
ability of this scenario due to Marcionite’s selective canon, combined
with the infrequent reassimilation of these manuscripts into orthodox
circles, makes alterations by Marcionism unlikely.
Gnosticism either had no united system of Christology or, if there
was one, would typically conceive of Jesus, however near to an emana­
So if the Gnostics
tion of God, as nevertheless distinct from the Father.6364
are implicated by known tampering with Scripture,6* perhaps they are
exonerated by the difficulty in finding a certain advantage and conse­
quent motivation to remove statements of Jesus’ ignorance.
Reservations about a movement altering the text drain away
with such modalistic expressions as Docetism, Sabellianism, and
Patripassionism (hereafter referred to collectively as Modalism) be­
cause they would have been motivated to remove any apparent differ­
entiation between the Father and Son.
Docetism was a very early strand of Gnostic presuppositions
that drove the notion that Jesus only appeared to be a man (but was
not). Seeds of this thinking began with the apparently convoluted and

“61:-Raisanat “Marcibn," 105.

62. Raisanen, “Marcion,” 101.
63. Valentinus (ca. 115-65 CE) taught the utter unknowable transcendence of the
Father (Brown, Heresies, 100-101). Gnostic thinkers such as Saturnilus (or
Saturninos) of Syria, Basilides of Egypt (specifically Alexandria), and Valentinus of
Rome/Alexandria would stress Christs lack of humanity. Saturninus was Docetic
and stressed that Christ was not really a man. Basilides theorized in the early
second century that Christ was an emanation from the Father He was apparently
involved in die Christian church in Alexandria and had great influence on it by
teaching in the schools.
64. Irenaeus charged the Valentinians of corrupting Scripture not only by reinterpreta­
tion but by adding to it (Against Heresies 3.11.9). The addition was not explicitly
interpolation into the current Gospel narratives per se but the creation of new ones
considered to be Gospels (e.g., the Gospel of Truth).

Adam G. Messer Chapter 4 / 143

 Patristic Theology and Recension in Matthew 24.36

self-aggrandizing personality of Simon Magus. He claimed to be both

the Son and the Father,65 and he apparently invoked the challenge of
the NT writers. The thread of Docetism that centered on Jesus (rather
than Simon) was undergirded by a dichotomy between the good, spiri­
tual realm and the bad, physical realm. Christ was heavenly and good
and therefore could not have a physical body. Diminutive notions of the
“Son" (for Simonites) and of "Jesus” (for all Docetists) that appeared to
Docetists to be rooted in the material realm would have been problem­
atic for their Christology. Further, those strains that equated Christ with
the “chief of the Aeons” would have been further motivated to strike
ουδέ ό υιός.66 Docetism incited the polemics of the apostle Paul, the
apostle John,67 Ignatius, Polycarp, Tertullian, Hippolytus, and others.68 It
appears to have remained in existence by evolving into (or clinging to)
Manichaeism and its derivatives in certain sectors and even in a pure
form as late as the fourth and fifth century in Spain.69 However, we may
assume its primary period of influence to overlap the period of extant
orthodox contestation (ca. 70-230 CE). This lengthy period is somewhat
mitigated by the difficulty in determining what proportion of Docetists
would have been motivated to make an alteration to Matthew 24.36.
Docetism was subsequently overshadowed by Adoptionism (dis­
cussed earlier under orthodox motivations for alteration), and the
Christo-diminutive conceptions inherent in Adoptionism nourished
the impetus behind another heretical suspect, Sabellianism. Noetus of
Smyrna and his disciple Praxeas of Rome,70 responding to Adoptionism

65. Acts 8.9-24; Justin, Apology L26; Irenaeus, Against Heresies 1.23.1-4. Simon was
active during the first century, as is evident from his interaction with Philip, Peter,
and John. Hultgren and Haggmark venture a date of 70 CE for his death [Ear­
liest Christian Heretics, 15). Apparently, a statue was erected in his honor with the
Latin epigraph SIMONI DEO SANCT0—which means, “To Simon the Holy God”
(Justin, Apology 1.26). Ignatius of Antioch said that Simon, claiming to be God,
should consequently be able to overcome death. Though Simon died, his followers
existed at least to the time of Origen (Origen, Against Celsus 1.57).
66. As mentioned, forerunners of Docetism such as Simon Magus and Menander saw
themselves as “the Son." These groups would have been motivated to remove di­
minutive references to “the Son” (e,g., Matt 24.36) to whatever extent they used NT
writings.
67. Paul may refer to a form of Docetism in Col 1.19 and 2.9. The elder combats it in 1
John 1.1-3 and 4.1-4 and in 2 John 7.
68. Ignatius, Trallians 10; Ephesians 7,18; Smymaeans 1-6; Polycarp, Epistle to the
Phiiippians 7; Tertullian, On the Flesh of Christ 5; Hippolytus, Refutation of All
Heresies 8.L
69. John Arendzen, 'Oocetae,” in TCE, vol. 5 (New York: Encyclopedia, 1913), 70-72.
70. Warren H. Carroll, The Founding of Christendom, vol. 1 of A History of Chris­
tendom Front Royal, VA: Christendom College Press, 1985), 467. Praxeas was
active in Rome at the very end of the second century onward, and Nofitus came
sometime afterward (Brown, Heresies, 100-101). Some chronologies list Noetus as
Praxeas's successor because Noetus came to Rome after Praxeas.

144 I Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

as well as a strand of Gnosticism that saw Christ as a lesser Aeon,

stressed Jesus’ divinity to the exclusion of his personal uniqueness
from God the Father. Although Praxeas repented and realigned with
the orthodox, Noetus’s disciple Epigonus began a school attended by
Sabellius, who ushered into full bloom the idea that the Son was the
Father.71 Such conceptions peaked in the early third century and were
subjugated by Pope Caliistus I (<L ca. 223 CE) sometime around 220
CE. These ideas influenced later Trinitarian formulations, particularly
the recognition of the divinity of the Holy Spirit.72 What we discover
from this short rumination of early Christological diversity is a lengthy
period for Docetist ideals, perhaps from 70 to 212 CE,73 and sharp
Sabellian influence for several decades, from 190 to 220 CE.74
Each of these heresies would have been motivated to strike ούδε
ό υιό? against developing orthodox theology. Against each heresy dis­
cussed in this section, orthodox scribes would have benefited by the ad­
dition of ουδέ ό υιό? had it been originally absent in Matthew. We can
speculate about both who would be motivated to change the text and
when, but given the substantial overlap of influence among these groups
(many with polar opposite Christologies), it is difficult to implicate one
group without embroiling another by extension of the arguments used.
Patristic evidence may be helpful in sorting out the who and when of
the omission in Matthew 24.36.

THE PATRISTIC EVIDENCE

At what point did the orthodox begin to become concerned with
the notion that Jesus did not know the day or hour of his return?75 Until

71. Carroll, Founding of Christendom, 467. Other historical accounts set up the rela­
tionships differently; e.g., Hippolytus saw Cleomenes, a follower of Epigonus, as
tiie founder of the school (Hippolytus, Philos. 9.7).
72. Brown, Heresies, 101
73. The date of 70 CE reflects the conviction that Simon Magus was the Simon from
Acts 8. The date of 200 CE is more indirect. We can surmise that Ignatius wrote
μ.

■hisietters'againstthe Docetists ca. 110 CE, during a set of years constrained by his
date of death somewhere in the reign of Trajan (98-117 CE). Although I sidelined
Mardon’s direct following, Tertullian was still writing against Marcion’s ideas in
Adversus Mardonem between 207 and 212 CE.
74. One who objects on the ground that such groups could not have altered manu­
scripts not in their possession should realize that (1) the complicated nature of
manuscript transmission would allow for it, (2) the heretics were often inside the
church ranks, and (3) one cannot have it both ways: either the orthodox controlled
their manuscripts or they did not, and if they did, the claims of wild copying must
be revisited.
75. Of the references listed in BP, one reference to Ambrose appears to be mismarked
(Ambrose, Defide 5.4.54), and the Pseudo-Clementine reference is vague, but the
author did not seem to have a problem with an alleged ignorance of Jesus. More
important, if a reference in BP is not listed in the following examples, it is because

Adam G. Messer Chapter 4 / 145

 Patristic Theology and Recension in Matthew 24.36

they are concerned with this, the motivation for alteration is difficult
to assert. In accordance with Origen’s solution, I distinguish between
a theological problem with ^re-resurrection ignorance and a problem
with j?o$t-resurrection ignorance.* 76 As the impact of Origen on later
fathers was substantial, his differentiation between the levels of Christ’s
knowledge before and after tire resurrection is a poignant issue when
considering whether they would remove the phrase to alleviate his pre­
resurrection ignorance, because the church fathers may have had a
problem with one and not the other. Further, since the idea behind Mark
13.32 is identical to that of Matthew 24.36, it is important to determine,
as often as possible, whether the verse quoted is indeed from Matthew.
If we mistake the source for a patristic quotation wherein ούδέ ό υιό?
is present to be from Matthew’s Gospel when it is truly from Mark, we
may falsely conclude that the church father in question used a version
of Matthew containing the phrase. So I will seek to determine the form
of Matthew a father is using by invoking the aforementioned method
against the differences in the Greek.

Irenaeus ofLyon
The earliest patristic discussion about Matthew 24.36 is by
Irenaeus of Lyon (ca. 130-200 CE). Irenaeus received the torch from
Polycarp, who received it from the apostle John, and he was not afraid
to invoke his heritage against Gnostics who were claiming that their
truth was of apostolic origin. In addition to Irenaeus’s early attestation
to the symmetry between the four Gospels and the “four zones of the
world ... and four principal winds,”77 we learn much about his the­
ology in his five-volume treatise entitled Adversus haereses (Against
Heresies), written to counteract Gnostic teaching. At one point,
Irenaeus is addressing the Gnostic practice of using Scripture to serve
their own ends, so he seeks to show the proper method of interpreting
Scripture—particularly obscure passages and the multifarious inter­
pretations of parables. He concludes that we should interpret these in
light of what we know to be true about God and not vice versa. In con­
trast to the Gnostic emphasis on secret knowledge, Irenaeus argues
that complete knowledge about every inquiry into divine Scripture
cannot be obtained in this life, and he points to Christ as an example
of ignorance to chastise those who claim to know the “unspeakable
mysteries of God":

it either did not help discern a father’s form of Matthew or did not appear to offer
additional insight beyond those examples explicitly discussed.
76. Origen, Commentary on Matthew 55.
77, Irenaeus, Against Heresies 3.11.8.

146 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

Ye presumptuously maintain that ye are acquainted with the unspeak­

able mysteries of God; while even the Lord, the very Son of God, allowed
that the Father alone knows the very day and hour of judgment.... If)
then, the Son was not ashamed to ascribe the knowledge of that day tn
the Father only,... neither let us be ashamed to reserve for God those
greater questions which may occur to us. For no man is superior to his
master.73

Irenaeus gives no indications of struggle regarding Christ’s igno­

rance; rather, he uses it as instructive for the Christian. It is difficult to
discern what his Matthean text looked like, particularly since the Greek
text is missing, but a Latin translation is extant. Because some of the
verse is missing, he may be quoting from memory, but we can at least
say that the presence of “alone” (solus), combined with the reality that
Matthew was, in general, the most read of the synoptics, would favor
the view that ifhe is recalling a verse to memory, it would be Matthew,
in which case our phrase is present in his text.7879 Whatever the case, we
find no evidence of the omission in Irenaeus.

Hippolytus ofRome

Assuming the authenticity of his work Philosophumean, Hippolytus of

Rome (ca. 180-230 CE) had an intriguing career. While his mentor com­
bated the heresy of Gnosticism, Hippolytus fought against the Iheodotians
(Adoptionism), on the one hand, and Noetus and Sabellius (Modalism),
on the other. When Pope Zephyrinus remained hesitant to condemn
Modalism, Hippolytus declared him incompetent. When that pope died
and Callistus, whom Hippolytus believed to be modalistic, took his place,
Hippolytus withdrew and declared himself not only a rival pope but the
antipope—that is, against the establishment and as the true apostolic suc­
cessor. Historians are uncertain whether Hippolytus’s objections against
Zephyrinus or Callistus are fair, but several of his personal characteristics
are evident. He was uncompromising and somewhat reactionary, was an
eloquent and prolific writer, and had some association with Origen. At one
point, he was banished from the Catholic Church, but he reconciled with
it prior to his death and was possibly even martyred for the faith.80

78. Irenaeus, Against Heresies2.28.6; translation from ANF 1:401.

79. The Latin of 2.28.6 appears to be a direct (albeit partial) quotation: De die autem
ilia et hora nemo scit, deque Filius, nisi Pater solus (Adelin Rousseau and Louis
Doutreleau, eds., Irenee de Lyon centre les Heresies, vol. 2, pt. 2, Sources Chre-
tiennes 294 [Paris: Les £ditions du Cerf, 1982], 282). Aside from leaving out "not
the angels of heaven,” its form matches exactly both the words and word order of
the Vulgate's form of Matthew. It retains the functional equivalent of μόνος· (solus).
80. Johann Peter Kirsch, “Saints Hippolytus," in TCE, vol. 7 (New York: Robert Ap­
pleton, 1910), 360-62.

Adam G. Messer Chapter 4 / 147

 Patristic Theology and Recension in Matthew 24.36

Hippolytus understands Christ as an amalgam between both the

human and divine traits, who acts in accordance with each nature at
different times and does not refuse the limitations of humanity (which
would seem to exclude foreknowledge):

For it was not in mere appearance or byconversion, but in truth, that He

became man. Thus then, too, though demonstrated as God, He does not
refuse the conditions proper to Him as man, since He hungers and toils
and thirsts in weariness, and flees in fear, and prays in trouble. And He
who as God has a sleepless nature, slumbers on a pillow. And He who
for this end came into the world, begs off from the cup of suffering. And
in agony He sweats blood, and is strengthened by an angel.81

In his commentary on Daniel that references Matthew 24.36, we

find these sentiments: “The disciples also sought to learn these things
[the time of end] from the Lord, but he concealed the day from them,
so that he might have people who watch [for it].”82 Here is a difference
between Hippolytus and Irenaeus; the former makes Jesus the concealer
of the mystery. Although Hippolytus makes the assumption that Christ
knew the time of the end at some point (since he is the concealer of it),
he was probably referring to Acts 1.7, wherein it seems to be Christ’s
explicit intention to deny the disciples this information. Therefore,
Hippolytus would probably have had a problem with the notion of post­
resurrection ignorance. Since his references to Matthew 24.36 are only
conceptual, no determination can be made about his form of the text.
Hippolytus lived in the late second and early third centuries, and he
battled against heresies on opposite sides of the theological spectrum—
both Adoptionism and Modalism. During this precise time frame when
some have conjectured an orthodox alteration to Matthew 24.36,83
church leaders were fighting heresies that would have benefited by the
removal of ούδέ ό υιός· (the Modalists). Therefore, it seems unlikely that
the orthodox would have made such a revision.

Tertullian of Carthage

Quintus Septimius Florens Tertullianus (ca. 160-220 CE), or

Tertullian, is arguably the most important Latin father after Saint
Augustine, and he is well known for shunning philosophy and yet
coining the term Trinity in opposition to early Sabellians. His training
in law, literature, and philosophy perhaps disillusioned him to the role
philosophy could play in the spiritual life, even though he was not averse

81. Hippolytus, Against Noetus 17-18; translation from ANF 5:230.

82. Hippolytus, Commentarium in Danielem 4.16.2.
83. Ehrman, Orthodox Corruption, 91-92.

148 / Chapter 4 Adam G. Messer

 Αλ Evaluation of Ehrman’s Text-Critical Methodology

to employing it as a jurist and later as a priest. He alludes to Matthew

24.36 in. three separate works. In On the Resurrection of the Flesh, he
spends time making the point that Christians are not resurrected im­
mediately upon death but must wait for the last day, known only to God
the Father:

Now, forasmuch as the seasons of our entire hope have been fixed in
the Holy Scripture, and since we are not permitted to place the ac­
complishment thereof; as I apprehend, previous to Christ’s coming,
our prayers are directed towards the end of this world, to the passing
away thereof at the great day of the Lord—of His wrath and ven­
geance—the last day, which is hidden (from all), and known to none
but the Father.8*

A second reference can be found in De anime (A Treatise on the

Soul), which is about the inherent conflict between judgment and the
transmigration of souls. In it, he makes a vague reference to Matthew
24.36 and affirms that only the Father knows the day of judgment:

Accordingly, God's judgment will be more full and complete, because

it will be pronounced at the very last, in an eternal irrevocable sen­
tence, both of punishment and of consolation, (on men whose) souls
are not to transmigrate into beasts, but are to return into their own
proper bodies. And all this once for all, and on “that day, too, of which
the Father only knoweth;"... in order that by her trembling expecta­
tion faith may make full trial of her anxious sincerity, keeping her gaze
ever fixed on that day, in her perpetual ignorance of it, daily fearing
that for which she yet daily hopes.84
85

In. a context about the uncertain time of death, the fact that he brings
in the Father’s singular knowledge of the end rather than the minimum
rhetorical idea needed (viz., that of the soul’s ignorance) suggests that
he has no conflict with Christ not knowing this same information.
' 'In Against' Praxeas, Christ’s ignorance of the last day and hour
exists in juxtaposition with reference to Christ’s deity.86 Tertullian
also addresses the distinctness of the Son from the Father by listing
a number of Jesus’ unique characteristics, one of which is simply that
“He is also ignorant of the last day and hour, which is known to the
Father only.”87 Although it appears that Tertullian can hold Jesus’ selec­
tive ignorance in a cohesive theological schema without any signs of

84. Tertullian, On the Resurrection ofthe Flesh 22; translation from ANF 3:560.
85. Tertullian, A Treatise on the Soul 33; translation from ANF 3:215.
86. Tertullian, Against Praxeas 26.
87. Ibid.

Adam G. Messer Chapter 4 / 149

 Patristic Theology and Recension in Matthew 24.36

conflict, the polemical nature of this treatise would preclude the likeli-
hood that he would express signs of conflict with this understanding
while arguing against Modalists. Had he held such internal reserva­
tions, he might just as easily have left out this example without any
violence to his argument. His importation of this idea into a treatise
against the transmigration of souls seems to suggest that Tertullian had
no problem with a pre-resurrection ignorance of Christ, if not also a
post-resurrection lack of knowledge, but the latter is speculative. As
far as his form of Matthew is concerned, no reference establishes its
certainly, but the presence of “only” suggests that Matthew was the
source and that Tertullian’s copy contained the phrase.88

Origen ofAlexandria
Origen of Alexandria (185-254 CE) was an astute textual critic, even
compiling a massive work known as the Hexapla, which consisted of six
columns comparing different versions of the OT for the purpose of es­
tablishing the original. Further, he discusses manuscript evidence in his
homilies and commentary on the NT. For example, in his Commentary
on Matthew, his discussion on Matthew 18.1 notes that some manu­
scripts use "in that hour” while others use “in that day” He then pro­
ceeds to establish what the original said.89 Origen truly had an acumen
for textual criticism that was extraordinary for the early third century,
but despite claims to the contrary,901 was unable to find any discussion
of this omission by Origen; rather, the extant Latin text from section 55
of his Commentary on Matthew attests to the presence of ούδέ ό υίό^,91
and he says about this text,

88. Tertullian, Against Praxeas 26. The Latin isignorans et ipse diem ethoram ultimam
soli Patri notam. The presence of soli reminisces of the same word solus in the
Latin version, of Matthew. See Hermann-Josef Sieben, ed., Tertullian Adversus
Praxean: Gegen Praxeas, Fontes Christiani (Freiburg: Herder, 2001), 232; Ernest
Evans, Tertulliani Adversus Praxean Liber {London: S.P.C.K., 1948), 123.
89. He states his intent: “Let us see if it is possible from them [the variants] to find a
way to understand, as being necessary, the addition, ‘in that day’ or 'hour'” (Origen,
Commentary on Matthew 2.13.14). If the production of the Hexapla and the pre­
ceding example are not enough to establish his interest, he is forthright why this
enterprise is important to him, "But now it is clear that there are many differences
in the manuscripts, whether from carelessness [ραθυμία] of various scribes, from
the audacity of some wicked scribes, from the neglect of correctors of the Scrip­
tures, and even from the supposition made by the corrector himself to add or take
away [words]” (Origen, Commentary on Matthew 2.15.14). This statement refer­
ences both a problem in Matthew and one in the OT.
90. Ehrman, Orthodox Corruption, 91-92.
91. The Latin is de die autem ilia et hora nemo scit, neque angeli caelorum, neque Fi-
lius, nisi Pater solus (MPG). This is an exact match in word form and word order to
the Vulgate, but with “nor the Son” {neque FiUus).

150 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman's Text-Critical Methodology

Just as Paul said, “So at that time the son will be subjected to the
Father"; not as one who was formerly independent, but as one whose
members are not yet fully perfected. Similarly, Christ did not know
that day or hour... according to the use of the word “know” in these
passages, “the one who keeps the commands will not know stum­
bling,” and, “the one who knew no sin was made sin on our behalf” In
this way [we understand] what is meant by “nor the son [knew] .”92

Origen, understands Jesus to be claiming that he does not have inti­

mate knowledge of the affair, like the experience of Adam lenowing Eve
intimately. His reason for why the Son claimed ignorance was that, al­
though he established the date prior to the incarnation, he had not fully
grown into all knowledge prior to his resurrection. This explains why
Jesus, during a post-resurrection occurrence in Acts 1.7, later changed
his response to the same question to “It is not for you to know the day
or hour.”93 Origen affirms this: “Because afterwards the Son also knew
by receiving knowledge from the Father even over the day and the hour
of the consummation, so that no longer would only the Father know, but
also the Son.”94
In summary, Origen does not appear to have a problem with Christ’s
pre-resurrection ignorance.95 He often contemplated the limitations

92. Origen, Commentariorum Series in Evangelium Matthaei 126-27. An interpretive

issue surrounds the phrase τό σώμα του Χρίστου, which is here taken to refer
to Christ himself (“Christ did not know”) and not to a group of believers or his
church (“the body of Christ did not know”). Although the immediately following
lacuna prevents additional verification of this interpretative decision, the context
is seeking to show what is meant by the phrase “neither the Son.”
93. Ihe emphatic placement of the pronoun in the verse brings legitimacy to the rhe­
torical stress indicated with italics. My appreciation goes to Dan Wallace for this
observation.
94. Origen, Commentary on Matthew 55. This section is only extant in Latin: Nam
postea etFilius cognovit scientiam a Patre suscipiens, etiam de die consummationls
et hora, utjam non solum Pater sciret de ea, sed etiam Filius (MPG).
95. Origen elsewhere entertains the notions that God has not even fixed the day of the
return (Origen, Commentary on Matthew 55) or that possibly Christ was speaking
on behalf of the church, to help its members realize that it is not for them to know
the time (Origen, Commentary on Matthew 55). These notions may indicate that
he was looking for reasons to avoid the implication that Jesus was ignorant, but it
is more likely that they represent Origen's focus and passion. When one reads his
works, what strikes the reader is his attention to detail and grasp of its implications.
He explores issues from a great variety of angles and demonstrates a willingness to
explore uncharted theological territory. He was truly an early Christian academic
in tilts regard. This predilection would explain why Origen disclosed multiple theo­
ries, along with the fact that the fathers were not averse to seeing multiple senses to
Scripture. Origen held to a threefold interpretation: a literal sense, a moral sense,
and an allegorical-mystical sense (Robert M. Grant and David Tracy, A Short His­
tory of the Interpretation of the Bible, 2nd ed. [Minneapolis: Fortress, 1984], 59).
He was particularly favorable to the allegorical sense, believing it to be preeminent.

Adam G. Messer Chapter 4 / 151

 Patristic Theology and Recension in Matthew 24.36

listed in Scripture about Jesus, and yet his systematic works found no in­
consistency between Jesus having ignorance while on earth and yet being
deity. He did, however, believe that Christ acquired all knowledge after
his resurrection. Of further interest, his great concern for properly estab­
lishing the text is valuable for the modern textual critic’s understanding of
what manuscripts Origen knew about in both Alexandria and Caesarea.

Eusebius of Caesarea
Eusebius of Caesarea (260-341 CE), the famous church historian of
the fourth century, may have referenced Matthew 24.36 in a fragmen­
tary work known as Supplementa ad quaestiones ad Marinum·.

He was able to attain the resurrection from the dead, according to which
hour no one knew, and according to which time no one was told by the
evangelists. And so one opportunely inquired, and to him was told every­
thing concerning the end by what had [already] been said, "Concerning
the day none knows, not even the angels of God.” Therefore, the Savior^
who has become the first fruits of the resurrection, according to which
hour none knew, was raised, escaping the notice of everyone, and
brought up in form out from the stone [grave] he had occupied.96

The possible reference to Matthew 24.36 has all distinguishing

Greek features eradicated (probably because it is a recollection from
memory), so there is no way to discern its source or his form of Matthew.
Theologically, the verse appears to be associating the uncertainty of the
day and hour to Christ’s resurrection. There are no indications that
Christ was ignorant, but given that other important parts, such as the
Father’s sole possession of this knowledge, is also absent, Eusebius’s evi­
dence is neutral at best.

Athanasius ofAlexandria
Athanasius (ca. 290-373 CE) was a staunch defender of the doc­
trine of the Trinity, so much so that he earned the title “the Father of
Orthodoxy." He records his thoughts on the matter \n Discourses Against

As such, Origen was unalterably committed to finding in every piece of Scripture

a spiritual nugget for the church. He is known to make even relatively mundane
details applicable to his listeners (Origen, Homilies on Luke 47). In this fight, it
is entirely unsurprising that he would posit Jesus speaking as the church in Matt
24.36. However, we also find that he did write a systematic presentation of his be­
liefs called Deprincipiis (On First Things) in which his language is less speculative,
and his section on the end clearly states that only the Father knows the time (De
principiis 1.6).
96. Eusebius, Supplementa ad quaestiones ad Marinum 985.

152 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman's Text-Critical Methodology

the Arians. He addresses the Arian point of view in multiple ways, but
one is hermeneutical—by referring to the canonical context:

“But of that day and that hour knoweth no man, neither the Angels
of God, nor the Son;’' for being in great ignorance as regards these
words, and being stupefied about them, they think they have in them
an important argument for their heresy.... and the Son who knows
the Father is said to be ignorant of an hour of a day; now what can be
spoken more contrary to sense, or what madness can be likened to
this? Through tire Word all tilings have been made, times and seasons
and night and day and the whole creation; and is the Framer of all said
to be ignorant of His work? And the very context of the lection shews
that the Son of God knows that hour and that day, though the Arians
fell headlong in their ignorance.97

Athanasius argues that since other portions of Scripture indicate

that Christ is the creator of all things, interpreters should allow this
truth to inform less clear verses. For him, the question is not, did the
Son know the day and hour? Instead, it is, given that the Son knows by
virtue of other clear passages, why would he say he did not? His answer
pertains to tine nature of humanity:

Now why it was that, though He knew, He did not tell His disciples
plainly at that time... [of] this I think none of the faithful is ignorant,
viz. that He made this as those other declarations as man by reason
of the flesh. For this as before is not the Word’s deficiency, but of that
human nature whose property it is to be ignorant.98

Athanasius has a definite problem with the idea that Jesus was not
omniscient, for he continues to give reasons, such as the following, for
why Jesus did know the time of his return: (1) the absence of the Holy
Spirit in Matthew 24.36 should clue us in that Jesus was not excluded
from the knowledge, because if the Spirit had it, so must have the one
onwhom-he-depended;99 (2) for him to express ignorance was to iden­
tify with humanity, even though he was much more;100 (3) Christ set the
proper example for how believers should not be negligent toward the
end;101 and (4) in context, Christ’s admonition to keep watch “because
you do not know on what day your Lord will come” does not say “be­
cause I do not know” (italics mine).102 It is worth mentioning that these

97. Athanasius, Discourses against the Arians 3.42; translation from NPFS 4:416.
98. Ibid., 3.43; translation from NPFS 4:417.
99. Ibid., 3.44.
100. Ibid., 3.46.
101. Ibid., 3.49.
102. Ibid., 3.45. The context is Matt 24.42. Cf. Mark 13.33.

Adam G. Messer Chapter 4 / 153

 Patristic Theology and Recension in Matthew 24.36

arguments by no means exhaust the arsenal of Athanasius, but they are

sufficient to demonstrate his problem with an ignorant Christ even be­
fore the resurrection. It is also noteworthy that his version of Matthew
does not contain the phrase ουδέ ό υιός·.103

Hilary ofPoitiers
Hilary of Poitiers (d. 368 CE) was exiled to Phrygia during the first
six of his final twelve years, for his steadfast stance against Arianism.
This sojourn provided opportunity to study the ways in which the East
and West used different vocabulary to refer to the same ideas. He discov­
ered that some differences between the East’s and West’s Christologies
lay not in their underlying ideas but in their phraseology. This insight
and his sharp rhetoric allowed him to be an influence toward reconcili­
ation between the Eastern and Western Trinitarian churches of the em­
pire for the short remainder of his exonerated life. Another advantage of
this banishment was the opportunity to write his De Trinitate (On the
Trinity) and De Synodis (On the Councils). In Hilary we can see some
of the sophistication with which these early fathers tackled the problem
of Christ’s ignorance, because much of Hilary’s writing on this issue is
extant. Whereas moderns may get involved in countering Arianism with
linguistic arguments, the fathers, such as Hilary, typically offered con­
ceptual rebuttals:

The Son is ignorant, then, of nothing which the Father knows, nor does
it follow because the Father alone knows, that the Son does not know.
Father and Son abide in unity of nature, and the ignorance of the Son
belongs to the divine Plan of silence, seeing that in Him are hidden all
the treasures of wisdom and knowledge. This the Lord Himself testi­
fied, when He answered the question of the Apostles concerning the
times, It is not yours to know times or moments, which the Father
has set within His own authority. The knowledge is denied them, and
not only that, but the anxiety to learn is forbidden, because it is not
theirs to know these times. Yet now that He is risen, they ask again,,
though their question on the former occasion had been met with the
reply, that not even the Son knew. They cannot possibly have under­
stood literally that the Son did not know, for they ask Him again as
though He did know. They perceived in the mystery of His ignorance
a divine Plan of silence, and now, after His resurrection, they renew

103. Ibid., 28.472.47-52. His Greek text of Matt 24.34-36 is almost identical to NA27,
except for the omission of tire phrase in question: Αμήν, λέγω ΰμΐν, otl ού μή
παρέλθη ή yevea αΰτη, έως αν πάντα ταΰτα γένηται. Ό ουρανός και ή γή

Adam G. Messer
 An Evaluation of Ehrman’s Text-Critical Methodology

the question, thinking that the time has come to speak. And the Son
no longer denies that He knows, but tells them that it is not theirs to
know, because the Father has set it within His own authority. If then,
the Apostles attributed it to the divine Plan, and not to weakness, that
the Son did not know the day, shall we say that the Son knew not the
day for the simple reason that He was not God?104

Here are Hilary’s basic arguments:

1. Christ’s assertion of his ignorance pertains to the divine

economy rather than to true ignorance.105

2. The apostles understood that Jesus’ words in Matthew 24.36

only indicated that it was not yet time to reveal the divine plan,106 which
is why they asked again in Acts 1.7.107

3. The ignorance is a means not to delude but to encourage in

perseverance.108

4. That the Father alone knows does not exclude the Son from
knowing, since he is one with the Father.

5. Jesus’ ignorance was true of him as a man but not as God.109

6. Christ spoke of this ignorance to demonstrate, against

Sabeilianism, that Christ submitted to the prerogative of the Father, who
held original initiative within himself.110

Did Hilary have problems with a genuine ignorance of Christ such

that his thought influenced scribes toward the removal of ουδέ ό υιός

104. Hilary, On the Trinity 9.75; translation from NPFS 9:181.

105. Ibid., 9.62.
-106.-HiIary-saw-thegreat challenge of Christ’s life as the gradual impartation of a truth
too big for the contemporary mind, particularly the contemporary Jewish mind.
For this reason, Christ was restricted in what he could say and when he could say
it, by his purpose to move a disbelieving people committed to monotheism into the
true fabric of reality, which centers on a Trinitarian understanding of his person.
This, he says, “is explained by a discrimination between gradual revelation and full
expression of His nature and power.... Christ’s true Godhead is no whit impaired
because, to form the mystery of the Gospel faith, die birth and Name of Christ
were revealed gradually, and under conditions which He chose of occasion and
time” (ibid., 1.30; translation from NPFS 9:48).
107. Ibid., 9.63.
108. Ibid., 9.67.
109. Ibid., 9.75.
110. Ibid., 10.8.

Adam G. Messer Chapter 4 ! 155

 Patristic Theology and Recension in Matthew 24.36

from the manuscript tradition? All of the preceding explanations as­

sume this difficulty, against which he offers additional counters: (1) the
apostle Paul’s declaration in Colossians 2.3 that all wisdom and knowl­
edge is contained in the Son,211 (2) the principle that overattention on
a single word or phrase divorced from conceptual contexts is greatly
dangerous,122 (3) that we should understand the words of Matthew 24.36
in light of (a) what manner of questions he was answering, (b) at what
times he was speaking, and (c) what knowledge he wished to impart at
that time,213 (4) that ignorance cannot defeat the “Lord of glory,”214 and
(5) Jesus’ desire to serve as an example to his disciples.215 Hilary dealt
with the issue of Christ’s ignorance on a conceptual level. Having said
this, his version of Matthew did appear to have contained ουδέ ό υΙός.111
115
114
113
112
116

Epiphanius ofSalamis

In 367 CE, Epiphanius117 became the bishop of Salamis, a town on

the east coast of Cyprus. He was intelligent and broadly studied, being
called “five-tongued” for his learning in Hebrew, Syriac, Egyptian,
Greek, and Latin. His work Panarion (Medicine Chest) deals with
around eighty ancient heresies he traveled widely to squelch. He appar­
ently even took issue with Origen and John Chrysostom. These factors,
along with his ascetic background, probably made him well known,
which means that his views were not parochial. He was an ardent de­
fender of Trinitarian theology, not unlike Athanasius.

111. Ibid., 9.62.

112. Ibid., 1.32. He says, "(The heretics’] folly being as great as their blasphemy, they
fail to mark that Christ's words, spoken under similar circumstances, are always
consistent; they cleave to the letter and ignore the purpose of His words."
113. Ibid., 1.30.
114. Ibid., 9.60. This is a sophisticated argument based on 1 Cor 2.8, namely, that unless
Christ participates in deity, how can he be glory’s Master?
115. Ibid., 9.71.
116. Ibid., 1.29-30. The reference in Hilary’s On the Councils (85) seems to quote from
Mark, since it is lacking “only.”
117. At least two men named Epiphanius are recorded in church history. .They.are actu­
ally over a hundred years apart, yet attribution of the works Ancoratus and Pan-
arion is made to both. BP attributes these works to Epiphanius of Constantinople,
a patriarch of this city in the early sixth century. Others attribute it to Epiphanius
of Salamis, an ascetic who lived through most of the fourth century, including the
Arian battles that accompanied this period. Interestingly, TLG attributes it to the
patriarch of Constantinople and yet dates it as though it were the other. Without
entering into this particular discussion, other than to note it, I assume the earlier
Epiphanius of Salamis for two reasons: (1) this position seems to have broader sup­
port, and (2) he is strategically placed within the Arian controversies of the fourth
century following the Nicene formulation (ca. 315-403 CE). Even more intriguing,
Epiphanius of Salamis was born in Judea, and after his conversion from Judaism
and sojourn as a monastic in Egypt, he returned to his birthplace and set up a
monastery.

156 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

Epiphanius had a definite problem with the idea that the Son did
not know the day and hour of his return. In his work Ancoratus (Well-
Anchored), he paraphrases Matthew 24.36 and Mark 13.32, saying:

For it is not said concerning angels [that] they examine the deep things
of God, nor about archangels. “For no one knows the day or hour," the
Son of God said, “not the angels of heaven nor the Son, except the
Father.” But the foolish who have not been adorned by the Holy Spirit
[need to] consider [that] one is not in the Father who is not in the di­
vinity of the Son. “For as the Father has life in himself, so also the Son
has life in himself? And, “All that belongs to my Father is mine," said
the holy Word of God himself. And as for the things which belong to
the Father: What belongs to God the Father is also of the Son. The life
of the Father, this also is of the Son.118

Here he argues that everything that the Father has, the Son also has—
including all knowledge. An ambiguity exists in the feet that Epiphanius is
speaking in the present tense, and this may refer to Christ’s current estate.
If a number of church fathers interpreted Christ’s statement of ignorance
as temporarily limited to his earthly sojourn prior to the resurrection (as
Origen earlier brought out), even if they believed that Christ in his ex­
alted state knew the day and hour, a scribe encountering Matthew 24.36
in the heat of controversy might have found reason to strike ούδε δ υιός
from his copy. He may have reasoned that such ignorance was liable to
misrepresent Christ’s present actuality and abilities. Though this may not
reflect a direct quotation, it uses a genitive for expressing “in heaven" (like
Matthew) and omits μόνος. If it is a quotation, it is probably from Mark.
Within this same work, Epiphanius later references Matthew 24.36
twice119 and adds two arguments to his arsenal. First, he argues that
since the Son knows the Father fully and since knowledge of the Father
is greater than knowledge of the day and hour, Christ must know the
time of his return.120 Second, he maintains that there are differing de­
grees of knowledge—namely, that of raw information and that of expe­
rience—and that Christ was only claiming ignorance of the latter.121

118. Epiphanius, Ancoratus 16.2-6. The candidate for quotation is ούδει? γάρ οΐδε τήν
ή μέραν ουδέ τήν ώραν ούτε οΐ άγγελοι του ουρανοί) ούτε δ υιό?, εΐ μή δ πατήρ.
119. Epiphanius, Ancoratus 19.6-7. The candidate for quotation is ούδεί? οιδε τήν
ημέραν και τήν ώραν ούτε οΐ άγγελοι οί εν τώ ούρανω ει μή μόνο? δ πατήρ.
Though this may not reflect a quotation, it uses a dative for expressing "in heaven”
(like Mark), omits ουδέ δ υιό?, and includes μόνο?. If it is a quotation, it is prob­
ably Matthew and would reflect die omission. The second reference is Epiphanius,
Ancoratus 22.1—4, which is just a snippet: εΐ μή δ πατήρ μόνο?, ούτε οι άγγελοι
ούτε δ υιό?.
120. Epiphanius, Ancoratus 17.1-4.
121. Ibid., 21.3-5.

Adam G. Messer Chapter 4 I 157

 Patristic Theology and Recension in Matthew 24.36

Epiphanius’s Panarion is built around a systematic attempt to

dismantle ancient heresies, from Ebionite to Essene, from Stoic to
Sebuaean, from Pepuzian to Pharisee. His discussion in Panarion ar­
gues that since the Father has entrusted all judgment to the Son, the Son
would know the day.122
Some interesting observations surface. The first of note is that
Epiphanius traveled widely throughout Judea, Egypt, and the region of
Cyprus to battle heresies and makes no formal mention of differing ver­
sions of Matthew. Second, although he takes vehement issue with the
idea that the Son did not know the time of his return, he may not have
had an issue with the Son’s pre-resurrection ignorance.123 Third, five of
Epiphanius’s references to Matthew 24.36/Mark 13.32 (presented here
with word order excepted) illustrate the difficulty in making a determi­
nation of what his form of Matthew looked like:

a. την ή μέραν καί την ώραν ούδεί? οίδε

b. περί τή? ημέρας εκείνη? καί τή? ώρα? ούδεί? οιδεν

Μ
c. περί δε τη? ημέρα? εκείνη? η τή? ώρα? ούδεί? οίδεν
d. γάρ την ημέραν ουδέ την ώραν ούδεί? οιδε
e.

a. ούτε οί άγγελοι οΐ εν τωι ούρανφ εί μή ό πατήρ μόνο?

b. ούτε οί άγγελοι έν ούρανφ ούτε ό υίό?, εί μή ό πατήρ μόνο?

c. ούτε οί άγγελοι ούτε ό υίό?, εί μή ό πατήρ μόνο?

ν e e/
d. ούτε οι άγγελοι τού ουρανού ουτ€ ο υιο$, εί μή ό πατήρ
e. ούτε οί άγγελοι ούτε ό υιό? εί μή ό πατήρ μόνο?

122. Epiphanius, Panarion 69.46.

123. Two indicators argue that he did not have a problem, with at least some difference
of knowledge between the Father and Son. Epiphanius’s breakdown of knowledge
into degrees allows him to maintain Jesus’ omniscience, but it still highlights a
difference between the Father’s knowledge and the Son’s (Epiphanius, Ancoratus
22.4). Second, all discussions seem to use present tense verbs (“the Son does not
know”) rather than past tense verbs. This may suggest that he is taking issue with
those who say Christ currently does not know the time of his return. As men­
tioned, Origen differentiated Christ’s knowledge before the resurrection with his
knowledge afterward. Epiphanius was no friend of Origen, for his condemnations
of Origen’s writings led to synodal censure. So for Epiphanius and Origen to share
this idea attests to its ingemination in patristic theology apparently across a diverse
spectrum of points of view.

158 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

No recognizable patterns truly emerge.124 However, if the presence of

μόνος is used as an indicator of Matthew’s text, Epiphanius would at­
test to the reading with ουδέ ό υιός. As it stands, the absence of certain
phrases and forms in any single example indicates that he is likely op­
erating primarily from memory—-perhaps completely so. That no refer­
ence matches our critically reconstructed text does not help decide the
case.

Amphilochius oflconium

Amphilochius of Iconium (ca. 340—400 CE), a contemporary and

member of the Cappadocian fathers, a friend to Gregory of Nazianzus,
and a theologian in league with Basil of Caesarea, interacts in an inter­
esting manner regarding the concept of the Son’s ignorance. In response
to such an idea, he argues that the Son was not making an emphatic
statement about Ms ignorance of the second coming as much as he was
asserting his dependence on his Father; that is, although Christ offers
foresight about a number of things, he is concerned to stress the utter
secrecy of the event. To support this assertion, Amphilochius makes
a case, from similar constructions in both testaments, that the el μή
(“except") should be taken in this instance—as it is taken in other places,
such as 1 Corinthians 2.11, Isaiah 1.9, and Psalms 93.17 and 123.2-3—
as not meaning “without exception”:

And when it [our passage] says, “neither the Son, except the Father,”
it does not falsely allege an ignorance of the Son, but it explains an
assertion about knowledge. For all things, as many as the Father has,
are clearly also of the Son. Not that the Son has this knowledge unless
(el μή) the Father has this knowledge. But [what] the Father has, truly,
[so] also does the Son similarly have. For Scripture has been known,
on occasion, to put the el μή in various places indeed, such as: no one
knows the things of man except (el μή) the Spirit of man which is in
him—similarly also the things of God no one knows, except (el μή) the
. Spinlwhichis from God [1 Cor 2.11]. But when el μή is placed beside
those things which are expected, then no longer does it assert what is
presented in the apodosis. Take, for example, when it says, “Unless (el

124. The Greek text appearing closest to direct quotation is περϊ δέ τής ή μέρας
εκείνης ή τής ώρας οΰδε'ις οΐδεν, ούτε οι άγγελοί ούτε ό υιός, εί μή ό πατήρ
μόνος (Epiphanius, Panarion 69.15.5). The conjunction is from Mark, the μόνος
from Matthew, and the angelic abode (“in heaven”) is missing. The Greek in the
other is περί τής ημέρας εκείνης και τής ώρας ούδεις οίδεν, ούτε οί άγγελοι
έν ούρανω ούτε ό υιός, εί μή ό πατήρ μόνος (Epiphanius, Panarion 69.43.1). The
conjunction is taken horn. Matthew, while the TSKS is absent as in Mark; and the
μόνος is from Matthew, while the dative εν ούρανω is from Mark. Because of this
mixture, it is difficult to discern which text he is quoting from—perhaps both.

