Speech-in-Character, Diatribe, and Romans 3:1–9

Biblical Interpretation Series

Editors in Chief

Paul Anderson (George Fox University)

Jennifer Koosed (Albright College, Reading)

Editorial Board

A.K.M. Adam (University of Oxford)

Colleen M. Conway (Seton Hall University)
Amy Kalmanofsky (Jewish Theological Seminary)
Vernon Robbins (Emory University)
Annette Schellenberg (Universität Wien)
Johanna Stiebert (University of Leeds)
Duane Watson (Malone University)
Christine Roy Yoder (Columbia Theological Seminary)
Ruben Zimmermann (Johannes Gutenberg-Universität Mainz)

volume 163

The titles published in this series are listed at brill.com/bins

 Speech-in-Character,
Diatribe, and Romans 3:1–9
Who’s Speaking When and Why It Matters

By

Justin King

LEIDEN | BOSTON
Library of Congress Cataloging-in-Publication Data

Names: King, Justin, author.

Title: Speech-in-character, diatribe, and Romans 3:1–9 : who’s speaking when
 and why it matters / by Justin King.
Description: Leiden ; Boston : Brill, 2018. | Series: Biblical interpretation
 series, ISSN 0928-0731 ; volume 163 | Includes bibliographical references
 and index.
Identifiers: LCCN 2018022183 (print) | LCCN 2018025548 (ebook) | ISBN
 9789004373297 (E-book) | ISBN 9789004373280 (hardback : alk. paper)
Subjects: LCSH: Bible. Romans, III, 1–9—Criticism, interpretation, etc. |
 Speech acts (Linguistics)
Classification: LCC BS2665.6.S6 (ebook) | LCC BS2665.6.S6 K56 2018 (print) |
 DDC 227/.1066—dc23
LC record available at https://lccn.loc.gov/2018022183

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface.

ISSN 0928-0731
ISBN 978-90-04-37328-0 (hardback)
ISBN 978-90-04-37329-7 (e-book)

Copyright 2018 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi,
Brill Sense and Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910,
Danvers, MA 01923, USA. Fees are subject to change.

This book is printed on acid-free paper and produced in a sustainable manner.

 Τῷ Παύλῳ
δούλῳ Χριστοῦ Ἰησοῦ
κλητῷ ἀποστόλῳ

To Hannah
a personification of patience and grace

∵
Contents

Acknowledgments xi
List of Illustrations and Tables xiii

1 Introduction 1
Staging the Project 2
Methodology 7
Significance 9
An Outline: A Preview to the Project 11

part 1
Speech-in-Character

Introduction to Part 1 15

2 Speech-in-Character in the Rhetorical Handbooks 19

Pseudo-Cicero: Rhetorica ad Herennium 19
Introduction 19
Analysis: Sermocinatio 20
Analysis: Conformatio 22
Synthesis: Speech-in-Character 24
Quintilian: Institutio Oratoria 26
Introduction 26
Analysis: Fictiones Personarum / Προσωποποιία 27
Summary: Ad Herennium and Quintilian on Speech-in-Character 36

3 Speech-in-Character in the Progymnasmata 38

Theon: Προγυμνάσματα 38
Introduction  38
Analysis: Προσωποποιία 40
[Hermogenes]: Προγυμνάσματα 45
Introduction 45
Analysis: ἠθοποιία, προσωποποιία, εἰδωλοποιία 46
Summary: Theon and [Hermogenes] on Speech-in-Character 51
viii Contents

4 Speech-in-Character: A Synthesis 54
Core Conventions of Speech-in-Character  54
Unique Features of Speech-in-Character 55

5 Examples of Speech-in-Character in Paul 58

Examples of Speech-in-Character in Paul 63
1 Thessalonians 5:3 63
Galatians 3:8 68
Galatians 4:6 71
1 Corinthians 1:12; 3:4 74
1 Corinthians 12:3 78
1 Corinthians 12:15–16, 21 81
2 Corinthians 12:9 85
Romans 10:6–8 88
Conclusion 95

part 2
Diatribe

Introduction to Part 2 101

6 Portrayals of Diatribe 103
Conclusion 124

7 Examples of Diatribal Dialogue 129

Introduction 129
Dialogue and Attributed Speech in Primary Sources for Diatribe 132
Teles, Fragment I: Περὶ τοῦ δοκεῖν καὶ τοῦ εἶναι 132
Epictetus, Discourses 2.20: Πρὸς Ἐπικουρείους καὶ
Ἀκαδημαϊκούς 138
Epictetus, Discourses 2.23: Περὶ τῆς τοῦ λέγειν δυνάμεως 149
Conclusions 157
Contents ix

part 3
Romans 3:1–9

Introduction to Part 3 163

8 Traditional Readings of the Dialogue in Rom 3:1–9 and Its Role in the
Letter 165
Pre-Bultmannian Readings of Romans 3:1–9 166
Diatribal Readings that Affirm the Traditional Script of
Romans 3:1–9 169
Preliminary Conclusions 187

9 Rescriptive Readings of the Dialogue in Romans 3:1–9 and

Its Role in the Letter 197
Non-Diatribal Critique of Traditional Readings 197
Rescriptive Readings of the Dialogue of Romans 3:1–9 199
Preliminary Conclusions 218

10 Romans 1–2: The Ethnically Inclusive and Impartial Gospel and the
Characterization of the Interlocutor 219
Romans 1:1–12 219
Romans 1:13–15 222
Romans 1:16–17 225
Romans 1:18–32 234
Romans 2 238
Romans 2:1–11 238
Romans 2:12–16 242
Romans 2:17–29 247

11 Romans 3:1–9 and the Argument of Romans 252

Romans 3:1–9 252
Romans 3:1–2 253
Romans 3:3–4 256
Romans 3:5–6 260
Romans 3:7–8 262
Romans 3:9 266
Romans 3:1–9 Rescripted 268
x Contents

Romans 3:1–9 and the Remainder of Romans 274

The Reoccurrence of Questions and Themes 274
God’s Inclusive and Universal Promises: Rom 3:1–9 and 9–11 278
A Purpose of Romans 288
Conclusion 292