Adam G. Messer Chapter 4 ! 159

 Patristic Theology and Recension in Matthew 24.36

μή) the Lord of the Sabbath had left to us a posterity, we would have
become like Sodom, and would have resembled Gomorrah” [Isa 1.9;
Rom 9.29]. Therefore, since he did leave for us a posterity, we have not
become like Sodom nor truly resembled Gomorrah.125

Amphilochius understood the el μή to reflect a less exclusive idea

than “except.” Rather, it would be used to mean, "No one knows the
day or hour, and no one could know—not the angels in heaven nor the
Son—if the Father didn’t know.” But since the Father knows, so also
does the Son; nevertheless, it is information locked away by virtue of
belonging exclusively to the divine realm. When Jesus brought the focus
back to the Father, he was asserting not his own ignorance but the do­
main of God.

Basil ofCaesarea
Basil of Caesarea (329-79 CE), arguably the most brilliant of the
Cappadocian fathers, spent time as a youth in Pontus and was schooled
in Caesarea, Constantinople, and Athens. He labored in arenas that the
modern American church might find politically unfeasible. He worked
to exclude “unfit candidates from the sacred ministry” and to deliver
“bishops from the temptation of simony.”126 Later in life, he was influ­
enced by his sister Macrina toward a more ascetic lifestyle and retraction
of earthly attachments, and he visited monasteries in Egypt, Palestine,
Coele-Syria, and Mesopotamia to gather information in order to start
one in Pontus, in which he implemented his cenobitic form. His foci
revolved around self-discipline, almsgiving, and avoiding the jealousy
his superior administrative and intellectual giftedness roused in more
established clergymen. He interacted not only with the Cappadocians
but also with Athanasius.
Basil was a powerful player in the Trinitarian batde, and he took up
our passage with these words:

So now also, if anyone takes the fact of ignorance as applying to Him

who had received all things according to the dispensation and was
advancing in wisdom and grace with God and men, will not be carried
beyond a conception consistent with piety. It would be a fitting task
for your diligence to set forth the words of the Gospel and to com­
pare those of Matthew and those of Mark with each other. For these
alone appear to have agreed with each other on this passage. Now

125. Amphilochius, Fragment 5 ofSer. Ecd. 5. This reference is from TLG.

126. Joseph McSorley, “Saint Basil the Great,” in TCE, voL 2 (New York: Encyclopedia,
1913), 330-34. Basil and Gregory of Nazianzus compiled a work known as Philo-
calia, which contained excerpts of Origen’s work.

160 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

the reading in Matthew is as follows: “But of that day and hour no

one knoweth, not the Angels of heaven, but the Father alone.”127 And
Mark's is: “But of that day or hour no man knoweth, neither the angels
in heaven, nor the Son, but the Father.”128 What, then, is worthy of note
in these words? That Matthew said nothing about the ignorance of the
Son, but seems to agree with Mark in meaning when he says: “But the
Father alone.” And we hold that “alone” was said in contradistinction
to the angels, and that the Son is not included with His own servants
in the matter of ignorance.129

There are a few things to notice here. Although Basil is fairly well-
traveled, he does not make mention of manuscripts of Matthew that
contain ουδέ ό υιός. Rather, he declares it noteworthy that Matthew
does not mention the ignorance of the Son but uses μόνος in con­
tradistinction to the angels only.130 He believed that the εί. μή did not
necessarily contain an exclusive idea but was here to be taken to mean
that God alone was preeminent; that is, Christ was asserting not his
ignorance but his dependence on the Father’s knowledge as the source
of his own. Beyond its interpretive implications, Basil’s comment is
also singularly interesting in light of the fact that Origen, who wrote
his commentary on Matthew while he was in Caesarea around 246 CE,
mentioned no awareness of its omission. Within 120 years, ουδέ ό υιός
apparently disappeared from the manuscript landscape available to
Origen in Caesarea in his day.131
Basil references the Son’s ignorance and our verse in other places as
well, incorporating these additional arguments:

1. It would be strange for the creator of all to be selectively igno­

rant of something that he created.132

127. The Matthean Greek is Περί 5e τη? ήμερα? εκείνη? ούδε'ι? οιδεν, ουδέ οί
άγγελοι των ουρανών, εί μή ό Πατήρ μόνο?.
128. The Markan Greek is Περ'ι δέ τη? ήμερα? εκείνη? ή ώρα? ούδε'ι? ο’ιδεν, ουδέ οί
άγγελοι οί έν οϋρανω, ουδέ δ Υίό?, εί μή ό Πατήρ.
,129..BasilleiieK,236.1—2;.translation froxn Roy J. Deferrari, Saint Basil: The Letters, vo\.
3, Loeb Classical Library (Cambridge, MA: Harvard University Press, 1926).
130. The Markan parallel has no μόνο? but uses an εί μή construction that can ap­
parently be taken in a preeminent sense instead of merely an exclusive manner
(Charles Edward Powell, “The Semantic Relationship between the Protasis and the
Apodosis of New Testament Conditional Constructions” [PhD diss., Dallas Theo­
logical Seminary, 2000], 180-207).
131. Although Origen may have brought manuscripts from Alexandria with him, he has
a penchant for comparing available manuscripts, and he establishes this practice
within his Commentary on Matthew in at least these places: Matt 4.17,16.20,18.1,
21.5, 24.19, and 27.16-17 (Bruce M. Metzger, “References in Origen to Variant
Readings;’ in Historical and Literary Studies: Pagan, Jewish, and Christian, New
Testament Tools and Studies 8 [Grand Rapids: Eerdmans, 1968], 92-94).
132. Basil, Adversus Eunomium 696-98. This reference is from TLG.

Adam G. Messer Chapter 4 / 161

 Patristic Theology and Recension in Matthew 24.36

2. The canonical context establishing "all things whatsoever the

Father hath are mine” should be an overruling consideration.133

3. The interpretive flexibility of the Greek phrase el μή can be fur­

ther corroborated by listing other scriptural examples.134'

The significance of these observations is that Basil has a problem

with an assertion of the Son’s ignorance. He does not explicitly mention
whether he takes issue with a pre-resurrection or post-resurrection ig­
norance (Basil was very familiar with Origen’s work). Yet it seems likely
that he would take issue with even a pre-resurrection ignorance, given
that he argues that this passage should be read in light of other passages
that discuss the Son’s present ownership of all that belongs to the Father.135

Gregory ofNazianzus

Fighting alonside Basil in the battle for the Son’s coequality was a
friend he made while he schooled in Athens, Gregory of Nazianzus136
(ca. 325-89 CE). He composed orations wherein he addressed the
Arian objection to the Son’s homoousian with the Father:

Their tenth objection is the ignorance [of Christ], and the statement that
Of the last day and hour knoweth no man, not even the Son Himself,
but the Father. And yet how can Wisdom be ignorant of anything—that

133. Basil, Letters 236.2.

134. Ibid., 236.1. He lists Matt 11.27 and Mark 10.18.
135. One cannot help but suspect that even had Matthew contained ουδέ ό υιό?, Basil
would have interpreted it just like he does Mark, namely, with a hermeneutic de­
pendent on a canonical context and a tradition that regarded the “rule of faith” (or
“rule of truth") as a passed-down lens through which a correct reading of Scripture
comes into focus. For Basil, this rule seems to have included Nicea (325 CE) and
the Son’s divinity.
136. Gregory was no stranger to Athens. He spent ten years there for school and met
the future apostate emperor Julian, even then discerning Julian's true character,
and distrusting him. Torn between a desire toward the monastic life of Basil and
the separate conviction that such was not in accordance with his reading of Scrip­
ture, he visited Basil on and off at the monastery during his period of reflection.
His bishop and father finally forced him to take on the priesthood around 361 CE
in Nazianzus. Gregory devoted his inheritance to the poor, refused the bishopric
offered him, and retreated to a monastery in 375 CE. Yet he emerged four years
later, upon invitation from persecuted Trinitarians, to minister to the arch city of
Constantinople, which had fallen into the darkness of Arianism for three decades.
Here he gave his celebrated treatises on the Trinity that earned him both fame and
persecution. During this period, St. Jerome became his disciple; the eastern em­
peror Theodosius was converted and baptized; and the consequent First Council
of Constantinople (381 CE) was convened, in which Gregory had considerable
influence.

162 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman s Text-Critical Methodology

is, Wisdom Who made the worlds, Who perfects them, Who remodels
them, Who is the Limit of all things that were made, Who knoweth the
tilings of God as the spirit of a man knows the things that are in him?
For what can be more perfect than this knowledge? How then can you
say that all things before that hour He knows accurately, and all things
that are to happen about the time of the end, but of the hour itself He
is ignorant? For such a thing would be like a riddle.137138

Gregory has essentially the same line of argumentation as his

Cappadocian brothers, namely, an a fortiori argument: if Christ knows
the greater, how much more should he know the lesser subset of it?
Further, some things can be known as God while not being known as
man. This paradox does not bother Gregory, but the controversies of
Monophysitism and Monothelitism, although germinated, had not yet
bloomed.

Gregory ofNyssa
Gregory of Nyssa (d. ca. 385 CE), Basil of Caesarea’s brother, entered
into the Christological controversies with argumentation similar to that
of his brother and Gregory of Nazianzus. Like the other Cappadocian
fathers, he had a problem with associating ignorance with Christ.158
Notable to this discussion is the fact that his version of Matthew does
not appear to have had ούδέ ό υιός.139

Ambrose ofMilan
Saint Ambrose (340-97 CE), beloved influencer of Saint Augustine,
also mentioned our verse:

137. Gregory of Nazianzus, Orationes 30.15; translation from NPFS 7:315. The Greek
text is not Matthew or Mark but reflects a conceptual parallel (καί τό μηδένα
γινώσκ€ΐν την τ€λ€υταίαν ή μέραν ή ώραν, μηδέ τόν υιόν αυτόν, ei μή τον
πατέρα).
138. “How [is it that] with everyone else the son does not know the day either—the one
who has in himself the Father and himself is in the Father?” (Gregory of Nyssa,
Adversus Arium etSabellium de patre etfilio 84-85).
139. Gregory of Nyssa, Adversus Arium et Sabellium de patre etfilio 76. The Greek text
is ούδείς οιδ€ την συντβ Χαοτικήν ημέραν και την ώραν ούδ1 οί άγγβλοι των
ουρανών ουδέ ό υιός έν τοΐς κατά Μάρκον ειρημένοις, el μή ό πατήρ μόνος.
The phrase έν τοΐς κατά Μάρκον βίρημένοις can be taken in two ways. Either the
whole verse is “according to Mark” (i.e., quoting Mark), or just ουδέ ό υιός· is ac­
cording to Mark. I take it in the latter sense, because of Gregory’s placement of the
attribution to Mark and the verse’s striking similarity in form with Matthew—note
the genitive plural ουρανών and the presence of καί and μόνος. Ihe implication of
this interpretation is that his copy of Matthew did not contain this phrase and that
Gregory was bringing in Mark’s emendation to address the principle behind it.

Adam G. Messer Chapter 4 / 163

 Patristic Theology and Recension in Matthew 24.36

It is written, they say: “But of that day and that hour lcnoweth no man,
no, not the angels which are in heaven, neither the Son, but the Father
only” First of all the ancient Greek manuscripts do not contain the
words, “neither the Son." But it is not to be wondered at if they who
have corrupted the sacred Scriptures, have also falsified this passage.
The reason for which it seems to have been inserted is perfectly plain,
so long as it is applied to unfold such blasphemy.140

Like all the other fathers since the beginning of the fourth century that
I have mentioned, Ambrose clearly had a problem with the Son’s igno­
rance before his resurrection. He even introduces a new twist by as­
serting that none of the ancient Greek manuscripts (plural) contain the
phrase “neither the Son.” He even ponders that its presence is due to
unorthodox corruption. Although he is writing in Latin, the fact that
he specifically mentions “ancient Greek manuscripts” as testifying to
the omission might suggest that current Greek manuscripts contained
the phrase.141

Didymus the Blind

Didymus the Blind (ca. 310-95 CE) of Alexandria is so named be­

cause he lost his sight at the age of four. Yet such was his attention to
oral learning and so acute his knowledge and wisdom that St. Jerome
referred to him as “the Seer.” Didymus may even have coined the
phrase treis hypostaseis, mia ousia (“three Persons, one substance”).
He has four works that reference the idea of the Son’s ignorance. His
arguments against it are similar to other fathers of the fourth cen­
tury. Notable is his mention of Basil and interpreting the el μή in the

140. Ambrose, Defide 5.192-93; translation from NPFS 10:308. We have two peculiari­
ties here. First, Migne's numbering (5.193) is different than NPFS (5.192), Second,
Migne or later editors of MPG believed that Ambrose was quoting from Mark
13.32, but he seems to be quoting from Matthew, because (1) his text bears all the
earmarks of Matthew (de die autem illo ethora nemo scit, neque angeli caelorurn,
nec Filius; nisi solus Pater) and (2) he mentions ancient Greek witnesses without
ουδέ δ vlos* (Mark’s more stable text makes Matthew a more likely source).
141. This is, of course, not a necessary deduction. However, that he has knowledge of
ancient Greek manuscripts suggests that he also has knowledge of current ones,
which he would not have been reluctant to include in bringing forth evidence for
the omission. He knew Greek very well, studied Greek authors such as Philo and
Origen, and even exchanged letters with Basil of Caesarea (e.g.s Basil, Letters 197).
It is possible that Basil, whose manuscripts had the omission, told Ambrose of
them, but one then must question Ambrose’s use of “ancientsince Origen does
not seem to have been aware of them while writing his Commentary on Matthew in
Caesarea. Ambrose and Basil’s relationship is imitable; in fact, some have asserted
that had the East and West always had the ability to communicate as intimately
as they did, the schism would never have happened (James Loughlin, "Saint Am­
brose," in TCE, vol. 1 [New York: Encyclopedia, 1913], 383-88),

164 / Chapter 4 Adam G, Messer

 An Evaluation of Ehrman's Text-Critic al Methodology

preeminent sense.142 Didymus apparently had multiple manuscripts at

his disposal, for his quotation of Matthew in De Trinitate is lacking
ουδέ ό υιός (384-92 CE), while it is present in his Commentaries on
Zechariah.1*5

SUMMARY AND SYNTHESIS OF PATRISTIC FINDINGS

Despite their attendant issues, the patristic testimonies provide a
historical understanding of why ούδέ ό υιός was omitted or added in
Matthew 24.36. Appendix A orders the fathers chronologically by the
dates of their work, their provenance, whether or not they were aware
of ουδέ ό υιός or its absence in Matthew, and whether or not they had a
theological problem with the idea of an ignorant Christ before and after
his resurrection.
By the time Athanasius comes onto the scene around 350 CE, not only
do we have unambiguous problems with an ignorance of Christ addressed
in patristic writings, but many of the fathers, if not all, seem to explain the
inclusion of ουδέ ό υιός as indicative of a pre-resurrection ignorance.144
At this same time, evidence of the omission begins to appear in their ref­
erences to Matthew 24.36. There exists no incontrovertible evidence that
the omission existed in the fathers’ writings about Matthew 24.36 prior
to about 350 CE, but there are hints of it in Ambrose.145 Additionally,

142. Didymus, De Trinitate 920.

143. Ibid., 916-17. His Mark reads, Περί δε τής ή μέρας εκείνης, ή τής ώρας, ούδεις
οιδεν, ούδέ οι άγγελοι οί εν ούρανω, ουδέ ό Υιός, εί μή δ Πατήρ. His Matthew
reads, Περί δε τής ήμέρας εκείνης, ή τής ώρας, ούδεις οιδεν, ούδέ οί άγγελοι
των ούρανών, εί μή ό Πατήρ μόνος. The other quotation of Matthew comes from
Didymus, Commentarii in Zacchariam 5.78: Περί τής ώρας καί τής ήμέρας
εκείνης ούδε'ις οιδεν, ούτε οί άγγελοι ούτε ό Υιός, εί μή ό Πατήρ μόνος.
144. Did Greek philosophy inform the fathers’ view of Christ’s omniscience? With
Plato’s theory of recollection, knowledge is recollected by the soul out of an in­
visible, eternal realm, often through reason (Plato, Meno). Such recollection is a
process, and immediate knowledge is consequently not available. Did the fathers
adopt a process-like view of recollecting knowledge whereby Christ could either
exercise recollection to yield knowledge or intentionally leave it unemployed to
remain in practical ignorance? Such a view has two interpretive strengths. First, it
would shed light on the patristic claim that Christ knew the time of his return in
his divinity but not in his humanity, because knowledge would only be selectively
present for Christ at a conscious level. Second, it would draw together a few seem­
ingly disparate reasons that the fathers choose over and over, namely, that in Jesus’
refusal to perpetually employ his omniscience, he was identifying with humanity
and serving as our model. Although later fathers such as Athanasius and Hilary
are predominantly occupied with rejecting the notion that Christ is ignorant, it
is difficult to determine if they would have had a problem with Christ's practical
ignorance» because such is not their milieu—they are reacting against Arian claims
that Jesus was inherently not omniscient.
145. Ambrose may not have had direct knowledge of ancient exemplars omitting ούδέ ό
ulos*. He and Basil corresponded by letter—Ambrose’s rare Greek knowledge being

Adam G. Messer Chapter 4 / 165

 Patristic 1 neology and Kecension tn Matthew 24.36

the geographic arrangement of their appearance raises questions. By this

time, its absence is the assumed correct reading in some sectors, and no
knowledge of its presence appears to exist in others. This suggests that the
omission existed prior to the Arian controversies of the fourth century,
and our writers understandably gravitated toward those manuscripts that
were lacking the phrase.
We have several lines of geographical evidence to consider. In
Alexandria, Origen, Athanasius, and Didymus inform our discussion,
but our conclusions depend largely on the assumptions we make about
Origen. His Commentary on Matthew (ca. 246 CE), normally noting
where textual variants existed, is silent on this point. In fact, Origen
never alludes to knowing about variants in Matthew 24.36, and it may
be difficult to reconcile his penchant for textual criticism with him not
knowing about a textual problem in this verse, though such a perspec­
tive may assume too much.146 Although Origen mentions variants in his
Commentary on Matthew, which was written from Caesarea, we do not
know whether or not he had Alexandrian exemplars. If he did and if those
exemplars differed, it would be a fair assumption that he would mention
the issue. If we can also assume that Origen either had or could recall to
memory Alexandrian forms of Matthew, then since he does not men­
tion any variants, manuscripts attesting the omission must appear some­
time after Origen (246 CE) but before Athanasius (357 CE). Whenever
it appeared, Didymus (381-98 CE) seems aware of both readings. The
omission thus may have made its way into Egypt sometime during the
century 250-350 CE.147 If Origen did not have and could not remember

an asset to East*West relations. It seems entirely conceivable that Basil, who was
only aware of the omission, informed Ambrose of the state of his exemplars. If true,
this sets limits on understanding Ambrose's modifier “ancient," because the manu­
scripts to which Basil would have referred, if dated too early, would have conflicted
with the manuscripts available to Origen when he wrote his Commentary on Mat­
thew in Caesarea. So “ancient" may mean documents written within 120 years or
less.
146. Origen's memory was apparently voluminous, and he had an eye for manuscript
variation. Had he brought no manuscripts from Caesarea (which is by no means
certain), would he have forgotten how his Alexandrian exemplars, over which he
had pored for many years, read? Origen may not have known about every manu­
script in Alexandria or Caesarea, but some confidence can be garnered by the fact
that had other manuscripts existed and been used, he would likely have encoun­
tered this discrepancy and was unlikely to have forgotten it Furfrier, he had earlier
(220-30 CE) referenced Matt 24.36 in De principiis 1.6, wherein he mentions that
only the Father knows the time (technically he uses “God,” but it is done in a con­
text where Christ is referred to separately).
147. Tlais might help explain the confluence of factors we see within Sinaiticus, namely,
that the correctors seemed to be undecided whether or not to include the phrase
ουδέ ό υιό?. According to die work of Milne and Skeat (1938) as well as Jong-
kind (2007), die first round of corrections happened before the manuscript left
the scriptorium (NA27, in light of the work of Tischendorf and Lake, places the

166 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

exemplars from Alexandria that attested the omission, it is possible that

manuscripts without the phrase existed there from the very beginning.
In Caesarea, Origen had multiple exemplars from which he wrote
his commentary on Matthew, and he attests directly to the presence of
ουδέ δ υιός·. Thus the omission seems to have entered Caesarea after 246
CE and to have obliterated all traces of ουδέ ό υ'ιό?, according to Basil’s
testimony in 370 CE, so we can suggest the omission becoming stan­
dard between 275 and 325 CE.148 This concurrence between Alexandria
and Caesarea is not all that surprising, since they are very near to each
other relative to the greater Roman Empire (less than four hundred
miles) and, being major cities, would have had many travelers between
them, particularly given that Alexandria was the most important center
for' trade in the empire.149
In France, both Irenaeus and Hilary were quoting (in ca. 185 and
359 CE, respectively) from a version containing the phrase. In Greece,
Epiphanius knew of both readings (377 CE). Pseudo-Clement, thought
to be of third- or fourth-century Rome, seemed to attest to ουδέ ό υιός.
Ambrose, twenty years after Hilary, said his Italian manuscripts were
without die phrase as were the ancient Greek manuscripts, yet he was
aware of those that still had the phrase (379 CE). If we take Ambrose’s
statements to mean that ancient Greek manuscripts in Italy attested the
omission, we have an omission existing at least by the late third century,
if “ancient” is to mean anything at all.150 How do we explain the ap­
pearance of the omission in distant geographical precincts (Alexandria,
Caesarea, and Italy) by the late third century? One of the following
might have happened:

first correction in either the fourth or sixth century). This initial indecision by the
scribes of Sinaiticus is consistent with our theory of the coexistence of exemplars
with and those without the phrase. It would also explain the oddity of a stricken
phrase that was readded: the readdition (by erasure of the diacritical dots) could
have happened later. Due to scribal evidence indicating time spent in Caesarea
------ (earlier noted by Lake), Milne and Skeat have asserted a Caesarean origin for
Sinaiticus (Milne and Skeat, Scribes and Correctors of the Codex Sinaiticus, 69). If
this is true, the manuscripts present within Alexandria at the time of the manu­
script’s creation would obviously be irrelevant. However, Jongkind undermines the
Caesarean hypothesis (Jongkind, Scribal Habits of Codex Sinaiticus, 252-54).
148. Of course, Caesarea was large, and there may have been multiple scriptoria. If so,
manuscripts attesting the omission may have existed alongside manuscripts at­
testing otherwise.
149. Cornelius Clifford, "Saint Athanasius," in TCE, vol. 2 (New York: Encyclopedia,
1913), 35-40.
150. Although Ambrose was qualified to speak to the state of tine Greek texts, he was
also in correspondence with Basil, who may have told him of the omission in
Caesarea. This would have been a relevant topic of discussion during the Arian
controversies.

Adam G. Messer Chapter 4 / 167

 Patristic Theology and Recension in Matthew 24.36

1. An unrecorded edict influenced the orthodox to alter the man­

uscripts all in the same way.

2. Scribes uniformly, but randomly, struck the phrase across the

empire.

3. The omission truly happened much earlier and naturally dis­

seminated over the decades, a fact not advertised until it was conve­
nient during the heresies of later centuries.

There are problems with each option. The first suffers from a lack
of any direct evidence. Further, why would the proposed ecclesiastical
edict not address Mark’s parallel phrase? Even worse, any such “edict”
during the fourth century could hardly have escaped attention, and this
strategy runs across the grain of the interpretive strategies employed by
the fathers.
Option 2 is not without issues either. It is not difficult to conjecture
that a scribe, fearing tire potential of misinterpretation, altered die text
to relieve this tension. But why would a scribe remove the phrase ou8e
ό υιός and leave die word μόνος intact? Would this truly alleviate the
problem? Further, why would such a scribe remove the problem from
Matthew and forget to remove it in Mark? Would it not have been easier
to simply remove μόνος to make the Matthean account match Mark?
There are several reasons why tiiis would have been better:

a. Any hermeneutical explanation found for Mark could likewise

be applied to Matthew.

b. A scribe would not have to remove a divine, personal reference.

c. A scribe could have removed one word rather than three.

d. There would consequently be no need to find an explanation for

the exclusive μόνος still present.151

Given diat there were better or at least multiple ways to alter the
text, it is oddly peculiar that when a variant occurs, it always occurs
in the same way in this verse—by removing ούδέ ό υιός. None of the

151. To believe otherwise would require that the scribe either did not believe Mark to
be a problem or had no opportunity to change Mark. Scribes sometimes cooper'
ated to finish manuscripts, so it is possible that the offending scribe was not able
to also change Mark. This possibility lessens as we move earlier into history, for
as Christianity became more organized, manuscript production was sometimes
collaborative. Another possibility is that, since Matthew was used and therefore
copied more often, the goal may just have been to transcribe Matthew.

163 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

other options for alleviating the potential confusion were exploited.

These factors make highly unlikely the theory that scribes in different
sections of the empire who were having problems with the phrase ran­
domly removed it from their respective manuscripts, for one would
have expected to see other solutions employed to assuage the theolog­
ical dilemma.
With the challenge to option 2 of how to get the omission across
the empire, option 3 seems better.152 Whether or not the change was
intentional, this dissemination would take some time. An early third-
century or prior archetypal omission is the best time frame to allow
its proliferation to foreign lands. If it is in reaction to adoptionistic
influences, we might even tentatively suggest the time frame of 190-
225 CE. If it is in reaction to modalistic influences (a point taken up
shortly), we might likewise suggest a late second-century alteration in
the West.

EVALUATION OF EHRMAN’S METHODOLOGY

This chapter has been an assessment of Ehrman’s thesis applied to
one concrete example (Matt 24.36), but our findings revolve around
two broader critiques that apply to his work as a whole. With the evi­
dence now available to us of a great diversity of points of view about
Christ in the first four centuries, it appears that there is always a reason
to postulate an orthodox corruption. As succulent as this is to our
critical taste, an unqualified application of this approach should leave
some distaste.
There are several general problems with the type of approach that
maps every variant to a theologically motivated orthodox alteration. Not
only does it fail to give ample respect to the level of reverence most scribes
within the tradition gave to the words being copied, but this approach
also fails to take into account the scribal habits of individual scribes.153
Ehrman has asserted that the change in Matthew 24.36 represents
an anti-adoptionistic alteration at least as early as the late second cen­
tury He ascertains this time frame from the idea that the Diatessaron

152. This follows unless we want to posit a champion ofthe shorter reading to speed the
dissemination, but this is unwarranted for any historical personality prior to Atha­
nasius in the fourth century. Athanasius himself may be a good candidate (Brogan,
“Another Look at Codex Sinaiticus,” 20).
153. One example is Ehrman’s postulate that a scribe purposely added an article to an
ancestor of Codex Regius, an eighth-century manuscript, in John 1.1c. As Morgan
concludes, die most probable reason for the presence of this "SabeUian” article
resides in the sporadic habits of the scribe, not in any theological agenda (Matthew
P. Morgan, "The Legacy of a Letter: Sabellianism or Scribal Blunder in John 1.1c?”
in this volume).

Adam G. Messer Chapter 4 / 169

 Patristic Theology and Recension in Matthew 24.36

and Origen testify to the omission of ουδέ δ υιός, as do a range of ver­

sions! witnesses.154 Several comments are in order.
There are a number of different versions of the Diatessaron (e.g.,
Syriac, Persian, Latin, Armenian, Arabic, Middle Dutch, etc.), with a
complex interrelationship.155 There are three reasons to conclude that
the Diatessaron does, contrary to Ehrman’s claim, attest to the phrase
ουδέ δ υιός.

1. Several, if not all, versions attest it: the Armenian version of

Saint Ephrem’s commentary, the Chester Beatty Syriac manuscript
709,156 and at least one strand of the Arabic version.157158

2. Peterson outlines a process by which one may settle on an

original Diatessaronic reading.188 Because both readings in Matthew
24.36 have attestation in Greek and Latin texts, which could have in­
fluenced the transmission of the Diatessaronic text, one could never, in
Peterson’s estimation, have full confidence that the omission is attested
in the Diatessaron.

3. Tatian’s Diatessaron was a harmony, and it was carefully crafted

from all four Gospels.159 It is unlikely that Tatian would have missed
ουδέ δ υΙός in Mark.160

Given these three factors, it is unlikely that the Diatessaron attests to

the omission.
What about the versional witnesses? The Latin manuscripts itgl
(eighth/ninth century) and it1 (thirteenth century) and the Vulgate
were born in the aftermath of the Arian controversies of the fourth
century. In fact, with the exception of the Sahidic Coptic version and
perhaps the Syriac Sinaitic version (third/fourth century), the Middle
Egyptian Coptic (fourth/fifth century), Bohairic (fourth century), pop­
ular Peshitta (fourth/fifth century), and Harclean (seventh century)

154. Ehrman, Orthodox Corruption, 92.

155. CirmelMcCartbtySaintEphrem'sCommentaryon Tatian’s Diatessaron, Journal of
Semitic Studies Supplement 2 (Oxford: Oxford University Press, 1993), 4, 8-9.
156. Ibid., 278-79.
157. ANF 10:35,108-9. This translates Vatican Arabic MS 14. It is also referenced by
the apparently nonstandardized label "Section 42.”
158. William Lawrence Petersen, Tatian’s Diatessaron: Jts Creation, Dissemination, Sig­
nificance, and History in Scholarship (Leiden: Brill, 1994), 373-74.
159. McCarthy, Saint Ephrem’s Commentary on Tatian’s Diatessaron, 7.
160. One argument for this possibility might be construed from Tatian's possible partic­
ipation in Encratism, which is believed to have affected the readings of the Diates­
saron in thirteen places. However, the Encratites were primarily known for beliefs
that would not have motivated a change in Matt 24.36, namely, strict asceticism
and sexual continence (Hultgren and Haggmark, Earliest Christian Heretics, 123).

170 ! Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

versions find their origin in the Arian controversies of the fourth cen­
tury or later, including the Georgian text (fourth century).161 Further,
they are not monolithic testimonies but have dissenting readings. For
example, the Palestinian Syriac version (sixth century), the Diatessaron,
various Georgian manuscripts, and a number of Old Latin and Vulgate
manuscripts contain ουδέ ό υιός·. Although our earliest manuscripts
of the Sahidic Coptic version are from the early fourth century, it may
originate in the late third century, and it does not offer mixed testimony.
Although we would not expect the Coptic tradition to be any more im­
mune from theological alteration than the Greek one,162 this may en­
able us to push the shorter reading back to the period just prior to the
fourth century.163 Nevertheless, the implication of this information is
that, contrary to the opinion of Ehrman, there is no clear reason to push
the origin of the omission back to the late second century based on the
versional testimony.
All references to Matthew 24.36 listed in BP were researched, in­
cluding all those in Origen’s writings. Contrary to Ehrman’s assertion,
no instances were found where Origen was aware of manuscripts at­
testing to the omission.164 There is, however, indirect evidence in­
dicating so. Jerome states in his Commentary on Matthew, "In some
Latin manuscripts is added: ‘nor the Son,' though in the Greek copies,
and especially those of Adamantius and of Pierius, this addition is not

161. Our earliest Georgian evidence supporting the omission comes from a tenth-cen­
tury revision of the Georgian version (UBS’, 28’-29‘, with dates in the insert). Ap­
parently, other manuscripts from this same revision differ at this point from those
that attest to the omission.
162. Indeed, it introduces the possibility of unintentional orthographic alterations,
one of which might occur by skipping from one Sahidic β to the next (or even
from ιιγβ to Hpe). An intriguing possibility revolves around an apparent lacuna
in Matt 24.36 surrounding the word eiMHT (“alone”). Although this observation
needs verification by a good Sahidic apparatus, perhaps the Sahidic synonym
ογα/τ (“alone”) was used instead. If so, a pregnant possibility would involve skip-
ping from one SOY to the ηκχί:ΉΜΠΜΥ©θγτ®πφΗρβογωττΊ6ΐωτ. These ex­
amples would cleanly skip over ογτθπφϋρ (“nor the Son’’)- The text used as the
Sahidic exemplar is taken from Rudolph Kasser, Papyrus Bodmer XIX: Evangile de
Mattieu XIV, 28-XXV11I, 20 Epitre aux Romains 1,1-1L 3 en Sahidique (Geneva:
Bibliotheca Bodmeriana, 1962). These possibilities require additional erudition to
determine their validity, but they demonstrate a field ripe for study.
163. There are other Greek manuscripts that support the omission: L (019) from the
eighth century: W (032) from the fourth/fifth century; f1 (Family 1), whose ear­
liest manuscript (MS 1582) dates from the tenth century (949 CE); and 33 from
the ninth century. None of these help establish a second-century entrance of the
omission.
164. Neither NA2’, UBS’, Tischendorf·, Legg, von Soden, Swanson, Lachmann, Scriv­
ener, or Merk listed Origen in support of the shorter reading. Likewise, in his
work dedicated to this topic, Metzger does not mention Origen’s awareness of this
variant (Metzger, “References in Origen to Variant Readings," 88-103).

Adam G. Messer Chapter 4 / 171

 Patristic Theology and Recension in Matthew 24.36

found.”165 According to Jerome, Origen knew of manuscripts lacking the

phrase,166 but the fact that Jerome implies that all (or at least the ma­
jority) of Origen’s manuscripts were missing the phrase seems, in. Light
of modern evidence to the contrary, to leave a bewildering number of
possibilities in its wake. Was Jerome unaware of the texts of Origen we
currently have? Was he mistaken in mentioning Origen? Did the in­
tervening church add the phrase into Origen’s Greek manuscripts as
they were transmitted?167168 Jerome’s mention of two witnesses adds cre­
dence to his claim.158 However, Jerome’s writings were no less immune
from ecclesiastical tampering than the biblical manuscripts, and a stra­
tegic interpolation recalling that Origen’s manuscripts were lacking the
phrase would have strengthened a claim for an original shorter reading.
If Origen did know about the omission, as Jerome has suggested,
it would mean that the shorter reading was attested by the late second
century. Although few would claim that our manuscript evidence is
comprehensive, it seems strange that Origen does not mention man­
uscripts attesting the absence of the phrase in his Commentary on
Matthew. This perplexity warrants the exercise of caution when sug­
gesting a second-century origination of the omission on the basis of
Origen’s testimony. Since the case is not as settled as Ehrman makes it
appear, one is left with a temporally unanchored variant that cannot be
established apart from other lines of evidence.
In Orthodox Corruption, Ehrman slices the early period into dif­
fering heresies and then puts a number of Christologically significant
NT variants in juxtaposition to the theology of the heresies. Of course,
most variants in Christologically significant verses can be construed to
favor some heresy, so it is fairly easy to find several good fits. He then
concludes that since the proto-orthodox would have benefited by al­
tering their manuscripts to combat that heresy, they therefore did. He
estimates that there are five or six dozen examples.169
Even if Ehrman’s assertions about Christologically significant verses
were correct, we do not have manuscript evidence from the first three

165. Jerome, Commentary on Matthew 4.24.36; translation from Saint Jerome, Com­
mentary on Matthew, trans. Thomas P, Scheclc, Fathers of the Church 117 (Wash­
ington, DC: Catholic University of America Press, 2008), 277.
166. Adamantius is Origen—to be distinguished from a later Adamantius of Alexandria
(fourth/fifth century).
167. As earlier mentioned, an intentional alteration of this sort is not good, since there
would not have been motivation to do so after the height of the Arian contro­
versies. A later harmonization to the text of Mark, however, still remains a viable
option,
168. Pierius was contemporaneous with the bishopric of Theonas (who ruled between
283 and 301) and apparently died in Rome sometime after 309 CE (Michael Ott,
"Pierius," in TCE, vol. 12 [New York; Encyclopedia, 1913], 79).
169. Ehrman, Orthodox Corruption, 281.

172 / Chapter 4 AdamG. Messer

 An Evaluation of Ehrman's Text-Critical Methodology

centuries to prove them—a fact he plainly acknowledges.170 Apart from

the patristic testimony (itself subject to tampering), it is very difficult to
date at what point these "corruptions” materialized. Thus, unless we find
definitive discussion by the fathers on the matter, Ehrman’s examples
each could have transpired at any time during those first three centu­
ries. Ihe significance of this feet is that during most of this time, oppo­
site heresies were struggling against the proto-orthodox. So Ehrman’s
theory is a double-edged sword. For every alleged orthodox motivation
that existed, an equally possible heretical motivation existed, sometimes
in conjunction with a less reverent attitude toward the text. With no
clear rubric for preferring orthodox or heretical tampering, Ehrman’s
book may just as easily have been entitled The Unorthodox Corruption
ofScripture.

THE LIKELIHOOD OF THEOLOGICAL MOTIVATION

Ehrman has used Matthew 24.36 as a key dement in most of his dis­
cussions of orthodox corruption; however, he has to make three basic
assumptions in order for orthodox corruption to stand:

1. Ούδέ ό υιός must originally exist in Matthew.

2. The alteration must be more likely to have been performed by

orthodox scribes than by heretical ones.

3. No accidental orthographic options must present themselves.

If any of these assumptions are upset, so is his claim about Matthew

24.36.1 here offer comment on the first two assumptions. As for the
third, although room disallows a perusal, several possibilities do present
themselves among the exemplars from several languages that, despite
traditional reservations, have potential for mutual interaction in the
early period (e.g., Greek, Aramaic, and Coptic). Nevertheless, their im­
probability campaigns against them.
Whatever was the case for the Matthean manuscripts of the fathers,
their Markan reading was fixed.171 If the orthodox responded to adop-
tionistic or Arian use of Scripture by changing the text for Matthew
24.36, not only would we expect issues with Mark,172 but we would

17a Ibid., 28.

171. Wallace notes that Ehrman neglects to share this counterbalancing information in
Misquoting Jesus and Orthodox Corruption (Wallace, “Gospel according to Bart,"
343).
172. Some believe that the Arians primarily used Mark 1,3.32 to support their con­
tention (Kevin Madigan, “Christus Nesciensl Was Christ Ignorant of the Day of
Judgment? Arian and Orthodox Interpretation of Mark 13:32 in the Ancient Latin

Adam G. Messer Chapter 4 / 173

 Patristic Theology and Recension in Matthew 24.36

anticipate similar alterations to places where Jesus could be construed

as limited in body or mind.173 Such verses include Matthew 26.39 (“let
ttus cup pass from me"), Mark 10.18 (*no one is good except for God
alone”)?74 Mark 15.34 (“my God, my God, why have you forsaken me?”),

West," HTR 96, no. 3 (2003): 258; Hanson, Search for Che Christian Doctrine of
God, 107). Since Matthew was read more often and contained the phrase ουδέ
ό vlos, the presence of μόνο$ would have made Arian use of Matthew more ad­
vantageous, but they may have used Mark due to mixed manuscript evidence in
Matthew. However tidy this argument might be for our purposes, Hanson offers no
corroborating evidence to support the assertion; in fact, when he makes the claim
that Mark was predominantly used, he quotes a verse containing μόνο?. Such an
observation introduces the suspicion that either proper discernment was not exer­
cised in making this attribution or such attribution was irrelevant to his purposes.
Madigan probably relied on Hanson’s claim.
173. Interestingly, in Ehrman’s anti-adoptionistic section, he spends the majority of his
space on alterations that made Christ’s deity explicit (Matt 3.3; Marie 1.3; Luke
3.5; John 1.18; 1033; 12.41; Acts 20.28; 1 Cor 10.5; Gal 2.20; 1 Tim 3.16; 1 Pet
5.1; 1 John 3.23; etc.). This angle does not truly bear the force of his argument,
because many of these texts already implied what was made explicit Further, the
motivation behind these changes may be attributed to simple devotion or Docetic
alterations. More pertinent would be verses that present problems for Christ's di­
vinity. His applicable list is much shorter—he mentions only eight. Three deal with
clarifying that Jesus was not merely a man (John 19.5; 1 Cor 15.47; Cot 1.22). AQ
three examples suffer from problems unmentioned in Orthodox Corruption· In
John 195, an article before "man'’ is removed in Vaticanus, and the resulting trans­
lation seems to support an anti-adoptionistic meaning, but no mention is made of
another omitted article before another substantive in the same verse—such that
an investigation of scribal habits/tendencies is warranted prior to drawing conclu­
sions, for these habits may offer greater explanatory power than theological moti­
vation. Further, the translation need not carry the angle Ehrman asserts, nor does
the alteration truly alleviate the difficulty in seeing Jesus as a man. In 1 Cor 15.47,
the interpolations are as easily devotional as reactionary; in fact, their intents ail
seem to clarify the referent (the Lord Jesus) or to express reverence, rather than to
destress his humanity (ίνθρωττο^ is not excised). Ehrman asserts that the removal
of the possessive pronoun αύτου from Col 1.22 in certain manuscripts changes the
meaning in such a way as to disassociate Jesus with a negative Pauline connotation
of σάρξ ("flesh”). While such a meaning is possible when the pronoun is removed,
the context (particularly v 20) necessitates that the referent be Jesus' body, making
αύτοΰ unnecessary and redundancy a greater motivation for intention^ change.
The article (in this case before σαρκός) sometimes signifies that an idea of posses­
sion is present, particularly when human anatomy is in view (Daniel B. Wallace,
Greek Grammar beyond the Basics: An Exegetical Syntax of the New Testament
(Grand Rapids: Zondervan, 1996], 215). Of the eight aforementioned examples,
three others represent clarification against a misconstruaJ of the text, which is
reason enough for die change, without positing any antiheretical influence (Heb
3, 2.18,10.29). The final two seem most relevant. One is Luke 2.40, which may
1.
just represent a harmonization to Luke 1.80, and the other is the primary text of
this chapter—Matt 24.36.
174. This verse's parallel in Matt 19.16-17 underwent scribal harmonization by or­
thodox scribes to match Mark and Luke. Wallace discusses an implication of this
feet, namely, that since the resulting Matthean text posed greater difficulty to an
orthodox Christology than Matthew's original, this example demonstrates that

174 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

Lul<e 2,52 (“Jesus increased in wisdom and maturity”), John 12.27

(“Father, save me from this hour”), John 14.28 (“the Father greater ls

than I”), John 17.3 (“in order that they might know You, the only true
God, and Jesus Christ whom You sent”), and John 20.17 (“my God and
your God”). No Christologicaliy significant variants for these verses
exist within our manuscript record prior to the fifth century.*175
We do see attempts by church fathers to deal with problematic pas­
sages through a hermeneutical grid; that is, they tended to clarify their
theology rather than change their texts. We see this in Hilary’s argu­
ment for a preeminent understanding of el μή, in Athanasius' argument
that Christ was exemplifying true humanity for us, and in Basil’s argu­
ment that Christ had already asserted that all things that were of the
Father were also of the Son.176 In fact, all the fathers were found to argue
— from the greater canonical context of Scripture. This raises a pertinent
question: if the leaders of the churches tended to deal with apparent
problems in Scripture using interpretive means, why would we expect
Christian scribes, by and large, to do otherwise?
A related observation surfaces. The idea of Jesus as “the Concealer”
of the mystery appears in Hippolytus. Slightly later, Origen’s assorted
interpretations of the Son’s ignorance initiates (or continues the earlier

harmonization could take precedence over a high ChristoLogy as a motivation for

textual emendation (Daniel B. Wallace, *Lost in Transmission: How Badly Did the
Scribes Corrupt the New Testament Mill Text?," in this volume). His observation vali­
dates by illustration what we might similarly deduce from a statistical comparison
between the frequency of scribal harmonizations and the number of alterations
resulting from an orthodox Christological bias, The significantly greater frequency
of the former supports Wallace’s contention that “narrative harmonization was a
stronger impetus [for alteration] than a high Christology” (ibid., 52). By way of
corollary, one should generally prefer the explanatory power of harmonization
over a high Christology (all else being equal), in light of the relative prevalence of
harmonization—a prevalence Ehrman describes with pizzaz by judging this scribal
tendency to be “ubiquitous” (MisquotingJesus, 97).
175. Two variants from NAW (post-fourth century) may be construed as Christologi­
caliy significant. In John 20.17 (“your God and my God”), two fifth-century manu­
scripts and a ninth-century manuscript remove "and my God" from Jesus’ words,
in Mark 15.34 (“forsaken” [έγκαταλβίπω]) is found as “reviled” (όνβιδίζω) in sev­
eral manuscripts—none prior to the fifth century. Ehrman interprets the latter as
an anti-separatist corruption (Orthodox Corruption, 143-45) and would probably
construe the former as anti-ado ptionistic. He also views anything found in later
manuscripts as primarily reflective of changes made during the first three centu­
ries (ibid., 28), but as 1 will later argue, this assumption needs to be revisited.
176. O’Keefe and Reno make the case that the fathers of this period, specifically Atha­
nasius among others, knew that many scriptural passages presented problems to
any theological system, but they strove for “a reading that maximizes the number
of unstrained interpretations of individual words, verses, and episodes” (John J.
O’Keefe and R. R. Reno, Sanctified Vision: An Introduction to Early Christian Inter­
pretation of the Bible [Baltimore: Johns Hopkins University Press, 2005], 61); that
is, they did not reword, they retheologized.

Adam G. Messer Chapter 4 / 175

 Patristic Theology and Recension in Matthew 24.36

initiation of) an exegetical tradition from which later authors would

draw and on which they would expand. This growing body of interpre­
tive commentary, ever widening in scope and geographical prolifera­
tion, would have served to diminish the motivation to omit ουδέ ό υιός·
177 Instead, Christians would have drawn on the her­
in Matthew 24.36.*176
meneutical tradition to understand the verse and dismantle heretical
misinterpretation. These considerations would tend to argue for an
origination of the variant in the centuries prior to an established ex­
egetical tradition surrounding Matthew 24.36.
The evidence from the fourth century suggests an earlier time frame
for the removal of ουδέ ό υιός. This would leave two good possibili­
ties for corruption: (1) the orthodox removed the phrase to combat
Adoptionism, or (2) Modalists removed the phrase to combat ortho­
doxy. Which of these groups more likely did it? To answer that, we must
assess which group would benefit more (strength of motive) and which
group’s theology would be better served by the change (usefulness of
strategy).
The orthodox would have had some reason to remove the phrase,
since Adoptionism was plaguing the scene from 190 to 275 CE. Yet
Modalism also affected the proto-orthodox from as early as 70 to 220
CE and even until Nicea. If the orthodox omitted the phrase to counter
diminutive Christologies, the change would have aided the position
of those on the opposite side of the spectrum. This dilemma would
have had a mitigating influence on any willingness to make alterations.
Modalists, in comparison, would only have benefited from the omission
against any opposing theology.
Might the actual alteration itself tell us anything about its em­
ployer? A Modalist who came to Matthew 24.36 would find one clear
way to alter it for benefit—he would remove ουδέ ό υιός. An orthodox
scribe would come to the same verse, notice the stark contrast between
the Father and Son devised by the presence of μόνος, and find several
options advantageous:

1. Strike ουδέ δ υιός.

2. Strike μόνος.

3. Strike both ουδέ ό υιός and μόνος.

4. Strike the whole verse.

5. Try any number of additive options.

177. My appreciation goes to Michael Svigel for this observation.

· V · V / Chapter
176 · ·♦ 4 Adam G, Messer
 An Evaluation of Ehrman’s Text-Critical Methodology

Option 1 has the disadvantage of catering to Sabellian claims. Options

3 and 4 would suffice but have the disadvantage of striking more words
than would be necessary. Option 5 is fascinating but probably requires too
much planning and pretension.178 Option 2 would be the most strategic,
because it would guard against Modalism, enact the least possible violence
to the text, have a greater chance of going undetected (since comparing to
Mark would raise no flags), prevent the deletion of a divine reference, and
allow whatever interpretive explanation existed for Mark to apply equally
to Matthew. Even more significant, if scribes were having problems with
this idea of the Son’s ignorance, one would expect manuscripts to appear
with some variation of the preceding possibilities. They do not.
Contemplation of this fact alone suggests greater significance than
might at first appear. If ουδέ ό ulds is original and if orthodox scribes
had problems with this phrase, why did every other remedial alternative
remain untapped? Why do they appear to have chosen and never devi­
atedfrom and option that (1) was not the most strategic,179 (2) left open
the same conflict they were trying to resolve,180 and (3) happens to be a
model example of what an early harmonization to Mark would look like?
Without alternative manuscript scenarios, we seem left with a vacuity.
Perhaps a solitary scribe omitted this phrase so early that its propaga­
tion was the introduction of the gospel in various other regions, a situa­
tion becoming increasingly impossible to evidentially differentiate from
an aboriginal lack of the phrase. Or perhaps we have to devise a way for
the omission to propagate against the exemplars of the known world.
This latter scenario would require an interesting paradox: it would show
a simultaneous use and eschewal of the change process. In other words,
the deviation from their exemplars in transcription would reflect a will­
ingness to change, an inclination that is difficult to reconcile with their
dogged refusal to alter their text in other strategic ways. Either we adopt
this somewhat convoluted theory of the variant’s propagation,181 or we

178. Ehrman concluded that scribes were more likely to correct than to interpolate no­
tions previously nonexistent (Orthodox Corruption, 277).
i79?Asalfeady mentioned, since orthodox alteration against one heresy would benefit
another heresy on die opposite side of the theological spectrum, it would often
not be strategic for the orthodox to make any Christologically significant changes.
Further, if the orthodox were inclined to omit text in order to combat a diminutive
Christology, two different alterations would have proven superior to the omission
of ουδέ όυίός: (1) the removal ofjust μόνος and (2) tire removal of both μ,όνος and
1 cs X C t /
ovoeovio?.
180. Basil’s interpretation of el μή in a preeminent sense was not employed until a cen­
tury after the alleged omission. Prior to this, hermeneutical factors that allowed ci
μή to mitigate the inherent contrast presented by μόνος did not exist, so a contrast
would have naturally existed between the Father and the speaker (the Son) whether
or not our phrase was present.
181. It is patently obvious that history can be, at times, very convoluted. However,
taking our cue from Ockham, we will at least look for a simpler explanation, if not

Adam G. Messer Chapter 4 / 177

 Patristic Theology and Recension in Matthew 24.36

are driven back to the possibility that Matthew never contained ουδέ ό
υιός in the first place.
The absence of ούδέ ό υιός in the original would easily account for
what we see in the manuscript tradition. Instead of postulating a com­
plex social scenario for how the omission was transported around the
empire, we would expect random scribes in random places to frequently
harmonize this verse to Mark, all in exactly the same way and without
collusion. This alleviates any need for a mechanism of transport. It
would also offer a reason for why the fathers almost constantly mix as­
pects of the two in their quotations of Matthew 24.36. It would explain
why the phrase is missing from some early versional evidence, such as
the Sahidic Coptic and the Sinaitic Syriac versions (fourth century).182 It
may do justice to Ambrose’s intent in using the word "ancient” {veteres)
of manuscripts in 380 CE.183
What we do with Origen is key in this patristic discussion. He is
crucial because he represents the earliest unambiguous evidence for
Greek manuscripts attesting ουδέ δ υιός, but there are multiple reasons
to doubt his testimony in both Caesarea and Alexandria. Did Origen
have manuscripts available to him in Caesarea from Alexandria? Were
the multiple exemplars in Caesarea even complete? Did he feel com­
pelled to mention every variant or just variants for those texts that he
could find a parenetic purpose? Did he, as Jerome states, attest to the
shorter reading in works not extant? The relevant Greek text is missing
from his Commentary on Matthew for Matthew 24.36, which is extant
only in Latin. In short, there are too many unknowns to be dogmatic
about this lone stronghold for the attestation of the ουδέ δ υίός. When
we get to a time period where we find multiple testimonies about
Matthew 24.36 against which we can compare, we find both variants
attested.
As mentioned earlier, if Matthew originally did not have ουδέ δ
υιός, it would require comment on two resulting problems. First, such
a position requires us to ignore our earliest and best witnesses. Second,
scribal harmonization of Mark to Matthew was much more likely to
occur than vice versa. In response to the first problem, I offer these
insights:

prefer it. This is precisely what Ehrman does when he argues for theological moti­
vation in Matt 24.36.
182. The Middle Egyptian (Coptic), Bohairic (Coptic), Peshitta (Syriac), Harclean
(Syriac), and Georgian versions also attest to the omission, but all these versions or
at least their manuscript evidence find their origin in the Arian controversies of the
fourth century or later (dates from. Metzger, Early Versions ofthe New Testament} >
so they are far less significant in impact.
183. The Latin word veins has a range of meanings from "old* to "ancient? but for Am­
brose to appeal to these manuscripts, they must have been of an age worthy of
referral.

178 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

1. We must acknowledge that our manuscript testimony is some­

what limited from the first few centuries of the church.

2. This type of variant would be of the type of alteration that

would readily require almost no time to occur, particularly given the
less rigorous methods of transcription known to have dominated the
early centuries and in conjunction with early efforts at harmonizing all
the Gospels into one (e.g., Tatian’s Diatessaron in ca. 170 CE). In fact,
no manuscript evidence exists for this text for three centuries, and even
then, both of our earliest manuscripts (Sinaiticus and Vaticanus) may be
from the same provenance—a provenance that was, based on the cor­
rections of Sinaiticus, itself struggling with the text of Matthew 24.36.
Finally, Jerome was also a textual critic; perhaps the decision he made in
the Vulgate at this juncture reflects more than just an anti-Arian bias.

In response to the second problem, pertaining to the likelihood of

the direction of the harmonization, the only consideration really posing
any form of issue is exemplified by the statistics calculated by Powell.
He figured that for 70 percent of the instances when there was a har­
monization between Matthew and Mark, Mark was harmonizing to
Matthew.184 It is here where statistics can easily lead us astray. In this
case, they only reflect what an author was likely to do, not what any
given text is likely to be; that is, they describe rather than prescribe, and
to make it otherwise fails to take into account the character of the indi­
vidual instances. In feet, it seems commandingly sufficient to know that
there were any cases where Matthew was harmonized to Mark, because
this instance alone is enough to conjecture it. For the specific instance
of Matthew 24.36, the proper perspective of the fact that it did happen
with some frequency should be not that the statistics make the chances
of harmonizing Matthew to Mark unlikely but that such harmoniza­
tions set a precedent within which our instance fits nicely.185

...... ' CONCLUSION

This chapter has addressed two main issues. First, did the theology
of the fathers and their practices make a theological motivation likely in
Matthew 24.36? If so, who most likely did it and in which time period?
*
184. Powell, Textual Problem of Ουδέ Ό Tios5
185. Powells argument that Matthew never harmonized to Mark in places where Mark
contained Christo-diminutive verses is specious, because it is unfair to equate the
situations. All of his examples were instances were Matthew did not even contain
the verse. It is much less likely for long segments of text (an entire verse) to be har­
monized from memory than for harmonization to occur with a short phrase that is
in the same verse and made memorable by its alliteration in Mark (ούδέ ... ουδέ).

Adam G. Messer Chapter 4.7 179

 Patristic Theology and Recension in Matthew 24.36

If not, how do we account for the omission? Second, does the testimony
of the fathers have anything to say to the text-critical problem?
It seems likely that if tire omission is tire result of a theologically
motivated change, it was most probably removed by Modalists, who
would have had no conflict of interest in doing so and whose theology
would strategically have warranted precisely this change—as opposed
to the mixed results that the employment of such a strategy would have
caused for tire orthodox. Given the paucity of early manuscript evi­
dence and the dearth of incontestable patristic witnesses, a time frame
is currently too difficult to resolve with certainty, but a late second-
century omission is as close as we can surmise. The omission might be
best traced to the West, perhaps to Rome, where Praxean and Sabellian
schools were in force at this time.
While a theologically motivated alteration is entirely possible, it is
more probable that ουδέ ό υιός never existed in Matthew 24.36, given
the distribution of patristic evidence and the singular form of the al­
leged theologically motivated textual remedy; that is, I feel that the geo­
graphical and temporal distribution of the variants, in both our patristic
writings and the manuscript evidence, along with the manuscripts' sin­
gular manner of alteration, preclude the likelihood that ουδέ ό υιός is
original, a stance commemorating Jerome’s text critical work,186 done
from a vantage point a millennium and a half closer to the original.
Although a late second-century adoptionistic influence may have
motivated the orthodox to remove ουδέ ό υιός as Ehrman posits, the
aforementioned historical alternatives better explain the confluence
of factors in the testimony of the fathers and manuscript evidence.
These alternatives are grounded on a conscientious exploration of the
competing motivations within orthodoxy as well as an analysis of the
geographic and temporal distribution of the variants. In light of these
alternatives, if the omission is either original or the result of heretical
alteration, the exoneration of the orthodox should correspondingly de­
tract from the marketed impression that the orthodox irretrievably cor­
rupted the text.
As the church was emerging during its first centuries with this bur­
geoning, yet not fully realized, notion of the hypostatic union of Christ
and his relation to the Father, the church wavered between two opposing
points of view: denigrating Christ’s identity and confusing him with the
Father. The pendulum began swinging. Perhaps most aboriginal was
confusing him with God without true flesh (Docetism), followed by
confusing him with mere man (Adoptionism), followed thereafter by

186. Jerome asserted that only the original text of the Scriptures were inspired and
without error. As a result, he sought to determine whether there had been altera­
tions. See Louis Saltet, “Saint Jerome” .in TCE, vol. 8 (New York: Encyclopedia,
1913), 341-43.

180 1 Chapter 4 Adam G-Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

confusing him again with God the Father enfleshed (Modalism), fol­
lowed yet later by confusing him with a godlike entity who was not
coequal with the Father (Arianism). To be sure, this oversimplification
ignores the fact that all these incomplete and false theologies coexisted
throughout the period leading up to Nicea, where the Great Council
pushed them to the fringe by embracing all opposing emphases, namely,
declaring Christ’s deity and humanity and distinctness from the Father.
Interlaced throughout were ripe opportunities by both orthodox and
heretic for textual emendation.
I have suggested that Adoptionism was most prevalent from 190 to
275 CE and may have existed from the very beginning. Docetism and
Modalism existed from 70 to 220 CE and Arian influence from 318 to
381 CE (with earlier roots). The significance of the dates corresponding
to the heresies is that with the exception of a few intervening decades,
from the end of the first century until the end of the fourth century
and even after, one heresy or another was attacking the church, often
several simultaneously. Although some of these groups may have found
the removal of ουδέ ό υιός· in Matthew 24.36 to be advantageous against
the orthodox view, others would have been hindered in their lesser
Christologies by its removal. These heresies fell on both sides of the
orthodox fence, each with varying levels of deviation from the truth and
requiring a response by the church to settle the matter in accordance
with the most biblical and historically conscious understanding as pos­
sible. These heresies revolved around the doctrine of Christ, of God,
and of the Holy Spirit for at least the first four centuries.
This means that no matter what textual problem relating to the
central theme and soul of the Bible (i.e., the Trinitarian God) may be
found in the manuscript tradition amid the first centuries—themselves
already characterized by a great diversity of variants due to a lack of
centralization, various persecutions, and a predominantly illiterate
populace—one can always postulate a motivation for an orthodox cor­
ruption, whether or not it is probable. This disingenuous method can
be applied because no matter whether an article is left off or added,
a word'slightly shifted or removed, due to orthographic errors or any
other unintentional type, it often changes the meaning just enough that
there is bound to be a heresy that would benefit from the change. If
an article is missing, it may seem that the unity of the Godhead is in
danger. If the article is present, it may appear to threaten their distinct
personalities. If a phrase exemplifying Jesus’ humanity is removed, it
was obviously to combat the heresy of Adoptionism. If it is added, it was
obviously to combat the heresy of Sabellianism. Realizing that hundreds
and perhaps thousands of variants can be construed to favor the or­
thodox against some heresy, the orthodox can be incriminated with im­
punity. This raises a bemused inquiry: has discrimination moved from
the ethnic and socioeconomic realms into the theological arena?

Adam G. Meeser Chapter 4 / 181

 Patristic Theology and Recension in Matthew 24.36

In each scenario, the orthodox are characterized as underhandedly

altering the text to counter whatever heresy would seem to benefit by
its anterior reading. This unfortunate but convenient state of affairs al­
lows one to posit an orthodox corruption for any given Christologically
significant variant. But when critics capitalize on the prevalence of al­
leged doctrinal alterations to sow distrust in the NT’s transcriptional
fidelity, an important consideration is often overlooked: namely, such a
point of view surreptitiously carries the assumption that these changes,
though intentional, actually corrupted the meaning of the text instead
of clarifying it. Although any change is a deviation from the original,
the difference in a clarification is that it typically better preserves the
meaning and buttresses it against heretical counterfesance.187
Christian orthodoxy, itself given contours by passages of Scripture
in tension, is like a razor. Beaders of Scripture will feel the sharpness of
its edge—the conundrum that this scriptural tension creates. To avoid
the pain of being cut, those predisposed against mystery allow this ten­
sion to force them to one side or the other—an easy theological posi­
tion, but one without the sharpness and vitality of truth. In response to
the heretic’s plea to pick a side or be torn asunder, the orthodox prayer
is, Though I be torn asunder, yet will I trust him.

187. Ehrman acknowledges this (Orthodox Corruption, 276).

182 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

APPENDIX A: PATRISTIC CHART

This appendix contains a chart outlining the basic findings from my

examination of the patristic writings listed in BP. It shows each father’s
approximate lifespan; the works in which he discusses Matthew 24.36;
whether Inis version of Matthew 24.36 attested to the omission or the
presence of ούδέ ό υΙός·, along with an indication of my confidence level
in this judgment; and two columns indicating whether or not he had a
theological problem with Christ’s ignorance regarding the day and hour
of his second coming—-both prior to and after the resurrection.
As for confidence levels, a “low” designation usually means that we
are dealing with an indirect reference. A “medium” designation typi­
cally means that a father’s form of the verse matches two of the three
primary Greek differences (mimicked in Latin) between Matthew and
Mark. A “high” designation usually means that we are dealing with a
direct Greek quotation, and “high-” means that the direct quotation is
in Latin and therefore that some caution is in order.
In the columns associated with ascertaining a father’s comfort
level with an ignorance belonging to Jesus, I use the following four
designators:

Y yes
L likely
P possibly
N no

When an asterisk attends a “possibly” (i.e., “P*”), it means that it

is difficult to determine on the basis of present tense verbs in the sur­
rounding context. For example, we may read a phrase like “of course
the Son of God knows the time and hour he will return.” Such discourse
immediately informs us that the writer in question has a definitive
problem with a post-resurrection ignorance, but it does not necessarily
mean that he has a corresponding problem with a pre-resurrection
lack of-k-nowledge. The chart188 offers boundaries around which we can
quickly survey and summarize early patristic evidence surrounding this
issue, but it also allows the reader to peer into some of the complexities
inherent in this study.

188. The question mark behind Pseudo-Clement of Rome, while explicitly indicating
that we don’t know the provenance for sure, is because, although it self-attests
to being written by Clement of Rome, its nature and dating make this unlikely
if not implausible. For Eusebius’s work, the question marks simply mean that we
don’t know (or at least that I wasn’t able to find) what date he wrote that work. For
Athanasius, the question mark next to the year indicates that some attempt has
been made to verify die year, but it is not real solid. The rest of the question marks
simply mean that we don’t have enough information to fill in the blanks.

Adam G. Messer Chapter 4 / 183

 Patristic Theology and Recension in Matthew 24.36

DATES j RELEVANT WORK(S) |

FATHER
(ca.)

130-200 \ Against Heresies (182-88)

Irenaeus of Lyon
Hippolytus of Rome 180-230 Commentary on Daniel (204)

Tertullian of Carthage 160-220 Against Praxeus (213)

Origen of Alexandria 185-254 Commentary on Matthew (246)

Recognitions (third/fourth
Pseudo-Clement of Rome? N/A
century)
Jjospgl Questions and Solutions to
Eusebius of Caesarea 260-341 ....
Marinus (??)
Discourses against the Arians
(351)
Athanasius of Alexandria 290-373

Disputes against Artus (357?)

On the Council (359)
Hilary of Poitiers (France) 300-368
On the Trinity (360)
Ancoratus (374)
Epiphanius of Salamis 310—403
Panarion (3ΊΤ)
Basil of Caesarea 329-379 Epistula (370-78)

Amphilochius of Iconium 340-400 Fragmtnia (370-80)

Gregory of Nazianzus 325-389 Orationes (379-81)

De fide ad Gratianum (377-80)

Ambrose of Milan 340-397

Exposito Evangelii secundam
Lucam (377-89)
On the Trinity (381-92)
Didymus the Blind of
313-398
Alexandria
Commentaries on Zechariah (398)
Antirrheticus adversus
Apollinarem (385)
Gregory of Nyssa 335-385
Adversus Arium et Sabellium de
patre etfilio

184 / Chapter 4 Adam G. Messer

 An Evaluation of Ehrman’s Text-Critical Methodology

1 I ISSUE WITH IGNORANCE?

READING ✓ ♦ · ·♦ · ♦* *
{OMISSION/ HOW SURE? '
14 PRE POST
PRESENCE)
* * s*
Presence High- N 7?

97
♦· N/A P Y
Presence Low N ??
| Presence High- N Y

?? 79
A4 N 99
e«
'•-ΰ

N/A ............ 99
•♦ 97
r » •*
1

• ♦. ' · ·
I

Omission High L Y

Presence High- Y Y

| Med
Presence P* Y

Omission Low
I High
Omission L Y
9?a
φ N/A P* Y
77
a> N/A P* Y

Both High Y Y

Both High P Y

Omission Low L Y

Adam G. Messer Chapter 4 / 185

 Patristic Theology and Recension in Matthew 24.36

APPENDIX B: ACCIDENTAL
ORTHOGRAPHIC CORRUPTIONS

This appendix outlines ways in which the Greek text may have be­
come unintentionally corrupted from its original form.

POSSIBILITY 1
There is the possibility that die overbar above the nomen sacrum for
“son” (fC) was faded to the degree that a novice or secular scribe may
have taken the text to say,

English: “Concerning that day and hour no one knows, not the angels
in heaven, nor the boar, except the Father.”

Greek: ITEPIAETHCHMEPACEKEINHCKAlfiPACOYAEICOIAENO
ΥΔΕΟίΑΓΓΕΛΟΙΤΩΝΟΥΡΑΝΩΝΟΥΔΕΟΤεΕΙΜΗΟΠΑΤΉΡΜΟΝΟΟ.
I
Note that the Greek text may have been intact enough to alert
the scribe to an obvious (and sacrilegious) error, which he prompdy
removed.189

ΟΥΔΕΙϋΟΙΔΕΝΟΪΔΕΟίΑ
ΓΤΕΛΟΓΓΩΝΟΤΡΑΝΩΝ
MteW^EIMHOnATH
PMONOCQCIIEPrAPAI

POSSIBILITY 2

The Greek text that the fathers used often had slight changes. For
instance, Epiphanius seemed to have a version of Matthew that read,

ΟΥΔΕΙΟΟΙΔΕΤΗΝΗΜΕΡΑΝΚΑΙΤΗΝΩΡΑΝΟΥΤΕΟΙΑΓΓΕΛΟΙΟΙΕ
ΝΤΩΙΟΤΡΑΝΩΙΕΙΜΗΜΟΝΟΟΟΠΑΤΗΡ.190

189. A secular or novice scribe is postulated because they would be less likely to im­
mediately recognize a nomen sacrum (particularly if nomina sacra are indeed
Christian phenomena), but any scribe working under very tiring conditions could
likewise have stumbled into this error, particularly under conditions of poor light.
190. lids type of text-critical work is highly dependent on the exact form of the verse,
and here is where value judgments about whether a father was quoting verbatim
or not really matter and precisely where the most care ought to be shown. For this
reason, conclusions must remain tentative where a particular "version" of Matthew
is invoked for orthographic analysis. In this case, 2A36 is embedded sporadically
within surrounding commentary, so it is also possible that it represents Mark or

186 / Chapter 4 Adam G. Messer

 An Evaluation ot Ehrman’s text-critical Metnoaoiogy

Such small changes introduce the real possibility of orthographic errors

that otherwise seem more remote.
The primary criteria for any of these orthographic errors to work
would be the presence of a scribe who does not know Greek well. Using
Epiphanius’s version of Matthew for our exploration of accidental
changes, we could postulate the following two-stage textual error from
Epiphanius’s text (or one like it):191

... ΟΥΔΕΚΖΟΙΔΕΤΗΝΗ
ΜΕΡΑΝΚΑΙΤΗΝΩΡΑΝΟΥΤΕ
ΟΙΑΓΓΕΛΟΙΟΙΕΝΤΩΙΟΥΡΑΝ
QfOYTEOYlSCEIMTLMONOC
ΟΠΑΤΗΡ...

Upon reaching the first “IO,” the scribe accidentally drops down
to the “IO” on the next line (the third instance), omitting everything in
between (haplography). This would result in the reading, “no one knows
the day and hour, not the angels who except the Father alone?”192 At a
later point, a more learned copyist recognized that the relative pronoun
“who” did not belong and struck it. This would make sense of why some
versions of the text appear to be missing “in heaven” and why its case
is sometimes genitive and sometimes dative. It is also possible that the
first scribe copied until the second “IO" and skipped past the third, re­
sulting in this text: “no one knows the day and hour, not the angels in
heaven who except the Father alone.” Again, a later scribe could have
removed the relative pronoun.
It is admittedly a more difficult conjecture to explain a variant by
positing a double corruption in a location. However, it likely happened
a few times within the manuscript tradition. Perhaps our reluctance to
explore this possibility is the reason this problem seems more likely to
be attributed to an intentional change.

POSSIBILITY 3
Another option would work with our NA27 critical text. A non-
Greek-speaking scribe may have committed an error of haplography

even an amalgam between the two gospels. However, even though the lack of ουδέ
ό vl0£ leans strongly toward a reference to, if not quotation of, Matthew, the main
point is that we are simply positing possibilities for the unintentional introduction
of the omission into the historical record.
191, His te>ct would actually be the first stages of corruption. These corruptions do not
struggle for credibility (given that Matthew and Mark’s accounts were divided over
case), so this type of minimal harmonization is definitely possible.
192. Here is the punctuated equivalent in minuscule: ouSeis oiSe την ήμέραι/ και την
ώραν οΰτ€ οΐ αγγβλοι os el μή μόνο$ ό πατήρ.

Adam G. Messer Chapter 4 / 187

 Patristic Theology and Recension in Matthew 24.36

with the iota (instance 1) or the omicron (instance 2), resulting in the
same nonsense relative pronoun, which was subsequently stricken.

Instance 1

... ΟΪΔΕίεΟΙΔΕΝΟΥΔΕΟΙ
ΑΓΓΕΛΟΪΤΩΝΟΥΡΑΝΩΝ
ΟΥΔΕΟΥ|θϋΕΙΜΗΟΠΑΤ
ΗΡΜΟΝΟΟΩεΠΕΡΓΑΡΑ...

Instance 2

...ΟΥΔ
ΕΐεΟΙΔΕΝΟΥΔΕΟΙΑΓΓΕΛΟΙΤΩΝΟ
ΥΡΑΝΩΝ^ΥΔΕΟΥΙ^ΟΕΙΜΗΟΠΑΤ
ΗΡΜΟΝΟΩΩΟΠΕΡΓΑΡΑ...

188 / Chapter 4 Adam G. Messer

 5

TRACKING THOMAS
A Text-Critical Look at the
Transmission of the Gospel of Thomas

Tim Ricchuiti1

ince its discovery over sixty years ago,2 the Gospel of Thomas has

S excited the minds of scholars3 and the public alike.4 It has proved
a malleable work, transforming from a primarily Gnostic5 and apoc­
ryphal6 text to a more theologically neutral work esteemed on a level

1. I thank Dr. Daniel B. Wallace for his initial thoughts on the transmission of the
Gospel of Thomas that pushed me to take up this subject in the first place, his
guidance through die initial stages of die formulation of the argument of the paper
on which diis chapter is based, and his consistent availability in pursuing the re­
sulting project through to its completion. Additionally many thanks go to Stazsek
Bialecki, Adam Messer, Philip Miller, and Matt Morgan, my σύνδουλοι, without
whose thoughts, criticisms, and encouragement I would be in the tall grass. Finally,
I would like to thank my lovely wife, Angel, who has put up with many cancelled
evenings through the completion of this work.
2. Technically speaking, though the Coptic manuscript was discovered approximately
sixty years ago, Thomas has been known to scholars in one form or another since
the late nineteenth century. See J. K. Elliott and M. R. James, The Apocryphal New
Testament: A Collection ofApocryphal Christian Literature in an English Transla­
tion (Oxford: Oxford University Press, 1994), 128-29, for additional information.
3. So says Stevan Davies in The Gospel of Thomas: Annotated and Explained (Wood-
stock, VT: Skylight Paths, 2002), xxvii: “For those interested in Jesus of Nazareth
and the origins of Christianity, the Gospel of Thomas is the most important manu­
script discovery ever made?
4. R. McL. Wilson's English translation appeared just before Christmas 1959. It sold
in excess of forty thousand copies, undoubtedly providing more than a few puz­
zling looks at the Christinas tree that morning (Stephen J. Patterson, James M.
Robinson, and Hans-Gebhard Bethge, The Fifth Gospel: The Gospel of Thomas
Comes ofAge [Harrisburg, PA: Trinity, 1998], 2).
5. See Robert M. Grant and William R. Schoedel, The Secret Sayings ofJesus (Garden
City, ΝΎ: Doubleday, 1960), for the strongest point of view in this direction.
6. “[The Gospel of Ihomas] is an apocryphal Gospel, and in no way can it enter the
canon as 'the Fifth Gospel"’ (Joseph A. Fitzmyer, “The Oxyrhynchus Logoi of Jesus
and the Coptic Gospel according to Thomas," in Essays on the Semitic Background
ofthe New Testament (London: Geoffrey Chapman, 1971), 419).

189 / Chapter 5
 TRACKING THOMAS

with the four New Testament Gospels,7 contributing, in the process,

to the demise of more than a few trees. Indeed, Thomas has suffered
its fair share (and then some) of abuses at the hands of idle scholars.
Nevertheless, I humbly request that it suffer one more.
Of the extensive bibliographies one could produce for the Gospel
of Thomas,8 one discipline that seldom pops up is textual criticism.9
More often than not, any textual criticism accomplished in relation to
Thomas occurs within a framework of replacing lacunae in order to
translate the Greek fragments, the Coptic manuscript, or both. This is
due to a number of foreseeable elements:

1. Unlike the case of the New Testament, external evidence plays

next to no part at ail in the discussion of Thomas, therefore all decisions
must be made on the basis of more subjective, internal considerations.

2. Those decisions must also be made on the basis of two dead

languages, with uncertainty as to the relationships between the manu­
scripts and as to the provenance of the text.

3. Being a much more particular subject than the New Testament,

the Gospel of Thomas resists the level of comprehensive treatment to
•which New Testament textual criticism lends itself (i.e., no one is going
to produce a whole volume on Thomasine textual criticism).10

It is my purpose here, however, to devote some energy to exploring

Thomasine textual criticism, in part to answer some questions scholars
raise about the reliability of the transmission of Thomas.

7. Or even esteemed above the other Gospels in some cases. See especially Robert
Funk, Roy W. Hoover, and the Jesus Seminar, The Five Gospels: The Search for the
Authentic Words of Jesus: New Translation and Commentary (New York: Mac­
millan, 1993).
8. They have been legion. Ernst Haenchen’s 'Literatur zum Thomasevangelium" {TRu
27 [19611: 147-78) represents scholarship leading up to 1960. Ktzmyer’s Oxy-
rhynchus Logoi” (420-33) takes over from there, cataloguing the next (extremely
fruitful) decade. Since then, there have been many additions to the scholarship,
and two recent works by April DeConick are about as comprehensive as can re­
alistically be expected: her Recovering the Original Gospel of Thomas: A History of
the Gospel and Its Growth (Library of New Testament Studies 286 [London: T&T
Clark, 2005]) contains a nice overview of the field, with particular emphasis on the
background of the Thomas community, while her The Original Gospel of Thomas
in Translation: With a Commentary and New English Translation ofthe Complete
Gospel (Library of New Testament Studies 287 [London: T&T Clark, 2006]) fea­
tures selected works for each logion.
9. The explicit exception to this is Miroslav Marcovich’s, “Textual Criticism on the
Gospel of Thomas’ (JTS 20 [1969J: 53-74), though, as shown in the previous notes,
there are many implicit exceptions.
10. Yet.

190 / Chapter 5 Um Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

Peter Head, for example, in. his 1993 article on the tendency toward
theological corruption in the NT, asserts about the transmission of
apocryphal works,

It is noteworthy that in the scribal tradition this "adaptation.” [as

observed in the article] is much more conservative than in die pro­
duction of apocryphal gospels. The scribes were interested in “trans­
mission” of texts, rather than in the creation of new texts.11

Larry Hurtado tackles Thomas’s role more explicitly, stating, “[lit is

important to note, however, that the Greek fragments indicate that
GThomas was transmitted with a noticeable fluidity in contents and
arrangement.”12 Finally, Nicholas Perrin makes similar comments,
though a bit more neutrally, when he observes,

What is particularly striking on any comparison between the Greek

fragments and the later discovered Coptic text is that, even ac­
counting for the obvious fact that the two recensions are in different
languages, there are dissimilarities in wording. At points the Greek
text is more expansive; at other places, it is more abbreviated than its
Coptic counterpart.13

What I wish to accomplish first here, therefore, is an exhaustive

comparison of the Greek to the Coptic.14 Secondly, I will assess the
"dissimilarities” by utilizing some universally recognized principles of
textual criticism, ultimately offering a conclusion on the matter of the
at­
tempt to ascertain whether or not scholars are correct in their represen­
tation of Thomas’s fluidity and what that fluidity means, both in light of
the communities that produced Thomas and in the broader stream of
early Christian backgrounds.

11. Peter M. Head, “Christology and Textual Transmission: Reverential Alterations in

__ „.ihe.Synoptic.Gospels,”WovT35 (1993): 128.
12. Larry W. Hurtado, 27ie Earliest Christian Artifacts: Manuscripts and Christian
Origins (Grand Rapids: Eerdmans, 2006), 34. Here, Hurtado cites an extended dis­
cussion of Thomas in another of his works, Lord Jesus Christ: Devotion to Jesus in
Earliest Christianity (Grand Rapids: Eerdmans, 2003), 425-79.
13. Nicholas Perrin, Thomas, the Other Gospel (Louisville: Westminster, 2007), 8.
14. Comparisons such as this are not without precedent. Though I hope to contribute
something new to this discussion (in format if nothing else), I recognize that I am
walking on well-tread ground. DeConick’s Original Gospel of Thomas in Trans­
lation and Uwe-ICarsten Plisch’s The Gospel of Thomas: Original Text with Com­
mentary (trans. Gesine Schenke Robinson [Stuttgart: Deutsche Bibelgesellschaft,
2009]) come to mind in terms of recent studies, but this comparison has also been
undertaken by Fitzmyer (“Oxyrhynchus Logoi”) and Marcovich (“Textual Criti­
cism”), among others.

Tim Ricchuiti Chapter 5 / 191

 TRACKING THOMAS

There are any number of things that we cannot know in the process
of analyzing and comparing the Greek fragments to the Coptic manu­
script: whether and which of the fragments are reliable transmissions of
the “original” text,1516
whether the eventual Greek manuscript that served
as the exemplar for the translation to Coptic maintained such reliability,
and whether the line of transmission from that initial translation to the
extant manuscript was commensurately reliable. Indeed, any study on
the textual support of the Gospel of Thomas must necessarily be full of
caveats. Perrin, for one, notes the source of such a problem:

[T]he Coptic text is probably only a translation of other texts which in

turn were presumably copies. The date of the original composition of
each of these texts must have been still earlier, perhaps considerably
earlier.15

The limited nature of the evidence requires that we extrapolate from

these meager beginnings and accept that the evidence can only go so
far in supporting a level of conclusiveness with which scholars are
comfortable. For that reason, I take to heart tire instructions of Harold
Attridge, who, in his excellent survey of the Greek papyri, points out
that “[with respect to Thomas] caution is required in drawing text-crit­
ical inferences,”17 a notation that shall here be my mantra.
With those qualifications in place, however, it is my desire to con­
duct just such an analysis of the text of the four extant manuscripts of
this gospel. I hope to demonstrate, as cautiously as possible, the rela­
tive merits and faults of the three scribes who copied the Greek text,

15. Indeed, if there was, at any point, what we would consider an original text, or, more
accurately, what we would consider an original text in the sense of the original NT
documents. I realize that even the NT usage of the term original is not without
controversy. Cf. E. J. Epp, “The Multivalence of the Term Original Text’ in New
Testament Textual Criticism,” HTR 92 (1999): 245-81; D. C. Parker, The Living Text
ofthe Gospels (Cambridge: Cambridge University Press, 1997). DeConick (Recov­
ering the Original Gospel of Thomas, 38-63) writes of the problems with deter­
mining an original text of Thomas. Because I. am. persuaded by much of DeConick’s
research (though, as will become clear, not all), I use the term original cautiously.
Sometime in the mid-second century, it appears likely that the text of Thomas
gained stability (see chart 1 of DeConick, Original Gospel of Thomas in Transla­
tion, 10). Whether that stability stems from the original composition of Thomas
with redaction and dependence on the Gospels and other traditions or from a con­
cretizing of the community that had been adjusting and molding Thomas over a
period of about one hundred years is an issue I will leave to other scholars. My
primary interest here is the question ofwhat the transmission history reflects once
the textgained stability, so it is in that sense that I am using tire term original.
16. Perrin, Thomas, the Other Gospel, 8.
17. Harold W. Attridge, "Appendix: The Greek Fragments,” in NagHammadi Codex II,
2-7: Together with XIII,2*, Brit. Lib. Or. 4926(1), andP. Oxy. 1, 654,655, ed. Bentley
Layton, Nag Hammadi Studies 20 (New York: Brill, 1989), 103.

192 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of die Gospel of Thomas

as well as those of the Coptic scribe. The method will be first to con­
duct the comparison previously mentioned and note the differences be­
tween the texts. In that comparison, I will make a determination as to
which reading is more likely to be original. My fundamental criterion is
simple enough: where there is variation and where the variation is not
nonsensical,18 the reading that best explains the rise of the other is prob­
ably authentic. This criterion will be expressed through two corollaries:
(1) the shorter reading is usually “original,” and (2) the more difficult
reading is usually “original.”
For that second canon, the Gospel of Thomas presents a special
challenge: what exactly is the more difficult reading in the case of a
noncanonical text? To the best of my ability, I will utilize that canon
only when it is clear that a particular reading is more difficult to the
worldview ofthe manuscript in question.19 There will, of course, be times
when these criteria do not apply or when they conflict with each other.
In those situations, I will either leave the case open or attempt to make
a determination on other grounds. It is my belief that such an analysis
will yield results in favor of the conclusion that the Greek fragments
represent an earlier strain of the Gospel of Thomas.
Certainly, this result would not be unexpected. Considering the
fact that the latest of the Greek fragments (probably P. Oxy. 655, ca.
250 CE) is about one hundred years senior to the Coptic manuscript
(ca. 350 CE), it appears eminently reasonably that scholars would as­
sume this position.20 Indeed, this is the default position of virtually
every scholar who tackles the question, and very few stray from those
beginnings.21

18. In cases where tire variation is nonsensical· involves misspellings, and so forth,
it will be classified inauthentic. This decision is a difficult one to make, yet I have
made it for the sake of consistency (both within this work and within the text-crit­
ical field), it is certainly true that a nonsensical reading can yet point to an earlier
form of the text. That did not happen to occur in this analysis.
19. As will be taken up in further detail in this chapter, the worldview of Thomas is a
-—... -bit-difficult-to establish. Generally, however, the strongly noticeable elements in
Thomas include an elevation of knowledge (gnosis), particularly for salvation, with
Christ as mediator to that knowledge. There are also a number of logia that feature
a reconciliation of dualisms (22,87,106,112), but quite a few more that do not fea­
ture such reconciliation—specifically, logia regarding the earthly versus heavenly
realms (3,27, 57, 76,82,109).
20. As does, e.g., Helmut Koester in “The Gospel according to Thomas: Introduction,"
in Layton, NagHammadi Codex II, 38.
21. This is a slightly different question from one of composition. On that end, James
M. Robinson represents the majority opinion rather boldly when he states, without
exception, “All [of the Nag Hammadi texts] were originally written in Greek and
translated into Coptic” (The NagHammadi Scriptures, ed. Marvin W. Meyer [New
York: HarperCollins, 2007], xi). For another scholarly affirmation of this opinion,
similar thoughts are echoed by Marvin Meyer a few pages later in the same volume,

Tim Ricchuiti Chapter 5 / 193

 TRACKING THOMAS

The next step in my method will be to determine what patterns exist

among the variations (essentially attempting to describe the scribal
habits behind each manuscript). Put most basically, are the changes
evidence of random accretions and deletions due to the degradation of
time, or do they represent a slightly more consistent character toward
a particular theological point of view? Fitzmyer did not shy away from
such characterizations, stating,

Though it is possible that another Greek recension existed, of which

the Coptic is a faithful rendering, it is much more likely that the Coptic
version is an adapted translation—most likely with adaptations made
to suit some of the theologoumena of the Gnostics who used or trans­
lated the Gospel.22

Presumably, he would be among those scholars Meyer pointed to as de­

luded or unduly influenced by the historical record of rejection of such
texts as Thomas due to their heretical character.23
As we may be entering more controversial waters, let me be clear
about what I do and do not mean. I am not resurrecting the blanket
assumption that Thomas represents full-blown Gnosticism. Nor do I
care to get into a debate regarding whether the text is fully Gnostic
or is, rather, a gospel too often misinterpreted as Gnostic.24 It will be
enough for our purposes to see if the alterations move discernibly
toward a Gnostic worldview25 or meander about more haphazardly.
On this matter, Head has already argued for this kind of noncanon-
ical theological alteration: “This type of activity can be seen in other
forms of Gospel-redaction: in the Gospel of Peter, the Gospel of

as he states a bit more softly, “The texts included [in this volume] were translated
into Coptic ... though they were lilcely composed in Greek" (1).
22. Fitzmyer, Oxyrhynchus Logoi/ 416.
23. Meyer, NagHammadi Scriptures, 6-7.
24. Koester’s affirmation that"[elements of gnostic theology are present in these pas­
sages” (“Introduction/ 44) represents well the cautious note I attempt to -strike
herein.
25. April DeConick has written a series of posts on her blog (Jlhe Forbidden Gospels,
http://forbiddengospels.blogspot.com [accessed April 1, 2009]) entitled “Trans­
theism or Supratheism?/ “Transtheism/Supratheism follow up/ and,l Transtheism
it is/ In this series, she moves away from the terms Gnostic and Gnosticism, be­
cause of their abuse, and prefers to use Transtheism instead to denote “groups of
religious people in the ancient world that worship a god who is spatially beyond our
universe and who is not identified as the immediate creator and ruler of our uni­
verse. Instead, these roles are attributed to subordinate powers who are not being
worshiped.” However, as she herself continues to affirm that there were Gnostics
in the early centuries of Christianity (“My reason for [changing terminology] is not
that I do not think that gnosticism existed in the ancient world—in fact I do"), I feel
I am on solid ground in continuing to speak of a “Gnostic worldview/

194 / Chapters Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

Thomas, and to some extent in Tatian’s redaction of the Diatessaron.”26

Once again, I note Perrin’s thoughts on the minefield we are getting
ourselves into:

While I agree that “Gnostic" and “Gnosticism” makes for a pretty un­
wieldy rug under which to sweep all those sects that are not ostensibly
proto-orthodox, tire term has its place, at least if defined accurately
enough. All the same, I disagree with those who say that the Gospel of
Thomas is Gnostic. To be sure, the sayings in the gospel shares many
elements with purported Gnostic texts (elements of anti-Judaism,
hatred of the body, secret knowledge, etc.), but there is no hint that
Thomas’s Creator God is the same sadistic deity or pompous idiot that
we meet in the Gnostic materials. Lacking these features, Thomas
must be judged to be non-Gnostic.27

Contra Perrin and, to a certain extent, DeConick,281 contend that

Thomas, either as it is now or as it existed “originally,” need not be clas­
sified either as Gnostic or non-Gnostic.29 Instead, I intend to explore
whether or not the alterations to Thomas, subtle though they were, were
such that lent themselves toward Gnostic thought, leading to the inclu­
sion of Thomas in the Nag Hammadi materials. Meyer makes a similar
distinction, carefully qualifying the extent to which Thomas is Gnostic,
saying, “the Gospel of Thomas may most appropriately be considered
a sayings gospel with an incipient Gnostic perspective.”30 So I submit
that the additions, omissions, and substitutions of material between the
Greekfragments and the Coptic manuscript represent, in at least some
cases, theological corruption of the text.