12 Conclusion 294

Bibliography 299
Modern Authors Index 315
Subject and Name Index 318
Ancient Sources Index 321
Acknowledgments

The apostle Paul’s description of the Church as a body, in which all members
work together to strengthen the whole, applies appropriately in the case of
this project. Every page bears the influence of countless teachers, colleagues,
friends, and family members to whom I owe the deepest level of gratitude. I
hope these brief words can somehow resemble my ineffable appreciation.
I discovered my intrigue with the academic study of early Christian litera-
ture during my undergraduate years at Oklahoma Baptist University. Though
many professors at OBU like Carolyn Cole challenged and encouraged me,
three biblical studies professors prepared and set me on this course that I am
currently traveling. Bobby Kelly imparted to me his love and aptitude for the
Greek language. C. Mack Roark was no less influential in linguistic matters,
introducing me to the broader scope of Hellenistic Greek literature. Finally,
Justin Hardin shared with me his passion for Paul’s life and letters. Hardin’s
fascination with Paul was contagious, and, though it manifests at times as a
blessing and at times as a curse, I completely caught it. To these professors who
were so influential in my early academic career, I am forever in debt.
My M.A. degree at the University of Chicago represented a period of sig-
nificant transformation and development. There, too, three professors deserve
mention. Under the tutelage of Margaret M. Mitchell, Hans Josef-Klauck, and
Jeffrey Stackert my historical-critical abilities grew significantly. Without the
influence of these scholars, this project would have been impossible. Mitchell
deserves a special note of thanks, as a conversation we shared concerning what
it means to “distrust Paul” in the interest of historical inquiry was particularly
formative. To these scholars as well, I am forever in debt.
The New Testament faculty at Baylor, however, has stored up a veritable
treasure trove in heaven for their sacrifices and contributions to my education.
Bruce W. Longenecker, Beverly R. Gaventa, Mikeal Parsons, Lidija Novakovic,
and Kelly Iverson each rank among the best the academy and humanity have
to offer. I have profited greatly from each of them, both regarding their aca-
demic rigor and their gracious interactions with students and one another. The
influence of Parsons’s rhetoric seminar in which I was a participant my first
semester at Baylor is surely visible throughout this project. Similarly, Gaventa’s
arrival at Baylor was a blessing, and her expertise on Paul and Romans was par-
ticularly influential in motivating me to define more concretely my own views
about this significant letter. Last but by no means least, Longenecker, my men-
tor, steered me from numerous pitfalls, and any that remain are surely due to
my occasional but rare decisions to argue against the grain. This project grew
xii Acknowledgments

out of a seminar Longenecker taught on Paul, and from the very beginning his
expertise, encouragement, and patience were the glue that held this project to-
gether. Again, without the guidance of Baylor’s New Testament faculty in these
ways and others, this project would have been impossible, and I will be forever
grateful. My dissertation committee and readers deserve an additional thank
you: Longenecker, Gaventa, Phillip J. Donnelly, Parsons, and Novakovic.
Countless colleagues, friends, and family members also contributed to
this project directly and indirectly. My colleagues, especially Scott Ryan, John
Duncan, Lindsey Trozzo, and Mike Whitenton made helpful insights and often
excelled at keeping me distracted. Beginning at OBU, my friends Jerry and
Diane Newsom have been a constant source of encouragement, good coffee,
and better conversation. My family, of course, has made sacrifices of inesti-
mable worth to assist in this project’s completion. My mom, Missy King, and
parents-in-law, Debbie and Carey Skinner, deserve nothing but my highest ap-
preciation. My brother and best friend, Jerrod King, has perhaps been my big-
gest source of inspiration and drive to complete this project. My wife, Hannah,
and our dogs, Darcy and Cooper, provided constant reminders of normalcy
often forgotten in the course of a Ph.D. program. Hannah’s care, support, love,
and willingness to allow me to devote so much time and mental energy to this
project have amassed a debt I will never be able to repay. To all mentioned
and so many more, this project is as much yours as it is mine, and I am forever
thankful.
List of Illustrations and Tables

Illustrations

1 Painting of a schoolboy being punished from the house of Julia Felix in

Pompeii 126
2 Drawing of the painting of a schoolboy being punished from the house of
Julia Felix in Pompeii 127

Tables

1.1 Arrangements of the script of Rom 3:1–9 in secondary scholarship based

on to which voice each line is attributed 6
11.1 Arrangements of the script of Rom 3:1–9 compared to my arrangement
informed by speech-in-character 269
 chapter 1

Introduction

Who’s speaking when and why does it matter? This question is applicable to
numerous passages in what is arguably the most influential Christian docu-
ment ever composed—Paul’s letter to the Romans. For instance, what is the
identity of the infamously agonized “I” who speaks in Rom 7:7b–12, 13b–24, 25b?
The “Christian” anachronism aside, is the speaker in these verses the Christian
Paul, the non-Christian Paul, an imaginary Christian speaker, or perhaps an
imaginary non-Christian speaker? Consider also Paul’s attribution of speech
to δικαιοσύνη based on πίστις in 10:6–8. What does it mean for δικαιοσύνη to be
characterized by πίστις? More to the point, what does it mean for δικαιοσύνη
based on πίστις to speak, and how does it advance the argument of Romans?
The question of who speaks when and why it matters is equally applicable
to Paul’s no less enigmatic consideration of Jewish advantage over non-Jews in
Rom 3:1–9. The text reads:

(3:1) Therefore, what advantage comes from being Jewish, or what benefit
accrues from circumcision? (3:2) There is much [advantage] in every way!
To begin, they were entrusted with the oracles of God. (3:3) To what end?
If some lacked πίστις, their lack of πίστις will not nullify God’s πίστις, will
it? (3:4) Absolutely not! Instead, let God be true but every human a liar,
as it is written, “So that you might be justified in your words, and you
will overcome when you are judged.” (3:5a) But if our unrighteousness
proves God’s righteousness, what shall we say? God is not unjust when he
brings wrath, is he? (3:5b) I speak in a human way. (3:6) Absolutely not!
Otherwise, how will God judge the world? (3:7) But if God’s truthfulness
is increased for his glory by my lie, why am I still being judged as a sinner?
(3:8a) Why not, (3:8b) as we are slandered and as some claim that we say,
(3:8c) “Let us do evil so that good might come?” (3:8d) Their judgment is
justly deserved. (3:9a) What then? Are we advantaged or disadvantaged?
(3:9b) By no means! (3:9c) For we have charged both Jews and Greeks all
to be under Sin, (3:10) as it is written …

Does Paul raise a series of rhetorical questions in 3:1–9 that he himself answers,
or does the passage represent imaginary discourse between Paul and a hypo-
thetical interlocutor? Assuming 3:1–9 represents fictitious dialogue, how is the
reader to understand the back-and-forth exchanges within the discourse? Paul

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_002

2 chapter 1

provides no overt indicators (such as a verb of speech) marking the transitions

between speakers. Does the interlocutor pose questions for Paul to answer, or
is it the other way around? Might the dialogue partners instead alternate be-
tween asking and answering questions? How would a reader even determine
which speaker is responsible for which lines? Besides, what difference does it
make who speaks which lines? Does it matter whether Paul asks the questions
in 3:1 or responds in 3:2, for example? Do certain arrangements of the script
create tension between 3:1–9 and other material in Romans or Paul’s theolo-
gizing elsewhere? If so, what do those tensions indicate about Paul’s thought?
Conversely, might alternative arrangements allow 3:1–9 to fit more harmoni-
ously in the holistic argument of the letter and Paul’s thought? Specifically,
does Paul, the apostle to the gentiles, think God privileges Jews over non-Jews
(cf. 3:2, 9; 9–11)? These are difficult questions to be sure, and this project en-
gages and proposes solutions for these important but complex questions con-
cerning Rom 3:1–9. Indeed, who is speaking when in Rom 3:1–9, and why does
it matter?

Staging the Project

This project is divided into three parts: Part 1, “Speech-in-Character,” Part 2,

“Diatribe,” and Part 3, “Romans 3:1–9.” Because relevant literature on any one of
these topics often fails to bridge the gaps between the other two, and the few
works that do are either incomplete or unpersuasive (see below), each Part
contains an introduction of relevant issues, as is suitable. As such, only a brief
survey is necessary here to introduce the topic(s) at hand and begin to create
space for this investigation’s contributions.
Rudolf Bultmann’s 1910 dissertation, Der Stil der paulinischen Predigt und
die kynisch-stoische Diatribe,1 established within Pauline studies the dialogical
nature of Paul’s letters and their relationship to the Greco-Roman diatribe. In
Bultmann’s view, Paul’s preaching and epistolary style approximates the Cynic-
Stoic street preachers who employ diatribe in their propagandistic and po-
lemical messages for the masses. Bultmann’s paradigm held sway over Pauline
scholarship for seventy years. With the publication of Stanley K. Stowers’s

1  Rudolf Bultmann, Der Stil der paulinischen Predigt und die kynisch-stoisch Diatribe (reprint;
Vandenhoeck & Ruprecht: Göttingen, 1910).
Introduction 3

dissertation in 1981, The Diatribe and Paul’s Letter to the Romans,2 the tide
turned on Bultmann’s prominence. Stowers redirected New Testament schol-
arship’s understanding of the diatribe, most significantly in terms of its scho-
lastic setting and the conventions for using it in argumentation. It is in the
wake of Stowers’s dissertation that scholars began to recognize the immense
relevance diatribe’s rhetorical conventions might have for explaining the dia-
logical structure of Rom 3:1–9.
Before Stowers, scholars traditionally read Rom 3:1–9 in one of two ways;
either they ignored or denied that dialogue was present,3 or they read the pas-
sage as though an imaginary interlocutor raised objections to Paul’s message
for the apostle to answer.4 According to this arrangement of the script, with
only minor nuances the interlocutor poses questions in Rom 3:1, 3, 5, and 7–8c,
and Paul responds in 3:2, 4, 6, 8d. This “traditional” reading is unsurprisingly
represented by Bultmann.5
Beginning with Stowers, however, scholars began to recognize that the gen-
eral conventions for diatribe allow one to revise, or “rescript,” the arrangement
of the dialogue in Rom 3:1–9. In these readings, scholars appeal to diatribal
conventions in order to attribute to Paul’s voice various lines traditionally
voiced by the interlocutor, and vice versa.6 Such revisions not only produce

2  Stanley Stowers, The Diatribe and Paul’s Letter to the Romans (Society of Biblical Literature
Dissertation Series 57; Chico: Scholars Press, 1981).
3  For example, though published two years after Stowers’s dissertation, David R. Hall,
“Romans 3:1–8 Reconsidered,” NTS 29.2 (1983): 183–97.
4  Almost all Romans commentaries neglect the dialogue of Rom 3:1–9 or read it tradition-
ally. To my knowledge, the only exceptions are Ben Witherington III with Darlene Hyatt,
Leander E. Keck, and Brendan Byrne. Witherington, Hyatt, and Keck uncritically follow
Stowers, and Byrne charts a slightly different course. Ben Witherington III with Darlene
Hyatt, Paul’s Letter to the Romans: A Socio-Rhetorical Commentary (Grand Rapids: Eerdmans,
2004), 93–4; Leander E. Keck, Romans (ANTC; Nashville: Abingdon, 2005), 89–96; Brendan
Byrne, Romans (SP 6; Collegeville: Liturgical Press, 1996), 106–15.
5  This is the case, though he barely addresses 3:1–9. Bultmann, Der Stil, 67, 94. On the sparse
treatment of 3:1–9, see Paul J. Achtemeier, “Romans 3:1–8: Structure and Argument,” ATR sup
11 (1990): 79.
6  Most relevant to my study, Stanley K. Stowers, “Paul’s Dialogue with a Fellow Jew in Romans
3:1–9,” CBQ 46 (1984): 707–22; idem., A Rereading of Romans: Justice, Jews, and Gentiles (New
Haven: Yale University Press, 1994); idem., “Apostrophe, Προσωποποιια, and Paul’s Rhetorical
Education,” in Early Christianity and Classical Culture: Comparative Studies in Honor of
Abraham J. Malherbe (eds. John T. Fitzgerald, Thomas H. Olbricht, and L. Michael White;
NovTSup 105; Leiden: Brill, 2003), 351–69; Neil Elliott, The Rhetoric of Romans Argumentative
Constraint and Strategy and Paul’s Dialogue with Judaism (JSNTS 45; Sheffield Academic
4 chapter 1