26. Head, “Christology and Textual Transmission,” 128.

27. Perrin, Thomas, the Other Gospel, 12-3.
28. April D. DeConick, Seek to See Him: Ascent and Vision Mysticism in the Gospel of
Thomas, Supplements to Vigiliae Christianas 33 (Leiden: Brill, 1996), 11-27.
29. Obviously, there are myriad studies that tackle just this question, though many
of them remain agnostic on it. There is, of course, DeConick's early work (ibid.),
which comes down in favor of a Jewish Apocalyptic model, as well as her later
work (Recovering the Original Gospel of Thomas), which represents a maturation/
evolution of that earlier view. A recent ThM thesis at Dallas Seminary (Christo­
pher Scott Geyer, “Is Thomas Gnostic? A Comparison of Doctrines in the Gospel
of Thomas to Early Gnosticism” (2007]) also engages in a comparison between
Jewish Apocalyptic and Gnostic doctrines (the author splits the difference). Be­
yond those comprehensive works, there are more particular studies into various
aspects of Thomasine theology (cf. Stevan Davies, “The Christology and Protology
of tire ‘Gospei of Thomas)’ JBL 111 [1992]: 663-82). Indeed, an important part of
the debate is to remember that the Gospel of Thomas was not produced in isola­
tion, and other Thomasine works can help shed light on, at the very least, where the
community that produced Thomas was headed.
30. Meyer, NagHammadi Scriptures, 133.

Tim Ricchuiti Chapter 5 / 195

 TRACKING THOMAS

THE MANUSCRIPT ATTESTATION

TO THE GOSPEL OF THOMAS

Thomas is preserved in four manuscripts (three of which are

merely fragments), composed within 100 and 150 years of each
other, in two different languages.31 The three Greek fragments
were discovered in Oxyrhynchus during the late nineteenth cen­
tury32 and were initially published separately33 within a collection
of the findings, as λόγια ιησού (P. Oxy. 1, dated shortly after 200
CE), “The New Sayings of Jesus” (P. Oxy. 654, dated to the mid-
third century), and “Fragments of a Lost Gospel” (P. Oxy. 655,
dated somewhere between the beginning and middle of the third
century).34 Not until the Nag Hammadi discovery in 1945,35 some
fifty years later, would scholars recognize the connection between
these works.36
The Nag Hammadi finding revealed a Coptic manuscript identified
as “The Gospel according to Thomas” (πβγ\ΓΓ6λΐοι:ΤικχτΛ. -©comxc:

Nag Hammadi Codices [NHC] II 2.51.27-28; dated to some time be­

fore the middle of the fourth century),37 containing 114 logia, some of
which were subsequently revealed to match up with ancient attestation
of tire Gospel of Thomas as well as the discovery half a century earlier
from Oxyrhynchus. That attestation (by numerous church fathers be­
tween the early third century and the eleventh century38) had always in­
terested scholars, and here, for the first time, was the discovery of a truly
“lost” gospel. To be sure, Thomas is not a gospel in the sense that we
understand the genre from the four canonical Gospels. It is largely de­
void of narrative material and instead consists mainly of sayings, usually

31. Attridge, "Greek Fragments,” 96-99.

32. Elliott and James, Apocryphal New Testament, 128.
33. P. Oxy. 1: Bernard P. Grenfell and Arthur S. Hunt, The Oxyrhynchus Papyri, vol. 1
(London: Egypt Exploration Fund, 1898). P. Oxy. 654 and 655: Bernard P. Grenfell
and Arthur S. Hunt, The Oxyrhynchus Papyri, vol. 4 (London: Egypt Exploration
Fund, 1904).
34. Attridge, "Greek Fragments,” 96-99.
35. Davies (Gospel of Thomas, ix) places discovery of the site in 1945, and Perrin tells
the tale of that discovery in Thomas, the Other Gospel, 1-2,15.
36. White (H. G. Evelyn White, The Sayings ofJesus from Oxyrhynchus [Cambridge:
Cambridge University Press, 1920]) and Grenfell and Hunt both felt the connec­
tions between the two sayings collections (P. Oxy. 1 and 654), but they could not be
expected to realize the nature and scope of the connections (and the connection to
P. Oxy. 655) without the Coptic manuscript.
37. See Koester, “Introduction,’’ 38, for more information on the dating. It is a rather
difficult question to answer, as there is the problem of dating the initial translation
from Greek to Coptic and then attempting to figure out how far down that line this
manuscript is.
38. Attridge, "Greek Fragments,” 103-9.

196 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

introduced by “Jesus said” (nexelc).39 Some of these would be recog­

nizable to the average Sunday school class, having close verbal and the­
matic parallels in the Synoptic Gospels, but many of them would defy
the modern Christian conception of Christ (not always to the scholar’s
dismay).40 The discovery nonetheless raised the possibility that we now
had in possession more material pointing back to the historical Jesus
than ever before.
The relationship of these four manuscripts to each other is difficult
to work out.4'1 Marcovich attempted one such explanation,42 hypoth­
esizing three separate recensions, but his attempt was unpersuasive to
Attridge, who noted that “it is methodologically unsound to construct
a stemma illustrating conjectured relationships among the witnesses
to the text of the GTh.... Such a construction simply goes beyond the
available evidence.”43 All of the Greek fragments are earlier than the
Coptic manuscript, and none of them overlap with each other. It is
impossible to know whether any of the Greek fragments, in their fuller
form, would have contained the same basic content as the eventual
Coptic manuscript or if the Coptic represented a vast expansion from
the original text. At the same time, the fragments do come in roughly
the same order as the Coptic manuscript, so it is best that our pre­
sumption regarding the transmission from Greek to Coptc not be too
negative. Ultimately, I do not here intend to present such a reconstruc­
tion of the relationship between the witnesses, choosing instead to
merely work out the relative merits of each manuscript in comparison
with each other.

39. Most translators (Davies in Gospel of Thomas and DeConick in Original Gospel of
Thomas, to name just a few) take this introductory formula as the aforementioned
“Jesus said." A few, however (cf. Marvin W. Meyer and Harold Bloom, The Gospel of
Thomas: The Hidden Sayings ofJesus [San Francisco: HarperCollins, 1992]; Plisch,
Gospel of Thomas), take both the Coptic and Greek to be historical presents. It
is an interesting problem, particularly because historical presents normally occur
in narrative material, which for the most part Thomas is not. It is also a problem
beyond the scope of the present study. For more, see Peter Nagel, “πβ-xelc—Zur
Einleitungder JesuslogienimThomasevangelium,” Gottinger Miszellen 195 (2003):
73-79.
40. See, e.g., Andrew Harvey’s foreword to Davies, Gospel of Thomas.
41. Richard Valantasis {The Gospel of Thomas [New York: Routledge, 1997], 4) ob­
serves, “The Coptic sayings comparable to the Greek do not seem to be a direct
translation of the same Greek text, and the Greek seems to witness to another ver­
sion of the gospel than the one on which the Coptic translation is based. So there
is not really a singular gospel, but two divergent textual traditions. This situation
makes a precise and well-delineated description of the Gospel of Thomas problem­
atic, because the Gospel of Thomas may refer to a number of different elements in
its textual historyf
42. Marcovich, “Textual Criticism.”
43. Attridge, “Greek Fragments,” 101.

Tim Ricchuiti Chapter 5 / 197

 TRACKING TMUMAS

COMPARISON OF THE COPTIC

THOMAS TO THE GREEK FRAGMENTS

The Greek papyri have here been placed roughly in order relative
to the occurrence of their content in the Coptic manuscript. The ac­
companying English translations have been completed with a word-for-
word equivalence in mind. I checked the Coptic text against Layton’s
scholarly text,44 as well as that of Meyer.45 The text for the Greek frag­
ments was based on Grenfell and Hunt’s text.46 Both texts were then
corrected as necessary against Plisch,47 Fitzmyer,48 Attridge,49 Aland,50
and DeConick.51 The symbols function as follows:

O An abbreviation (usually nomina sacra) or otherwise omitted

text has been supplied. Text in parentheses would not have
been in the original manuscripts and is supplied for the ease of
the reader.

11 Lacunae in the manuscript and any emendations proposed to

fill in those lacunae.

Black Text not present in the corresponding Greek or Coptic logion

on gray (i.e., in addition to what is present in the corresponding logion).

White Greek and Coptic phrases being used in each other’s place (i.e.,
on black in substitution of each other)

Wavy Text missing from the corresponding Greek or Coptic logion,

line usually due to lacunae, and without proposed emendations.

44. Layton, Nag Hammadi Codex II, 52-93.

45. Meyer and Bloom, Gospel of Thomas: The Hidden Sayings ofJesus.
46. Grenfell and Hunt, Oxyrhynchus Papyri, vols. 1 and 4.
47. Plisch, Gospel of Thomas.
48. Fitzmyer, Oxyrhynchus Logoi.”
49. Attridge, “Greek Fragments.”
50. Synopsis Quattuor Evangeliorum: Locis ParalleUs Evangeliorum Apocryphorum et
Patrum Adhibitis, ed. Kurt Aland, 15th rev. ed. (Stuttgart: Deutsche Bibelgesell-
schaft, 1996).
51. DeConick, Original Gospel of Thomas in Translation.

198 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

P. Oxy. 654—Prologue
NHC11232.10-12 P. Oxy. 654.1-5
we θ-οηπ θντ&.
ουτοι oi Hλόγοι οί [απόκρυφοι οΰ$

l(hcoy)c ©τον2 χοογ κγω έλα] λησ^ν Ίη(σού)? ό £ών κ[αι

xqc2*ACOY h6i
έγραψαν Ιούδα ό] και Θωμά

Ϊουλ^ο -OWM^C
These are the hidden words that the These are the hidden words that the
living Jesus spoke and living Jesus spoke and Judas Thomas
Judas Thomas wrote. wrote.

The prologue to the first saying is identical in all but the name of the
recipient of Jesus’ words.52 The Greek fragment clearly has “Thomas”
(Θωμά/·β·ωΜ2>.ο),53 but its lacuna would be unable to fit both “Didymos”
(Δίδυμος/ΛίΛΥΜΟο) and “Judas” (’Ιούδα/ ϊογΛ^ο). This text plays rather
nicely into our first canon: the shorter text of the Greek is most certainly
the earlier form. The scribal tendency toward explicitness (and, indeed,
harmonization with other traditions)54 asserts itself here in the longer
name of the disciple.55 This also illustrates our general principle rather
well: it is more likely that the shorter Greek text would give rise to the
longer Coptic text (through the aforementioned scribal expansion) than
it is that a longer and more explicit Coptic text (or Greek forbearer to
that text) would lead to a deletion. It is, of course, possible that the scribe
would have deleted Δίδυμος from the forbearer to his copy, but absent
any compelling evidence for such a deletion, I will not entertain the
possibility.

52. Ignoring the easy-to-detect dittography (ούτοι ol οί λόγοι οί) occurring at the be-
ginning of the logion, it is an unintentional scribal error.
53. Most commentators are unconcerned by the lack of a final ς in P. Oxy. 654. Marc-
ovich (“Textual Criticism,” 53) passed it over fairly easily with a reference to its
“common enough” nature, pointing, for attestation, to E. Mayser, Grammatik der
griechischen Papyri aus der Ptolemaerzeit, 3 vols. (Berlin: Walter de Gruyter, 1926),
1:205.
54. A. E J. Klijn ("John XIV 22 and the Name Judas Thomas” in Studies in John Pre­
sented to Professor Dr. J. N, Sevenster on the Occasion of His Seventieth Birthday,
Supplements to Novum Testamentum 24 [Leiden: Brill, 1970], 271-78) goes into
detail on the series of Ihomasine works (Gospel of Thomas, Acts of Thomas, Book
of Thomas) that feature “Judas the twin” and on the traditions from which they
stem. Marvin W. Meyer (“The Beginning of the Gospel of Thomas,” Semeia 52
(1990): 161-73) also comments a bit on the ascription.
55. While it is by no means certain that the scribe of this particular Coptic manuscript
was the one who originally made the addition (indeed, it is rather doubtful that
he is), that is not really our concern. I will here be treating the Coptic manuscript
representatively for the Coptic tradition.

Tim Ricchuiti Chapter 5 / 199

 TRACKING THOMAS

P. Oxy. 654—Logion 1
NHC ll232.12-14 P. Oxy. 6543-5

χγα> nexxq xe πβτχρβ και €ΐπ€ν· [fis δν τήν έρμηνβι]αν των

eeepMHHsix'HHSSi^xe λόγων τούτ[ων eCpn, Βανάτου] ού

fne Μΐ“ΜΠΗΟΥ μή γευσηται.

And he said, “Whoeverfinds the And he said, “Whoever finds the

meaning of these words will not meaning ofthese words will not taste

taste death.” death, ”

There are no significant differences between the Greek and Coptic

in logion 1.

P. Oxy. 654—Logion 2
NHCII232.14-19 P. Oxy. 654.5-9

nss<s l(rtCOY)C ΗΝΤρβ<4Λ0’Ή& [λέγβι Ίη(σοΟ)$]· μή παυσάσθω ό

πβτφίΜβ equine tpKHTeqSme χγα> ζη[τών τοΰ {ητεΐν ίω£ δν] €υρη, και

2OTM4 eqip^H&Ne
όταν βυρρ [θαμβηθήσ€ται, και

qNip φπΗρβ *γω βασιλεύσβ, [igfj

7Μ<
qi^pppo sswiryRc

Jesus said, “Do not let the one who Jesus said, “Do not let the one who

seeks stop seeking until hefinds. And seeks stop seeking until hefinds. And

when he finds, SB®#: Λ when he finds, he will be amazed;

zwiz/kwte

hfi?5. he will be S/sOjS'dii, he will rule;

amazed. And he will rule

e&w&nniL

200 / Chapter 5 Tun Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

Logion 2 illustrates well the difficulty of a textual comparison span­

ning two languages. Despite overlap at the edges, there are four dis­
tinct additions within the text—two in Coptic and two in Greek. We will
deal with the last two first, as they are significantly easier to sort out.
Internally, Thomas is familiar with “rest” as a final stage in the spiritual
journey (1. 90). Externally, the text has many parallels, both themati­
cally and verbally. We find thematic parallels among wisdom literature
recommending the pursuit of knowledge as a method to achieve peace
(Sir 51.13-14; Wisd Sol 1.1-2; 6.12-14).56 Therefore, we are on solid
ground when it comes to accepting the Greek reading καί βασιλεύσα?
άυαπαήσεται as original. However, verbal parallels to logion 2 (such
as Gos Heb 4)57 do not contain the final Coptic addition (exn πτι-ipq),
making it quite likely that the Coptic scribe expanded the text, possibly
for clarity, possibly for some sort of theological reason. Bammel posits
that the Coptic is actually less Gnostic than the Greek,58 but DeConick
dismisses such an interpretation out of hand.59
Indeed, what is going on with the first two additions is much harder
to work out. At first glance, it appears that the Coptic qNxcpfpfp χγω
equ>M40?TopTp ("he will be troubled, and when he is troubled^ is par­
allel to the Greek θαμβηθήσεται, και θαμβηθεις (“he will be amazed,
and when he is amazed’’).60 However, the Coptic verb φτρτρ (“to be
troubled/disturbed”) does not match up well to the Greek θαμβεω (“to
be astounded/amazed”). If this were the end of the story, making an
evaluation as to whether or not they represent a scribal corruption
of the text or just a very loose translation would be next to impos­
sible.61 In this case, however, I think I can offer an explanation. The
Greek represents the original text, with a certain process toward el­
evation (seek-^find, find-^be amazed, be amazed->rule, rule->rest).
The Coptic has inserted another step early in the process (find->be
troubled, be troubled-^be amazed) and deleted a step at the end
(rufe“>rest) to maintain symmetry in the steps. There does not ap­
pear to be a compelling theological motivation for adding and deleting
those particular steps, but that does not mean that there is not one.

56. Stevan Davies, The Gospel of Thomas and Christian Wisdom (New York: Seabury,
1983), 38-39.
57. As found in Clement of Alexandria, Stromata 2.9.45 and 5.14.96.
58. Ernst Bammel, “Rest and Rule,” VC 23 (1969): 88-90.
59. DeConick, Original Gospel of Thomas in Translation, 49-50.
60. Plisch {Gospel of Thomas, 41) does see this as a possibility but does not wholly
endorse it, preferring instead to simply recognize that the original Thomas would
have had some multistep process toward salvation.
61. In other cases such as this at first appears to be, it will here be my philosophy to
give the scribe the benefit of the doubt, particularly in semantic and lexical consid­
erations. Only where there is compelling evidence that the exemplar would have
had a different word at some point in the transmission history will I discuss the
possibility of textual corruption.

Tim Ricchuiti Chapter 5 / 201

 TRACKING THOMAS

P. Oxy. 654—Logion 3
NHC 11232.19-335 Ρ. Oxy. 654.9-21
πβχβ i(hcoy)c xe βγιι^χοοο
λέγει Ί[η(σου)$... .tt έάυ] οί eXxovres

ΝΗΤΝ'Ηώ NBTCttK 2ΗΤΤΗγΥΠ Κβ

ήμδ$ [εϊπωσιυ ύμϊν- Ιδού] ή βασιλεία

61C2HHTB βΤΜΠτβρΟ ΤΠβ GSI© ίν ούρα[υψ, ύμα? φθήσεται] τά

Μ2Μ.ΗΤ Ν^ρφορπ βραχΛΓκτβ τηβ

ftCTeiva του ούρ[ανοΰ· έάν δ* ειπωσιν

βγυ^ΗΧΟΟΟ ΝΤΠΉ
δ)τι έστ[ιν.

e\A2xCCA.
ββΙβ*ΝΤΒΤ κΧρφορπ
είσελεύσονται] οί ιχθύες τη$

βρίΓΓΗ ΤΜΝΤβρΟ θαλάίοσης προφθάσαν]τ€2 ύμα$· καί

οΈπ©τΠ2ογΗ *.γω οηπ^τΈβ^α

ή βασ[ιλ€ΐα kciiOjffl έυτάδ υμών

2OTa.N erefriu^HCoYOJN τι-ιγτΉ [4s]ti (κάκτος

αντην € 5 δτε ύμε 12]

τβΤΗΧβίΜβ χβ Nrarfw πβΊίφΗρβ

έαυτού? γνώσεσθα[ι, εισεσθε δτι

Ήπβιωτ βΤ0Ν2 βιρωπβ Λ6

υΙοί] έστβ νμ€Ϊ£ του πατρό^ του
J ΤΙίγΛϊ λίΐ ββι© £(ώι/το2· el 54 μή] MBfffffll έαυτούς,

τβτιι^οογιυι

τβΤΝψοοη 2“Η ογΗκτ^ΜΚβ χγα>

έν [τή πτωχεία έστέ] καί ύμβΐς έστε ή

ΉτωΥΗ πβ ΤΜΝΤ2ΗΚβ *τω[χ€ία].

Jesus scud, “If those who lead you Jesus said, “If those who lead you

say to you. Behold, the kingdom is in say to you, "Behold, the kingdom is in

heaven, * then the birds in the sky will heaven, ’ then the birds in the sky will

come before you. If they say to you. come before you. If they say to you,

It is 1 lhen the fish 1 then (he fish

62. DeConick (Original Gospel of Thomas in Translation, 51) suggests that Attridge's
reconstruction is wanting here, as the Greek manuscript has room for another "five
to eight” characters. She tentatively proposes something like αύτοΐ$ but correctly
notes that any such proposal "would be purely conjecture, since there U no parallel

202 / Chapter 5

Tta Ricchuiti
 A Text-Critical Look at the Transmission of the Gospel of Thomas

the
come before you. Rather,61*
63 will come before you. Indeed, the

kingdom is inside ofyou and outside kingdom is inside of you and

ofyou. When you know yourselves, outside ofyou.

you will

realize that you are the children of know yourselves, you will realize that

the livingfather. But ifyou will not you are the children of the living

yourselves, then you are in father. But ifyou will not know

poverty, and you are that poverty. yourselves, then you are in poverty,

and you are that poverty.

One of the five variants in logion 3 is easily explainable: the oc­

currence of γνώσθε for what should be γνώσεσθε is due to uninten­
tional scribal error (haplography of the -σεσ-). From there, however,
the road gets significantly more winding. The initial substitution of
xe csH for υπό την γην is likely another place where the
Coptic scribe or translator was attempting to make sense of a diffi­
cult passage. The Greek υπό την γην—the more confusing, though
still sensical, reading64—has parallels in early literature65 and is likely
original. The addition of a modifier following ή βασιλεία presents
an interesting problem, not least because the modifier is lost to us.
Some have proposed του ούρανοΰ to fill the lacuna, noting that it is

in the Coptic to aid us? Such an addition would likely be inauthentic, but due to
the uncertainty on all ends, I remain agnostic on the matter and have elected not
to include the possible variant in any count.
63. Note the differing translations at this point due to a slight difference in the Coptic
.and the. Greek kAA*· for καί). I decided that it was best to interpret this difference
not as a variant (for which there would appear to be little motivation) but as a case
where the syntax of a particular term (in this case καί) could not be replicated
in the recipient language. So another, more appropriate term was chosen. Dieter
Mueller ("Kingdom of Heaven or Kingdom of God?,w VC Z7 [1973]: 271) argues
that the alteration would have been intentionally clarifying the "feeble ‘and’”
64. Indeed, the feet that it is confusing is a point in its favor. It is easy to see a scribe
reading "If drey say to you, ‘It is under the earth,’ then the fish will come before
you" and thinking that the original must surely have read something more like ‘It
is in the sea? It is much more difficult to see a scribe altering an internally coherent
phrase ("If they say to you, It is in the sea,' then the fish will come before you")—
one with well-attested parallels (see n. 65)—to something nearly nonsensical·
65. Mueller ("Kingdom of Heaven or Kingdom of God?”) notes the parallel to Job
12.7-8. T. F. Glasson ("The Gospel of Thomas, Saying 3, and Deuteronomy xxx.ll-
14? ExpTint 78 [1977]: 151-52) notes the parallel to Deut 30.11-14.

Tim Ricchuiti Chapter 5 / 203

 TRACKING THOMAS

a much more common phrase within the gospel66 and fills out the
space more fully.67 Mueller demonstrates a compelling case that του
θεού ought to fill the lacuna.68 Logion 27 uses the same modifier for
“kingdom,” and the Coptic redactor appears to exhibit a tendency
toward replacing other instances of ή βασιλεία τοΟ θεού (cf. 1. 54,
where Luke 6.20 has been altered just slightly). As with other places
in Thomas, a conclusion cannot be stated too strongly due to the
number of factors up in the air. However, I favor an original reading
of ή βασιλεία τοΟ θεού, making this a probable case of theological
alteration.
The other two expansions of the text (τοτβ csn^coyu>(n) thug,

“then/when you will become known"; os du εαυτόν γυω, ταύτην

εύρήσει, “whoever knows himself will find it”) represent yet another
special problem in this logion. Regarding the Coptic addition, the
phrase awkwardly interrupts a sequence similar to what we just saw
in logion 2. It is so awkwardly worded that Mueller suggests it might
have originated as a marginal note.69 Regarding the Greek addition,
we could argue that the addition bears similarity to that same se­
quence in logion 2 that the secondary Coptic addition interrupts—
and so is original. But we could just as easily argue that it is because
of that similarity that the phrase is likely a secondary expansion from
an overzealous scribe. A third possibility is that the two additions are
not actually independent, instead being in corrupted transposition of
each other. In lieu of any significant evidence to illuminate the situ­
ation, I conservatively fall back on the canon of the shorter text in
each case.

66. "Much more common” being a relative term. "Kingdom of Heaven” is an ascription
that occurs three times in the Coptic Thomas (II. 20, 54, and 114). "Kingdom of
God” does not occur in Coptic Thomas and occurs only once in the Greek frag­
ments (L 27).
67. So DeConick, Original Gospel of Thomas in Translation, 52. It does not appear;
however, that DeConick considers the possibility that του ούρανοΰ could have
been replaced by a nomen sacrum, thus rendering it too short for the allotted space
as well. Each line of the Coptic manuscript has room for between twenty-eight
and thirty-three characters. DeConick Is correct that if τοΰ ούρανοΰ were written
out fully, it plus the last five characters of βασιλεία would make twenty-nine total
characters in the line. If, however, it were not written out fully, the shortened ver­
sion of τού ούραυοΰ would make it too short, meaning that length and number of
characters should not be the only (or even the most significant) factor in deciding
how to fill in the lacuna.
68. Mueller (“Kingdom of Heaven or Kingdom of God?,” 271-74) ultimately decides
otherwise, but Fitzmyer (“Oxyrhynchus Logoi,” 376-77) sticks with τοΟ θεοΰ.
69. Mueller, "Kingdom of Heaven or Kingdom of God?,” 268, following Grant and
Schoedel [Secret Sayings ofJesus, 70) and Otfried Hofius (“Das koptische Thoma-
sevangelium und die Oxyrhynchus Papyri Nr 1,654, und 655,ӣvT20 [I960]: 31).

204 ! Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

P. Oxy. 654—Logion 4
NHC1I233.5-10 P. Oxy. 654.21-27
[λέγει Ίη(σοΟ)?]· ούκ άποκνήσει
nexe i(hcoy)c qwxxNkY s.nT>6i

ανθ[ρωττοε παλαιοί ήμε]ρών

Πρα>Μ®·Η2λλΟ 2N Nec|2OCY Θ25Ν6

έπερωτήσαι πα[ιδίου επτά ήμε]ρώυ

ογκογβιΤϊιιιιιρβ ujhm eqgR caxpq

περί του τόπου τη[ε ίωής, και

Έ^οογ βτβθ ητοποοΤ?ΠΦΗ2 4.γω

ίή]σετε·ότι πολλοί έσονται π[ρώτοι

qN*.iun2 xe οϋν 2^2"Rq?opn Hi.”

έσχατοι Iqaffi όι έσχατοί π/ιωήρι!, και

24.® 4.γα>Nceujoine oya. ογωτ™

[els γε νήσου ]σιυ.

Jesus said, ‘The man old in days will Jesus said, The man old in days will

not hesitate to ask a child of seven not hesitate to ask a child of seven

days about the place of life, and he days about the place of life, and he

will live. For many who are first will will live. For many who are first will

be last, and they will become one.

they will become one.

Climbing down from the mountain of variants in logion 3, we have

only the addition of ol έσχατοι πρώτοι to deal with in logion 4. This is yet
another variant that could easily go either way: we could argue that it is
a predictable addition to the text by the Greek scribe, on the grounds of
maintaining parallelism within the text or an attempt to harmonize with
the canonical Gospels (cf. Matt 19.30; 20.16; Mark 10.31; Luke 13.30). We
could also argue for its originality and attribute its deletion in the Coptic
to either an accidental omission or a purposeful attempt to blunt the force
of thfe'sayihgfthough it is unclear what kind of understanding or desire
would need to be present in the community to generate the deliberate
deletion of “the last will be first” while leaving in “the first will be last").
Indeed, it is quite possible to see the accidental deletion of the phrase
from the Coptic. The text in the manuscript currently reads like so:

...4.Y0)
qN4.a)N2X60Y“2^2HQ)0Pn|i^2i.
β4.γα>Ποβφωπ@ογ4.ογωτ*

70. See A. F. J. Klijn, “The ‘Single One’ In the Gospel of Thomas," JBL 81 (1962): 271-
78, on the equivalence of ογί. ογωτ and els.

Chapter 5 / 205

Tim Ricchuiti
 TRACKING THOMAS

The original text, containing the phrase “and the last will be first”
(&.γωΈ2λβ φορπ), could have looked something like this:
nap

QNMUN2xeoY'H2a,2Nmopnnxp2i,
κγωΗ2κβιικρψορπ j.YU)Nceqjw
πβογ^ογα>τ

As the scribe's eye jumped ahead from the second ^γω in the pas­
sage to the third—

...κγω

^^Hcecpo)
πβογκογωτ

—he would have accidentally deleted the phrase in question, leaving

only

...^γω
4ΝλΦΗ2Χ©ογ'Ν2^2ΝφορηΝχ'ρ2^
β^γωΗοβα)ωπβογ^ογα)Τ

This is only one possibility,71 but it suggests that the Greek contains
the original.

P. Oxy. 654—Logion 5
NHC 11233.10-14 P. Oxy. 654.27-31

rxexe i(hcoy)c coycdh nefHrfHTO

λέγει Ίη(σοΟ)ς· γ[νώθι τδ δυ

Τϊπ<3Κ2θ θβολ κγα> πβο+ιπ βροκ

έ μπρος )θευ τής δψεώς σου, καί (τό

qiiKtoAii βΒΟλ NKK >4H AKKY ΓΚρ

κβκαλυμμένου] από σου

eq2Hn eqNKOYlMNS βΒΟλ KN

αποκαλυφήσετ[αί σόι· ού γάρ ές]τιν

71. It is a possibility that works particularly well with Attridge’s reconstruction (*Greek
Fragments,” 115) but not so well with DeConick’s (Original Gospel of Thomas in
Translation, 57). DeConick notes that Attridge’s reconstruction would require fif­
teen characters, while the line appears to have room for only twelve.

206 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

κρυπτόν δ ού φανείρόν γβνήσεται],

Jesus said, “Know what is before Jesus said, “Know what is before

yourface, and what is hidden from your face, and what is hidden from

you will be revealed to you, for there you will be revealed to you, for there

is nothing hidden that will not be is nothing hidden that will not be

manifested." manifested,

The manuscript of logion 5 reads άποκαλυφήσ€τ[αι, while it

should read άποκαλυφθήσετ[αι. That is a fairly straightforward mis­
spelling on the part of the Greek scribe (he omitted the Θ). The second
variant—the addition of the phrase και θ€θαμ.|ΐένου δ ούκ έγερθήσβται
(“and (nothing) buried that will not be raised”)—presents a bit more of .!

a problem. Apparently, this phrase became a rather common burial in­ I

I

scription in Egypt, making it a predictable addition.78 Once again, it looks i

as if the canon of tire shorter text holds here and as if the Greek is inau­
I
thentic. However, we should also note that the burial inscription (fifth
or sixth century CE) is known from at least a few hundred years later.
It seems odd to not at least recognize the possibility that the burial in­ I
i
scription stemmed-from recognition that Jesus said something like this
(perhaps in original Thomas), rather than the other way around. The pos­
sibilities do not end even there, as logion 6 contains a similar Coptic ad­
dition (a.yiD λλΛ.γ e<i2Ofic ©ΥΝλόΦ ογβφΗ 6OAnq, “and nothing that
hn

will remain covered without being revealed”), raising the possibility of a

transposition similar in land, though not degree, to the transposition of
logion 30+77b.7274 Looking at the dissimilarities between the two sayings, I
73 ί
I

72. Likely a misspelling of re θαμμένου.

73. Fitzmyer, Oxyrhynchus Logoi,* 383. See also Grenfell and Hunt, Oxyrhynchus Pa- I
pyri, 4:18; DeConick, Original Gospel of Thomas in Translation, 60.
74. Taken up in more detail shortly.
I

Tim Ricchuiti Chapters / 207

 TRACKING THOMAS

think we can shed light on the original text here. If και θεθαμμένον [stc]
δ ονκ €γ€ρθήσ€ται is an addition to the Greek of logion 5, it ought to
be an addition to the almost identically ending logion 6. But we do not
find it there. Instead, we find the altered Coptic addition, which shifts
emphasis from a physical resurrection to the more common refrain of
gnosis and revelation. Indeed, recognizing the theological motivation for
such a deletion, it is most likely that P Oxy. 654 preserves the original text
of Thomas 5.

NHC 11233.14-23 P. Oxy. 65432-40

^YXHOYqTSi NsqH^etiTHc πβχ^γ [€ξ]ετά£ουσιν αυτόν ofi μαθηταί

αύτου και λέίγουσιυ· J

I, και πώς
λγιο βφ τβ -ee βΗλφΛΗΑ euxf

προσευξό]μεθα, και πώς

6λβ«ΗοογΝΜ ^γα> ebuC?m.p^Trtpsi

[Ελεημοσύνην ποιήσομεν κ)αϊ τι

βογ η£ ογαΗ nexe i(hcoy)c zg

παρατηρήσ[ομεν περί των

Hrip Χ6 6οα χγω ηβτ^ιΈΗΟΟΤβ

βρωμάτω]ν· λέγει Ίη(σοϋ)ς· [μή

“RMoqTTrip^Kq xe csSoAn τιιρογ

ψεύδεσβε και δη μις)εϊται

ττοιεΐτ[ε· δτι πάντα ενώπιον

γα»ρ sqgnn equ^OYoms βιοα

έστίν ά[π)οκεκρ[υμμένον δ ού

φανερόν Εσται].

His disciples questioned him and said His disciples questioned him and

how shall we pray? Shall we give should we pray?And how should we

alms?Andfrom whatfoods shall we give alms? And what should we

abstain? ” Jesus said, “Do not tell lies observe concerning ourfood? Jesus

208 / Chapters Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Ihomas

and do not do that which you hate. said, “Do not lie, and that which you

For everything is revealed before hate, do not do. For everything is

heaven . For there is nothing hidden revealed before For there is

that will not be manifested, nothing hidden that will not be made

known."

Each variant of logion 6 is a clear instance of the Coptic’s secondary

character. The Coptic addition of Ni.q appears to be a scribal alteration
toward clarity. Regarding the substitution of πώς νηστεύσουν (“How
should we fast?”) for χβκογωφ βτρΠρΝίιοτβγ© (“Do you want us to
fast?”), DeConick explains particularly well the difference in and sig­
nificance of these questions beginning the saying: “The questions have
a more parallel structure in the Greek than in die Coptic. . . . [T]he
Coptic questions appear to have been revised to reflect the practices
of later Christians who no longer wished to continue obligatory fasting
practices.”75 The substitution οίΉπβΜΤο Έτη© (“in the face of
©boa

heaven/the sky”) for ενώπιον τής άληθείας (“in the face of truth”) is
relatively simple to explain once one realizes that the Coptic for "truth”
ism©, only one character off from the Coptic for “sky” (ne).76 The scribe
unintentionally substituted π for what should have been Finally, the
m.

additional statement χγω λέ^γ ©qsosc ©γΜλ.6α> ογβφκ 6oxnq

μη

(“and there is nothing covered that will not be revealed”) is almost cer­
tainly an expansion and alteration of the original like that previously
described for logion 5.77
Finally, with logion 7 bringing us to the close of P. Oxy. 654, we
have our first example of a text wherein conjecture leaves the arena of
external probabilities completely (such probabilities being unlimited
by the sparse text available). With only seventeen visible characters re­
maining in the Greek fragment, it is of little value here to guess about
what mistakes the scribe may or may not have made.

Analysis

Overall, there are eighteen variants to deal with between the text
of P. Oxy. 654 and logia 1-7 of the Coptic Thomas. There does not

75. DeConick, Original Gospel of Thomas in Translation, 62.

76. It is also worth noting that the two characters’ similar look may have contributed
to the confusion.
77. The alteration also shares some verbal familiarity with Matt 10.26 and Luke 12.2.

Chapter 5 / 209

Tim Ricchuiti
 TRACKING THOMAS

appear to be any tendency toward a particular type of variant: five

are substitutions, seven are additions within the Greek text, and six
are additions within the Coptic text. Nor does any particular logion
stand out: logion 3 has the highest number of variants, at five; but
logia 2 and 6 are just behind, at four variants each. What does stand
out is the clearly secondary character of the Coptic text. Of the eigh­
teen variations, I judged the text of P. Oxy. 654 to be original in thir­
teen cases (72.2%). I judged the Coptic text to be original in just three
cases (17%), each of which was an easily detectable spelling variation.
Scholars are clearly on solid ground when they assume the primacy of
the text of P. Oxy. 654.

Table 5.1. Variation of P. Oxy. 654 by Logion

Substitution 1•; Greek Addition Coptic Addition Total

I
Prologue 0 1 1 2

1 0 0 0 0

2 0 2 2 4

3 2 2 1 5

4 0 1 0 1

5 1 1 0 2

6 2 0 2 4

7
1 ° 0 0 0

Total 1 5 7 6

The nature of the alterations also stands out. While I remain ag­
nostic on the motivation of the alteration of logion 2, logia 3, 5, and
6 each display theologically motivated alterations. Logion 3 appears
unwilling to ascribe the “kingdom” (to which adherents of the text
were taught to aspire) to “God” (θεός), who would have been associ­
ated with an evil demiurge, rather than the benevolent “Father” seen
elsewhere in Thomas. Logia 5 and 6 alter a phrase that would have
glorified and endorsed bodily resurrection. To be sure, most variation
between the texts can be attributed to natural, neutral, and uninten­
tional alteration; but we must also recognize those places where that
is not the case.

210 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of die Gospel of Thomas

P. Oxy. l—Logion 26
NHCII2.38.12-17 P. Oxy. 1.1-4

βΡ9^-Π 9-9 9Ι Λβ

ΠβΚΒΧλ TOT© ©BOA τότε διαβλέψεις εκβαλεΐυ τό

©Noyxe μπχη 6KO 2H ΠΒλλ κόρφος το όφθαλμφ του

MflSKCOH αδελφού σου.

... Then you will see clearly to cast 'Arid then you will see clearly to cast

the twig out of your brother’s eye. out the twig 22 your brother's eye.
The beginning of P. Oxy. 1 agrees almost exactly with the Coptic man­
uscript, where they overlap.78 There are two small disagreements: (1) the
addition of a καί at the beginning of the verse and (2) the slightly different
nuance of the prepositions used in description of the “twig” (eeox, “from,”
vs. ev, “in”). The text is parallel to Matthew 7.3-5 and Luke 6.41-42:

τί § .T9g-V-T9- Τί δέ βλέπεις τό κόρφος τό έν τφ

ρφ 8 αλ.μ_φ του ρjg λφρύLgov ?_Tflv_ S έ όφθαλμώ τ^

?Φ,θΦθόλ]ΐψ>δρκόυ ου

κατανοείς·; ή πώς έρεΐς τφ άδελφώ

jT0 Kq,pjbo^g^K TOv

όφθολμρΰ gov Jji τό κόρφος τό εν τω όφθαλμω σου,

αυτός την έν τώ όφθαλμω σου δοκόν

δοκόν, και τότε διαβλέψεις εκβαλεΐν τ^ώ toj^jqv j^SSK^K-

τό κόρφος εκ τού οφθαλμού τού σου, και τότε διαβλέψεις τό κόρφος

αδελφού σου. τό έυ τώ όφθαλμω τού αδελφού σου

έκβαλεΐν.

78. The extant papyrus page begins with και τότε διαβλέψεις.

Tim Ricchuiti Chapters / 211

 TRACKING THOMAS

The Luke text is in. complete verbal parallel save for the location
of έκβαλεΐν. The Matthew text is in complete verbal parallel save for
the prepositional phrase, which more closely matches the Coptic. Ihe
best explanation for these parallels is that the Coptic represents the
original text or stems from an earlier tradition that also lies behind
the Matthean text.79 Following this explanation, it appears that the
scribe of P. Oxy. 1 attempted to correct his text utilizing the verbal
parallel in Luke, which introduced both the και and the alteration of
the original prepositional phrase (which would have looked more like
Matt7.5).80

P. Oxy. 1—Logion. 27
NHCII 2.38.17-20 P.Oxy. 1.4-11

βΤΘΤΜρΗΗΟΤβγβ
θπκοομοο
εάν μή νηστεύοηται

ΤβΤΝΧ2θ βΤΜΪΪΤβρΟ τον κόσμον, ού μή εύρηται τήν

βΤβΤΪΓΓΜβί ρβ ΜΠΟ^Μ Β^Τ ΟΝ βασιλείαν €qv μή

NCi.BB^TON“NTeTN^N\Y βΠβίωΤ σαββατίσητε τό σάββατον, ούκ

δψεσθε τό(ν) π(ατε)ρα.

“If you do notfastfrom the world., “Ifyou do not fastfrom

you will notfind the Kingdom. Ifyou the world, you will notfind the

do not keep the Sabbath as Sabbath, Kingdom if you do not

you will not see the Father.” keep the Sabbath the Sabbath, you

will not see the Father.n

79. Stephen J. Patterson, The Gospel of Thomas and Jesus (Sonoma, CA: Polebridge,
1993), 30-31.
80. DeConick {Original Gospel of Thomas in Translation, 128) notes another possi­
bility proposed by John H. Sieber (“A Redactional Analysis of the Synoptic Gospels
with Regard to the Question of the Sources of die Gospel according to Thomas”
[PhD diss., Claremont Graduate School, 1966], 72—74)—namely, that this type of
variant appears to stem from oral tradition, rather than literary development of the
gospel.

212 / Chapters Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

Logion 27 gives us our first identifiable spelling variation in the

Coptic Thomas: the Coptic manuscript reads βτβτΠρΝΗΟΤβγβ where
it should read βτβτΰτΜρΝΗΟτβγβ. Additionally, P. Oxy. 1 has three ad­
ditions: (1) the introductory “Jesus said,” (2) the genitive modifier του
θεού, and (3) a καί introducing the second half of the logion.81 Variants
1 and 3 are almost certainly cases where the shorter text (the Coptic
text in each case) lines up with the original text. The addition of the in­
troductory formula is a completely predictable adjustment on the part
of the scribe.82 Beyond this, it is not as if logia lacking the introductory
formula are unknown (cf. logia 62, 69,101).83 The additional καί could
be merely a translational issue,84 but along with the additional καί of
logion 26, we can detect the beginnings of a scribal tendency toward
additional conjunctions in P. Oxy. 1.
This leaves us only with the addition of του θεού. This is essen­
tially the same problem seen in logion 3. It is true that “kingdom” is
generally found without any modifiers (logia 3,22,27, 46, 49, 82,107,
109,113),85 but if anything, that would push us in the direction of af­
firming the originality of του θεού.86 Beyond this, it is easy to detect a
motivation to delete του θεού from the manuscript. A pair of parallel
conditional statements form logion 27:

If you do not fast from the world, you will not find the Kingdom.
If you do not keep the Sabbath as Sabbath, you will not see the Father.