new readings of Rom 3:1–9 but of Paul’s overall argument in Romans as well.
These are the “rescriptive” readings, since they rearrange the script7 of the dia-
logue between Paul and his interlocutor.
At the time of his dissertation, Stowers apparently did not completely rec-
ognize the ramifications his reassessment of diatribe might have for the script
of Rom 3:1–9, as he seems to maintain the traditional reading. Stowers soon
returned to Rom 3:1–9, however, in an attempt to solve four problems he as-
sociates with traditional readings of the pericope: (1) the difficulty in account-
ing for its place in the letter, (2) the view that it lacks unity and coherence,
(3) the inability to make sense of its dialogical nature, and (4) the production
of readings in which Paul speaks in a plethora of voices.8 Relying on introduc-
tory formulas and the observation that diatribal teachers typically ask ques-
tions and guide discussions, Stowers attempts to resolve these problems by
revising the script of 3:1–9 for the first time. In Stowers’s reconfiguration, the
interlocutor speaks in 3:1, 4, 6, 9a, and Paul speaks in 3:2, 3, 5, 7–8, 9b. Stowers,
therefore, maintains the frame of the traditional reading, but he inverts the
middle exchanges so that Paul poses the questions in 3:3 and 5 for the inter-
locutor to answer. Stowers also attributes to Paul the questions and response
in 3:7–8. Ten years later, Stowers maintained this script and argument, (largely)
reprinting it in A Rereading of Romans, his third attempt at the pericope.9
Stowers’s rescriptive agenda for Rom 3:1–9 did not fall flat, but neither did
it fully persuade. On the one hand, Neil Elliott revises the script of 3:1–9 even
more exhaustively than Stowers. Elliott accepts Stowers’s internal adjustments,
but he identifies inconsistencies in Stowers’s arrangement of 3:1–2 and 7–9.
Elliott does not engage the primary diatribal literature as Stowers. Rather,
Elliott follows Stowers’s presentation of diatribe and uses it against him. Elliott
argues that 3:1 and 9a should be read as recapitulative leading questions in

Press, 1990; reprint; Minneapolis: Fortress Press, 2007); idem., The Arrogance of Nations:
Reading Romans in the Shadow of Empire (Paul in Critical Contexts; Minneapolis: Fortress
Press, 2008); Douglas A. Campbell, The Deliverance of God: An Apocalyptic Rereading of
Justification in Paul (Grand Rapids: Eerdmans, 2009).
7  By “script,” I mean to evoke the concept of a dramatic script, such as for a play or movie,
in which the various lines of a discourse are attributed to one character or another. For an
ancient reference, as depicted by Plato, many of Socrates’ dialogues are representative of this
style. As already noted, the dilemma with the script of the dialogue in Rom 3:1–9, and thus
the catalyst and justification for this project, is that Paul does not—overtly—indicate which
lines of the discourse belong to each speaker.
8  Stowers, “Paul’s Dialogue,” 707–22.
9  Idem., Rereading, 159–75.
Introduction 5

Paul’s voice, which the interlocutor answers in 3:2 and 9b. Thus, Elliott com-
pletely inverts the traditional script, so that Paul takes on the role of Socratic
questioner in 3:1, 3, 5, 7–8c and 9a, and the interlocutor responds in 3:2, 4, 6, 8d
and 9b.10 Elliott also maintains this script in his later work, The Arrogance of
Nations,11 and Douglas A. Campbell follows him completely.12
On the other hand, not all (not even most) works on Romans and diatribe
since Stowers’s publications result in rescriptive readings of 3:1–9. For exam-
ple, Changwon Song continues to endorse the traditional reading.13 Song’s ar-
gument for the traditional script, however, fails to convince. Relying solely on
Epictetus’ Discourses as evidence for diatribe, Song focuses mainly on the use
of μὴ γένοιτο by Epictetus and Paul. Unfortunately, Song’s only argument is that,
in Discourses, μὴ γένοιτο “as a statement of rejection … [is] usually attributed
to the Teacher.”14 But Song immediately cites numerous exceptions to the rule
and allows that Stowers’s reading “may be possible also.”15 Thomas H. Tobin
similarly supports the traditional script of Rom 3:1–9, but he too fails to provide
substantive arguments for his dialogical arrangement. Like Song, Tobin only
demonstrates engagement with Epictetus. Further, Tobin offers no argument
for why he reads 3:1–9 in the traditional sense. Rather, Tobin simply asserts the
traditional reading without evidentiary support; Tobin assumes, “Once Paul
has made his argument … he then deals with objections that might be raised
against his position.”16
Generally speaking, therefore, three arrangements of the dialogue in 3:1–9
exist: (1) traditional readings; (2) Stowers’s rescription; and (3) Elliott’s rescrip-
tion. To view the various options at a glance, refer to the translation above and
Table 1.1.