If του θεού is original, logion 27 makes an implicit connection between

“the Father” and “God.” But this would have been unacceptable for ad­
herents to a theology that taught that “God” was to be associated with
the demiurge, while the “Father” represented the true Supreme Being.
Therefore, the most likely scenario is that του θεοΰ is original and was
deleted due to theological considerations.

81. It is a bit of bad form on my part to separate του 0eov καί into two variants when
they appear as one addition in the manuscript. However, I believe that I can. dem­
onstrate good cause to treat them separately.
82. Interestingly enough, DeConick (Original Gospel of Thomas in Translation, 129)
views this as the basis for why the Greek is likely authentic here, noting that “this
introductory clause is consistent with the manner in which the majority of logia are
introduced in the Gospel" However, it is precisely because of that consistency that
I find it more likely that the λέγει Ιησούς is secondary.
83. Plisch, Gospel of Thomas, 207-8.
84. Indeed, DeConick ignores the καί in her translation (Original Gospel of Thomas in
Translation, 131).
85. Ibid., 129.
$6. It is more likely that a scribe would delete του θεού to keep the manuscript in line
with other logia than that a scribe would add του θεού and break such parallel.

Chapter 5 / 213

Tim Ricchuiti
 TRACKING THOMAS

Ρ. Oxy. 1—Logion 28
NHCII2.38.20-31 P.Oxy. 1.11-21

nSXG |(|1COY)C ^6iu>23 ®P*.T λέγει Ί(ησοΰ)ς· έ[?]την ev μέσω του

2Ν TMHTeTinKocHoc λγιο κόσμου και έν ώφθην αύτοι?

λ.βιογΦΝ2 ββθλ ΗΑ.γ 2~ ^g^eupov πάντα? μεθυοντα?

βρΟΟΥ ΤΗρογ ®γΤ4.2βΤίΠ12®

ούδένα εΰρον δειψώ(ν)τα εν αυτοί?

βλλ^γΤΪ2ΗΤογ eqoBG *γω καί ττονεΐ ή ψυχή μου επί τοΐ? υιό*?

κτΜΎΧΗ t tk^c exRThpupe

των άν(θρώπ)ωυ δτι τυφλοί είσιν τή

"HppCDMG ΧΘ 2ΝΒΑΑβ©Υ@ HS 2M καρδίμ αύτώ[ν] και [οΰ] βλεπ[ουσι.ν

πογρητ ^γω οβΗλ.γ βεολ χβ δτι...]

Έΐ^ϊ§1£Π!1221ί22^2Χ3ί2Χ£2

§XSllJ§^2!i®I22X§L§S2i2t!

TIKOCMOC θΥ_φ0Υβ1Τ ΠλΗΝ,ΤβΝΟΥ

9^Τ°_2θ-_2?Τ_^^ θΥ^ΗΗθ2 ΗΡΥΰΡΠ

TOTe ΟβΝΧΡΗΘΤ&,ΝΟβΙ

Jesus said, "I stood in the midst of Jesus said, “I stood in the midst of

the world and I appeared to them in the world and I appeared to them in

the flesh. Ifound all of them drunk; I the flesh Ifound all of them

found none of them thirsty. And my drunk a%4! I found none of them

soul suffered in pain over the sons of thirsty. And my soul suffered in pain

men, for they are blind in their hearts over the sons of men, for they are

and they do not see .. . ” blind in their hearts and they do not

see..."
There are no significant differences between the Greek and Coptic
in logion 2S.87 However, in keeping with the scribal tendency for P. Oxy.
1, the Greek has two additional instances of και. One other scribal error
has P. Oxy. 1 reading σαρκεί88 instead of σαρκί.

87. To the point that, in keeping with the rest of P. Oxy. 1, most translations leave out
each καί entirely.
88. While this could possibly be a spelling variation rather than a scribal error, the fact
that the σαρκεί has been corrected within the manuscript itself (σαρκεί) pushes us
strongly in the direction of error. Of course,’it is impossible to know whether a later
corrector or the scribe made the correction.

Tim Ricchuiti
 A Text-Critical Look at the Transmission of the Gospel of Thomas

P. Oxy. 1—Logion 29
NHCII23831-392 P. Oxy. 1.22
j(iiqqy)oTCj<p^

igwne .gT_sg. riHg* gyojnMPg T9

gtpgg nBKAg eyse nc<wu>

QYgnMPejgfflirtpg ng. j?

SffilES30ii§LSSJ2i22^ISS!i^

MHjjiPpHKo ά,οογω? aw ... evoucjei [ταύτ]η[ν τ]ήν

πτωχέ ία(υ),

TGGiHbiT2HKe

... dwelled in this poverty. ... dwells in this poverty.

There are no significant differences between the Greek and Coptic

in logion 29.81

P. Oxy. 1—Logion 30+77b

NHC Π2392-5 P. Oxy. 123-27
nexe i(hcoy)c se πμ* βγπ φοΜτ
[λέγJet Π(ησοΰ)ε· δπ]ου έάν ώσιν

^MMKY 2ΗΝθγτβ Νβ ΠΗ^ (τρ]δ[ΐς], €[ί?Γιυ θεοί· [δ]ττου 3ΒΙ

eyn cii^y u ova, ?w n έγώ είμι μετ’

NMMdX[ αύτ[οϋ]...

Jesus said, ‘"Where there are three Jesus said, “Where there are three,

they are gods. Where they are gods. Magi where HBCTH

two or onemEEESlSSISU alon

NHC II 2.46.22-28

89, As can be seen from the provided translation, there is a possible difference in tense,
but the various ways the Greek tense could be interpreted syntactically make this
one of those places where we have to assume that semantic difference between
Coptic and Greek is a more likely explanation of any difference than is scribal
error/textual corruption.

Chapters / 215

Tim Ricchuiti
 TRACKING THOMAS

n Ig n THP q NTX πτιιρ q β t gB OA

... 6γει[ρ]ον τον λίθο(ν) κάκ€Ϊ

ΠΦ2 ΗΗΟΥφβ λΗΟΚ Μ1 ΜΠ0)Νβ

βνρήσ€ΐ$ μ€· σχίσου τό ξύλου ί^γώ

*γω τβΤΗ^2β ©ροβι μμ^υ

€K€L €ΐμι.

Split a piece of wood, I am there. Lift up the stone and you will find me

Lift up the stone and you will find me there. Split the wood and I am there.

there.

The combination of logia 30 and 77 represents the most difficult,

or at the very least the most elaborate, of all Ihomasine text-critical
problems. To be dealt with are (1) a highly corrupted text in logion 30,
in which the original text is in all likelihood no longer maintained in
either the Greek or the Coptic; (2) the only occurrence50 of transposi­
tion among any of the variants in any of the manuscripts; and (3) the
peculiar problem of the disparate Coptic sayings’ combination in the
Greek manuscript.
Our first difficulty is in attempting to get behind both the Coptic
and Greek of logion 30, as neither appears to preserve the original text.
The Coptic contains an additionalTmoYTS (“gods”) in the initial clause,
apparently because of a scribe’s attempt to make sense of the following
clause (“they are gods”). The Greek contains an additional καί, a pattern
that we have seen emerge with respect to P. Oxy. 1 and that is probably
inauthentic. It also includes an emphatic λέγω (“I say”) in the middle
of Jesus’ statement, which we will come back to in a moment. Finally,
there is a substitution between the Coptic and the Greek. Where the
Greek has only “where one is alone,” the Coptic has “where there are
two or one.” The closest biblical parallel is Matthew 18.20 (“For where
there are two or three having come together in my name, I am there
in their midst”).91 It is possible that the Coptic scribe, remembering
that the biblical parallel had two items (“two or three”), simply got the
two items wrong (substituting “two or one”). It is also possible that he
simply wanted to be more explicit (i.e., answering the question of what
happens when there are two gathered).
A final possibility has to do with the extra λέγω. Plisch proposes a
scenario wherein the Coptic is closer to the original text.92 The original
Greek Vorlage would have therefore read something like όπου εις ή δύο

90. Excepting the possible instance of transposition within 11.5 and 6.

αυτών.
92. Plisch, Gospel of Thomas, 98-99 n. 2.

216 1 Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

ecTLV μόνο?.93 The text would have become corrupted at the point of ή
δύο, and a later scribe would have reconstructed the text according to
Matthew 18.19-20 (replacing the four characters with λέγω). This last
scenario actually works quite well with an assumption that should be
growing more obvious: contrary to the assumptions of most scholars,
the Greek text ofP. Oxy. 1 appears to be secondary in nearly every case.
There is no doubt that it was produced earlier than the Coptic, but the
Coptic was copied (or translated) from a better exemplar than the one
that produced P. Oxy. 1.
Moving on to logion 77, there are two variants to deal with. The
first is the extra καί (in this case hidden in the crasis κάγώ). The second
is the transposition of the statement “Split the wood, I am there” from
the end of the logion in Greek to the beginning of the Coptic state­
ment.9495 It is unlikely that the scribe of either manuscript would have
made this mistake unintentionally: the Coptic phrase (nu>2 ϊϊΝογίρ©
ajiok has four words totaling eighteen characters, while the
Greek phrase (σχΐσον τό ξύλον κάγώ <=Κ€Ϊ ε’ιμι.) has six words totaling
twenty-five characters. The question, then, is what kind of intentional
change would bring about the corruption of the text as we have it now,
particularly considering the movement of logion 77b.
It appears that this transposition is best explained by the Greek
text representing the original more faithfully. This would have come
about as a scribe, either in translating from Greek to Coptic or in
copying the Coptic text, wished to introduce or reinforce (depending
on one’s perspective) the Christological assertions of logion 77
(“Jesus said, Ί am the light that is above everything. I am everything.
Everything has come from me, and everything has reached [nws] up
to me.’”) To do so, he moved the “stone-wood” phrase from logion
30 to logion 77. Because the verb of logion 77’s last statement was
a homonym to the verb of the “stone-wood” phrase’s last statement
(“split [no>2] a piece of wood, I am there”), the order was reversed on
stylistic grounds. What remains unexplained, however, is why that
scribe would also delete the “stone-wood” saying from logion 30. The
.Coptic Thomas is clearly comfortable with repetition (e.g., logia 5/6,
6/14, 11/22, etc.). While this may not be the most elegant solution,
it is at least widespread: DeConick and Kuhn come to the same con­
clusion, though DeConick explicitly states her assumption of P. Oxy.
l’s primacy.55 Kuhn, while not explicitly endorsing the same, argues

93. Marcovlch ("Textual Criticism/ 67-68) actually proposes δττου els εστιν μόνοε
ή δύο έγώ ίίμι μβτ* αυτών as a reconstruction of the corrupted text of P. Oxy.
1, though that reconstruction was criticized by C. H. Roberts (”Ihe Gospel of
Thomas: Logion 30a,” JTS 21 [1970]: 91-92) and Benedict Englezakis ("Thomas,
Logion 30,” NTS25 [1978]: 262-75).
94. Not die beginning of the logion, but the beginning of this particular statement.
95. DeConick, Original Gospel of Thomas bi Translation, 138-39.

Tim Ricchuiti Chapter 5 / 217

 TRACKING THOMAS

for the Coptic’s secondary nature on the basis of the use of nct>2 as a
Coptic catchword linking the end of what would have been the orig­
inal logion 77.96

P. Oxy. 1—-Logion 31
NHCII 239.5-7 P. Oxy. 30-35

ΠΒΧΒ l(MCOY)C MN ΓψΟφΗΤΗΟ Ο)ΗΠ λέγει Ί(ησοϋ)ί· ούκ εστιν δεκτοί

2M πβα COBIN προφήτης· εν τη π(ατ)ρίδι αύτ[ο]ΰ,

p-eep^neye nnbtcooyn mmom

ουδέ ιατρό? ποιεί θεραπεία? εί? τού?

γειυώσκοντα? αύτό(ν).

Jesus said, “No prophet is accepted Jesus said, “No prophet is accepted

in his no physician heals in his homeland ; no physician heals

those who know him. ” those who know him. ”

Logion 31 has only one difference, and while categorizing it as

a variant is a bit of a close call, the difference is sufficiently large to
merit such a categorization. That sufficiency is not due solely to the
semantic range between “village” and “homeland/country”
(+hs)

(πατρίδι)—to be sure, a difference large enough to label this as a

variant rather than dismiss it as a translational difference. It is also
due to the striking manner in which the NT parallels shed light on the
discussion. Being that there is certainly a variant here, the problem
is in determining if the Greek originally had κώμη (“village”) or if
the Coptic originally had cwipe (“country”). The NT parallels (Matt
13.57; Mark 6.4; Luke 4.23-24; John 4.44) exclusively contain πατρίδι,
in agreement with P. Oxy. 1. The Coptic NT, however, does not match
the Coptic Thomas, lining up instead with the Greek NT (ci-ιφ, al­
ternate form of cttcpe).971 find it much more likely that the Coptic
Thomas is preserving the original reading here, as it goes against its
own tradition, whereas the Greek appears to be harmonizing within
its own tradition.

96. K. H. Kuhn, “Some Observations on the Coptic Gospel according to Thomas,” Mus
73 (1960): 317-19.
97. The Sahidic Coptic NT of Mark 6 also provides evidence that this is not a simple
semantic issue. Both cwcpe and fne occur within the space of a few verses, in par­
allel with their NT counterparts, indicating that these two words are sufficiently
separated semantically within Coptic to not merely evidence a translation issue.

218 / Chapter 5 Tim Ricchuiti

 A lext-Cntical Look at the transmission ot the Gospel of Ahomas

P. Oxy. 1—Logion 32
NHCII 239.7-10 P. Oxy. 136-41

nex© i(hcoy)c x© ογποχιο λέγει Ί(ησοδ)$· πόλις* οίκοδομημέυη

βγκωτΉκοο 2iXH ογτοογ έττ* [ojpous υώηλου?ΒοΒ

eqxooe βοτ^χρκγ MN 6OM“NC2© ^στηριγμένη ούτε Te[s]eiv δυνατοί

ΟγΛΘ ΧΝ οΰτ€ κρυ[β]ήναι..

Jesus said, “A city built on a high Jesus said, “A city that has been built

mountain and fortified can neither upon of a high mountain and

fall nor be hidden. ” that has been fortified can neither fall

nor be hidden. ”

The Greek scribe inserts a final ς where one should not be (υψηλούς
for υψηλού). Other than that, logion 32 has only one addition: an expan­
sion in the Greek text from “on a high mountain" (2ΐκΰ ογτοογ eqxoce)
to “upon the top of a high mountain” (επ ’ άκρου [δ ]ρους ύψηλου). Lacking
any rationale for the longer text, we should prefer the shorter Coptic here.

P. Oxy. 1—Logion 33
NHCII 239.10-18 P. Oxy. 1.41-42

nexe i(hcoy)c ||||κΝχοωτΉ ©poq λέγα Ί(ησοϋ)^· άκούβις [eJis τό ||

ί

2M ΠΘΚΜΚΚΧ© ώτίον σου το [ντο κήρυζον ...]

Jesus said, will hear in Jesus said, “You hear in your g|jg

your ear... ” ear... ”

Tim Ricchuiti Chapter 5 / 219

 TRACKING THOMAS

This is one of those logia that could easily be interpreted in a number

of ways, without any concrete solutions.59 It is possible that the addi­
tion of Zv (“one”) in the Greek could simply be implied in the Coptic
phrase (“your ear ... other ear"). It is also possible that it was deleted
from the Coptic or that it is a secondary expansion. In this case, we will
remain without a conclusion.

Analysis

Overall, there are nineteen variants to deal with between the text
of P. Oxy. 1 and logia 26-33 and 77b of the Coptic Thomas. While
there is relatively the same amount of variants within these first two
papyri, the similarity between the two ends there. Whereas P. Oxy. 654
did not betray an affinity with any particular type of variant, P. Oxy. 1
has a clear tendency toward additions (ten additions within the Greek
compared to just six substitutions, two additions within the Coptic,
and one transposition).

Table 5.2. Variation of P. Oxy. 1 by Logion

Greek Coptic
Substitution Transposition Total
Addition Addition
26 1 0 1 0 2
27 1 0 3 0 4

28 1 0 2 0 3

29 0 0 0 0 ()

30+77b 1 1 2 1 5

31 1 0 0 0 1

32 1 0 1 0 2

33 0 0 1 1 2

Total 6 1 10 2

98. Most scholars (Attridge, “Greek Fragments," 121; DeConick, Original Gospel of
Thomas in Translation, 142-43; Plisch, Gospel of Thomas, 101) supply the missing
o to the Greek text, as it is certainly an unintentional deletion.
99. Plisch, Gospel of Thomas, 102: “The first sentence of Gas. Thom. 33 contains a text-
critical problem that one can try to solve in different ways without ever attaining
absolute certainty?

220 / Chapters Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

This weight toward expansion would, by itself, engender questions

regarding the Greek’s authenticity; as the preceding analysis shows,
those questions would be well founded. Of the nineteen variants, the
Coptic preserves the more primitive reading fourteen times (74%), while
the Greek preserves the original in only four cases (21%). “This is a re­
verse of the case for P. Oxy. 654. This in itself should be surprising, con­
sidering the unanimous regard in which the Greek fragments are held
as a whole, but it is even more surprising that the main motivation for
the Greek text’s preservation of the original wording appears to be theo­
logical (the deletion of του 6eoC from L 27 and the transposition of 1.
77b). Admittedly, this is a small sample (two out of four instances), but it
is nevertheless compelling because of the stark contrast between P. Oxy.
1 and the other two papyri.

P. Oxy. 655—Logion 24
NHC U 2.383-10 P. Oxy. 655 (Fragment d)

enyonoc stkmh^y enei τ^η^γκη

NAY ngTPYH

MAPgQ 0-ΟΠ*η ογη ογοβίΝ cpoon

μΨουηΊϊΝουρηουοθπί λ.γω qp
... φώ$ €sJtlv [ev άνθρώπφ

ογοβικ βπκοοΜΟΟ τι-ipq eq-rnp

φ]ωτ€ΐυω, [έν δλω τω κ]6σμψ

oyoem ογκ^κβ ηβ
[φωτίζει· cl δέ μ)η, [σκοτξίνόε €]στιν.

.there isli^rwitluntheman of ... there is light in the man full of

light and he becomes light to the light, he gives light to the whole

whole world. If he does not become world. But if not, he is darkness.

light, he is darkness.

The fragmentary nature of P. Oxy. 655 (eighteen visible characters)

prevents any analysis of the differences between the Greek and die
Coptic in logion 24.

Tim Ricchuiti Chapter 5 / 221

 TRACKING THOMAS

P. Oxy. 655—Logion 36
NHCII 2.3924-27 P. Oxy. 655, col. 1.1-17

nexe i(hcoy)c Mwqi ροογα; xin

[λέγει Ί(ησοΰ)?- μή μεριμνάτε ά]πό

2τοογβ ρογ2β &γα> xin

ιτρωί ε[ως όψέ, μήτ]ε άφ’ έσπ[έρας·

2ΐρογ2β φχ 2τοογβ χβ ογ

neeniM’^ci 12ΐωτ τκγτΉ μήτ]€-Γ στϊολήΰμώ^-'ι

Λ
λη,χ
: κρέί [σσρ

Jesus said, “Do not be concerned Jesus said, “Do not be concerned

from morning until evening andfrom from morning until evening, nor from

evening until morning about evening until morning,

you will war·

•&e^ou^oia-,c&fy3iQ

There are two additions and a substitution to resolve for logion

36. The substitution is a misspelling within the Coptic manuscript:
it reads neeTN2<T^q where it should read The two
πθτθτνλ.τχλχι.

222 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

additions represent, by far, the biggest chunk present in the Greek but
missing in the Coptic. However, this is another instance where the
canon on the primacy of the shorter text will be set aside. While the
Greek does line up better with the biblical parallels (Matt 6.25-30;
Luke 12.22, 27-28), it only just manages to recall the voice of those
parallels, to say nothing of the very words. Therefore, it is not a pre­
dictable adjustment on the part of a scribe or a purposeful expansion/
reconstruction of tire text behind the Coptic. Instead, it makes more
sense to view the expanded text as original to Thomas and to consider
the deletion an act of clarification/correction. The rationale for such
a deletion appears to be an attempt by the Coptic scribe to resolve
an apparent discrepancy between the recommendation of this logion
(“don’t worry so much, God will provide your clothing”) and the rec­
ommendation of the next (“you will not see the Son until you can strip
naked, getting rid of your clothing’’).100 If the readers of Thomas un­
derstood the “garments/clothing" to represent the flesh (as in 1. 37),
the motivation behind such an alteration becomes clear: we have yet
another instance of theological alteration.101

P. Oxy. 655—Logion 37
NHCII 2.39.27-40.2 P. Oxy. 655, col. I.17-col2.1

nsxe weqw^HTHC xe Η2θογ λέγουσιν αύτφ οι μαθητα 1 αύτου·

βΚΝ^ογωκ^ esoA. hkn λ.γω πότβ ήμβΐν βμφανή^ έσ€ΐ, καί ττότ€ σ€

Ν2θογ βΗΜ4χγ epox nexe όψόμβθα;

xe βΤΘΤΗ^ΚβΚ λέγει· όταν έκδύσησθ€ καί μή

ΤΙΐγΤΗ ©2UY Ηπβτπφιπ© λ,γα) ίσχυνθήτ€

100. As 1 37 exists in P. Oxy. 655 as well, this same problem would have been present for
the Greek scribe. This may offer a glimpse into Thomas in transition: 1.37 had ac­
creted at this point (within DeConick’s model, e.g. [Recovering the Original Gospel
of Thomas]), but the longer text of 1. 36 had not yet been adjusted to make the
two logia cohere. Perhaps, though, the existence oil. 37 in the Greek fragment is
evidence against my view, in which case it would be better to go with Plisch’s sug­
gestion that the shorter Coptic text is more in line with other sayings (Gospel of
Thomas, 105-7).
101. It would go against the theology of Coptic Thomas to have the flesh as "a gift from
God that one had to put on” (DeConick, Original Gospel of Thomas in Translation,
149).

Tim Ricchuiti Chapters / 223

 TRACKING THOMAS

ΈΗ99Υ_Τ9Τ9ίΤ$Ρ^ΛΧ]-9ΠΦΰΕ9

Ήπβχρ Hj jsY<Q t stn x? 2OTG kN

[ουδέ φοβη]θ[ήσβσθβ].

His disciples said, ‘When will you His disciples said “When will

appear to us? And when will we see you become visible to us, and when

you? ” will we see you?

said, ‘When you strip He said, ‘When you disrobe and are

yourselves naked without being not ashamed,..

ashamed...

... you will not be afraid.” ... you will not be crfraid, ”

Much has been written about the theology and meaning of logion
37, particularly as it relates to the view of physicality for the Ihomasine
community.102 We will not be interacting with such works here, as the
variation between the Greek and Coptic is easily explained on other
grounds. Logion 37 contains a Coptic addition and a Greek addition,
though neither appears to be authentic. The addition in the Greek of
αύτω and the addition in the Coptic of each appear to be an
ihcoyc

expansion toward clarity.

P. Oxy. 655—Logion 38
NHCII 2.40.2-7 P. Oxy. 655, col. 2.2-11

Π6Χβ l(HCOY)C Χβ 2A.2 NCOH λέ[γ€ΐ Ίζησοΰ)?· πολλάια^] ο[ύν

kTsTHpeni-o-YMei cccdth έιτ€θυμ£ΐτ€ ] τ[ούτου$ τού? Xo}y[ous

kNeeitpkxe Nkei β-[χω“ΗΜθογ μου άκοΰσαι] κα[ι έχ€Τ€ ov8e]v[0s

NHTN χγω MNTNTN KCOYk SCOTHOY άλλου άκοδσαι] κα[ι έλ€υσονται]

ΈτοοτΖϊ ογΗ ^κ^οογ ήμ[έραί δτ€ ζητή]σ€[τέ μ.€ και ού μή

102, See, e.g., April D. DeConick and Jari Fossum, "Stripped before God: A New Inter­
pretation of Logion 37 in the Gospel of Thomas,” VC 45 (1991): 123-50; Marvin
W. Meyer, "Seeing or Coming to the Child of the Living One? More on Gospel of
Thomas’ Saying 37" HTR 91 (1998): 413-16; Gregory J. Riley, "A Note on the Text
of Gospel of Thomas 377 HTR 88 (1995): 179-81.

224 / Chapter 5 Tim Ricchuiti

 A Text- Critical Look at the Transmission of the Gospel of Thomas

“Πτβϊκφι He Ήοω ei τβτΝ^2© xh ευρήσετεμε].

βροβι

Jesus said, “You desired many times [Jesus] said ...

to listen to these words that I speak

to you, and you do not have another

from whom to hear them. There will

be days when you seek me, (but) you

will notfind me. ”

There are no significant differences between the fourteen charac­

ters of Greek and the Coptic in logion 38.

P. Oxy. 655—Logion 39
NHC II 2.40.7-13 Ρ. Oxy. 655, col. 2.11-23

nexe i(hooy)c Μψ^ριο^ιοο μη

[λέγει Ί(ησου)ς· οί Φαρισαίοι και οι

ΈΓρ^ΗΜ^τβγο ά.γχιΉφχφτ γραμματείς] ελ[αβουτάς κλείδας]

Ήτιίόοιο ^Υ2οπογ ογτβΤιπογΒωκ τής [γνώσεως. αυτοί έ]κρυψ[αν

β2ογπ ^γιυ Η©τογα>φ eeu)K β£ογΝ αύτάς. ούτε] εισήλ[θον, ούτε τούς]

*ΜΠογκλ^γΈτωτΉ Λβ φωπβ είσερ [χαμένους άφή]καν [είσελθε ιυ.

“<bpoNiMOCTteeTiH2oq λ.γω υμείς] δε γεί[νεσθε φρόνι]μοι ώ[ς

ΉλΚβρΜΟΟΪΪ-^θΈΝόρΟΜΠβ δφεις κα'ι ά]κέραι[οι ώς

περιστε]ρα[ί].

Jesus said, “The Pharisees and the Jesus said, ‘The Pharisees and the

scribes took the keys of knowledge. scribes took the keys of knowledge.

They hid them. Neither did they go in, They hid them. Neither did they go in,

nor did they pernut those who wished nor did they permit those who were

to go in to do so. You, however, going in to go in. But you become

become wise like serpents and wise like serpents and innocent like

innocent like doves. ” doves. ”

Tim Ricchuiti Chapter 5 / 225

 TRACKING THOMAS

Although there is some slight variation between, the Greek present

participle τούς εισερχόμενους and the Coptic phrase Νβτογωιρ θβμκ

®2ΟΥΝ in logion 38, it does not rise to the level of variation. It is much
more likely to be due to translational issues than to scribal alterations.

Analysis
With only six variants to analyze, it is a bit precocious to charac­
terize the merit of P. Oxy. 655. While the variants of logion 37 were both
secondary, the variants of logion 36 demonstrate the more primary
character of P. Oxy. 655 well. Additionally, it is probably for theological
considerations that such a large amount of material was deleted from
logion 36, and so, in this, even the especially small sample of P. Oxy. 655
demonstrates our thesis concerning Thomas's theological corruption.

Table 5.3. Variation of P. Oxy. 655 by Logion

Substitution Greek Addition Coptic Addition Total

24 0 0 0 0

36 1 2 0 3

37 0 1 1 2 1
38 0 0 0 0

39 0 0 0 0

Total 1 3 1

CONCLUSION

It was my purpose here (1) to conduct a comparison of the Greek

fragments of the Gospel of Thomas to the full Coptic manuscript, (2)
to assess the merit of the four manuscripts containing Thomas with
respect to their originality, and (3) should the assessment yield fruit,
to draw a few conclusions on the scholarly consensus regarding both
the character of the Greek fragments versus the Coptic text and the
amount of theological alteration present, particularly in Thomas, but
more generally in noncanonical works as a whole. Considering the
preceding analysis of the Gospel of Thomas in both Greek and Coptic,
it appears that scholars are on fairly solid ground when it comes to the
assumption that the Greek represents an earlier strain of Thomas. The

226 / Chapter 5 Tim Ricchuiti

 A Text-Critical Look at the Transmission of the Gospel of Thomas

data behind P. Oxy. 654 and 655 certainly bear this out. Of the twenty-
three variants, I judged the Coptic to represent the earliest form of the
text only three times (13%), while the Greek text preserved the earlier
text sixteen times (70%). The Greek text is slightly more expansive, but
even that meager tendency disappears when considering the massive
addition of logion 36.

Table 5.4. Variation among P. Oxy. 654,655

Substitution Greek Addition Coptic Addition Total

P. Oxy. 654 5 7 6 18

P. Oxy. 655 1 1 3 1 5

Total 6 10 7 23

However, my initial hypothesis (and some of those scholarly as­

sumptions) must be adjusted in light of P. Oxy. l's clearly secondary
character. As has already been noted, the Coptic text preserves the
original reading nearly four times as often as the Greek text (fourteen
instances compared to four instances). There is a strong tendency
here for the Greek to be more expansive—the text of P. Oxy. 1 ex­
pands on the Coptic text ten times, while the reverse happens only
twice.

Table 5.5. Variation w ithin P. Oxy. 1

Substitution Transposition Greek Addition Coptic Addition Total

1 6 1 10 2 19

Many of those expansions are due to the tendency of the scribe of P.

Oxy. I103 to add the connective καί. However, the secondary nature of
the Greek fragment is not limited to extra conjunctions: where there is
a substitution, the Coptic tends to preserve the original (five out of six
times).
As far as conclusions on the fluidity of Thomas goes, it should be
noted that the scribes who copied the gospel appear to fall victim to
many of the exact same traps that corrupted canonical texts. Beyond the

103. Or, of course, some precursor to P. Oxy. 1.

Tim Ricchuiti Chapter 5 / 227

 TRACKING THOMAS

obvious places for overlap, such as haplography, dittography, and so on

(errors due to physical limitations), a significant number of the variants
were due to harmonization with the relevant biblical parallels or other
tradition. At the same time, the Coptic scribe prefers, in at least some
cases, secondary readings to theologically difficult readings. That prefer­
ence accounted for two of the four places in P. Oxy. 1 where the Greek text
was judged to be original (11. 27, 77b). Additionally, it can be argued that
the changes demonstrate some theological alterations due to praxis (e.g.,
the alteration of dietary restraints in 1.6) and others due to pistis (e.g., the
massive deletion in the case of 1. 36, the alterations of "Kingdom of G od”
phrases in 11. 3 and 27, and possibly the alteration of 1. 2). Altogether, al­
terations in the Coptic text evidence theological alteration in at least six
places (11.3,5-6,27,36,77b),104 irrespective of which fragment it occurs in
or the fragment's overall character. It does indeed appear that the Coptic
scribe altered Thomas in such a way as to make it more amenable to the
community that eventually decided to include it in the Nag Hammadi
writings.
Ultimately, there were limitations in the present study, with only
a little over a thousand words to analyze where the four manuscripts
overlap. But there remain hidden avenues and alleyways yet unex­
plored. In the first place, a more in-depth comparison of the transmis­
sion of Thomas in these four manuscripts to the transmission of certain
NT passages in comparable manuscripts (e.g., a few early fragments
checked against a later, more complete manuscript) could generate
substantive conclusions about the relative level of reverence the works
were held in by the early Christian communities, notwithstanding the
work of some church fathers to discourage their use or ban certain
texts outright.105 The land of comparison undertaken here of tire extant
manuscripts of Thomas could be performed on other early Christian
literature to determine if Thomas is an outlier in its transmission or if it
is in line with what we would expect for noncanonical texts (or, indeed,
canonical texts). In any case, the present study ought to demonstrate
that textual criticism has a role to play beyond just the New Testament,
illuminating issues that might otherwise be missed.

104. The process toward “rest” of 1.2 accounting for the "at least."
105. E. J. Epp ("The Oxyrhynchus New Testament Papyri: ‘Not without Honor except in
Their Hometown’?," JBL 123 [2004]: 245-81) proposes something similar to this,
though without the same explicit purpose.

228 / Chapter 5 Tim Ricchuiti

 6

JESUS AS ΘΕΟΣ
A Textual Examination

Brian. J. Wright1

rom Aland to Zuntz, every major NT scholar has explored the

F canon of the NT for texts that call Jesus θβό?.2 While this may seem
like a painless pursuit with plenty of “proof passages," several stumbling
blocks quickly emerge? No author of a Synoptic Gospel explicitly

1. I dedicate this work to several people. First, I dedicate it to my mom, Debra L.

Wright, whose selfless love and perseverance continues to humble me. Second,
I dedicate it to a group of men—Ereke Bruce, Steve Llewellyn, Grant Mayfield,
Dustin Walker, et al.— to whom I am eternally grateful for first sharing the gospel
with me during my junior year of college and purposefully discipling me with an
interdenominational love in/through Christ. Finally, yet importantly, I also dedi­
cate this work to John R. Brown. Although lean point to many books, events, and
people that have helped form my cognitive view of Christianity, his life continues
to model for me the practicality of the Christian life as Christ commanded in Matt
5.16: "Let your light shine before others, so that they may see your good works and
glorify your Father who is in heaven.3
..Special.thanks are due to J. K. Elliot, Gordon D. Fee, P. J. Williams, Daniel B.
Wallace, Tommy Wasserman, Darrell Bock, and Chrys Carago unis for looking at a
preliminary draft of this manuscript and making valuable suggestions.
2. For a detailed list of many such views, see Daniel B. Wallace, Granville Sharp's
Canon and Its Kin: Semantics and Significance (Bern: Peter Lang, 2009), 27-28.
3. Bart Ehrman, in at least three published books and one published lecture se­
ries, even suggests that the Ausgangstext does not necessarily teach the deity
of Christ. He bases these allegations on alleged textual problems that he attri­
butes to manipulative scribal activity, most often pointing to textual problems
behind such verses. He almost exclusively leans toward the manipulation of
early proto-orthodox scribes in the development of a high Christology in his
book The Orthodox Corruption of Scripture: The Effect of Early Christological
Controversies on the Text of the New Testament (Oxford: Oxford University
Press, 1993).

229
 JESUS AS ΘΕΟΣ

ascribes the title θεό? to Jesus.4 Jesus never uses the term θεός for
himself.5 No sermon in the book of Acts attributes the title θεός to
Jesus.6 No extant Christian confession7 of Jesus as θεός exists earlier
than the late 50s.8 Prior to the fourth-century Arian controversy, no­
ticeably few Greek MSS attest to such “Jesus-θεός” passages.9 And
possibly the biggest problem for NT Christology regarding this topic

4. As Raymond Brown hypothesizes, “The slow development of the usage of the title
‘God’ for Jesus requires explanation.... The most plausible explanation is that in
the earliest stage of Christianity the Old Testament heritage dominated the use
of God’; hence, ’God’ was a title too narrow to be applied to Jesus.” I am uncon­
vinced that this is the “most” plausible explanation, given the predominately Jewish
context that may have dictated the early evangelistic terminology (e.g., Matthew’s
"kingdom of heaven”). Nevertheless, Brown adds, “[W]e do maintain that in gen­
eral the NT authors were aware that Jesus was being given a title which in the LXX
referred to the God of Israel” (Raymond Brown, "Does the New Testament Call
Jesus GodT TS 26 [1965]: 545-73). At the same time, "To reconstruct the history
of titles as if this were the study of christology is like trying to understand the win­
dows of Chartres cathedral by studying the history of coloured glass” (Leander E.
Keck, “Toward the Renewal of NT Christologyf NTS 32 [1986]: 368; whole article
on 362-77).
5. In fact, Mark 10.18 records that Jesus differentiates himself from God (= the Father;
cf. Matt 19.17; Luke 18.19; Mark 15.34; Matt 27.46; John 20.17). H. W. Montefiore,
in his essay "Toward a Christology for Today,” notices this as he postulates that
Jesus seems to have explicitly denied that he was God (Soundings 45 [1962]: 158),
In addition, R. H. Fuller believes, similar to Bultmann, that Jesus understood him­
self as an eschatological prophet (Reginald H. Fuller, The Foundations ofNew Tes­
tament Christology [New York: Charles Scribner’s Sons, 1965], 130). While none
of these texts or interpretations portray a complete NT Christology (Jesus does
identify himself with God [e.g., John 10.30; 14.9], he never explicitly rejects that he
is God, and he understood himself to be more than an eschatological prophet), it is
true that Jesus never uses the term 0€0? for himself.
6. Acts 20.28 is in a speech (and the only one) addressed to a Christian audience. "All
the others,” John Stott explains, “are either evangelistic sermons,... legal defenses,
... or the five speeches before the Jewish and Roman authorities” (John R. W.
Stott, The Message ofActs: The Son, the Church, and the World [Downers Grove, IL:
InterVarsity, 1990], 323). Cf. Richard I. Pervo, Acts: A Commentary (Minneapolis:
Fortress, 2009), 515-31.
7. Raymond Brown, however, insightfully notes a danger in judging usage from oc­
currence, because NT occurrence does not create a usage but testifies to a usage al­
ready existing (Raymond E. Brown, Jesus: God and Man [Milwaukee: Bruce, 1967]}.
None of the passages considered in this chapter give any evidence of innovating.
8. With Rom 9.5 probably occurring first, if one could be certain of its punctuation/
grammar (see discussion following).
9. In a recent popular book, ReinventingJesus, the authors note that “there are at least
forty-eight (and as many as fifty-nine) Greek New Testament manuscripts that
predate the fourth-century.” In an endnote, the authors go on to explain that these
are only Greek New Testament MSS and do not include the early versions or the
pre-fourth-century patristic writers. Even so, only four "Jesus-9eo$w passages (Rom
9.5; John 1.1,18; 20.28) are included in these MSS (J. Ed Komoszewski, M. James
Sawyer, and Daniel B. Wallace, Reinventing Jesus: How Contemporary Skeptics Miss
the Real Jesus and Mislead Popular Culture [Grand Rapids: Kregel, 2006], 116).

230 / Chapter 6 Brian J. Wright

 A Textual Examination

is that textual variants exist in every potential passage where Jesus is

explicitly referred to as θβός.10
This plethora of issues may provoke one to repeat, for different rea­
sons, what a Gnostic document once confessed about Jesus, “Whether
a god or an angel or what I should call him, I do not know.”11 Yet “it
was the Christians’ habit on a fixed day to assemble before daylight and
recite by turns a form of words to Christ as a god,” Pliny the Younger
wrote in a letter to Emperor Trajan about Christians.12 “We must think
about Christ as we think about God,” the author of 2 Clement opens his
homily. “I bid you farewell always in our God Jesus Christ,” concludes
Ignatius in his letter to Polycarp.13 “True God from true God,” the first
ecumenical council ultimately dogmatized concerning Jesus.14 When,
then, did this boldness to call Jesus 6eos begin?15

10. The authors of ReinventingJesus note (114), “If a particular verse does not teach the
deity of Christ in some of the manuscripts, does this mean that that doctrine is sus­
pect? It would only be suspect if all the verses that affirm Christ’s deity are textually
suspect.” Unfortunately, regarding the explicit “Jesus-Qeos” passages, that may be
the case here. At the same time, the authors continue, “And even then the variants
would have to be plausible/ This further reveals the importance of this study.
11. Infancy Gospel of Thomas 7.4, from tire Greek text of Constantin von Tischendorf,
Evangelia Apocrypha (Hildesheim: George Olms, 1987; original, Leipzig, 1867).
For a more recent text-critical work on it, see Tony Chartrand-Burke, “The Greek
Manuscript Tradition of die Infancy Gospel of Thomas,” Apocrypha 14 (2003):
129-51.
1.2. Pliny: Letters and Panegyricus, vol. 2, Letters Books VIIl-X, Panegyricus, trans.
Betty Radice, Loeb Classical Library (Cambridge, MA: Harvard University Press,
1969), 288-89.
13. "Ignatius effortlessly and spontaneously wove within his understanding of the re­
lationship between the Father and the Son the simple and unequivocal proclama­
tion that Jesus Christ is God” (Thomas Weinandy, “The Apostolic Christology of
Ignatius of Antioch: The Road to Chalcedon," in trajectories through the New Tes­
tament and the Apostolic Fathers [New York Oxford University Press, 2005], 76).
Here are fourteen such occurrences in Ignatius: Eph. prol., 1.1,7.2,15.3,18.2,19.3;
Rom. prol. (2x), 3.3,6.3; Smyrn. 1.1,10.1; Trail. 7.1; Pol. 8.3.
14. θεόν αληθινόν έκ θεοΰ αληθινού (Philip Schaff, lhe Creeds of Christendom: A His­
tory and Critical Notes, vol. 2 [New York: Harper and Brothers, 1877], 57).
15. -I am-discussing the origin of the title Seos as applied to Jesus and not the origin
of understanding Jesus as divine. That understanding was early and expressed in
various ways (see, among others, C. F. D. Moule, The Origin of Christology [Cam­
bridge: Cambridge University Press, 1977]; Larry Hurtado, Lord Jesus Christ: De­
votion to Jesus in Earliest Christianity [Grand Rapids: Eerdmans, 2003]). As for
the title Qeos, "On the one hand, the dominant Greco-Roman ethos assumed that
there were many gods and that human beings could be deified. Many emperors
refused to be called gods during their lifetimes, yet were named gods after their
deaths. The term‘god’ was also used for living rulers, like Agrippa (Acts 12:21-22;
Josephus, Ant. 19.345) and Nero (Tacitus, Annals 14.15). On the other hand, the
Jewish tradition centered on faith in one God (Deut 6:4), who was not to be por­
trayed in human form or to be identified with, a human being (Exod 20:4; Deut 5:8;
2 Macc 9:12; cf. John 5:18; 10:33)” (Craig R. Koester, Hebrews: A New Translation
with Introduction and Commentary [New York: Doubleday, 2001], 202). Further,

Chapter 6 / 231
 JESUS AS ΘΕΟΣ
i

CONDENSED EXAMINATION

Although this work will examine the textual certainty of every po­
tential NT ascription of θεός to Jesus,16 ten of the possible seventeen
passages will be dismissed up front for the following reasons:17

1. Punctuation. Romans 9.5 involves a punctuation issue “which

our earliest manuscripts do not answer.”18 Moreover, even if the absence
of any discernible type of standardized punctuation cannot be definitively
traced back to the earliest Greek NT MSS, "the presence of punctuation in
Greek manuscripts, as well as in versional and patristic sources, cannot be

one should note that the majority of passages in which Jesus is potentially called
fieos appear in writings attributed to Jewish settings, whereas only a few might be
Pauline (see, e.g., Richard N. Longenecker, The Christology ofEarly Jewish Christi­
anity (Naperville, IL: Allenson, 1970], 139).
16. 1 will employ a reasoned eclecticism method, the currently reigning view among
textual critics. Several limitations exist, however, on the scope of my research. For
example, 1 did not exhaustively examine each critical apparatus to find other vari­
ants that potentially affirm Jesus as fteos. I did not work extensively with foreign
literature. I relied heavily on the manuscript collations of others. I created no com­
prehensive comparative analysis of the manuscript relationships for the Pauline
corpus or for any individual book(s). I did not determine the scribal habits of every
MS or witness cited. I also depended heavily on those whose academic acumen
regarding textual criticism far exceeds mine and whose scholarly contributions are
highly regarded.
17. A handful of other verses are sometimes used to implicitly equate Jesus with 6e6s
(Luke 8.39; 9.43; 1 Ihess 4.9; 1 Tim 1.1; 5.21; 2 Tim 4.1; Titus 13; 3.4$ Heb 3.4; James
1.1), yet 1 did not think enough academic support existed to merit their inclusion
in this work
18. Douglas J. Moo, "The Christology of the Early Pauline Letters/ in Contours of
Christology in the Afew Testament (Grand Rapids: Herdmans, 2005), 190. Simi­
larly, Ehrman concludes, "Nor will 1 take into account variant modes of punc­
tuation that prove christologically significant, as these cannot be traced back to
the period of our concern, when most manuscripts were not punctuated” (Or­
thodox Corruption of Scripture, 31). C£ Robert Jewett, Romans (Minneapolis:
Fortress, 2007), 555, 566-69; Bruce M. Metzger, A Textual Commentary on the
Greek New Testament, 2nd ed. (Stuttgart: German Bible Society, 1994), 459-62;
Murray J. Harris, Jesus as God: The New Testament Use of Theos in Reference to
Jesus (Grand Rapids: Baker, 1992), 150-51; and Walter Bauer; A Greek-English
Lexicon ofthe New Testament and Other Early Christian Literature, ed. Frederick
William Danker, 3rd ed. (Chicago: University of Chicago Press, 2000 [hereinafter
BDAGJ), s.v. “θ€0ε."
Ihe earliest MS of Romans to date (i|54*, ca. 200 [cf. Kurt Aland, Kurzgejasste
ListedergriechischenHandschriften desNeuen Testaments, 2nd ed. (Berlin: Gruytei;
1994), 31-32]) does not contain any punctuation here. Nevertheless, Lattey shows
that a fifth-century codex (C/04) contains a small cross between σάρκα and q ών to
designate some form of a stop, which the NA27 and UBS4 texts reflect with a comma
(Cuthbert Lattey, “Ihe Codex Ephraemi Reseriptus in Romans bu 5,” 35
[1923-24]: 42-43).