10  Elliott, Rhetoric, 132–41.

11  Elliott, Arrogance, 105–7, 205n74.
12  Campbell, Deliverance, 572–4, 1088n117.
13  Changwon Song, Reading Romans as a Diatribe (Studies in Biblical Literature 59; New
York: Peter Land, 2004), 94–95.
14  Ibid., 94–95, 112n5.
15  Ibid.
16  Thomas H. Tobin, Paul’s Rhetoric in its Contexts: The Argument of Romans (Peabody:
Hendrickson, 2004), 118–22, see also 120n44. Similarly, Stanley E. Porter, The Letter to the
Romans: A Linguistic and Literary Commentary (NTM 37; Sheffield: Sheffield Phoenix
Press, 2015), 84–87.
6 chapter 1

table 1.1 Arrangements of the script of Rom 3:1–9 in secondary scholarship based on to which
voice each line is attributed

Line in script Traditional Stowers Elliott

3:1 Interlocutor Interlocutor Paul

3:2 Paul Paul Interlocutor
3:3 Interlocutor Paul Paul
3:4 Paul Interlocutor Interlocutor
3:5a Interlocutor Paul Paul
(3:5b) Paul; authorial aside
3:6 Paul Interlocutor Interlocutor
3:7 Interlocutor Paul Paul
3:8a
(3:8b) Paul; authorial aside
3:8c Interlocutor
3:8d Paul Interlocutor
3:9a Interlocutor Interlocutor Paul
3:9b Paul Paul Interlocutor

One of the primary problems plaguing treatments of Rom 3:1–9, however, is

one and the same for both traditional and rescriptive readings. This underlying
and overarching problem is the reliance on and acceptance of assumptions in
lieu of argumentation supported with valid evidence. The reality is, despite the
amount of discussion trying to explicate Rom 3:1–9’s dialogue, little persuasive
argumentation actually exists on either side of the spectrum. As we will see, this
holds true in multiple ways not only for the scholars already mentioned but also
for the works of Abraham J. Malherbe,17 Rafael Rodríguez,18 James D. G. Dunn,19

17  Abraham Malherbe, “Μη Γενοιτο in the Diatribe and in Paul,” HTR 73 (1980): 231–40.
18  Rafael Rodríguez, If You Call Yourself a Jew: Reappraising Paul’s Letter to the Romans
(Eugene: Cascade, 2014).
19  James D. G. Dunn, Romans 1–8 (WBC 38a; Dallas: Word Books, 1988), 128–44; idem., The
Theology of Paul the Apostle (Grand Rapids: Eerdmans, 1998), 117–19.
Introduction 7

N. T. Wright,20 Douglas J. Moo,21 Robert Jewett,22 John M. G. Barclay,23 and

Stanley E. Porter,24 amongst others.
Though this argumentative and methodological neglect represents an un-
fortunate state of affairs in biblical scholarship, it is simultaneously something
of a paradoxically fortunate opportunity. Neither reading begins with a privi-
leged foot forward; traditional and rescriptive readings are on equal—albeit
unsupported—footings. This means there is no room to show initial partiality
to one reading over another simply based on some scholar’s argument on its
behalf. Consequently, what is required is not a simple weighing of the merits
and demerits of relevant views (though this is important), but a reassessment
of method and a thoroughgoing application of it to Rom 3:1–9. In this way, to
the degree possible, argument and evidence shall precede and validate conclu-
sions rather than be assumed to fit within a set of presupposed views. Such
a need is brought into even sharper relief by the fact that so many scholars
draw from a common pool of diatribal tradition but inevitably come to dif-
ferent conclusions concerning the shape and meaning of Rom 3:1–9. If under-
standing 3:1–9 is to gain clarity in its own right and concerning its function
holistically in Romans, then evidence, argument, and greater methodological
constraint are required.

Methodology

This naturally brings us to the question of method. This monograph primarily

draws on tools within the historical-critical arsenal, especially those relevant
for analyzing Romans in light of first-century rhetorical conventions. For over a
century, scholars have recognized the validity of reading Romans in relation to

20  N. T. Wright, “Romans 2:17–3:9: A Hidden Clue to the Meaning of Romans?,” Journal for
the Study of Paul and His Letters 2.1 (2012): 1–25; idem., The Letter to the Romans (NIB 10;
Nashville: Abingdon, 2002), 452–55; idem., “The Law in Romans 2,” in Paul and the Mosaic
Law (ed. James D. G. Dunn; WUNT 89; Tübingen: J.C.B. Mohr (Paul Siebeck), 1996), 131–50;
idem., Paul and the Faithfulness of God: Book II (Christian Origins and the Question of
God 4; Minneapolis: Fortress Press, 2013), 836–39.
21  Douglas J. Moo, The Epistle to the Romans (NICNT; Grand Rapids: Eerdmans, 1996), 177–97.
22  Robert Jewett, Romans: A Commentary (Hermeneia; Minneapolis: Fortress Press, 2007),
238–52.
23  John M. G. Barclay, Paul and the Gift (Grand Rapids: Eerdmans, 2015), 471–74, see also
483n.89.
24  Stanley E. Porter, The Letter to the Romans: A Linguistic and Literary Commentary (New
Testament Monographs 37; Sheffield: Sheffield Phoenix, 2015).
8 chapter 1

diatribe, either as a formal diatribe or as a text evincing diatribal elements. As

Stowers emphasizes, one of the most salient features of diatribe is its dialogical
aspect,25 and it is this feature that is so important for Rom 3:1–9. This project
advances this tradition, with one caveat. As shown below, the primary sourc-
es for diatribe display a remarkable degree of diversity. This variation in part
accounts for why so many scholars appeal to the corpus of primary diatribal
literature as evidence for understanding Rom 3:1–9 but nevertheless produce
divergent interpretations. Stated plainly, due to its inherent instability, relying
solely on diatribal literature as evidence for explicating Rom 3:1–9’s dialogue is
unable to produce a consistent or persuasive reading. Additional, more stable
evidence is necessary if one hopes to improve comprehension of Paul’s staged
discourse in Rom 3:1–9.
Fortunately, such a body of evidence stands at the ready, namely, the litera-
ture related to the rhetorical practice of attributing speech to speakers other
than oneself, “speech-in-character.”26 Because in diatribe every instance of dia-
logue with an imaginary interlocutor requires the creation and attribution of
speech, every occurrence of such dialogue necessarily engages in speech-in-
character.27 What is more, numerous primary sources share and teach speech-
in-character’s core conventions of attributing speech to an imaginary speaker
that is appropriate to that speaker’s character. Thus, speech-in-character is
both directly relevant to diatribe and represents a significantly more consis-
tent body of evidence from which to draw conclusions about Rom 3:1–9 (or any
other ancient, Greco-Roman, dialogical text).
The premise is that if—as with Rom 3:1–9—a dialogical text does not iden-
tify imaginary speakers by name, verbs of speech, introductory formulas, or
other overt means, then speech-in-character’s complementary conventions
of characterization and appropriateness to that characterization can help
identify which lines in a discourse belong to which speaker. For example, if an
imaginary speaker is characterized as uneducated, speech that thoughtfully

25  Stowers, Diatribe.

26  I am aware of the varying terms for the practice in the Progymnasmata and broader
rhetorical tradition, which distinguish between προσωποποιία, ἠθοποιία, and εἰδολοποιία.
See Part 1 for my decision to translate the various technical terms as “speech-in-character.”
See George A. Kennedy, Progymnasmata: Greek Textbooks of Prose Composition and
Rhetoric (Leiden: Brill, 2003), 47–49, 84–85, 115–17, 164–66, 213–17; Rhetorica Ad Herennium
4.55, 65; Quintillian, Institutio Oratoria 9.2.
27  Note that, though diatribe and speech-in-character are linked, they are not identical.
In staging a diatribal dialogue, the speaker or writer necessarily engages in speech-in-
character when inventing and/or attributing words to an interlocutor. Thus, speech-in-
character is only one part of the diatribal dialogue.
Introduction 9

engages Aristotle’s Ars Rhetorica or Cicero’s De Oratore would probably not be

appropriately read or heard in that imaginary speaker’s voice. On the contrary,
if the interlocutor is characterized as well educated, such speech could be alto-
gether appropriate and can serve as an aid to identify the speaker by retrospec-
tively identifying a matching characterization. In this way, the conventions for
speech-in-character bring new light to diatribal dialogues that otherwise am-
biguously transition between speakers. Parts 1 and 2 contribute to developing
this approach, and Part 3 illustrates how this method is particularly appropri-
ate and helpful for understanding the script of Roman’s imaginary dialogue.

Significance

Romans 3:1–9 is perhaps the most obviously dialogical and most vexing peri-
cope in the entirety of Romans. Because of the difficulty in determining who
speaks which lines in the dialogue, and because certain arrangements of the
script confuse rather than clarify Paul’s logic, the passage has been touted as
“one of the most difficult, perhaps, in the epistle”28 and as “obscure and feeble,”
to the extent that the whole epistle would make better sense if the pericope
was omitted.29 More recently and more optimistically, Rom 3:1–9 has been
proclaimed as a key to understanding the whole of Romans.30 Given such
vastly differing perceptions, Rom 3:1–9 is certainly a text deserving of fresh
inspection.
It is surprising that current scholarship lacks an extensive treatment of
Rom 3:1–9 that compiles and analyzes the diversity of approaches to and re-
visions of the staged dialogue in the passage and its function in the epistle.
This monograph, first of all, fills this lacuna by creating a readily accessible and
critical compendium of the diverse evidence and arguments offered by New
Testament scholars.
More importantly, though the traditional reading can be identified as the
majority view at this time, the increase in rescriptive readers illustrates that a
general consensus does not exist. Again, this is in large part due to the nature of
the evidence and arguments (or lack thereof) employed to support either view.

28  Frédéric Louis Godet, Commentary on Romans (translated by A. Cusin; Grand Rapids:
Zondervan, 1969), 131.
29  C. H. Dodd, The Epistle of Paul to the Romans (Moffatt New Testament Commentary;
London: Hodder and Stroughton Limited, 1932), 46.
30  N. T. Wright, “Romans 2:17–3:9.”
10 chapter 1

This project secondly provides what is perhaps the first, and at least the most
thoroughgoing, methodologically focused argument concerning the arrange-
ment of the dialogue in 3:1–9 and its import. What makes this possible is the re-
alization that the conventions for speech-in-character are inherently relevant
to the dialogical portions of diatribal literature, including Paul’s letter to the
Romans. As discussed below, though a few scholars begin to make this observa-
tion, none allow it to make its full or appropriate contribution to Rom 3:1–9. By
allowing the conventions for speech-in-character to exert their due influence,
this project draws on both speech-in-character and diatribal evidence in order
to provide the evidentiary stability that diatribe alone is unable to offer.
Third, as demonstrated in Part 2, little progress has been made in our un-
derstanding of diatribe since Stowers’s work(s). There has been no thorough
reassessment of the primary sources or their application to Rom 3. Instead,
scholars seem content to rely heavily on Stowers’s analyses (to the degree dia-
tribe is concerned, this work does too). This project, however, addresses the
problem of over-reliance on Stowers by rebalancing the focus of the conversa-
tion from diverse diatribal sources to the much more stable primary literature
on speech-in-character. The relevant primary sources for speech-in-character
are engaged extensively and exhaustively, both as individual texts and in con-
junction with one another. This allows one to identify the core conventions of
the rhetorical figure that all or most sources share and the elements uniquely
attested by each individual source. These core conventions constitute what
one would expect to find in almost any concrete example of attributed speech
in antiquity. Conversely, the uniquely attested features in each source amount
to elements that an author like Paul might employ when composing speech-
in-character but that are not central or necessary to the proper implementa-
tion of the rhetorical figure. This project, therefore, makes a contribution in
the area of speech-in-character, but, because of the close relationship between
speech-in-character and diatribal dialogue, it also makes significant advances
in diatribal studies.
Fourth, and finally, this work seeks to answer the perennially problematic
questions of the meaning and function of Rom 3:1–9, both on its own terms
and in the scheme of Romans. This is accomplished in three ways: (1) by defin-
ing whose voice is responsible for each line of the dialogue’s script, providing
the evidence for and adhering more closely to the rhetorical conventions than
previous scholarship has achieved; (2) by demonstrating how a correct dif-
ferentiation of the speakers significantly influences our understanding of the
pericope; and (3) by properly situating the passage into the larger argumenta-
tive context of Romans. Though this investigation touches on countless points
in Pauline studies, it has immense bearing for questions of Paul’s rhetorical
Introduction 11

acumen, his endorsement of divine impartiality, and his understanding of

the relationship between Jews and non-Jews in God’s eschatological, salvific
agenda.