232 / Chapter 6

Brian }. Wright
 A Textual Examination

regarded as more than the reflection of current exegetical understanding

of the meaning of the passage.”**

2. Extenuating circumstances. Although Colossians 2.2 contains

“no fewer than fifteen textual options,”19
20 the issue is syntax rather than
textual pedigree and is outside the scope of this investigation. Other
such extenuating circumstances hold true for the following:

Matthew 1.2321
John 17322
Ephesians S.52324
2 Thessalonians 1.12*
1 Timothy 3.1625

19. Metzger» Textual Commentary, 167. C£ H.-C. Kammler, "Die Pradikation Jesu
Christi als ‘Gott' und die paulinische Christologie; Erwagungen zur Exegese von
R6m 9,5b,” ZNW94 (2003): 164-80.
20. Douglas J. Moo, The Letters to the Colossians and to Philemon (Grand Rapids:
Eerdmans, 2008), 168. These options are listed conveniently in Bruce M. Metzger
and Bart D. Ehrman, The Text ofthe New Testament: Its Transmission, Corrup­
tion, and Restoration, 4th ed. (New York: Oxford University Press, 2005), 334.
21. The text is overwhelmingly certain here, as the author cites ha 7.14 in relation to
die birth of Jesus. Yet, despite its textual certainty, we cannot be sure that the evan­
gelist takes "God with us” literally and attempts to call Jesus Seos (as J. C. Fenton
concludes in “Matthew and the Divinity of Jesus: Three Questions concerning
Matthew 1:20-23,” in Studio, Biblica 1978, vol 2, Papers on the Gospels, ed, E. A.
Livingstone [Sheffield: JSOT Press, 1980], 79-82). See, among others, R. T. France,
The Gospel ofMatthew (Grand Rapids: Eerdmans, 2007), esp. 49-50,56-58.
22. Note the discussion of the grammatical issues relating to this phrase in Harris,
Jesus as God, 258-59. The text, nonetheless, should be considered certain.
23. The textual evidence is solid here. Ehrman accurately explains, flIn the text
that is almost certainly original (‘the Kingdom of Christ and God’), Christ
appears to be given a certain kind of priority over God himself. This problem
is resolved by all of the changes, whether attested early or late” (Ehrman, Or­
thodox Corruption, 269). See Harris, Jesus as God, 261-63, for grammatical
issues.
24. The textual issue in this verse does not pertain to the clause in question, leaving
• one with “two possible Greek genitive translations: (1) "according to the grace of
our God and Lord, namely Jesus Christ” or (2) "according to the grace of our God
and the Lord Jesus Christ.” 1 favor the latter (which does not attribute the title
(teos to Jesus), primarily for the following reason, "Second Thessalonians 1:12 does
not have merely ‘Lord’ in the equation, but Tord Jesus Christ.’ Only by detaching
κυρίου from ’Ιησού Χριστού could one apply [Granville] Sharp’s rule to this con­
struction” (Wallace, Sharp’s Canon, 236).
25. The attestation for the variants here is not strong enough to warrant serious con­
sideration (contra Stephen W. Frary, “Who Was Manifested in the Flesh? A Con­
sideration of Internal Evidence in Support of a Variant in 1 Tim 3:16a,” EFN 16
[2003]: 3-18). Towner notes, "the change to δ (D* and Vg plus some Latin Fa­
thers) was a gender adjustment to accord with τό μυστήριον; another late solution
was the change to Oeos (a1 A' C? D2 Ψ 1739 1881 TR vg""), which supplies the
antecedent thought to be lacking in os’ (Philip Towner, The Letters to Timothy

Chapter 6 ! 233

Erian J. Wright
 JESUS AS ΘΕΟΣ
>

Titus 2.13J6
1 John 5.2027
Jude 426

and Titus (Grand Rapids: Eerdmans, 2006], 278). C£ W. M. Zoba, "When Manu­
scripts Collide," CT 39, no. 12 (1995): 30-31. Cf. also Robert H. Gundry, “The Form,
Meaning, and Background of the Hymn Quoted in 1 Timothy 3:16/ in Apostolic
History and the Gospel: Biblical and Historical Essays Presented to F F. Bruce, ed.
W. Ward Gasque and Ralph P. Martin (Exeter: Paternoster, 1970), 203-22.
26. Though I strongly feel that this verse attributes the title θεός Co Jesus, a textual
examination is unnecessary, since the only viable variant concerns the order of the
last two words; Ίησου Χριστού or Χριστού ’Ιησού. The debate, then, will have to
continue congregating around syntax. See Gordon Fee, Pauline Christology: An
Exegetical-Theological Study (Peabody, MA: Hendrickson, 2007), esp. 442-46.
Against Fee’s position, see Wallace, Sharp's Canon, 256-64; Robert M. Bowman Jr.,
“Jesus Christ, God Manifest: Titus 2:13 Revisited,” JETS 51 (2008): 733-52; Robert
W. Yarbrough, 2-3 John (Grand Rapids: Baker, 2008), 320.
Several NT scholars put an asterisk by this book because they consider it deu-
tero-Pauline. Yet even if one assumes that Paul did not write Titus, it still would have
been written in the first century and, therefore, would be impervious to some of the
critiques often given for such texts, such as orthodox corruption^) due to the third-
century Arian controversy. As a matter of fact, although Ehrman did not mention
Titus 2.13 specifically in Orthodox Corruption, by his own argument regarding 2 Pet
1.1, Titus 2.13 would explicitly equate Jesus with θίός: "Because the artide is not re­
peated before ’Ιησού (in 2 Pet 1:1), it would be natural to understand both bur God’
and ‘Savior1 in reference to Jesus [our 'God and Savior’]” (Orthodox Corruption, 267).
In other words, Ehrman recognizes that one artide with two nouns joined by καί
refers to the same person, making Titus 2.13 an explicit reference to Jesus as
27. Of the two notable variants in this verse, neither of them effectually touches our
present topic. The crux interpretum is the antecedent of ούτος, but it is not dear
whether it represents a reference to God the Father or Jesus Christ (see Wallace,
Sharp's Canon, 273-77 for a discussion of the syntax of 1 John 5.20). Even so, Au­
gustine used this verse to support his argument that Jesus was "not only God, but
also true God” (Ike Trinity: Introduction, Translation, and Notes [New York: New
City, 2000), 71). Likewise, Rudolf Schnackenburg argues strongly, from die logic of
the context and the flow of the argument, that “This is the true God* refers to Jesus
Christ (John E. Rotelie, Die Johannesbriefe, in Herders theologischer Kommentar,
2nd ed. [Freiburg: Herder, 1963], 291). Stephen Smalley notes, "But even if we do
not accept the equation (Jesus as God) as explicitly present in this verse, it remains
true that there is an association between God and his Son that is articulated here
more clearly than anywhere else in 1 John” (Stephen S. Smalley, 1,2,3 John [Nash­
ville: Thomas Nelson, 2007], 295). Cf. also Judith M. Lieu,X II, and III John: A Com·
mentary (Louisville: Westminster John Knox, 2008), 233-34$ Peter Rhea Jones, 1,2
& 3 John (Macon, GA: Smyth and Helwys, 2009), 231-35.
28. I kept tliis text In the condensed list primarily because several MSS contain the word
θβόν (e.g., K L P S Ψ 049 104 syrh,pb). Landon persuasively argues that the internal
evidence supports δεσπότην θ€0ν rather than simply δεσπότην and that the expres­
sion refers only to God (Charles Landon, The Text ofJude and a Text-Critical Study
of the Epistle of Jude, JSNT Supplement 135 [Sheffield: Academic, 1996], 63-67).
What makes his argument strong is that if Ehrman is correct about die direction
of corruption away from adoptionistic heresies, noting the text cf 2 Pet 1.2 in p72,
this reading alone resists orthodox interference (i.e., shortening by scribes who wish
to show God and Jesus as the same entity, thereby stressing Christ's divinity). Yet

234 / Chapter6

Brian J. Wright
 A Textual Examination

Ibis leaves seven texts warranting extended examination.

EXTENDED EXAMINATION

John LI*29

According to Aland’s Kurzgefasste Liste, the Gospel of John has

even with Landon’s well thought-out thesis, of which 1 did not list all his perceptive
reasons, I still reject the longer reading for the following reasons: (1) the earliest and
best MSS support die shorter reading [e.g., P72 ϊ*78 R A B C 0251 33 1739 Lectpt it*
vg cop”,bo geo], (2) it is probable that a scribe sought to clarify the shorter reading
and/or stay within the NT’s normal pattern (i.e„ Luke 2.29; Acts 4.24; 2 Um 2.21;
Rev 6.10)» and (3) it is the more difficult reading. Therefore, my preference is for the
shorter reading: δεσπότην (used ofGod in Luke 229, Acts 4.24, and. Rev 6.10 and of
Christ in 2 Pet 2.1 and here). For exhaustive MS evidence, see Tommy Wasserman,
The Epistle of Jude: Its Text and Transmission (Stockholm: Ahnqvist und Wlksell,
2006), esp. 251-54. Cf. also G A. Albin, Judasbrevet: Traditionen, Texten Tblkningen
(Stockholm, 1962), 148,596.
In addition, the shorter reading in Jude 4 (where Christ is described as the
ruling Master, δεσπότην) would comport well with Jude 5 if*Jesus" is indeed the
original reading. This would clearly highlight the pre-existence of Christ and thus
implicitly argue for his deity. Therefore, both verses taken together make a compel­
ling argument for the pre-existence, as well as the deity, of Jesus Christ (without
giving the tide θεό? to Jesus). For in-depth textual discussion of Jude 5, in which
the author argues for Ιησού? here, see Philipp E Bartholoma, "Did Jesus Save the
People out of Egypt? A Re-Examination of a Textual Problem in Jude 5,” NovT 50
(2008): 143-58. For an opposing view on Jude 5, see James R. Royse, Scribal Habits
in Early Greek New Testament Papyri, New Testament Tools, Studies, and Docu­
ments 36 (Boston: Brill, 2008), 610-12.
29. I recognize that the anarthrous (teos denotes the pre-existent λόγο? and not explic­
itly Jesus (yet?). I also acknowledge that some scholars have argued well that John LI
is a part of the hymn exalting God’s σοφία (the ?todh of Prov 8; c£ Sir 1.1-10) and/
or have shown that Philo periodically uses the term θεό$ without the definite article
for λόγο? (e.g., Sown. 1.230). Nevertheless, without taking the referent for λόγο$ for
granted (even though, e.g., σοφία is never designated die title θεός and though Phi­
lo’s over 1,300 uses of λόγο^ are systematically different from John’s meaning), I still
believe the pre-existent λόγο? eventually points to Jesus, the λόγο$ incarnate (i.e.,
John 1.14,17; cf. Rev 19.13), and therefore pertains to this chapter’s examination.
For similar (recent) conclusions about the pre-existent λόγο? eventually pointing
to Jesus, see, among others, Martin HengeL "The Prologue of the Gospel of John as
the Gateway to ChristologicaJ Truth,” in The Gospel ofJohn and Christian Theology,
ed. Richard Bauckham and Carl Moss er (Grand Rapids: Eerdmans, 2008), 271$ Uwe-
Karsten Plisch, The Gospel of Thomas: Original Text with Commentary (Stuttgart:
Deutsche Bibelgesellschaft, 2008), 76-77; Moo, Colossians and the Philemon, 118;
Andreas J. Kostenberger and Scott R. Swain, Father, Son, and Spirit: The Trinity and
John's Gospel, ed. D. A. Carson (Downers Grove, IL: InterVarsity, 2008), 113; Douglas
W. Kennard, Messiah Jesus: Christology in His Day and Ours (New York: Peter Lang,
2008), 503; Petr Pokorny, A Commentary on the Gospel of Thomas: From Interpreta-
dons to the Interpreted (New York: T&T Clark, 2009), 16.

Chapters / 235

Brian J. Wright
 JESUS AS ΘΕΟΣ

more papyrus fragments than any other book of the NT.30 Surprisingly,
though, neither UBS* nor NA27 list any variants for John 1.1c. Only three
major published NT Greek texts—those of TischendorfR, Merk, and von
Soden—even list textual variants in their apparatus (with 100 percent una­
nimity as to its Ausgangstext: και 0e6s ήν ό Xoyos).31 No textual debates
on John 1.1c exist in any standard work on Jesus-9e0s passages, and until
eighteen years ago,32 NT textual critics were unanimous in their certainty
of John 1:1c. This scholarly agreement continues today even though one
textual critic, Bart Ehrman, stated his reluctance to dismiss the testimony
of a single eighth-century Alexandrian manuscript, L.33 To Ehrman, an
articular θεός gives him the “distinct impression” that the orthodox party
changed it due to the Arian controversies.34 In other words, Ehrman
points out that an articular Geos possibly makes this otherwise implicit
identification (Jesus as simply divine) an explicit one (God himselfJ.35
Although the most probable understanding of the anarthrous 0e6s is
qualitative (the Word has the same nature as God),36 three points concern
us here textually:

30. Aland, Kurzgefasste Lists. 29-33. This statistic was without the benefit of many
more John papyrus fragment discoveries to date (see, e.g., J. K. Elliott, “Five New
Papyri of the New Testament,’ NovT41 (1999]: 209-13; Elliott, “Four New Papyri
Containing the Fourth Gospel and Their Relevance for the Apparatus Criticus,"
JETS 59 [2008]: 674-78; Peter Head, Ψ Bodmer II (J)66): Three Fragments Identi­
fied; A Correction,” NovT 50 [2008]: 78-80).
31. This unanimity continues today, e.g., in such specialized (Le., single book) text-
critical works as the IGNTP edition of the Gospel of John (i.e., The American and
British Committees of the International Greek New Testament Project, The New
Testament in Greek IV, The Gospel according to St, John, voL 1, The Papyri [New
York: BrilL 1995], 123; vol 2, The Majuscules [Leiden: Brill, 2007], 189).
32. Reference is here made to the publication year (1993) of Bart Ehrman’s Orthodox
Corruption.
33. Ehrman, Orthodox Corruption, 179 n< 187.
34. Ibid.
35. Ibid. John could have used (e.g., Acts 17.29; 2 Pet 1.3,4) or some other word
meaning “divine," had he wished to convey Jesus as simply divine. Keener helpfully
points out, “Regarding Jesus as merely 'divine' but not deity violates the context;
identifying him with the Father does the same. For this reason, John might thus
have avoided the article even had grammatical convention not suggested it; as a
nineteenth-century exegete argued, an articular θεός would have distorted the
sense of the passage, ‘for then there would be an assertion of the entire identity of
the Logos and of God, while the writer is in the very act of bringing to view some
distinction between them... .' Scholars from across the contemporary theological
spectrum recognize that, although Father and Son are distinct in this text, they
share deity in the same way” (Craig S. Keener, The Gospel ofJohn: A Commentary
[Peabody, MA: Hendrickson, 2003], 374).
36. Contra Modalism/Sabellianism (and the Jehovah’s Witnesses rendering of John
1.1c in their New World Translation). Philip Harner, after probing the fourth
Gospel for passages that use predicate nouns, points out that the qualitative force
of the predicate is more prominent than its definiteness or indefiniteness in forty
of the fifty-three cases that use anarthrous predicates preceding the verb. He notes

Brian J. Wright
236 / Chapter 6
 A Textual Examination

1. Both ϊ>75 and Codex B attest to the absence of the article in John
1.1c. This is significant since “[tjhese MSS seem to represent a 'relatively
pure’ form of preservation of a 'relatively pure’ line of descent from the
original text.”37 Kenneth W. Clark concludes, “(I]t is our judgment that
$>7S appears to have the best textual character in the third century.”38
Ehrman concurs, “Among all the witnesses, $>75 is generally understood
to be the strongest.”3940Thus this evidence significantly strengthens our
initial external examination in favor of an anarthrous θβός.

2. Only two MSS (L and Ws) contain the articular 0e0s in και Q
0eos ήν ό λόγος·.*3 In addition, these two MSS are late (eighth century)41
and have never produced a reading that has found acceptance into the
base text of the NA27 or UBS4 without the support of better and earlier
MSS. In fact, as Matthew P. Morgan points out regarding Regius (L),
the article with θεός in John 1.1c represents the only sensible variant
involving a single letter in all (53) of this scribe’s singular readings. The
best explanation for the addition of the article is the sloppy scribal be­

specifically, "In John 1:1 I think that the qualitative force of the predicate is so
prominent that the noun cannot be regarded as definite” He also suggests, “[TJhe
English language Is not as versatile at this point as Greek, and we can avoid misun­
derstanding the English phrase only if we are aware of the particular force of the
Greek expression (hat it represents* (“Qualitative Anarthrous Predicate Nouns:
Marik 1539 and John l:lf 92 (1973): 87; whole article on 75-87). CL J. G.
Griffiths, "A Note on die Anarthrous Predicate in Hellenistic Greek,” ExpTlm 62
(1950-51), 314-16; Robert W. Funk, "The Syntax of the Greek Article: Its Impor­
tance for Critical Pauline Problems” (PhD diss., Vanderbilt University, 1953), 148;
Robertson, (A. T. Robertson, A Grammar ofthe Greek New Testament in the Light
ofHistorical Research, 4th ed. (New York: Hodder and Stoughton, 1923), 767-68;
Daniel B. Wallace, Greek Grammar beyond the Basics: An Exegetical Syntax ofthe
New Testament (Grand Rapids: Zondervan, 1996), 266-69.
37. Gordon D. Fee, φ66, and Origen: The Myth of Early Textual Recension in Al­
exandria,* in New Dimensions in New Testament Study, ed. R. N. Longenecker and
M. C. Tenney (Grand Rapids: Eerdmans, 1974], 44). Cf. also Peter M. Head, “Chris-
tology and Textual Transmission: Reverential Alterations in the Synoptic Gospels,”
NovT35 (1993): 105-29, esp. 112-13.
38. Kenneth W. Clark, “The Gospel of John in Third-Century Egypt” NovTZ (1962): 24.
Cf. also S. A. Edwards, “ϊ>75 under the Magnifying Glass/’ NoVT 18 (1976): 190-212.
39. Ehrman, Orthodox Corruption, 112.
40. Only Merk's critical NT text contains Codex Freerianus (W[032-S]). Then again, as
Daniel B. Wallace reminded me. Codex W was not discovered until after Tischen-
dorf wrote his critical work and while von Soden was producing his work (l.e., its
publication was shortly before von Soden’s final volume).
41. “[TJhe first quire of John... is a later (probably eighth-century) replacement quire
that bears no relation to the rest of the manuscript and made up for the (presum­
ably) lost original portion” (James R. Royse, "The Corrections in the Freer Gospels
Codex,” in The Freer Biblical Manuscripts: Fresh Studies of an American Treasure
Trove, ed. Larry W. Hurtado [Atlanta: Society of Biblical Literature, 2006], 186).
Cf. Metzger and Ehrman, Text ofthe New Testament, 77-81; Edgar J. Goodspeed,
“Notes on the Freer Gospels7A/T13 (1909): 597-603, esp. 599.

Brian J. Wright Chapter 6 / 237

 JESUS AS ΘΕΟΣ

evident in every aspect of this manuscript (i.e., the Gospel of John por­
As for W®, Morgan points out the following:
tion of Regius).4243

1. There is no evidence to establish a direct relationship be­

tween these two eighth-century manuscripts. As a result, the
occurrences of the article with 9eos found in John 1.1c in
both MSS should be considered isolated readings.

2. Alignment of Codex L and W® never merits inclusion in the

accepted text of NA27 without support from other key MSS
(K, B,C,

3. There are no known instances where a combination of Ws

with a single other witness finds credibility as a potentially
“original” reading.

Therefore, the inclusion of W8 as a subsingular reading in John

1.1c does not negate the egregious nature of the scribal behavior
in Codex L, and it further demonstrates that this combination
possesses insufficient testimony to consider the reading και ό
θεός- ήν ό λόγο? to be a plausible original.42

This scant evidence, at best, struggles to gain any viability in going

back to the Ausgangstext. In addition, it is highly improbable that this
was a deliberate corruption by the Orthodox Churchfive centuries after
the Arian controversy.

Sahidic Coptic MSS,44 usually considered decent representatives of

the Alexandrian form of text,45 offer an intriguing clue to the textual
certainty in John 1.1c. In short, Sahidic has both an indefinite and a

42. Matthew P. Morgan, "The Legacy of a Letter: Sabellianism or Scribal Blunder in

John 1:1c?” in this volume.
43. Ibid, 116.
44·. According to Schussler, “Today we count about 182 Coptic MSS of the Gospel of
John in the Sahidic dialect” (Karlheinz Schussler, "Some Pecularities of the Coptic
(Sahidic) Translations of the Gospel of John)’ journal ofCoptic Studies 10 [2008]:
41-62). That number, Schdssler continues, includes five complete MSS of John’s
Gospel (Le., sa 505, 506,508, 561,600), 38 lectionaries, and three other liturgical
MSS. His recent MS calculation helps explain the “1057 Coptic citations of John’s
gospel in the 27th edition of the Nestle-Aland Novum Testamentum Greece”
(Christian Askeland, “Has the Coptic Tradition Been Properly Used in New Testa­
ment Textual Criticism?” [paper presented at the annual meeting of the Society of
Biblical Literature, Boston, MA, November 22,2008], 1).
45. Frederik Wisse, "The Coptic Versions of the New Testament,” in The Text of the
New Testament in Contemporary Research: Essays on the Status Quaestionis, ed.
Bart D. Ehrman and Michael W. Holmes (Grand Rapids: Eerdmans, 1995), 137.
Ci. Metzger, Textual Commentary, 15; Metzger, The Early Versions of the New

238 / Chapter 6 Brian J. Wright

 A Textual Examination

definite article (whereas Koine Greek only has a definite article). What
gives this fact significance is that John 1.1c has the indefinite article in
Sahidic (and Bohairic) MSS46: \rw ΗβγΝογτβ ne ntpxxe.47 It should
come as no surprise, then, that the occurrence of the indefinite article
(ογ, which has contracted) before “God” (uoyTe) in this passage sug­
gests that the Coptic translator was looking at a Greek Vorlage with an
anarthrous θεό?. In other words, the fact that Seo? was translated into
Sahidic (and Bohairic) as an indefinite noun strongly suggests that the
translator was translating a Greek text without the article.

To flesh this out a little more, Horner translates John 1.1c into
English as follows: “and [a] God was the Word.”48 The apparatus, how­
ever, states, “Square brackets imply words used by the Coptic and not
required by the English, while curved brackets supply words which are
necessary to the English idiom."49 Unlike English, the Sahidic indefinite
article is used with abstract nouns (e.g., truth, love, hate) and nouns of
substance (e.g., water, bread, meat).50 An example of this can be seen
in Horner’s translation of John 19.34b (και έξήλθεν εύθύ? αίμα καί
ΰδωρ), where there are no Greek articles: “and immediately came out
[a] blood and [a] water."51 None of the words in brackets are necessary
in English, but they are noted by Horner due to the presence of the in­
definite article in the Coptic MSS.
Circling back to the textual assessment, the question we must now
answer is, did Coptic translators uniformly translate the nominative

Testament: Their Origin) Transmission) and Limitations (Oxford: Clarendon,

1977), esp. 132-37; Metzger and Ehrman, Text of the New Testament, 110-15.
46. Bohairic was a different (but new) Coptic translation from Greek. More impor­
tantly, though, it is an important witness to die secondary Alexandrian type of text
(see, e.g., Metzger, Textual Commentary, 15).
47. George W. Horney ed., The Coptic Version of the New Testament in the Southern
Dialect, Otherwise Called Sahidic and Thebaic, with Critical Apparatus, Literal
English Translation, Register of Fragments, and Estimate of the Version, 7 vols.
(Oxford: Clarendon, 1911-24), 3:2. Cf. Hans Quecke, Das /ohannesevangelium
’ ~sai'di$chrText der Handschrift PPalau Rib. Inv^Nr, 183 mit den Varianten der
Handschriften 813 und 814 der Chester Beatty Library und der Handscrift MS69
(Barcelona: Papyrologica Castroctaviana, 1984), 73. For Bohairic, see George W.
Homer, ed., The Coptic Version ofthe New Testament in the Northern Dialect, Oth­
erwise Called Memphitic and Bohairic, with Introduction, Critical Apparatus, and
Literal English Translation, 4 vols. (Oxford: Clarendon, 1898-1905), 3:2.
48. Horner, Coptic Version of the New Testament in the Southern Dialect, 3:3.
49. Ibid., 3:376 (italics added).
50. Thomas Lambdin, Introduction to Sahidic Coptic (bfacon, GA Mercer, 1983), 5.
Cf. also Bentley Layton, A Coptic Grammar: Sahidic Dialect (Wiesbaden: Harras-
sowitz, 2000).
51. Horner, Coptic Version of the New Testament in the Southern Dialect, 3:307. A few
other examples from the Gospel of John include 1.16, 26, 33; 3.5, 6; 5.39; 6.10;
16.33.

Chapter 6 / 239

Brian J. Wright
 JESUS AS ΘΕΟΣ

singular θεός?52 To answer this, I examined every occurrence of the

nominative singular θεός in every potential Johannine writing (i.e.,
John, 1 John, 2 John, 3 John, Revelation).53 This examination revealed
that John 1.1c was the only time the nominative singular θεός (ar­
ticular or anarthrous) was translated with a Coptic indefinite article.
Putting this in further perspective, of the five NT books examined,
there were only four other anarthrous uses of θεός (if one includes the
textual variant in Rev 21.3),5* Still, despite whatever one understands
the Coptic translator to have done with the other four potential in­
stances (assuming their contained them), John 1.1c is the only
text for which we can be certain that the Coptic translator was in fact
looking at a Vorlage that contained an anarthrous θεός (Le., no evi­
dence to the contrary exists to date). As stated already, only two, late
eighth-century MSS contain the articular θεός, and both the Sahidic
and Bohairic versions were composed prior to then. In other words,
until (or unless) new evidence is discovered to the contrary, it is highly
probable that the Coptic translator(s) were looking at a Greek Vorlage
with an anarthrous θεός, as reflected by the only Coptic indefinite ar­
ticle with a nominative singular θεός in the five NT books previously
mentioned.
In sum, it is highly improbable that the Coptic translator was trans­
lating a Greek Vorlage containing an articular θεός. Internally (and
syntactically), the absence of the article does not necessarily deny the
full deity of Jesus. "Neither in LXX Greek nor in secular Greek,” Harris
explains, “is a firm or a fine distinction drawn between the articular and
the anarthrous θεός. This judgment is confirmed, as far as Hellenistic
Greek writings contemporaneous with the NT are concerned, by
Meecham, who cites specific examples from the Epistle to Diognetus.”55

52. For a more in-depth work dealing specifically with this question, see Timothy Ric-
chuiti and Brian J. Wright, “From ‘God’ (θεόδ) to 'God' (κογτβ): A New Discussion
and Proposal regarding John 1:1c and the Sahidic Coptic Version of the New Testa­
ment; JTS 62 (2012): 494-512.
53. The following statistics were produced via the base text of the NA27 and UBS4 in
Bibleworks 8.0. For the Sahidic Coptic version, I examined Hornet Coptic Version
of the New Testament in the Southern Dialect; Quecke, Das Johamiesevangelium
saldisch; and Herbert Thomas, ed., The Coptic Version of the Ads of the Apostles
and the Pauline Epistles in the Sahidic Dialect (Cambridge: Cambridge University
Press, 1932) (Chester Beatty MSS). Admittedly, a slight distortion of die database
might occur due to text-critical issues (e.g., John 8.54). I did not exhaustively ex­
amine each critical apparatus, MS, or witness, to find other viable variants that
attest to a nominative singular Θεός/Νογτ©. My purpose here was merely to obtain
a highly probable snapshot of occurrences and patterns via several modern Greek
NT editions.
54. θ£0$· occurs twice in this verse (with the second occurrence placed in brackets in
both the NA17 and UBS*), but only the second (anarthrous) one is reflected in the
Coptic. The other three are John 1.18, John 854, and Rev 21.7.
55. Harris, Jesus as God, 29.

240 / Chapter 6 Brian J. Wright

 A Textual Examination

Another critic puts it more specifically: “The term θεός· appears in some
form 83 times. Of these 63 are articular and 20 anarthrous. Still, it is
highly improbable that the Fourth Evangelist intends any consistent dis­
tinction to be drawn between θεό? and ό θεό?.”56 At any rate, the schol­
arly consensus is correct that the text is certain and that every viable MS
ascribes the title θεό? to Jesus.

John 1.18
At least thirteen variant readings57 exist for John 1.18, of which three
are viable.5859All three viable variants divide into two distinct groups,
reading with either υιό? or θεό?.® If the latter group is chosen, the final
categorization ultimately depends on the presence or absence of the
article:

1. μονογενή? θεό?—B C* LS*423 Diatessaron-^^syri, «^geo2

Apostolic Constitutions Ariusacc·t0 EP>Phanius Basil Clement^1 Cyril1'4
Didymus Epiphanius Gregory-Nyssa Heracleon Hilary Irenaeus1’’
1/3 Jerome Origen6'2/4 Pseudo-Ignatius Ptolemy Synesius3" *° EPiphiBius
Iheodotus*®·t0 aemen* Vdentinians“c·tolrenacui Qement

2. 6 μονογενή? θεό?—J}75 N133 copsab° Basil1'2 Clement2'3 CIement&om

iheodotus ία Cyril2'4 Epiphanius Eusebius3'7 Gregory-Nyssa Origen8'2/4
Serapion1'2

ό μονογενή? υιό?—A C3 G Θ K T X Ψ Π 063 0141 0211

1 13 22 24 63 69 79 106 114 118 124 131 138 152 154 157 158
160 165 168 173 178 180 185 191 205 209 213 220 222 228 245 265
268 270 280 295 333 345 346 348 352e 357 370 377 382 389 391
397 401 423 430 472 482 489 508 513 515 537 543 544 555 557 565
579 589 597 649 679 683 700 709 713 716 720 726 731 732 733 736
740 744 747 775 787 788 792 799 807 809 821 826 827 828 829 833
841 851 863 865 873 874 878 883 884 888 889 891 892 899 904 931
968 969 979 982 983 989 992 994 1006 1009 1010 1014 1021 1026
1029 1038 1043 1071 1079 1085 1087 1093 1113 1118 1128 1187

56. Daniel Rathnakara Sadananda, Ute Johannine Exegesis ofGod: An Exploration into
thejohannine Understanding of God (New York: Gruyter, 2004), 177.
57. Kurt Aland, Barbara Aland, and Klaus Wachtel, Text und Textwertdergriechischen
Handschriften des Neuen Testaments: Jokannesevangetium (New York: Gruyter,
2005), 3-5.
58. Several exegetical and historical details exist that will not be canvassed here.
59. John 1.18 is actually the only verse listed under textual issues in both major works
on this topic: the standard work by Murray Harris, Jesus as God, lists only three
problems as “textual” (Heb 1.8; 2 Pet 1.1; John 1.18), and Raymond Brown’s An in­
troduction to New Testament Christology lists three under “textual·: Gal 2.20; Acts
20.28; John 1.18.

Brian J-Wright Chapter 6 / 241

 JESUS AS ΘΕΟΣ

1188 1195 1200 1216 1230 1241 1242 1243 1253 1292 1342 1344
1365 1424 1505 1546 1646 2148 Byz [E F G H] Leet it3·aut·* '·e·ί «2)·1
vg syrc· n*pal arm eth geo1 slav Alexander Ambrose10'11 Ambrosiaster
Athanasius Augustine Basil1'2 Caesarius Irenaeus1311/3 Irenaeus1312/3
Clement*01’1 TI,eodQtU5 1/2 Clement1'3 Cyril1'4 Chrysostom Hippolytus
Origen1311/2 Letter of Hymenaeus Eustathius Eusebius477 Serapion1'2
Gregory-Nazianzus Proclus Theodoret John-Damascus Tertullian
Hegemonius Victorinus-Rome Hilary5'7 Pseudo-Priscillian Faustinus
Fulgentius Gregory-Elvira Phoebadius Jerome Varimadum Letter of
Hymenaeus Nonnus Synesius Titus of Bostra Victorinus of Rome

External Evidence?0
Θεός- is attested in the best Alexandrian majuscule (B) and in the
61 The significance of this is that if the
earliest available MSS ($>66 j£75).60
Alexandrian witnesses for υίός (e.g., T Δ Ψ 8921241) cannot reasonably
go back to the Alexandrian archetype, its attestation therein is almost a
moot point.62 Ehrman rightly concludes that the semirecent discovery
of and ^>75 did “very little (in this instance) to change the character
of the documentary alignment” and has “done nothing to change the
Granted, no scholar, to my knowledge, argues against this
picture.”6364
fact. Nevertheless, these two MSS continue to persuade certain scholars
(particularly in evangelical circles) that θεός· is now the superior reading.
For example, Kostenberger and Swain recently concluded, “With the
acquisition of 5>66 and both of which read monogenes theos, the
preponderance of the evidence now leans in the direction of the latter
reading [monogenes theos]’.’6* This evidence, albeit strong, has not really
changed the picture. That is why scholars who opt for υΙός consistently
point out the apparent isolation of θεός in the Alexandrian form of text.
In fact, Ehrman argues that because "virtually every other representa­

60. Several major published Greek NT texts are evenly divided here as to the Ausgang-
stext (e.g., von Soden, Bover, and TischendorF choose ό μονογενής υιός [though
it should be noted that the discoveries of and £)75 occurred after two of these texts
were published], while UBS4, NA27, and Merk favored μονογενής θεός).
61. For the chief characteristics regarding the copying activity of tire scribes of both p66
and ξ|575, consult Royse, Scribal Habits, esp. 544,704.
62. Nevertheless, as Clark admonishes, “We are mindful that these papyri cannot
claim unquestioned priority on the ground alone of their greater antiquity... [nor
can we] blindly follow their textual testimony even when the two are in agreement
with one another” (“Gospel of John in Third-Century Egypt,” 23). Cf. also Eldon
J. Epp, “A Dynamic View of Textual Transmission,’’ in Studies in the Theory and
Method ofNew Testament Textual Criticism, ed. Eldon J. Epp and Gordon D. Fee
(Grand Rapids: Eerdmans, 1993), 274.
63. Ehrman, Orthodox Corruption, 112.
64. Kostenberger and Swain, Father, Son, and Spirit, 78. ,Cf. also Andreas J. Kosten­
berger, John (Grand Rapids: Balter, 2004), 50.

242 / Chapter 6 Brian J. Wright

 A Textual Examination

tive of every other textual grouping—Western, Caesarean, Byzantine—

attests to υΙό?,” θεός does not “fare well at alL”65 Let us assess, then,
the remaining textual groupings Ehrman mentioned, since no one, to
my knowledge, is arguing against θεό? going back to the Alexandrian
archetype.
Three main issues require comment concerning the Western
tradition:

1. The quality of the Western MS supporting θεό? (N)66 is com­

paratively greater than all Alexandrian MSS supporting υιό?. Therefore,
unlike the overwhelming improbability of υιό? going back to the
Alexandrian archetype, θεό? does have a viable possibility of doing so
in the Western tradition.

2. K is the earliest Western MS containing this passage (thus

strengrhening its possible connection with the Western archetype). In
the least, this demonstrates that θεό? is not isolated in the Alexandrian
form of text, with weak attestation elsewhere.

3. Although υιό? has relatively stronger support in the Western

form of text (e.g., Ws it vg syrc Irenaeus), one could still argue that “in
the early period [pre-180] there was no textual tradition in the West
that was not shared with the East.”67 In other words, “the origin of
the ‘Western’ text lies anywhere but in the direction its name would
suggest.”68 Ehrman concludes, “Above all, it is significant in saying
something about the transmission of the so-called ‘Western’ text of
the Fourth Gospel. To be sure, we have not uncovered any evidence
of a consolidated form of this text that could match the carefully

65. Ehrman, Orthodox Corruption, 79 (italics added). Later, we shall see that he re­
verses the same external appraisal he employs here (see the discussion of Heb 1.8
in this chapter).
66. Sinaiticus (N) aligns with the "West” in John 1.1-8.38. See Gordon Fee, “Codex
Sinaiticus in the Gospel of John: A Contribution to Methodology in Establishing
Textual Relationships,” in Epp and Fee, Studies in the Theory and Method of New
Testament Textual Criticism, 221-43.
67. Kurt Aland and Barbara Aland, The Text of the New Testament: An Introduction
to the Critical Editions and to the Theory and Practice of Modern Textual Criti­
cism, 2nd ed. (Grand Rapids: Eerdmans, 1989), 54. Aland and Aland argue, “Hardly
anyone today refers to this putative Western text without placing the term in quo­
tation marks, i.e., as the ‘Western text’” (ibid.). Likewise, Scrivener concludes that
“the text of Codex Bezae, as it stands at present, is in the main identical with one
that was current both in the East and West" (Frederick H. Scrivener, Bezae Codex
Cantabrigiensis [London: Bell and Daldy, 1864], xlv).
68. Aland and Aland, Text ofthe New Testament, 67. Cf. Roger L. Omanson, A Textual
Guide to the Greek New Testament: An Adaptation ofBruce M. Metzger’s “Textual
Commentary”for the Needs of Translators (Peabody, MA: Hendrickson, 2006), 22.

Chapter 6 / 243

Brian J. Wright
 JESUS AS ΘΕΟΣ

controlled tradition of Alexandria.”69 The possible implication of this,

then, is that even the attestation of υιός in the majority of MSS in the
Western form of text does not necessarily add a lot of textual weight
to its authenticity (especially without R or stronger Western support).

Does the Caesarean textual grouping strengthen the argument

in support for υιός? Again, the overwhelming majority read υιός (Θ
565 579 700 fl f23 geo1). This, however, is problematic for at least two

1. More recent nomenclature moves away from the label

“Caesarean,” since strong argumentation exists against it being a fourth
form of text.70 Admittedly, some merit might exist in using the label
with the result that further geographical distribution might be exposed.

2. Assuming the Caesarean group does reveal further geograph­

ical distribution, θεός is attested in it, albeit scarcely (geo2): “Like
the Armenian version, it [Georgian] is an important witness to the
Caesarean type of text. Among the oldest known Gospel manuscripts
are the Adysh manuscript of A.D. 897, the Opiza manuscript of 913,
and the Tbet’ manuscript of 995. In most apparatus critici, the Adysh
manuscript is cited as Geo1 and the testimony of the other two, as
Geo2.”71 Again, die evidence here shows that θεός is present outside an
exclusively Alexandrian tradition, with a viable witness to an archetype
(increasing its geographical distribution).

Regarding the Latin and Syriac traditions (aligning with the

“Western” type of text), υΙός occurs most frequently, with θεός still
present in some Syriac MSS (syrh(rng! syrp).72 At first glance, this scant
evidence seems irrelevant. Impressive here, diough, is that θεός is at­
tested again outside the Alexandrian tradition (e.g., the Peshitta [syrv]
was “transmitted with remarkable fidelity]’ and syrh(mg) is close to the

69. Bart Ehrman, “Heracleon and the ‘Western’ Textual Tradition,” NTS 40 (1994):
178-79.
70. See Bruce Metzger, “The Caesarean Text of the Gospels," in Chapters in the History
ofNew Testament Textual Criticism (Grand Rapids: Eerdmans, 1963), 42-72; Larry
Hurtado, Text-Critical Methodology and the Pre-Caesarean Text: Codex W in the
Gospel ofMark (Grand Rapids: Eerdmans, 1981), 24-45; Eldon Epp, "Issues in New
Testament Criticism," in RethinkingNew Testament Textual Criticism, ed. David A.
Black (Grand Rapids: Baker, 2002), 39.
71. Metzger and Ehrman, Text of the New Testament, 118-19. Cf. Robert Blake and
Maurice Briere, “The Old Georgian Version of the Gospel of John" Patrologia Ori-
entalis 26, no. 4 (Paris: Firmin-Didot, 1950).
72. For certain cautions when using Syriac, see P. J. Williams, Early Syriac Translation
Technique and the Textual Criticism ofthe Greek Gospels (Piscataway, NJ: Gorgias,
2004).

Brian J. Wright
 A Textual Examination

“Western” type of text).73 At the same time, θεός is the exclusive reading
in both the Arabic and Coptic traditions.74 Geos , then, is attested in one
of the earliest versions of the NT where υΙός is completely absent (the
Coptic versions).

date of υιός by listing three specific fathers “who were writing be­
fore our earliest surviving manuscripts were produced” (Irenaeus,
Clement, and Tertullian).75 Regrettably, he does this without acknowl-

period. We can, however, list three: Irenaeus, Clement, and Eusebius.

One may notice that two of the names appear on both sides of the de­
bate. This redundancy reveals the fact that many fathers (both Greek
and Latin) use υιός as well as θεός in their writings on John 1.18.
The point is that there are many names that could be used to support
either reading. In fact, here are three more: Basil, Cyril, and Origen.
At the risk of sounding repetitive, θεός· here shows up again outside
the Alexandrian tradition (e.g., early Latin fathers in the Gospels
are Western witnesses),76 with relatively strong textual weight (per
Ehrman’s argument).
At least two more issues, though, are critical regarding the church
fathers. First, McReynolds warns us that any reference to ό μονογενής
υιός by a father is unsubstantiated unless it specifically denotes John
1.18. The citation or allusion alone could equally apply to any of the
other passages in John (1.14; 3.16) or in the NT (Luke 7.12; Heb 11.17;
1 John 4.9) where μονογενής refers to the “Son.” The same problem
does not apply to μονογενής θεός, since it occurs nowhere else. Thus
one can be sure that John 1.18 is in view if μονογενής θεός, with or
without the article, is read (e.g., Arius, Basil, Clement, Cyril, Didymus,
Epiphanius, Eusebius, Gregory of Nyssa, Heracleon, Hilary, Irenaeus,
Jerome, Origen, Pseudo-Ignatius, Ptolemy, Serapion, Synesius, Tatian,
Theodotus, Valentinius). McReynolds concludes “that patristic evi­
dence for various readings needs to be used much more carefully, and
with a full view of the context of the Father being quoted.”77

73. Metzger and Ehrman, Text ofthe New Testament, 98. For a recent discussion on the
genetic relationship between the Old Syriac and the Peshitta, see Andreas Juckel,
"Research on the Old Syriac Heritage of the Peshitta Gospels/ in Journal ofSyriac
Studies 12, no. 1 (2009): 41-115.
74. See, e.g., Quecke, Das Johannesevangelium sal'disch, 75.
75. Ehrman, Orthodox Corruption, 79.
76. Contra Sadananda, Johannine Exegesis of God, 210. Cf. Metzger, Textual Commen­
tary, 15.
77. Paul McReynolds, "John 1:18 in Textual Variation and Translation/ in New Tes­
tament Textual Criticism (Oxford: Clarendon, 1981), 118. Cf. Carroll Osburn,
"Methodology in Identifying Patristic Citations in NT Textual Criticism," in NovT
47 (2005): 313-43.