An Outline: A Preview to the Project

My argument progresses from the broadest sphere (historical and rhetorical

backgrounds), through various scholars’ approaches to Paul and Rom 3:1–9,
and finally down to the narrowest sphere in which I engage Rom 3:1–9 and
its epistolary context. Part 1 concerns speech-in-character. Chapter 2 inves-
tigates the way two rhetorical handbooks address the practice of attributing
speech to other characters. These handbooks are the pseudonymous Rhetorica
ad Herennium and Quintilian’s Institutio Oratoria. Chapter 3 similarly sur-
veys the presentation of speech-in-character in the earliest two collections of
Progymnasmata, those of Theon and [Pseudo-]Hermogenes. Chapter 4 catego-
rizes the findings from Chapters 2 and 3 in order to create a composite picture
of the rhetorical figure in antiquity. Finally, Chapter 5 looks at examples of
Paul’s use of speech-in-character throughout his literary corpus.
Part 2 takes up the issue of diatribe. Chapter 6 reviews the approaches to
diatribe in secondary literature. Chapter 7 analyzes concrete examples of di-
atribal dialogue in light of the conventions for both diatribe and speech-in-
character. Thus, Chapter 7 (and Chapter 5) sets the methodological course for
the investigation of Rom 3:1–9 to follow.
Part 3 applies the findings from Parts 1 and 2 to the dialogue in Rom 3:1–9
and considers its fit and function in the letter as a whole. Chapter 8 provides a
history of research that reviews and begins to analyze traditional readings of
Rom 3:1–9, and Chapter 9 is the corresponding material on rescriptive readings
of the dialogue. Chapter 10 investigates the beginning of Romans and Paul’s
apostrophic characterization of his interlocutor in Rom 2. Chapter 11 addresses
the dialogue in 3:1–9 and situates the pericope in the broader argumentative
scope of Romans.
Chapter 12 draws conclusions from the project as a whole and considers fu-
ture research trajectories. The stage thusly set, it is now time to engage the fas-
cinating rhetorical figure of speech-in-character—the appropriate attribution
of speech to a speaker other than one’s self.
 part 1
Speech-in-Character

∵
Introduction to Part 1

The following four chapters examine the rhetorical practice of speech-in-char-

acter. This term, “speech-in-character,” conveys most accurately the core ele-
ments of the concept, namely, writing or giving a speech that coheres with the
character of another speaker.1 In the course of this examination, the detailed

1  This solves the terminological problem in which various ancient authorities utilize diverse
terms to discuss a single exercise. James R. Butts, “The Progymnasmata of Theon: A New
Text with Translation and Commentary” (Ph.D. diss, The Claremont Graduate School, 1987),
459–60. Stowers also follows Butts’s translation of “speech-in-character.” Stowers, Rereading,
16–17, 333n.40; idem., “Romans 7:7–25 as a Speech-in-Character (προσωποποιία),” in Paul
in His Hellenistic Context (edited by Troels Engberg-Pederson; Minneapolis: Fortress Press,
1995), 180n.1.
The issue regarding what to call this rhetorical technique is both technical and transla-
tional. It is technical in the sense that one must first decide whether or not προσωποποιία
and ἐθοποιία (amongst others) constitute different or identical exercises. The analyses below
strongly indicate that it is reasonable to think about προσωποποιία and ἐθοποιία as essen-
tially describing a single technique, namely, the attribution of speech to another character.
To begin, the first-century writers Quintilian and Theon use the single term προσωποποιία to
express the whole concept of attributing speech to a diversity of character types. Moreover,
even among the writers who differentiate between the terms, the differentiation only has to
do with the type of character in whose mouth words are placed and not with an entirely dif-
ferent technique altogether.
The issue is also translational, as one must decide what modern-language translation(s)
of the technical terms best expresses the practice described by them. Since the working
understanding is that both προσωποποιία and ἐθοποιία ultimately pertain to the composition
of speech in the voice of another character, what translation best expresses the concept?
The range of translations is broad: (1) Donald A. Russell opts for “impersonation.” Quintilian,
Institutio Oratoria (ed. and trans. Donald A. Russell; LCL 127; Cambridge: Harvard University
Press, 2001), 9.2.29. (2) Harry Caplan and George Kennedy employ “personification” for
προσωποποιία. [Cicero], Rhetorica ad Herennium (trans. Harry Caplan; LCL 403; reprint;
Cambridge: Harvard University Press, 2004), 4.66; Kennedy, Progymnasmata, xiii, 47.
(3) George Kenendy, Ray Nadeau, and Charles Baldwin translate ἐθοποιία as “characterization.”
Kennedy, Progymnasmata, xiii, 84, 115; Ray Nadeau, “The Progymnasmata of Aphthonius
in Translation,” SM 19 (1952), 278–79; Charles Sears Baldwin, “The Elementary Exercises
(ΠΡΟΓΥΜΝΑΣΜΑΤΑ) of Hermogenes,” in Medieval Rhetoric and Poetic (New York: Macmillan,
1928), 34–35. Finally, (4) Stanley Bonner and Butts translate “speech-in-character,” though
Bonner nevertheless defines it in terms of impersonation. Stanley F. Bonner, Education in
Ancient Rome: From the elder Cato to the younger Pliny (Berkeley: University of California
Press, 1977), 253; Butts, “Theon,” 459–60. To problematize several of these translations,
the translation of the concept as “impersonation” potentially suggests the kind of mimicry