Brian J. Wright Chapter 6 / 245

 JESUS AS ΘΕΟΣ

Second, the reading μονογενής· θεός is not an anti-Arian polemic.

Arians did not balk at giving the title θεός to Jesus.78 In fact, Arius sup­
ports the reading θεός here (according to Epiphanius)79 and even called
Jesus “God” in a letter he wrote to Eusebius bishop of Nicomedia:

But what do we say and think? What have we taught and what do
we teach? That the Son is not unbegotten or a portion of the unbe­
gotten in any manner or from any substratum, but that by the will and
counsel of the Father he subsisted before times and ages, full of grace
and truth, God, only-begotten, unchangeable.80

If this is true, it throws into doubt that an orthodox scribe would change
the text away from Arius if θεός bolsters the complete deity of Christ.
Even if the reverse is true (if Epiphanius’s testimony is wrong and/or
if Arius never wrote that letter), one would have to assume that each
scribe that changed υιός to θεός knew about the Arian controversy. But
it did not change the text to the higher Christology (examples of which
would be many, given the MS evidence previously listed). Even then,
the evidence shows inconsistency in their alleged corruption(s) given
John 1.1 and 20.28. On top of all that, it would also have to be shown
that all the textual evidence originated during or subsequent to this
Arian controversy (which it does not). One might still argue, though,
that there only needed to be one extremely early scribe who generated
θεός. The real question would then become how early this occurred.
To answer this objection, the evidence reveals that earlier MSS (in fact,
the earliest, well before the Arian controversy) attest to θεός. This indi­
cates that the objection would remain highly speculative and against the
clearer testimony of earlier and better MSS. In other words, the earliest
and best MSS (as well as the fact that both sides of this Christological

78. “ ‘Ήν ποτέ δτε ούκ ήν’—Άί one time he did not exist’—became the slogan that
best expresses die core of Arius’s theology, which he shaped in die Thalia (θάλεια
= 'banquet’), his main work, which is cited in almost all of the sources dealing with
Arius” (Hubertus R. Drobner, The Fathers of the Church: A Comprehensive Intro­
duction [Peabody, MA: Hendrickson, 2007], 236). Cf. also Raymond E. Brown, The
Gospel according to John (i-xii): Introduction, Translation, and Notes (Garden City,
NY: Doubleday, 1966), 17.
79. Valentinus (another theologian deemed heretical) also accepted the same reading:
Valentiniansacc ,e Irenaeus aemanc. furthermore, no church father accuses him of
changing the text. Hort argued here that μονογενής Beds was original because the
Gnostics (such as Valentinus) did not invent this phrase; instead, they quoted it (E
]. A. Hort, Two Dissertations [Cambridge: Macmillan, 1876]).
80. William Rusch, The Trinitarian Controversy (Philadelphia: Fortress, 1983), 29-30.
For Greek text, see Urkunden zur Geschichte des arianischen Streites, ed. H. G.
Opta (Berlin: Gruyter, 1934). Cf. R. P. C. Hanson, The Search for the Christian
Doctrine of God: The Arian Controversy, 318-381 (Grand Rapids: Baker, 2005), 6;
Lewis Ayres, Nicaea and Its Legacy: An Approach to Fourth-Century Trinitarian
Theology (New York: Oxford University Press, 2004), 105-26.

246 J Chapter 6. Brian J. Wright

 A Textual Examination

controversy use/quote θεός) heighten the argument away from the al­
legation that this is an orthodox corruption.81
Two other plausible reasons might explain the mainstream survival
of υΙός. One, “Son” may have prevailed as the easier (more predictable)
reading before the composition of most extant versions. In support of
this, “Son” has universal agreement in later copies, with no observable
evidence of scribes to alter it. Two, given the preceding external argu­
ments, even though θεός has wide geographical distribution, it remains
weak compared to tire distribution of υίός in other non-Alexandrian
forms of text. A probable explanation is that θεός is by far the more
difficult reading theologically, statistically, and stylistically, which gen­
erally produces various textual variants (see the following discussion of
the internal evidence).
In sum, externally, both readings enjoy wide geographical distri­
bution, even though υιός is relatively stronger in non-Alexandrian
forms of text. Both readings coexisted in the second century, although
weightier MSS support θεός.82 As a whole, then, θεός is more probable
due to the quality, antiquity, and transmissional history of the witnesses
previously listed. Nevertheless, this external evidence alone does not
make θεός the exclusive heir to the throne.

Internal Evidence

Several internal observations initially seem convincing in support

of υιός. For starters, statistically, μονογενής refers to the “Son” else­
where in John (1.14; 3.16) and in the NT (Luke 7.12; Heb 11.17; 1 John
4.9). “The only occasion in the NT where μονογενής is not used of an
‘only son? Harris observes, “is Luke 8:42, where it qualifies θυγάτηρ.”83
Stylistically, the reading "Son” is more natural with the mention of “God”
earlier in the verse as well as the mention of “Father” later in the verse.
Otherwise, why would “God” be repeated twice, and how could God
reside in the bosom of another God (“the Father”)? Theologically, the
NT rarely calls Jesus θεός, making the reading almost too difficult. All
of these-obser-vations seem to point one in the direction of an original
reading of ό μονογενής υιός.84

81. See also Royse, Scribal Habits, esp. 19—27.

82. Given my methodology previously stated, merely counting MSS is inadequate; one
must also weigh them.
83. Harris, Jesus as God, 92.
84. For a more extensive grammatical examination, especially whether the adjective
μονογ€υής can be used substantivally when it immediately precedes a noun of the
same inflection, see Stratton L. Ladewig, “An Examination of the Orthodoxy of the
Variants in Light of Bart Ehrman’s The Orthodox Corruption of Scripture” (ThM
thesis, Dallas Theological Seminary, 2000), 51-62; Daniel B. Wallace, “lhe Gospel
according to Bart: A Review Article of Misquoting Jesus by Bart Ehrman,” JETS 49
(2006): 344-46.

Brian J. Wright Chapter6 / 247

 JESUS AS ΘΕΟΣ

In response, the offense of using Geos probably drove a scribe to the

less offensive Christology of υιός, which comports well with the scribal
tendency to simplify the text (substituting “God" for “Son” is highly im­
probable, perhaps best explaining the absence ofGeos in later Greek MSS).
Even more, μονογενής Geos is never used elsewhere.85 One must ask,
then, why here and only here do we have the textual variant μονογενής
θεός (with or without the article)? The answer, given this scenario alone,
seems to be that θεός best explains the rise of the other variants.
Stylistically, θεός closes the inclusio begun in 1.1c, also possibly
providing a parallel with 20.28 (the gospel as a whole). Perhaps the in­
tention was to shock the reader. If this phrase occurred frequently, the
author may have failed in achieving his desired result. The reference
“who is in the bosom of the Father" is an anthropomorphic metaphor
for intimacy and fellowship.8687 In other words, it is an idiom for close­
ness and does not truly affect either reading. Lastly, the author of John’s
Gospel has a penchant for varying Christological designations (c£, e.g.,
1.49; 4.42; 6.69; 9.38; 11.27; 20.16).
Another internal argument sometimes given is that a scribe could
have easily erred, since only one Greek majuscule letter differentiates
"Son” from “God”:TC or 0C. One problem with this option, however,
is that υιός was not one of the original (or earliest) nomina sacra.67 At
the same time, though, θεός (OC) was one of the four earliest (i.e.,
’Ιησούς, Χριστός, κύριος, and θεός) and most consistently rendered
nomina sacra from the second century onward.88 To state this differ­
ently, although this option is not impossible, it is highly improbable
given the transmissional evidence we have.
What variant, then, best explains the rise of the others? The subtle
meaning of the two words in their original apposition, μονογενής
θεός, may have caused an early misconception. Thus an article was as­
signed to the original reading, now ό μονογενής θεός, as early as )£75,
X, and copb°'sa. Ironically, this change wound up alleviating nothing and
was inconsistent with other Johannine and NT usage. Accordingly, the
next stage of evolution changed “God” to “Son”: ό μονογενής υιός.
Finally, although a few other variants arose that either combined the
two readings (ό μονογενής υίός θεός)89 or simply omitted both (o

85. Certain texts (John 5.44; 17.3; Rom 16.27; 1 Tim 1.17; Jude 25) do not legitimately
belong here since they all use μόνο? and not μονογενής.
86. See BDAG, 556-57; L&N, 34.18 {The Greek/English Lexicon of the NT Based on
Semantic Domains. Edited by Johannes P. Nouw and Eugene A. Nida [New York,
United Bible Societies, 1989]),
87. See, e.g., Larry W. Hurtado, "The Origin of the Nomina Sacra: A Proposal,” JBL
117, no. 4 (1998): 655-73; Dirk Jongkind, Scribal Habits of Codex Sinaiticus, Text
and Studies: Contributions to Biblical and Patristic Literature, 3rd ser., vol. 5 (Pis­
cataway, NJ: Gorgias, 2007), 62-84.
88. Hurtado, Origin of the Nomina Sacra? 655,657.
89. E.g., if Ambrose1'11 Irenaeus lat1/3 Origen**.

248 / Chapter 6 Brian J. Wright

 A Textual Examination

μονογενής),90 ό μονογενή? υιός became the majority reading, with no

viable evidence of change in later Greek MSS.91
In retrospect, μονογενή? θεό? is the best reading given all the evi­
dence we have internally and externally. As a result, it is highly probable
that the text of John 1.18 calls Jesus θεό?.

John 20.28
Far beyond the confession of Nathanael in John 1.49 (“Rabbi, you
are the Son of God; you are the king of Israel!”),92 the Gospel of John
ends with the fullest Christological confession of faith in the entire
Gospel of John:93 άττεκρίθη Θωμά? και ειπεν αύτω· δ κύριό? μου και ό
θεό? μου (“Thomas answered and said to him, ‘My Lord and my God’”).
While this chapter does not seek to demonstrate or articulate tihe sense
in which Jesus was understood to be θεό?,94 the aim here, again, is to
find out whether a textual analysis will reveal a particular degree of tex­
tual certainty that this title was even ascribed to him.
Externally, a single fifth-century Western manuscript, D (05), omits
the second article in this verse, thus rendering θεό? μου instead of ό
θεό? μου.95 While this changes nothing contextually,96 D is arguably
one of the most important Western MSS textually. As Aland and Aland
note, “When D supports die early tradition the manuscript has a gen­
uine significance, but it (as well as its precursors and followers) should
be examined most carefully when It opposes the early tradition.”97 In

90. I consider ό μονογενή? so poorly attested externally (vg1"5 Ambrose Aphrahat Cyril
of Jerusalem Diatessaron Ephraem Jacob of Nlsibis Nestorius Ps-Athanasius Ps-
Ignatius Ps-Vigilius1'2 Victorinus-Rome) and too easily explainable transmission-
ally to necessitate the reverse hypothesis of starting with it.
91. For similar conclusions, see Jack Finegan, Encountering New Testament Manu­
scripts: A Working Introduction to Textual Criticism (Grand Rapids: Eerdmans,
1974), esp. 174-77; Lieu, I, II, and IIIJohn, 182 n. 113,233 n. 163.
92. άπεκρίθη αύτω Ναθαναήλ- ραββί, σύ ε ί δ υιό? τοΰ θεοΰ, σύ βασιλεύς el τοΰ ’Ισραήλ.
93. Ν. Τ. Wright, Ihe Resurrection ofthe Son ofGod (Minneapolis·. Fortress, 2003), 664.
94. For example, some have felt that Jesus allowed this statement in order not to “ruin
—— the momentl'YetJesus quotes Deut 6:13, “You are to worship the Lord your God
and serve only him,’’ in Matt 4:10 and Luke 4:8. Therefore, his teachings and con­
victions seem to strongly negate this option. Likewise, it is important to note that
Jesus is the sole object of Thomas’s interjection (αύτω), while, at the same time,
the two exclamations are impossible to unlink due to the conjunction καί. Nev­
ertheless, for the most plausible interpretive options of Thomas’s confession, see
Sadananda, Johannine Exegesis of God, 11-44. Cf. also Bruce M. Metzger, “Jeho­
vah’s Witnesses and Jesus Christ,” IhTo 10 (1953): 65-85, esp. 71.
95. Cf. Frederick H. A. Scrivener, Bezae Codex Cantabrigiensis (London: Bell and
Daldy, 1864), 156; IGNTP, Gospel according to St John, 2:541.
96. Harris lists four solid reasons in Jesus as God (109), though I believe his third
reason can be stated much stronger, since δ κύριο? is never used of God the Father
in John’s Gospel except in two OT quotations (12.13,38).
97. Aland and Aland, Text oftheNew Testament, 110.

Brian J. Wright Chapter 6 / 249

 JESUS AS ΘΕΟΣ

this case, title latter is true. Furthermore, D is an eccentric MS and regu­

larly drops the article.98 Yet even if D is original and the second article is
absent, this verse (ό κύριό? μου και θεό? μου) grammatically falls under
the criteria of Granville Sharp's rule:

In native Greek constructions (i.e., not translation Greek), when a

single article modifies two substantives connected by καί (thus, article-
substantive-καί-substantive), when both substantives are (1) singular
(both grammatically and semantically), (2) personal, (3) and common
nouns (not proper names or ordinals), they have the same referent.”99*

In other words, if D is correct and there is no article before θεό?,

both “Lord” and “God” in this verse explicitly refer to Jesus because of
this grammatical construction (cf. also 2 Pet 1.1, discussed later in this
chapter). Thus Granville Sharp’s rule makes the phrase even more ex­
plicit and leaves “no wiggle room for doubt.”105 John 20.28, no matter
which variant or MS one chooses, is categorically secure for referring
to Jesus as θεό?.101*

98. “By actual count, there is a parsimonious use of the article in D; in fact, this situ­
ation obtains in each book except Luke” (James D. Yoder, “The Language of the
Greek Variants of Codex Bezae,” NovT 3 [1959]: 245).
99. Wallace, Sharp's Canon, 132. For brief discussion of personal pronouns within
these constructions (e.g., μου following θεό?), see n. 144 in the present chapter.
In addition, regarding “my God,” Schnabel points out, “[Ajccording to Aristotle,
'It would be an absurdity to profess a friend's affection for Zeus' (φίλε τον Δία
ip

[Afag. mor. 1208 b 30]). The exclamation of Euripides’ choir upon seeing the sculp­
tures at the temple in Delphi, T see Pallas, my own goddess' (Euripides, Ion 211),
is one of the very few references in Greek literature that uses the phrase ‘my God?
Burkert [‘“Mein Gott?’ Personliche Frommigkeit und unverfugbare Gctteif in Ge-
sckichte—Tradition—Reflexion: Festschrift filr Martin Hengel zum 70 (Tubingen:
Mohr Siebeck, 1996), 3-14] points out, however, that this exclamation must be
understood as the ‘aesthetic wow-experience’ of a collective. There is very little
archaeological evidence for Greek cults of a ‘personal' god who was interested in
or connected with the individual person. In die context of the Anatolian cult of
Men, dedications refer to, for example, 'the Men of Artemidoros’ (CMRDM III 67-
70), but such formulations do not imply a particular ‘pact’ with the god. In a Greek
polls the gods of the city were important, not the god worshiped by the individual.
In everyday life people established contacts with gods only when needed. Burkert
concludes that ‘insofar a person fulfills his religious obligations, there remained
normally a realm of freedom, of the οσίου [‘profane'], in which religious concerns
vanished. This would be contradicted by a unique or comprehensive obligation or
affiliation. Resort to the gods becomes important in a time of need, however....
The pious person was prepared for being saved, but he does not have a revelation
on a document and no treaty with ‘his' god. Gods are not at his disposal” (Eckhard
J. Schnabel, Early Christian Mission: Jesus and the Twelve, vol. 1 [Downers Grove,
IL: InterVarsity, 2004], 615).
100. Wallace, Sharp's Canon, 132.
101. Moreover, these two OT expressions are frequently juxtaposed when referring
to Israel's one God. Waltke further defines, “The distinct meaning of these two

250 Brian}. Wright

 A Textual Examination

Acts 20.28

Acts 20.28 involves two distinct textual problems, of which at least

nine variants (seven and two, respectively) exist. The viable options are
as follows:

Textual Problem 1
τήν εκκλησίαν του θεού—X Β Η Μ S V W Θ 056 0142 4 104 218
257 312 314 322 383 424 454 459 614 621 629 917 1175 1409 1495
1505 1522 1611 1758 1831 2138 2147 2298 2412 2495 160 1592
1598 Ζ603 Ζ1021 Ζ1439 copb0 vg it"c-dem-ph-ί0·w syrh-p geo Ambrose
Athanasius Basil Chrysostom Cyril-Alexandria Epiphanius

τήν εκκλησίαν του κυρίου—ip74 A C* D Ε S Τ Ψ 13 33 36 40 81

94 104 181 206 209 307 337 429 431 436 453 522 610 630 623 945
1678 1739 1829 1891 2344 2464 Z164 Z599 arm copsa it1-e· p syrh,ns
Ambrosiaster Didymusdub- Irenaeuslat Jerome Lucifer Pelagius
Theodoret1'2

Textual Problem 2
του αίματος- του ίδιου—£41 £>74Κ* A Β C* D Ε Ψ 33 36 69181307
326 453 610 945 1175 1611 1678 1739 1837 1891 2464 160 arm geo
syr1,gr Cyril lheodoret

του ίδιου αίματος—H L Ρ 049 056 0142 1 88 104 226 323 330 440
547 614 618 927 1241 1243 1245 1270 1409 1505 1646 1828 1854
2147 2344 2412 2492 2495 Byz Leet slav Athanasius Chrysostom
Didymusdubvid

With the external evidence geographically and genealogically pro­

portionate on the first textual problem (the other five readings lack

names [the appellative Elohim and Yahweh] is widely recognized: whereas the tide
Elohim contrasts God with man. in their natures, the name Yahweh presents God
as entering into a personal relationship with man and revealing Himself to him’’
(Bruce K. Waltke, “The Book of Proverbs and Old Testament Theology” BibSac 136
[1979]: 305). Barrett also notes that John’s “My Lord and My God” directed to Jesus
reflects the LXX, where it represents D’nbs nirt’ and similar expressions, but also
makes contact with an expression fairly common in pagan religious literature (C.
K. Barrett, The Gospel according to John: An Introduction with Commentary and
Notes on the Greek Text [Philadelphia: Westminster, 19781, 572). Cf. B. A. Mastin,
“Theos in the Christology of John: A Neglected Feature of the Christology of the
Fourth Gospel,” NTS 22 [1975—76]: 32-51, esp. 37—41; G. Deissmann, Lightfrom
the Ancient East (New York: George H. Doran, 1927), 366-67; H. D. Betz, Lukian
von Samosata and das Neue Testament (Leiden: Brill, 1961), 102.

Brian J. Wright Chapter 6 / 251

 JESUS AS ΘΕΟΣ

sufficient external support, are obvious conflations, or both),102 the only

thing a textual critic can do is appeal to the internal evidence. Yet this,
too, is equally balanced.103104
Of course, it must be noted that “church of the
Lord” is absent from the NT and apostolic fathers, while “church of God”
occurs eleven other times in the NT (1 Cor 1.2; 10.32; 11.16, 22; 15.9; 2
Cor 1.1; Gal 1.3; 1 Thess 2.14; 2 Thess 1.4; 1 Tim 3.5,15) and twelve times
in the apostolic fathers (i Clem. 1.1; Ign., Eph. 17.1; Ign., Trail. 2.3; 12.1;
Ign., Phld. 1.1; 10:1; Ign., Smyrn. 1.1; Pol., Phil. 1.1; Mart. Pol. 1.1; Herm.
Sim. 18.2,3,4).1M
Most scholars accept Oeou as original not merely because of its
difficulty but also because of their confidence that, in the second tex­
tual problem, the authentic reading is του αίματος του ιδίου (“the
blood of his own [Son]” or "his own blood”). Transmissionally, Lars
Aejmelaeus proposes an actual literary dependence of Acts 20.28
on 1 Thessalonians 5.9-10 and Ephesians 1.7. This is in keeping
with his overall thesis that Pauline allusions in Acts are invariably
due to Luke’s knowledge of the Pauline letters.105 In addition, του
αίματος του Ιδίου is undeniably superior externally (£>41 ip74 R* A
B C’ D E Ψ 33 1739 geo syr). Its strength also rests on the logic
that it is the harder reading and best explains the rise of the others.
"That God suffered was acceptable language,” Hamack notes, "be­
fore criticism required some refinement of the conviction that God
(or God’s Son) had become man and died on the cross.”106 In ad­
dition, τοΰ αίματος τοΰ ιδίου received an “A” rating by the UBS4
editorial committee,107 and most major published NT Greek texts
agree (e.g., NA27, Tischendorf8, Bover, Merk, von Soden, Westcott
and Hort, Vogels, and Weiss).
The text most likely originally read την εκκλησίαν τοΰ θεού, ήν

102. For a better understanding of the individual witnesses for the book of Acts, see
Joseph A. Fitzmyer, The Acts of the Apostles: A New Translation with Introduction
and Commentary (New York: Doubleday, 1998), 66-79; C. K. Barrett, A Critical
and Exegetical Commentary on the Acts of the Apostles, vol. 1 (New York: T&T
Clark, 2004), 2-29; Darrell L. Bock, Acts (Grand Rapids: Baker, 2007), 28-29.
103. Metzger, Textual Commentary, 425-27. Cf. Pervo, Acts, 523.
104. The phrase “assembly of the Lord," (εκκλησία κυρίου), however, does occur in the
LXX (e.g., Deut 23.2, 3, Abis, 9; 1 Chron 28.8; Mic 2.5), but never as an articular
construction (i.e., ή εκκλησία τοΰ κυρίου) or with εκκλησία having the semitech-
nical sense that it does in the NT.
105. Lars Aejmelaeus, Die Rezeption der Paulusbriefe in der Miletrede: Apg 20:18-35
(Helsinki: Suomalainen Tiedeakatemia, 1987), 132-42. Cf. Pervo, Acts, 524.
106. Adolf Harnack, History ofDogma (London: Constable, [ca. 1900]; repr., New York:
Dover, 1961), 1:187 n. 1. For further discussion on the imagery of “the blood" in die
history of the church, as related to the work and person of Christ as God, see Jaro­
slav Pdikan, Acts (Grand Rapids: Brazos, 2005), 221—22. Cf. also Charles F. Devine,
“The ‘Blood of God’ in Acts 20:28," CBQ 9 (1947): 381-408.
107. Metzger, Textual Commentary, 427.

252 / Chapter 6

Brian J. Wright
 A Textual Examination

περιεποιήσατο διά του αίματος του Ιδίου.108 The reading θεοΰ soon
changed to κυρίου because of the difficulty in reconciling it with
αίματος του Ιδιου.109 This conclusion seems verifiable and reinforced
by the combination of variants in the witnesses. Here are a few ex­
amples of such combinations (in order of the proposed transmissional
history):

1. Witnesses that read both originals (θεού and αίματος του Ιδιου):
Κ· Β 1175 Ζ60

2. Witnesses that changed θεού to κυρίου because of the second orig­

inal (αίματος του Ιδίου): A C*D ΕΨ 33 453 945 1739 1891 36
181 307 610 1678 arm Theodoret

3. Witnesses that kept θεοΰ because of the second nonoriginal

(ιδίου αίματος): Η 056 104 614 1409 1505 2412 2495 Athanasius
Chrysostom

4. Witnesses that support both nonoriginals (κυρίου and ίδιου

αίματος): 2344 Didymus

To summarize, the variants that best explain the rise of the others
are θεού and αίματος του ιδίου. Acts 20.28, therefore, does not neces­
sarily equate Jesus with θεός (especially with solid evidence for both of
these readings). The decision ultimately comes down to one’s under­
standing and interpretation of the phrase διά του αίματος του Ιδίου:
“with the blood of his own [Son]” or “with his own blood.110

108. Also, “εκκλησία του θεού occurs eleven times in Paul; εκκλησία κυρίου occurs
seven times in die LXX but never without v.L θεού in the New Testament" (J.
Keith Elliott, “An Eclectic Textual Study of the Book of Acts," in 27/e Book ofActs as
Church History: Text, Textual Traditions, and Ancient Interpretations, ed. Tobias
Nicklas and Michael Tilly [New York: Gruyter, 2003], 29).
109. “The text [Acts 20.28] caused such puzzlement (God’s own blood?) that some of
the scribes responsible for making copies of Luke’s book evidently attempted to
improve or clarify it-—particularly by reading 'the church of the Lord, which he
obtained through his own blood’ (cf. Heb. 9.12)” (James D. G. Dunn, The Acts ofthe
Apostles [London: Epworth, 1996], 272).
110. Barrett suggests, ftIt is very unlikely that a trained theologian would write ’his own
blood’; but Luke was not such a theologian, and the natural way of reading the
Greek should probably be adopted [“with his own blood”]. It was enough for Luke
that when Jesus Christ shed his blood on the cross he was acting as the representa­
tive of God; he was God’s way of giving life, blood, for the world” (C. K. Barrett, A
Critical andExegetical Commentary on the Acts ofthe Apostles, voL 2, Introduction
and Commentary on Acts XV-XXVJII [New York: T&T Clark, 1998], 977). Cf. also
Harris, Jesus as God, 131-41; "blood of God” as used in the apostolic fathers: Ign.,
Eph. 1.1; Ign., Rom. 6.3 (cf. Tertullian, Ad uxor. 2.3.1 [sanguine dei]).

Brian J. Wright Chapter 6 / 253

 JESUS AS ΘΕΟΣ

Galatians 2.20111

The original text of Galatians 2.20, according to Metzger, Ehrman,

and others, must have read, του υίοΰ του θεοΰ (“the Son of God”). In fact,
all major published NT Greek texts have this reading with the exception
of Bover (who reads 9eo0 και Χριστοί). The UBS4 committee continues
their support and certainty of it. In fact, the committee agreed to increase
their rating from a “B” (found in the third edition) to an “A” (found in
the fourth edition). Additionally, the authors of the text-critical notes in
the New English Translation, using different arguments (e.g., progressive
revelation),112 came to the same textual conclusion.113 Yet after consid­
ering the internal and external evidence (as well as the arguments from
many secondary sources), I still think that several stones have been left
unturned and that discourse has been left unsaid regarding the reading
θεοΰ και Χρίστου. Therefore, although I ultimately accept the reading
υίοΰ του θεού, I will give several reasons why I am reluctant to give it an
“A” rating or exclude it among the list of passages potentially proclaiming
Jesus as θεός.
The four noted variants for this passage, in no particular order, are

1. του θεού—1985

2. του θεού του υίοΰ—330

3. του υίοΰ του θεοΰ—Ν A C D2 Η Κ L Ρ S Τ V Ψ 056 075 01510278

6 33 69 81 88 104131 205 209 226 256 263 323 326 365 424 436 440
459 460 489 517547 614 618 796910 927 945 999 1175 1241 1242
1243 1245 1270 1315 13191352 1424 1448 1505 1573 1611 1646
' 1734 1735 1738 1739 1827 1836 1837 1852 1854 1874 1881 1891
1912 1962 1982 2125 21272147 2200 2400 2412 2464 2495 2815
ϊί“-f> c cop54'b0 syrh-ρ Leet vg arm eth geo slav Ambrosiaster Augustine
Chrysostom Clement Cyril Didymusdub Jerome Mardon"®toAdamantius
Pelagius Severian Theodoret Varimadum

111. Two primary reasons encouraged me to include Gal 2.20 in this study: (1) most
standard works on this topic include this passage (e.g.; Brown, "Does the New Tes­
tament Call Jesus God?”; Harris, Jesus as God, 259-61; A. W. Wainwright, “The
Confession ‘Jesus as God’ in the New Testament,” SJT10 [1957]: 274-99), and (2)
it is possible (though I do not think highly probable) to translate two of the textual
variants as either “God even Christ” (θεοΰ και Χριστοΰ) or “God the Son” (του
θεοΰ τοΰ υίοΰ) (see, e.g., Ehrman, Orthodox Corruption, 86).
112. "Although Paul certainly has an elevated Christology, explicit ‘God-talld with ref­
erence to Jesus does not normally appear until the later books” (New Testament:
New English Translation, Novum Testamentum Graece, ed. Michael H. Burer, W.
Hall Harris, Daniel B. Wallace [Stuttgart: Deutsche Bibelgesellschaft; Dallas: NET
Bible, 2004], 860).
113. Ibid.

254 / Chapter 6

Brian J. Wright
 A Textual Examination

4. θεού και Χρίστου—154£ BD*G itlb),d-B Marius Victorinus-Rome

Pelagius

Externally, although the majority of witnesses favor τοΰ υΐοΰ του θεού,
the two oldest MSS support θεού και Χρίστου (jp46 B) along with several
other important witnesses (D* G it(b)·d-6 Marius Victorinus-Rome Pelagius)
. Furthermore, along with its early "proto-Alexandrian” support B),114115
a strong group of Western witnesses concur (D* F G it0*1, d-g Victorinus-
Rome). This variant, then, is relatively early and possesses agreement be­
tween good Western and Alexandrian witnesses (though it is not attested
in the Byzantine, Caesarean, or secondary Alexandrian form of text).
Next, two main internal arguments against this reading exist: (1)
Paul nowhere else expressly speaks of God as the object of a Christian’s
faith; and (2) during the copying process, a scribe’s eye probably passed
over the first article to the second, so that only του θεού was written (as
in MS 330).us In response to the former, God is the object of a believer’s
faith in Romans 4.24. Moo writes, “It is typical for Paul to designate God
as the one who raised Jesus from the dead (cf. [Rom] 8.11; 10.9; 1 Cor
6.14; 15.15; 2 Cor 4.14), but it is somewhat unusual for him to designate
God himself as the object of Christian faith. Undoubtedly he does so
here [Rom 4.24] to bring Christian faith into the closest possible rela­
tionship to Abraham’s faith.”116 As to the latter, that theory best explains
only one of the four previously noted variants, του θεού, not all of them.
Furthermore, θεού και Χρίστου does find some syntactical parallel
in the corpus Paulinum: 1 Timothy 5.21 and 2 Timothy 4.1 (cf. also 1
Tim 6.13).117 Beyond this, “Son of God” is the easier reading and pos­

114. Cf. Philip W. Comfort and David P. Barrett, The Text of the Earliest New Testa­
ment Greek Manuscripts: New and Complete Transcriptions with Photographs
(Wheaton, IL: Tyndale House, 2001), esp. 27-29; E. C. Colwell, “Method in Es­
tablishing the Nature of Text-Types of New Testament Manuscripts," in Studies in
Methodology in Textual Criticism ofthe New Testament, New Testament Tools and
Studies 9 (Leiden: Brill, 1969), 45-55, esp. 48.
115. Metzger, Textual Commentary, 524. For other scribal possibilities, see Colwell,
,. . StudiesdnMethodology in Textual Criticism ofthe New Testament, 106—24. Cf. J. R.
Royse, "Scribal Tendencies in the Transmission of the Text of the New Testament,”
in Ehrman and Holmes, The Text of the New Testament in Contemporary Research,
239-52; J. R. Royse, "The Treatment of Scribal Leaps in Metzger’s Textual Com-
' mentaryf NTS 29 (1983): 539-51; Metzger and Ehrman, Text of the New Testa-
ment, 250-71.
116. Douglas Moo, The Epistle to the Romans (Grand Rapids: Eerdmans, 1996), 287.
Cf. Moo, Colossians and the Philemon, 84 n. 13. The question remains, however,
whether Paul, anywhere in his writings, speaks of both God and Christ Jesus to­
gether as the object of faith, which is the case in Gal 2.20 if the authentic reading is
θεού και Χρίστου,
117. Contra Ehrman, "neither of the other expressions (“God even Christ,” “God the
Son”) occurs in this way in Paul” (Orthodox Corruption, 86). The position of the
pronoun does not affect the sense. 1 Tim 5.21 is surely not ascribing the title Geos

Brian I Wright Chapter 6 / 255

 JESUS AS ΘΕΟΣ

sibly explains why a scribe preferred it. It is also possible that there is
a contextual harmonization of verses 19 (“live to God”) and 20 ("Christ
lives in me”), keeping with the Western tradition and Pauline theology.*118119
120
Of course, textually speaking, harmonization seems to be more literal
than conceptual.
Externally and internally, several issues still need more clarification
and resolution; nevertheless, the traditional readinguiou του θεοί is the
best of all probable scenarios. At the same time, this reading may not
merit an "A" rating (despite the assessment of the UBS4 committee) and
needs further study.

Hebrews 1.8139

Two main interconnected textual issues exist in Hebrews 1.8 that

possibly resolve the broader grammatical dilemma of how to interpret
and translate b θεός in verses 8 and 9:12D

προς· δέ τον υιόν,

Ό θρόνος σου δ θεός εις τον αιώνα του α’ιώνος,
καί ή ράβδος τής εΰθύτητος ράβδος τής βασιλείας σου (ν 8)

ήγάπησας δικαιοσύνην και εμίσησας ανομίαν

διά τούτο εχρισέν σε ό θεός δ θεός σου
έλαων άγαλλιάσεως παρά τούς μετόχους σου (ν 9).
>
1. The first textual variant involves the presence or absence of του
αίώνος· after εις τον αΙώνα:

t
ο θρόι/o? σου ό θβό? els τόυ αΙώνα

ό θρόνο? σου δ θβός el? τον αιώνα_______

to the chosen angels as well by adding και των εκλεκτών αγγέλων after θεού και
Χρίστου Ιησού.
118. Paul seems to adhere to a bidirectional life for the believer, with the two foci being
God and Christ.
119. Although Hebrews’ author is anonymous, the author was at least a male (11.32)
contemporary of the apostle Paul’s protege Timothy (Heb 13:23), placing Hebrews
in the first century
120. There are two other variants in this verse that do not need further discussion
here: the omission of the conjunction καί and the word order of ή ράβδο? τη?
ευθυτητο?. The second one in no way affects our question of whether Jesus is ex­
plicitly called θεό?, and the first one, according to Metzger and others, would only
slightly reduce the difficulty of the last variant if it were to read αυτού. Still, for
clarity’s sake, 1 feel confident that these two variants together should read και ή
ράβδο? τή? ευθύτητα? (maintaining the καί and subsequent word order).

256 / Chapter 6 Brian J. Wright

 A Textual Examination

Externally, the absence of του α’ιώνος is significantly inferior, with

only a small handfill of concentrated MSS omitting it (B 33 t vg™).
Although it is true that scribes often expanded readings, that is not the
situation here, for several reasons. First, του αϊώνος is a direct quotation
from the LXX (Ps 44.7; MT Ps 45.7). Second, this reading is supported
by some of the best and earliest MSS (only a few omit it: B 33 t vgn13).
Third, almost every place □bw occurs in the Psalms according to the
MT, the LXX translates it with τοϋ αϊώνος (e.g., Ps 10.16; 45.7; 48.15;
52.10; 104.5). Even when this is not the case (e.g., Ps 20.5), it maintains
a resemblance (ε’ις αιώνα αϊώνος).
Putting it another way, if one accepts the shorter Greek rendering
of the OT quotation in Hebrews 1.8 (simply by els τόν αιώνα) and
does not include του α’ιώνος, it goes against all the ancient versions
of the OT. This variant’s potential implication for our study, though
not directly determinative on the Jesus-θεός issue, is to establish all
possible links to a Vorlage, best understand the grammatical struc­
ture, and assess every possible textual alignment (i.e., character count)
in various MSS. In this case, its OT reference (or even Vorlage) was
probably the LXX.

2. The second main textual issue in verse 8 is whether the last word
in the verse should read αύτοΰ or σου (Le., προς δέ τον υιόν ό θρόνος
σου ό θεός εις τον αιώνα του αιώνος, και ή ράβδος τής εύθύτητος
ράβδος τής βασιλείας [σου/αύτοΰ?]). The outcome, simply put, will
help determine whether ό θεός is a case of the nominative used for die
vocative (if σου) or a subject-predicate nominative (S-PN) construction
(if αύτου):

Nominative for vocative = “Your throne, O God, is forever and

ever, and a righteous scepter is the scepter of your kingdom.”

S-PN = “God is your throne [or, more likely, Your throne is God]
forever and ever, and a righteous scepter is the scepter of his [i.e., God’s]
kingdom.”!21....................121

121. Harris writes, "Grammatically, no valid objection maybe raised against these ren­
derings, but conceptually they are harsh”; “To render ό θρόνος σου δ θεός by'Your
throne is God’ is implausible in light of die articular θεός. ... No more prob­
able is the translation 'God is your throne’” (Murray Harris, “The Translation of
Elohim In Psalm 45,* TynBul 35 [1984]: 72, 89; whole article on 65-89). Even more,
though, nowhere else is the phrase “God is your throne” ever used. The expression,
according to T. K. Cheyne, is not “consistent with the religion of the psalmists”
(The Book ofPsalms: A New Translation with Commentary [London: Kegan, Paul,
Trench, Trubner, 1888], 127). Cf. Peter Craigie, Psalms ISO (Waco, TX: Word.
1983), 33-37. For an opposing view, see K. J. Thomas, "The Old Testament Cita­
tions in the Epistle to the Hebrews,” NTS 11 (1965): 303-25, esp. 305; A. Nairne,
The Epistle to the Hebrews (Cambridge: Cambridge University Press, 1917), 31-34.

Brian J. Wright Chapter 6 / 257

 JESUS AS ΘΕΟΣ

Internally, whereas they are both grammatically possible,122 only the

first construction resonates with the central theme of the section and book
(i.e., the exalted Christ). Ehrman believes, however, that the orthodox
party corrupted this text, because of their “need to differentiate Christ
from God.’’123 He concludes by saying, “(W]e are now dealing not with
a corruption of the original text but with a corruption of a corruption.”124
What Ehrman may be missing is that the author of Hebrews stands in
the exegetical tradition of the quoted Psalm. “That Jewish exegetes regu­
larly understood the text as an address is clear?' Abridge points out, “both
from the Targum and from the revision of the LXX by Aquila.”125 Little
doubt remains, then, that the LXX translator construed it so, suggesting
that ό θεός in Hebrews 1.8 points to Jesus’ essential unity with God while
preserving his functional subordination (see ό θεός σου in v 9).126 “It is
not impossible that the uniform testimony of the ancient versions in sup­
port of the vocative may reflect a messianic re-reading which stresses
the transcendence of the King-Messiah,” Harris writes, “but it is at least
equally possible that all these versions testify to the most natural way of
construing □’Tibs, whether they understood the word in reference to the
Messiah, or, as Mulder believes {Psalm 45 48), to God.”127128
Caragounis summarizes several other salient points regarding the
use of the nominative for vocative: (1) it occurred very early in clas­
sical Greek, (2) it originally applied to deities, (3) it was more frequent
in poetry than prose, (4) it gave greater emphasis, and (5) its usage
increased substantially in the NT from classical Greek.129 In addition,
after probing the centuries (from ancient to modern times) for the use
of the nominative ό θεός in lieu of the vocative, he concludes, “[T]he
articular nominative ό θεός when used as vocative has a more exalted,

122. E.g., Mitchell claims, “The predicate nominative is preferred here to the nomina­
tive as a vocative, so that God is not directly addressing the son as ‘God’" (Alan C.
Mitchell, Hebrews, ed. Daniel J. Harrington [Collegeville, MN: Liturgical, 2007],
49). Harris (Jesus as God, 212) lists two other commentators, Hort and Nairne, who
hold to this view. Wallace points out, though, “As to which of these two options
is better [subject or predicate nominative], we have already argued that with two
articular nouns, the first in order is the subject.... Hence, ό Θρόνος· σου would be
the subject rather than ό θ€0ς· (contra most NT scholars who opt for either of these
views)” (Exegetical Syntax, 59).
123. Ehrman, Orthodox Corruption, 265.
124. Ibid.
125. Harold W. Attridge and Helmut Koester, ed., The Epistle to the Hebrews (Philadel-
. phia: Fortress, 1989), 58. Cf. Cheyne, Psalms, 127.
126. See, e.g., Murray Harris, "The Translation and Significance of Ό ΘΕΟΣ in Hebrews
1:8-9,” TynBul 36 (1985): 129-62.
127. Harris, ‘'Elohim in Psalm 45,” 77-78. Cf. Gert J. Steyn, "The Vorlage of Psalm 45:6-7
(44:7-8) in Hebrews 1:8-9;’ HTS 60, no. 3 (2004): 1088.
128. Chrys C. Caragounis, The Development of Greek and the New Testament: Mor­
phology, Syntax, Phonology, and Textual Transmission (Grand Rapids: Baker,
2006), 141-43.

258 / Chapter 6 Brian J. Wright

 A Textual Examination

a more distanced, tone belonging to a more formal, solemn and el­

evated diction . . . [and h]ardly a more solemn or dignified context
could be imagined than the one in which this address is placed [Heb
1.8] .”129 In other words, all of these observations together strengthen
the internal probability of understanding ό θεός as a nominative for
vocative and thus supporting the more natural reading, σου.
With all that in mind, there is also a μέυ . . . δε construction in
verses 7-8. Wallace feels that the syntax of nominative for vocative ad­
equately handles this construction; the predicate nominative does not:
“Specifically, if we read v 8 as 'your throne is God' the δέ loses its ad­
versative force, for such a statement could also be made of the angels,
viz., that God reigns over them.”130 To sum this up another way, if one
holds to the predicate nominative view, there is no clear distinction here
between the angels (subordinate; ephemeral; servants) and Christ (supe­
rior; eternal; deity).
Lastly, various translators handle the preposition πρός differently
throughout this pericope (namely, 1.7, 8, 13). Several translations
render it “of” (e.g., ESV, NASB, NET, RSV), some “to” (e.g., KJV, NJB,
NLT), and still others “about” (e.g. CSB, NIV); with varying combi­
nations of all three instances. However, the translations with “of” or
“about” reflect a "misconstrual of the citation as a word about [of] the
Son, not to him.”131 In other words, πρός in verses 8 and 13 “must be
translated ‘to.’”132 This pertains to the present internal investigation
because it strengthens the preceding μέν . . . δέ discussion toward
translating the nominative as for vocative. I agree with Attridge, then,
that “the variant ‘his’ was probably occasioned by the ambiguity of the
preposition used to introduce the citations and the failure to construe
the whole citation as an address.”133
Externally, though both are well attested in the Alexandrian tra­
dition, the pronoun σου has more impressive weight and variety than
αύτου.134 Here is a snapshot of the witnesses supporting each:

129. Chrys C. Caragounis, “The Use of the Nominative ό Θεό? as Vocative in the Septu-
agint and the New Testament,” in Holy Scripture and the Ancient Word: Festschrift
for Professor loannis Galanis (Thessalonica: P.Pournaras, 2011).
130. Wallace, Exegetical Syntax, 59. Similarly, F. F. Bruce says, "Whatever be said of the
force of δέ in v. 6, there is no doubt about its strongly adversative force here, where
it harks back to μέι· in v. 7 (κα’ι προς μέν τούς άγγέλουε ... πρός δέ τον υιόν)”
{The Epistle to the Hebrews, rev.ed. [Grand Rapids: Eerdmans, 1990], 59).
131. Attridge, Hebrews, 57.
132. Ibid. Cf. George H. Guthrie, “Hebrews,” in Commentary on the New Testament Use
ofthe Old Testament, ed. G. K. Beale and D. A. Carson (Grand Rapids: Baker, 2007),
936; L. Timothy Swinson, "‘Wind' and Tire' in Hebrews 1:7 A Reflection upon the
Use of Psalm 104 (103),” 77 28 (2007): 218 n. 17; Wallace, Exegetical Syntax, 59.
133. Ibid., 59.
134. This assessment was kept even after recognizing that the combination of [β46 R B
"has the original reading in eleven other cases of minority readings in Hebrews”

Brian J. Wright Chapter 6 / 259

 JESUS AS ΘΕΟΣ

1. αύτου—ξ)46 8 Β Η S

2. σου—A D F Κ L Ρ Ψ 075 01210150 0243 0278 6 33 81104 256 263

326 365 424 436 459 1175 1241 1319 1739 1852 1881 1912 1962
2127 2200 2464 arm Byz copsa·bo' geo Leet it5>b· “mp· d> <· Vslav syrh·p-
pai(ms) Vg Chrysostom Cyril Gregory-Nyssa Jerome Theodoret

Outside the Alexandrian tradition (primary, [β46 K B; secondary, H),

αύτου is almost nonexistent to date (one majuscule, S). And of these six
MSS, they are only present in one class of NT witnesses (Greek MSS),
two categories of Greek MSS (papyri and majuscules), and four centu­
ries (i.e., £>46 [III] Κ [IV] B [IV] Η [VI] S [949]). In comparison, σου is
ubiquitous. Every possible geographic area—-Alexandrian (e.g., primary,
1739; secondary, 0243), Western (e.g., D), Byzantine (e.g., K), and other
important MSS (e.g., Ψ)—and every century from the third to the four­
teenth century (e.g., cop53 [III] Gregory-Nyssa [IV] A [V] D [VI] syr1'
[616] Ψ [VIII] 33 [IX] 1739 [X] 424 [XI] 365 [XII] 263 [XIII] 2200 [XIV])
contains at least one witness to σου. What is more, σου is present in
every class of NT witnesses (Greek MSS, ancient translations into other
languages, and quotations by early ecclesiastical writers) and virtually
every category of Greek MSS (papyri, majuscules, and minuscules).
One more external issue requires a response. Ehrman remarks, “It
is interesting to observe that the same MSS that evidence corruption
in Hebrews 1:8 do so in John 1:18 as well, one of the other passages.”* 135
While this brief statement is basically correct, he leaves the reader
with a distorted view of scribal activity and transmissional history.
Indeed, many examples of the reverse exist. Here are five examples
from the MSS he used numerous times regarding our present topic:

1.
a. Corrupted text according to Ehrman (i.e., calls Jesus θεός1):
Gal 2.20.
b. Texts that supports the reading that Ehrman’s entertains
(i.e., does not call Jesus 6e0s): Heb 1.8.

2. X(01)
a. Corrupted texts according to Ehrman (i.e., calls Jesus 6eos):
John 1.18; 20.28.
b. Texts that support Ehrman’s reading (i.e., does not call Jesus
θεό?): Acts 20.28; Gal 2.20; Heb 1.8; 2 Pet 1.1.

(Harris, Jesus as God, 210). For detailed understanding of the MSS for Hebrews,
see Attridge, Hebrews, 31-32. Cf. Frank W. Beare, "The Text of the Epistle to the
Hebrews inq5“,” JBL 63 (1944): 379-96; Ceslas Spicq, L^pitre auxHebreux, 3rd ed.
(Paris: Gabalda, 1952), 1:412-32.
135. Ehrman, Orthodox Corruption, 265.

260 / Chapter 6 Brian J. Wright

 A Textual Examination

X L(019)
a. Corrupted texts according to Ehrman (i.e., calls Jesus θεός):
John 1.18; 20.28.
b. Text that supports the reading that Ehrman entertains (i.e.,
does not call Jesus θεός): John 1.1.

4. L(020)
a. Corrupted texts according to Ehrman (i.e., calls Jesus θεός·):
Heb 1.8; Jude 4 (Ehrman does not mention this text directly,
but see n. 28 above).
b. Text that supports Ehrman’s reading (i.e., does not call Jesus
θεός): Gal 2.20.

5. W(032)
a. Corrupted texts according to Ehrman (i.e., calls Jesus θεός):
John 1.1; John 20.28.
b. Text that supports Ehrman’s reading (i.e., does not call Jesus
θεός): John 1.18.

In light of these five examples, which are only a small sampling,

much more work needs to be done in the realm of transmissional
history. More important, though, just given the preceding examples,
no one would have received a purely truncated view of the deity of
Christ if they only received their manuscript. This means that each
manuscript in the preceding list has at least one “Jesus-θεός” verse
that affirms the deity of Christ. It is inconsequential, then, that every
potential “Jesus-θεός” passage in every manuscript affirm the same.
This evidential conclusion causes another major problem in Ehrman’s
overall orthodox corruption thesis.136
In the end, the preponderance of evidence (geographically, genea­
logically, and internally) points to the true textual reading “but to the
Son [he declares], ‘Your throne, O God, is forever and ever, and a righ­
teous scepter is the scepter of your kingdom.’” The probability, then, is
high that Hebrews 1.8 explicitly calls Jesus θεός.

2 Peter 1.1
Second Peter 1.1 is another NT verse potentially calling Jesus θεός.
Some MSS (X Ψ 398 442 621Z596 syrph vg1’135 copsa)137 read κυρίου instead
of θεού in verse 1:

136. “It is rare to find a single manuscript consistently supporting a particular type of
reading in Ehrman’s categories,” concludes Parker (David C. Parker, An Introduc­
tion to the New Testament Manuscripts and Their Texts [New York: Cambridge
University Press, 2008], 302).
137. NA27 and Tischendorft differ on 2 Pet 1.1 regarding X. Nevertheless, after person­
ally checking a high-resolution digital photograph posted online by the Center for

Brian J. Wright Chapter 6 / 261

 JESUS AS ΘΕΟΣ
1

ev δικαιοσύνη του βςου ήμων και σωτήρος ’Ιησού Χριστού

ev δικαιοσύνη τού ίίό£ ήμών και σωτήρος ’Ιησού Χριστού

The external support, however, overwhelmingly favors θεού. In fact,

the NA27 and the Editio Critica Maior together only list nine witnesses
for κυρίου (mentioned previously, with only the NA27 listing vg^). This
means that virtually all other witnesses support θεού.
Nevertheless, there are several factors that appear to support κυρίου:

1. θεού could have arisen due to a scribal oversight of the nomen

sacrum·. ΚΫ vs. GY.*138

2. The phrase “Lord and Savior” is statistically superior when re­

ferring to Christ in 2 Peter. It reads “Lord and Savior” four times (1.11;
2.20; 3.2, 18), while it reads “God and Savior” only once (if one accepts
it in 1.1).

3. A shift to θεού could have been a motivated orthodox corrup­

tion to make the text speak unambiguously of Jesus as θεός· due to the
Christological controversies during the early centuries.

4. Κυρίου maintains the alleged parallelism between 1.1 and 1.2,

distinguishing θεός and Jesus.

5. θεός is rarely used of Jesus in the NT.

As for θεού, most of the preceding critiques can be justifiably re­

versed, while adding a few more arguments.

1. Although κύριος and θεός are among the earliest nomina

sacra,139 no other viable variants for κύριος or θεός exist in 2 Peter (1.14;
2.4, 9; 3.9).

2. “Lord and Savior” is to the NT (and 2 Peter) norm, and a scribe

could have easily harmonized to it.

the Study of New Testament Manuscripts (http://images.csntm.org/Manuscripts/

GAJ)1/GAOl_122a.jpg [accessed July 27,2011]) I determined that the NA27 is cor­
rect In other words, X attests to κυρίου. Cf. also Novum Testamentum Graecum:
Editio Critica Maior, vol. 4, Catholic Letters, installment 2, The Letters ofPeter, ed.
Barbara Aland, Kurt Aland, Gerd Mink, and Klaus Wachtel (Stuttgart: Deutsche
Bibelgesellschaft, 2000), part 1 (Text), 204.
138. These forms are reflecting their textual inflection in the verse, since that is the way
they would have been written (i.e., the last letter changes to reflect the form).
139. Hurtado, Origin of the Nomina Sacra? 655, 657.

262 / Chapter 6 Brian J. Wright

 A Textual Examination

3. Κυρίου might have been sought to maintain the alleged paral­

lelism between 1.1 and 1.2 (even though the alleged parallelism would
be extremely rare in the NT).

4. θεοΰ is the harder reading, as the opposing critiques reveal.

5. The construction is different when an author desires to distin­

guish two persons (e.g., 2 Pet 1.2:του θεού και Ίησου του κυρίου ήμών).140

6. The doxology in 2 Peter 3.18 and the phrase in 1.1 are attesting
to Jesus’ exalted status and are both consistent Christologically with the
rest of the NT.141 It should not be argued that tire differing words (“God”
in 1.1 and “Lord” in 3.18) refute this concept, since similar parallels can
be shown elsewhere with dififering words (e.g., Matt 1.23 and 28.20;
Mark 1.1 and 15.39; John 1.1 and 20.28).

7. This phrase might be in sync with Hellenistic religious language

in order to communicate the gospel meaningfully to Gentile converts.142

8. The external evidence is far better and earlier (not to mention

the existing unanimity among all major published NT Greek texts, e.g.,
NA27, Tischendorf8, UBS4, Bover, Merk, von Soden, Westcott and Hort,
Vogels, and Weiss).

9. The identification of Jesus as θεός· here is entirely realistic in

light of progressive revelation (2 Peter being one of the last NT books
written).143

10. The Granville Sharp rule undoubtedly applies to this construc­

tion, thereby referring both titles (“God” and “Savior”) to Jesus Christ.144

140. The Granville Sharp rule does not include proper names, and thus 2 Pet 1.2 does
not fit the rule (“Jesus” and “Lord Jesus Christ" are both proper names). Cf. Wallace,
Sharp's Canon, 159-62. On the use of the article with the name of God, see B. Weiss,
....... ''Der-GebrauchdesArtikelsbeidenGottesnamen,” TSK84(1911): 319-92,503-38.
141. See, e.g., Thomas R. Schreiner, 1, 2 Peter, Jude (Nashville: Broadman and Holman,
2003), 287.
142. See Tord Fornberg, “An Early Church in a Pluralistic Society: A Study of 2 Peter”
(PhD diss., Uppsala University, 1977), 143. Cf. Michael Amaladoss, Making All
Things New (Maryknoll: Orbis, 1990).
143. In addition, although 1.1 is the only explicit place Jesus is called θεό? in 2 Peter,
“other things 2 Peter says about Jesus more or less imply this same understanding.
One of the clearest instances is 1.3 where the author of 2 Peter speaks of τη? θεία?
δυι/άμεω? αυτού, and the antecedent of αυτού is probably Jesus, the last named
substantive (in v. 2)” (Terrance Callan, “The Christology of the Second Letter of
Peter," Bib 82 [2001]: 253; whole article on 253—63).
144. Furthermore, "The construction occurs elsewhere in 2 Peter [cf. 1.11 and 2.20],
strongly suggesting that the author's idiom was the same as the rest of the NT

Brian J. Wright Chapter 6 / 263

 JESUS AS ΘΕΟΣ

Michael Green argues, “It is hardly open for anyone to translate 1 Pet
1:3 ‘the God and Father’ and yet here decline to translate ‘the God and
Saviour’’145146
Likewise, Hiebert concludes, “Elsewhere, this epistle never
uses the word Savior alone but always coupled with another name under
the same article (cf. 1:11; 2:20; 3:2,18)."14S

At the end of the day, θεοί best accounts for all the evidence. If this
verdict is correct, it is highly probable that Jesus is explicitly called θεό?
in 2 Peter 1.1.

CONCLUSION
No one contests that the NT usually reserves the title θεό? for God
the Father. Yet this usage, though dominant, is not exclusive.147 The tex­
tual proof of the designation θεό? as applied to Jesus in the NT merely
confirms what other grounds have already established. In fact, the title
θεό? only makes explicit what is implied by other Christological titles,
such as κύριο? and υΙό? θεού. Harris adds,

Even if the early Church had never applied the title θεό? to Jesus, his
deity would still be apparent in his being the object of human and
angelic worship and of saving faith; the exerciser of exclusively divine
functions such as creatorial agency, the forgiveness of sins, and the
final judgment; the addressee in petitionary prayer; the possessor of all
divine attributes; the bearer of numerous titles used of Yahweh in the
OT; and the co-author of divine blessing. Faith in the deity of Christ

authors’” (New Testament: New English Translation, Novum Testamentum Graece,

608 n. 10). Of course, as some scholars note, one can hardly overlook the signifi­
cance of the personal pronoun ημών added to θεός in 2 Pet 1.1 (arguably disrupting
the Granville Sharp construction). Yet, after exhaustively examining 2 Peter, the
NT, and hundreds of nonbiblical papyri, Wallace states, "In all such instances the
possessive pronoun had no effect on breaking the construction. The fact, then, that
a possessive pronoun attached only to the first substantive never nullifies Sharp’s
principle—either in 2 Peter or in the NT or in the papyri that I have examined—is
strong confirmation of the validity of the rule in 2 Pet 1:1. In this case, as always,
presumption must give way to evidence” (Wallace, Sharp's Canon, 266). Cf. also
Gene L. Green, Jude and 2 Peter (Grand Rapids: Balter, 2008), 175.
145. Michael Green, 2 Peter and Jude (Grand Rapids: Eerdmans, 1993), 69.
146. D. Edmond Hiebert, Second Peter and Jude: An Exposltional Commentary (Green­
ville, SC: Unusual Publications, 1989), 37.
147. An argument based on the NT’s usage or nonusage of the title θεός for Jesus is
different from the claim that the NT authors were so entrenched in Jewish mono­
theism that they could not have thought of Jesus as θεός. Such a claim assumes
that they could not reconcile two truths or break away from their prior presup­
positions. Even though they may use "contradictory” terminology, they appear to
believe in the divinity of Jesus, sometimes even using preexistent categories (cf.,
e.g., 1 Cor 8.6; Col 1.15-17; Phil 2.6-11).

264 / Chapter 6 Brian J. Wright

 A Textual Examination

does not rest on the evidence or validity of a series of "proof-texts" in

which Jesus may receive the title θεός but on the general testimony of
the NT corroborated at the bar of personal experience.148

The question now before us is not whether the NT explicitly as­

cribes the tide Geos to Jesus but how many times he is thus identified
and by whom.149 Therefore, with at least one text that undoubtedly calls
Jesus θεό? in every respect (John 20.28) and with several others that
we can assume as much of with a similar degree of certainty (John 1.1,
18; Rom 9.5; Titus 2.13; Heb 1.8; 2 Pet 1.1; 1 John 5.20), we come back
to our initial question: when did this boldness to call Jesus θεός begin?
It began in the first century. It was not a creation of Constantine in the
fourth century. It was not a doctrinal innovation to combat Arianism in
the third century. Nor was it a subapostolic distortion of the apostolic
kerygma in the second century. Rather, the church’s confession of Christ
as θεός began in the first century, with the apostles themselves and/or
their closest followers. Such an affirmation therefore reaches back, in
some sense, to Jesus himself.

148. Murray Harris, “Titas 2:13 and the Deity of Christ,’ in Pauline Studies: Essays Pre­
sented to F. F. Bruce, ed. Donald A. Hagner and Murray J. Harris (Grand Rapids:
Eerdmans, 1980), 271.
149. A conceptual fallacy exists for any scholar to reject every possible text to show that
the original author(s) did not support this concept. Nevertheless, the answer to
this question will inevitably boil down to the presuppositions of each scholar (cf.
Robert H. Stein, Jesus the Messiah: A Survey of the Life of Christ [Downers Grove,
IL: InterVarsity, 1996], 17).

Brian J. Wright Chapter 6 / 265

 JESUS AS ΘΕΟΣ

Table 6.1—Degree of Certainty for Jesus as Θεό?

Too
A High A Lower
No Reason Uncertain to
Passage Degree of Degree of
to Doubt Allow Any
Probability* Probability
Reliance

Matt 1.23 X

John 1.1 X

John 1.18 X

John 17.3 ...... — — .. e .. ♦·♦·♦· ..... - · · ·■ --X·-™ t-- ------- .—··—<

John 20.28 X

Acts 20.28 X

Rom 9.5 X

Gal 2.20 X

Eph 5.5 X

Col 2.2 X

2Thess 1.12 X

1 Tim 3.16 X

Titus 2.13 X

Heb 1.8 X

2 Pet 1.1 X

1 John 5.20 X

Jude 4 • X
......
* While it is still possible to interpret the text another way, it is not highly
probable. •

266 / Chapter 6 Brian J. Wright

 Scripture Index

Exodus 104.5 257

20.4 231 123.2-3 159

Deuteronomy Proverbs
4.1-2 59 8 235
5.8 231
6.4 101,231 Isaiah
6.13 249 1.9 159,160
12.32 59
23.2 252 Micah
23.3 252 2.5 252
23Abis 252
23.9 252 Matthew
29.19-20 59 1.20-23 233
30.11-14 203 1.23 233,263,267
33 174
1 Chronicles 4.10 249
28.8 252 4.17 161
6.22 119,120
Job 6.25-30 223
12.7-8 203 6.28 135
7.3-5 211
Psalms 75 212
8.6 79 10.26 209
10.16 257 11.27 162
20:5 257 12,8 122
44.7 257 12.13 134
44.7-8 (MT) 258 13.57 218
45 258 13.58 134
45.6-7 258 16.4 134
45.7 257 16.20 161
45.7 (MT) 257 18.1 150,161
48.15 257 18.19-20 217
52.10 257 18.20 216
93.17 159 19.14 134
103 (MT) 259 19.16 53
104- 259 19.16-17 53,174

267*
 Scripture Index

19.17 53,230 10.31 205

1930 205 13.32 47,68, 69, 70,71,131,133,
20.16 205 135,136,137,139,146,157,
21.5 161 158,164,173
21.32 135 13.33 153
24.19 161 14.14 135
24.36 44,45, 46, 47, 48,50,60, 64, 15.34 82,83,175,230
68,70,71, 76,85,88,128, 15.39 118, 237,264
129,130,131,133,135,136, 16.8 138
137,138,139,141,143,144, 16.9-:20 21,44
145,146,148,149,152,153,
155,156,157,158,165,166, Luke
169,170,171,173,174,176, 1.80 174
178,179,180,181,183,186 2.16 41
24.42 153 2.29 235
25.45 135 2.33 88
26.39 174 2.38 88
27.14 135 2.40 174
27.16-•17 60,64,88,161 2.52 175
27.46 230 3.5 174
28.19 63 3.22 36,82
28.20 263 4.8 249
4.16 63
Mark 4.23-24 218
1.1 82, 263 6.5 122
1.3 174 6.20 204
1.41 20,21, 36,44,45, 49,88 6.41-42 211
2.28 122 7.12 245, 247
3.5 134 8.3 88
5.9 135 8.39 232
5.30 135 8.42 247
6 218 9.43 232
6.4 218 11.4 88,119,120
6.5-6 134 11.34 120
6.38 135 12.2 209
7.35 50 12.22 223
8.12 134,135 12.27--28 223
8.23 135 13.30 205
9.16 135 18.19 230
9.21 135 22.19--20 82
9.33 135 22.43 88
10.14 134 22.43--44 60,63, 82, 83
10.17 53 23.34 63
10.17- 18 53 23.45 I 60
10.18 162,174,230 24.12 82

268
 Scripture Index

24.51 88 3.3 107,115

24.51-52 82 3.4 115 '
24.53 88 3.5 107,239
3.6 239
John 3.13 88
1.1 75,92, 98,101,105,111, 3.16 74, 245, 247
118,122,123, 230, 235, 237, 3.18 74,109
246, 261,263,265, 266 3.21 115
1.1b 92,101,118,122 3.23 115
1.1c 48, 75, 93,95, 99,100,101, 3.28 115
102,103,106, 111, 113,114, 4.1 115
115,116,117,118,119,120, 4.5 115
121,122,123,124,125,126, 4.14 115
169,236,237,238, 4.17 115
239,240,248 4.23 115
1.1-5.11 105,106,113,114 4.24 115
1.1-8.38 243 4.25 115
1.1-18 101 4.27 115
1.3 92 4.30 115
1-8 114 4.31 115
1.14 75,235,245,247 4.34 115
1.16 239 4.35 115
1.17 235 4.36 115
1.18 44, 45, 72, 74,75, 76,82, 85, 4.37 115
86,112,113,174, 230, 240, 4.38 115
241, 245, 249, 4.42 115,248
261,265,266 4.42£>is 115
1.19 114,115 4.43 115
1.20 115 4.44 218
1.26 239 4.45 115
1.30 41,110 4.46 115
1.33 106, 240 4.48 115
1.37 115 4.52 115
1.39 114 5.2 109,115
1.44 115 5.3 111
1.47 114 5.9 115
1.49 115, 248,249 5.10 115
1.49&S 115 5.11 106
2.4 115 5.U&S 115
2.15&is 115 5.12 105,116
2.17 115 5.18 231
2.18 115 5.39 239
2.19 106 5.44 248
2.24 115 6.10 239
2.25 107 6.22 109

269
 Scripture Index

6.23 108 21.12 107

6.27 108
6.51 107 Acts
6.63&is 120 1.7 148,151,155
6.69 248 1.14 88
7.8 111 2.41 88
7.20 107 4.24 88,235
7.30 106 5.32 88
7.36 111 8.9-24 144
7.53-8.11 21,44 9.22 88
8.13 106 12.21-22 231
8.20 110 15.29 88
8.45 110 17.29 236
8.54 240 20.28 174, 230, 241, 251, 252, 253,
9.14 40 254, 261, 267
9.14-21 40 22.26 122
9.17 40 22.27 122
9.21 40 28.4 122
9.38 248
10.30 101, 111, 230 Romans
1031 106 1.9 122
10.33 174,231 4.24 255,
11.27 248 8.11 255
11.33 64,88 9.4 88
12.13 106, 249 9.5 230,232, 265, 266
12.27 175 9.29 160
12.34 110,111 10.9 255
12.38 250 16.27 248
12.41 174
14.8-10 101 1 Corinthians
14.9-11 111 1.2 252
14.19 230 2.8 156
14.28 175 2.11 159
15.1 120 6.14 255
16.30 107 8.6 264
16.33 239 10.5 174
17.2 106 10.32 252
17.3 175, 233, 248,267 11.3 120
17.14 107 11.16 252
19.5 174 11.22 252
19.34b 240 11.25 120
20.16 248 14.34-35 137
20.17 175, 230 15.9 252
20.28 36,111, 230, 246,248,249, 15.15 255
250, 260,263, 266 15.27-28 79

270
 15.47 174 5.21 232, 255
6.10 122
2 Corinthians 6.13 255
1.1 252
3.17 120 2 Timothy
4.14 256 2.21 235
4.1 232, 256
Galatians
1.3 88,252 Titus
2.19 256 1.3 232
2.20 174,241,254,255,261,266 2.13 234, 265, 266
3.4 232
Ephesians
1.7 252 Hebrews
5.5 233,267 1.3 174
1.6 259
Philippians 1.7 259
2.6-11 264 1.7-8 259
1.8 241,243,256,257,258,259,
i Colossians 260, 261, 265, 266
1.15-17 264 1.8-9 258
1.19 144 1.9 258
1.22 174 1.13 259
2.2 233, 266 2.8b 79
2.3 156 2.9 44, 45,78, 79, 81
2.9 144 2.9b 77,79, 80, 81, 82
2.18 174
| 1 Thessalonians 3.4 232
4 2.7 41 9.12 253
?4 2.9 42 10.29 174
** 2.14 252 11.17 245, 247
4.9 232 11.32 256
4
5.9-10 252 13.23 256
✓
2 Thessalonians James
)
t 1.4 252 1.1 232
)
( 233, 267
< 1.12
1 Peter
1 1 Timothy 1.3 264
I 1.1 232 1.22 88
I 248 5.1 174
1.17
3.5 252
3.15 252 2 Peter
3,16 82,83,174, 233,234, 266 1.1 234, 241,250, 261, 262,263,
3.16a 233 264, 265,266
 Scripture Index

1.2 234,263,263 5.20 234, 265, 266

1.3 236, 263
1.4 236 2 John
1.11 262,263,264 7 144
1.14 262
2.1 235 Jude
2.4 262 4 234, 235,261,266
2.9 262 5 235
2.20 262, 263,264 25 248
3.2 262,264
3.9 262 Revelation
3.18 262, 263, 264 2.14 59
2.20-23 59
1 John 6.10 235
1.1-3 144 13 42
3.23 174 13.18 42
4.1-4 144 18.23 121
4.3 82 19.13 235
4.9 74, 245,247 21.3 240
5.7 21,44 21.7 240
5.7-8 44 22.18-19 59

272
 Ancient Sources Index

Anastasius Sinaita Didymus

Hodegos Commentarii in Zacchariam
148 60 5.78 165

Ambrose De Trinitate
Defide 916-917 165
5.4.54 145 920 165
5.16 60
5.192-93 164' Enoch
1 Enoch
Athanasius 104.10-13 59
Discourses against the Arians 2 Enoch
3.42 153 48.6-9 59
3.43 153
3.44 153 Epiphanius
3.45 153 Ancoratus
3.46 153 16.2-6 157
3.49 153 17.1-4 157
28.472.47-52 154 19.6-7 157
21.3-5 157
Basil 22.1-4 157
Adversus Eunomium 22.4 158
696-98 161 31 60
Letters
197 164 Panarion
- - 236.1 162 ,30.1.1 141
236.1-2 161 30.2.1-8 141
236.2 162 30.3.7 141
30.16.1 141
Clement of Alexandria 30.26.1 141
Stromata 30.34.6 141
2.9.45 201 34 140
5.14.96 201 65.1.5-10 141
65.3.2-4 141
Clement of Rome 69.15.5 159
1'Clement 69.43.1 159
1.1 252 69.46 158

273
 Ancient Sources Index

Euripides 28 214,220
Ion 29 215, 220
211 250 30 216, 217
30+77b 207, 215, 220
Eusebius 30a 217
Ecclesiastical History 31 218, 220
3.27-28 141 32 219, 220
4.23.12 60 33 219, 220
5.20.2 59 36 222, 223, 226,
5.28 138,141 227, 228
5.28.3 142 37 223, 224, 226
5.28.16 60 38 224, 225, 226
6.17.1 138,141 39 225, 226
7.30 141 46 213
Ecclesiastical Theology 49 213
2.14.3 100 54 ' 204
2.17.1-2 100 57 193
Martyrdom ofPolycarp 62 213
1.1 252 69 213
Supplementa ad quaestiones ad 76 193
Marinum 77 216,217, 218
985 152 77b 217, 220, 221, 228
82 193, 213
Gospel of Thomas 87 193
1 200 90 201
1-7 209 101 213
2 200,201, 204, 210,228 106 193
3 193, 202, 203, 205, 107 213
210, 213, 228 109 193, 213
4 205 112 193
5 206, 207, 208, 209, 113 213
210, 216, 217 114 204
5-6 228 Nag Hammadi Codex II
6 207,208, 209, 210, 2.32.10-12 199
216, 217, 228 2.32.12-14 200
7 209 2.32.19-33.5 202
11 217 2.33.5-10 205
14 217 2.33.10-14 206
20 204 2.33.14-23 208
22 193, 213, 217 2.38.3-10 221
24 221, 226 2.38.12-17 211
26 211,213,220 2.38.17-20 212
26-33 220 2.38.20-31 214
27 193,204, 212, 213, 2.38.31-39.2 215
220,221,228 2.39.2-5 215

274
 Ancient Sources Index

2.39.5-7 218 18.3 252

2.39.7-10 219 18.4 252
2.39.10-18 219
2.39.24-27 222 Hilary
2.39.27-40.2 223 On the Councils
2.40.2-7 224 85 156
2.40.7-13 225 On the Trinity
2.46.22-28 215 1.29-30 156
2.51.27-28 196 1.30 155,156
Oxyrhynchus Papyri 1.32 156
1.1-4 211 7.7 142
1.4-11 212 9.60 156
1.11-21 214 9.62 155,156
1.22 215 9.63 155
1.23-27 215 9.67 155
1.27-30 215 9.71 156
1.36-41 219 9.75 155
1.41-42 219 10.8 155
30-35 218
654.1-5 199 Hippolytus
654.3-5 200 Against Noetus
654.9-21 202 17-18 148
654.21-27 205 Commentarium in Danielem
654.27-31 206 4.16.2 148
654.32-40 208 Philos
655, col. 1.1-17 222 9.1 145
655, col. 1.17-col. 2.1 223 Refutation ofAll Heresies
655, col. 2.2-11 224 7.21-22 141
655, col. 2.11-23 225 8.1 144
655 (Fragment d) 221
Ignatius
Gregory of Nazianzus Ephesians
Orationes prologue 231
"’30.15 163 1.1 231, 253
34.8 142 7 144
7.2 231
Gregory of Nyssa 15.3 231
Adversus Arium etSabellium de 17.1 252
patre etfilio 18 144
76 163 18.2 231
84-85 163 19.3 231

Hermas

Philadelphians
Shepherd ofHermas, Similitude 1.1
18.2 252 10.1

275
 Ancient Sources Index

Romans Justin
prologue 231 Apology
3.3 231 1.26 144
6.3 231, 254
Letter of Aristeas
Smyrnaeans 310-11 59
1.1 231, 252
1-6 144 Origen
10.1 231 Against Celsus
1.57 144
To Polycarp 2.27 60
8.3 231 Commentary on John
2.2.13-14 99
Trallians 2.2.17-18 99
2.3 252 2.2.17-20 99
7.1 231 Commentary on Matthew
10 144 2.13.14 150
12.1 252 2.15.14 150
15.14 60
Infancy Gospel of Thomas 55 146,150,151
7.4 231 121 60
126-27 151
Irenaeus 134 60
Against Heresies De principiis
1.23.1-4 144 1.6 152,166
1.26.1-2 141 Homilies on Luke
1.26.2 141,142 47 152
2.28.6 147
3.11.7 141 Philo
3.11.8 47,146 De Somniis
3.11.9 138,143 1.230 235
3.21.1 141
4.33.3 141 Photius
5.1.3 141 Epistle 138 to Theodore 60

Jerome Polycarp
Commentary on Matthew Epistle to the Philippians
4.24.36 172 1.1 252
7 144
Josephus
Against Apian Second Maccabees
1.42-43 59 9.12 231

Antiquities Sirach
Λ
19.345 231 1.1-10 235

276
 Ancient Sources Index

51.13-14 201 27.11 98

A Treatise on the Soul
Tacitus 33 149
Armais On the Flesh of Christ
14.15 231 5 144
On the Resurrection ofthe Flesh
Tatian 22 149
Compendium haereticarum The Prescription Against Heretics
fabularum 36 32
1.20 32
Wisdom of Solomon
Tertullian 1.1-2 201
Ad uxor 6.12-14 201
2.3.1 253
Adversus Praxeam (Against Praxeas)
26 149, 150

I
I

277
i

i
 Person and Subject Index

A Bengel, Johann Albrecht 42, 86

absolute certainty 22-26 Bentley, Richard 27
Adoptionism 46,48,66,68,72,74, Blomberg, Craig 27,45
140-142,144,147,176,180 Bludau, August 62
Adversus haereses (Irenaeus) 146 Brown, T. A. E. 125,126
Adversus Noetus (Hippolytus) 97 Burgon, John 62
Adversus Praxeam (Tertullian) Byzantine texts 53,74,104,130,133,
See Against Praxeas (Tertullian) 243,255, 260
Aejmelaeus, Lars 252
Aetians 140 C
Against Praxeas (Tertullian) 98,149 Caesarean texts 133,243,244,255
Alexandrian texts 36, 38, 51, Callistus 97,145,147
69, 72, 73, 99,104,114,130, canon of unorthodoxy
133, 238-239, 255, 260 See also textual variants and
and John 1:18 242-243 transmission of texts
allegorical interpretation 100 and Hebrews 2:9b 77-81
Ambrose of Milan 60,163 and John 1:18 72-77
Amphilochius of Iconium and Matthew 24:36 68-72
on Matthew 24:36 159-160 critique of 86-88
Ancoratus (Epiphanius) 157 Ehrman's methodology 71-72,
Anomeans 140 76-77, 80-81, 82-84
apocryphal works 191 in modern textual criticism
apostolic fathers 252, 254 61-64
Arianism 140,142,153,162,174, in the early church 59-60
181,246 orthodox corruption of texts 64-67
Ariiis—- ------------- 246 other texts 81-82
Asclepiodotus 142 Caragounis, Chrys C. 258
Athanasius of Alexandria 156,160 Celsus 60
on Matthew 24:36 152-154 Cerinthus 141
Attridge, Harold 192,197, 258, 259 Chester Beatty papyri 35
Augustine of Hippo 148,163 Christ and divine name
in 2 Peter 1:1 262-265
B in Acts 20:28 251-253
Basilides of Egypt 141,143 in Galatians 2:20 254-256
Basil of Caesarea 47,159,164 in Hebrews 1:8 256-261
on Matthew 24:36 160-162 in John 1:1 235-241
begging the question 81,102 in John 1:18 241-249

279
 Person and Subject Index

in John 20:28 249-250 Codex Sinaiticus 34,35,104,130,

as mediator 193 133,137,166,179, 243, 260
deity of 46, 72, 84, 89, Codex Vaticanus 33 -34,104,125,
128,141 137,179, 237
doctrine of 52 Codex W8 See Codex Freerianus
historicity of 52 Colwell, E. C. 102-103
omniscience of 146,152,165 Colwell’s rule 117-118,122
relationship to the Father 95-97, Commentary on Matthew
180 (Jerome) 171-172
Christology 59,229, 230, Commentary on Matthew
246, 248, 254 (Origen) 150-151,172
and textual variants 111 Commentary on Zechariah
early views of 66 (Didymus) 165
Eastern and Western 154 Conybeare, F. C.” 63
emerging 53-55 copying, nature of 35,36,51
heterodox 76 See also scribes, activity of
high 52,74 Council ofNicea 142,181
in Gospel of Thomas 217
in Matthew 50,134-136 D
of Marcion 143 Dead Sea Scrolls 35
orthodox 84-85,174 De anime (Tertullian) 149
church fathers 196 DeConick, April 195, 201,209,
and Matthew 24:36 (chart) 184-185 217-218
NT commentaries 28 definite article 239-241
on John 1:18 245 De Synodis (Hilary) 154
on Matthew 24:36 175-176 De Trinitate (Didymus) 165
on Matthew 24:36 De Trinitate (Hilary) 154
(summary) 165-169 Diatessaron 32,169,179
reaction to Sabellianism 97-100 Didymus the Blind of Alexandria
Clark, Kenneth W. 237 on Matthew 24:36 164-165
classical texts 27, 29-30 Dionysius 60
Clemens, Samuel 27,28 Discourses Against the
Qeomenes 145 Arians 152-153
Codex 1739 78, 79 dispensational literature'■ 43
Codex B See Codex Vaticanus Docetism 66,142, 143-144,180
Codex Bezae 64 dualisms 193
Codex Freerianus 92, 102,105-106,
124-126, E
237, 261 Ebionites 47,141-142
scribal profile of 113-116 economic trinity 98-99
Codex L See Codex Regius Ehrman, Bart 28,236
Codex Regius 92, critique of his thesis 84-86
102,104-105, 116-117,124, text-critical methodology
237--238, 261-262 71-72, 76-77, 80-81, 82-84,
scribal profile of 4
106-113 169-173

280
 Person and Subject Index

theologically motivated alterations 213-214,221-222,

57-58, 65-67,128-130, 258,260 223-224, 228
view of original text 23-26 transmission of texts 191-193,
view of textual variants 19-22, 196-197, 228
54-55 Granville Sharp rule 250, 263
Encratism 170 Gregory of Nazianzus 159
Epigonus 145 on Matthew 24:36 162-163
Epiphanius of Salamis 60,140 Gregory of Nyssa
on Matthew 24:36 156-159 on Matthew 24:36 163
Epp, Eldon 64 Griesbach, Johann J. 61
Erasmus, Desiderius 57
eschatology 42 H
Eunomians 140 harmonization 52- 54, 70-71,110,
Eusebius of Caesarea 60 134·-135,178-179,
on Matthew 24:36 152 199, 205,218, 228, 256,262
reaction to Sabellianism 100 Harnack, Adolf 252
evangelists, the 50 Harris, J. Rendell 63
Harris, Murray 240, 247, 258
F Head, Peter M. 64,191,194
Fee, Gordon 49,86 Hengel, Martin 87
Fitzmyer, Joseph A. 194 Heterousians 140
Five Gospels, The (Funk et aL) 22 Hexapla (Origen) 150
Freer, Charles 125 Hilary of Poitiers
Freer Gallery 126 on Matthew 24:36 154-156
fundamentalism 88 Hippolytus of Rome
Funk, Robert 22-23 on Matthew 24:36 147-148
reaction to Sabellianism 97-98
G homoousian 162
Gnosticism 77, 95,141,142,143, Horner, George W. 239
146,194-195 Hort, F. J. A. 62,128
Goodspeed, Edgar J. 125,126 Howard, Wilbert F. 63
“Gospel according to Bart, Ihe” Hurtado, Larry 125,191
(Wallace) 44 hypostatic union 84,180
Gospel.ofThomas_
analysis of textual variants 209, I
220-221, 226-228 Ignatius of Rome 231
challenges in textual indefinite article 239-241
criticism 193-194 interpretation of the text 21
scribal activity in 199,201, Introduction to the New Testament
(Ehrman) 129
222-223,227-228 Irenaeus of Lyon 46
textual criticism of 190-192 on Matthew 24:36 146-147
textual variants and
theological alteration 203-204, J
207, 210-211, Jehovah’s Witnesses 236

: 281
IJ*
 Person and Subject Index

Jerome 140 Monophysitism 163

Jesus Seminar 22 Monothelitism 163
Johannine style 74, 248 Moo, Douglas 255
John (apostle) 146 Morgan, Matthew P. 48,237, 238
John Chrysostom 156 Mueller, Dieter♦Λ* 204
mystery religions 85
K
XV
King James Only advocates 22, 26 N
King James Version 130 Nag Hammadi 196
knowledge (gnosis) 193 Nag Hammadi manuscripts 35
Kostenberger, Andreas J. 242 New Finds manuscripts 35
Kuhn, K. H. 217-218 Noetus of Smyrna 95,144,147
nomina sacra 115,143, 248,262
L nonsense readings 41,103,109,
Lake, Kirsopp 63 115-116,193
“Leper in the Hands of an Angry Novum Instrumentum-Omne
Jesus, A” (Ehrman) 21 (Erasmus) 57
Lewis, C. S. 40
liberalism 88 O
logical fallacy 81 Olivet Discourse 45,68,128
Logos Christology 97,100 On the Resurrection ofthe Flesh
Lost Christianities (Ehrman) 24 (Tertullian) 149
Origen of Alexandria 46,60-61,156
M on Matthew 24:36 150-152
Macrina 160 reaction to Sabellianism 98-99
Manichaeism 144 Orthodox Corruption ofScripture,
manuscript production 31--32,101 The (Ehrman) 20, 43,47, 58,65,
manuscripts, NT 81,129,130
reliability of 49-54 Oxyrhynchus 196
Marcion 139,143-144
Marcionism 142 P
Marcovich, Miroslav 197 260
McReynolds, Paul M. 245 ^75 33,104, 236
meaningful variants 42-43 Panarion (Epiphanius) 156,158
Merk, Augustinus 124 parallelism 205, 211,212,216, 218,
Messer, Adam 45 223, 262, 263
Metzger, Bruce 28, 54,64,83,254 P atrip assianism 66, 85, 95,96,142
Meyer, Marvin 195 Paul of Samosata 142
Middleton, T. F. 117 Perrin, Nicholas 191,192,195
MisquotingJesus (Ehrman) 20, Philosophumean (Hippolytus) 147
24, 31, 43, 49, 54, 58,129 philosophy 85, 97,148,165
Modalism 75,142,143,148,150, physicality 224
176-177,181, 236 Plato 165
modernism 22 Pliny the Younger 231
Monarchianism 95, 96,142 Plisch, Uwe-Karsten 216

282
 Person and Subject Index

Polycarp See Polycarp of Smyrna scriptorium 107,137,166

Polycarp of Smyrna 146, 231 Scrivener, Frederick 62-63
postmodernism 23 Separationism 66, 84
post-resurrection ignorance 146,165 Silva, Moises 49
Praxeas of Rome 95,98,144 Simon Magus 144-145
predicate nominative Simon, Richard 61
constructions 118-122 singular readings 103
pre-resurrection ignorance 146,165 skepticism 22-23, 54
presuppositions 87,102,264,265 Small, Keith 38-39
Prior, J. Bruce 125,126 sociohistorical setting 84,85,100
spelling differences 40
R St. Catherine’s Monastery 35
Rabbula of Edessa 32 Studies in the Textual Criticism of
reasoned eclecticism 21,81,232 the New Testament (Ehrman) 129
redaction criticism 50 subordinationism 96
Reformation, the 37 sub singular readings 103,114
rigorous eclecticism 87 substitution 216
Robertson, A. T. 92,120 Supplementa ad quaestiones ad
Robinson-Pierpont Majority Martnum (Eusebius) 152
Text 102,106,114 synonyms 41
Synoptic Gospels 49-50,197, 229
s
Sabellianism 48, 92-93,120, T
142,177,181,236 Tatian 32,170,179
influence on transmission Taylor, Vincent 53
of NT texts 100-102 Tertullian of Alexandria 32
reaction of church Tertullian of Carthage 46
fathers to on Matthew 24:36 148-150
roots of reaction to Sabellianism 98
Sabellius Text of the New Testament, The
Sackler Gallery (Metzger and Ehrman) 52,129
Sanders, H. A. Textual Commentary on the Greek
Saturnilus of Syria New Testament (Metzger) 64
scribes textual variants
activity of See also canon of unorthodoxy
analysis of, in Gospel
255, 260 of Thomas 210,220-221,
in Codex Freerianus 113-116 226-228
in Codex Regius 106-113, and grammatical issues 117-123
116-117 dates of manuscripts 28-31
in Gospel of Thomas 199,201, definition of 26
209, 212,213-214, heretical corruption 142-145
222--223,227-228 in Hebrews 1:8 257
analysis of activity 102-103 in Matthew 24:36 133-138,
harmonization by 70 186-188

283
 rerson and Subject Index

nature of 40-43 of apocryphal works 191

number of 26-40 Qur’an, the 34-40
number of manuscripts 26-28 transposition 41,207,216,217-218
orthodox corruption 139-142
significance of 21 U
theological issues 43-49, Uthman (calif) 35,37-38
111-113,116-117,122-123,
173-179 V
theologically motivated alterations Valentinus of Rome 143
in the Gospel of Thomas 203-204, von Soden, Hermann 124,126
213-•214,221-222,
223-224, 228 W
Theodoret of Cyrrhus 32 Wachtel, Klaus 126
Theodotians 147 Wallace, Daniel B. 74,78,259
Theodotus the Money-Changer 142 Westcott, B. F 37-38,128
Theodotus the Tanner 140 Western texts 36, 38,51,69,
Tischendorf, Constantin von 126 105,114,130,255, 260
total despair 22-26 and John 1:18 243-244
Trajan 231 and John 20:28 249-250
translations, vernacular 37 Wettstein, Johann 61
transmission of texts 59-67,92 Williams, C. S. C. 64,139
and Acts 20:28 252-253 wording 42-43
Gospel of Thomas 191-193,
196-197,228 Z
NT manuscripts 31-40, Zephyrinus 97,142,147
104-106, 260 Zuntz, Gunther 81
NT manuscripts and Sabellian
influence 100-102,123-124

284
1