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_003

16 part 1

intricacies of speech-in-character will find fuller form than the limited defi-
nition provided above suggests, both with respect to its core elements and
its unique variations in presentation by its advocates. Therefore, the primary
goals of Part 1, “Speech-in-Character,” are twofold: (1) identify from the primary
literature the elements that are central to the practice of speech-in-character,
and (2) highlight any differences or developments in the various treatments of
speech-in-character.
The methodological assumption uniting this twofold goal is that the core
elements of speech-in-character provide the surest foothold when analyzing
Paul’s application of this exercise, whereas the differences advanced by vari-
ous ancient proponents represent unessential but potential elements Paul
may or may not follow in any given occurrence of speech-in-character in his
letters. This of course contains the caveat that the form of speech-in-character
presented here must be a form of speech-in-character that could have been
relevant to Paul. So, the primary sources examined must not only discuss
speech-in-character, but they must also pre-date or be in close chronological
proximity to Paul. For this reason, exhaustive engagements with the fourth-
century Progymnasmata of Aphthonius and the fifth-century Progymnasmata
of Nicolaus are excluded, though their differences are documented in the
footnotes.
Secondary scholarship on speech-in-character has tended to take one of
three basic forms: (1) general and brief treatments of speech-in-character,
usually as one monolithic concept in Greco-Roman rhetoric which ignores
or downplays the differences between various authors, (2) dictionary or en-
cyclopedia style articles that isolate various elements of speech-in-character,
and (3) discussions of tightly focused aspects within the broader practice of
speech-in-character and pertinent to a given scholar’s arguments about New
Testament texts (or other ancient documents).2 Given the significant degree of

and sarcasm Quintilian specifically wishes to avoid (Inst. 1.8.3), “personification” fails to
apply to the attribution of speech to human persons who, by definition, do not need to be
“personified,” and all of the translations with the exception of “speech-in-character” fail to
communicate the core element of speaking in another character’s voice. Thus, though it
might be somewhat cumbersome, translating the concept of προσωποποιία and ἐθοποιία as
“speech-in-character” is by far the best, as it adequately expresses the central element of
speech that occurs in the technique.
2  For treatments of speech-in-character, see Stanley K. Stowers, “Romans 7:7–25,” 180–88; idem.,
Rereading, 16–21; Charles Sears Baldwin, Ancient Rhetoric and Poetic (Gloucester: Peter Smith,
1924, reprinted 1959), 71–73; Tobin, Paul’s Rhetoric, 227; George A. Kennedy, A New History of
Classical Rhetoric (Princeton: Princeton University Press, 1994), 205–6; idem., Greek Rhetoric
Under Christian Emperors (A History of Rhetoric 3; Princeton: Princeton University Press,
Speech-in-Character 17

overlap between such summations of speech-in-character, no extensive survey

of research on the technique is necessary, and any points of contention with
other scholarship will be addressed as they arise.
The approach to speech-in-character presented here differs from these gen-
eral trends in several important ways. First, instead of picking and choosing
which portions of theoretical texts to discuss, texts are engaged exhaustively.
Second, this analysis is not isolated to a single type of writing but is triangu-
lated between the treatments of rhetorical theory in the so-called handbooks
and the earliest two Progymnasmata, the classical rhetorical texts cited by
these rhetorical theorists, and with concrete examples in the New Testament.
Third, this examination indicates which elements are the most salient features
of speech-in-character and which elements represent subsidiary or potentially
optional aspects. Fourth, as an aide to future work on speech-in-character and

1983), 64; Heinrich Lausberg, Handbook of Literary Rhetoric: A Foundation for Literary Study
(Trans. Matthew T. Bliss, Annemiek Jansen, and David E. Orton; eds. David E. Orton and
R. Dean Anderson; Leiden: Brill, 1998), §820–829, 840, 1131–1132, 1137.3; Christine Heusch, “Die
Ethopoiie in der griechischen und lateinischen Antike: von der rhetorischen Progymnasma-
Theorie zur literarischen Form,” in ‘ΗΘΟΠΟΙΙΑ: La représentation de caracteres entre fiction
scolaire et réalité vivante a l’époque imperial et tardive (edited by Eugenio Amato and Jacques
Schamp; Cardo 3; Salerno: Helios, 2005), 11–33; Ronald F. Hock, “The Rhetoric of Romance,” in
Handbook of Classical Rhetoric in the Hellenistic Period: 330 B.C.–A.D. 400 (ed. Stanley E. Porter;
Leiden: Brill, 1997), 445–65; David E. Aune, The Westminster Dictionary of New Testament and
Early Christian Literature and Rhetoric (Louisville: Westminster John Knox, 2003), 383; R. Dean
Anderson Jr., Glossary of Greek Rhetorical Terms Connected to Methods of Argumentation,
Figures and Tropes, From Anaxamenes to Quintilian (Contributions to Biblical Exegesis and
Theology 24; Leuven: Peeters, 2000), 106–7; Donald Lemen Clark, Rhetoric in Greco-Roman
Education (Westport: Greenwood Press, 1957), 199–201; Laurent Pernot, Rhetoric in Antiquity
(trans. W. E. Higgins; Washington D.C.: Catholic University of America Press, 2005), 148;
Witherington and Hyatt, Romans, 179–80; Charles H. Talbert, Romans (Smyth & Helwys
Bible Commentary; Macon: Smyth & Helwys, 2002), 186–87; Campbell, Deliverance, 532–
33; Alicia D. Myers, Characterizing Jesus: A Rhetorical Analysis on the Fourth Gospel’s Use of
Scripture in its Presentation of Jesus (LNTS 458; London: T & T Clark, 2012), 51–55; Michel
Patillon and Giancarlo Bolognesi, eds., Aelius Theon: Progymnasmata (Paris: Les Belles Lettres,
2002), xxxiv–xxxviii; Matthew W. Bates, The Hermeneutics of the Apostolic Proclamation: The
Center of Paul’s Method of Scriptural Interpretation (Waco: Baylor University Press, 2012), 194–
99. For a treatment of speech-in-character based largely on [Hermogenes] and Aphthonius,
see Craig A. Gibson, “Prosopopoeia in the New Testament: Where should we look and what
should we expect to find?” (paper presented at the annual meeting of the Society of Biblical
Literature, Philadelphia, PA, November, 2005), used with permission. For a tightly focused
treatment of certain aspects of speech-in-character, see Stowers, “Apostrophe,” 351–69.
18 part 1

early Christian writings, this work compiles as much of the methodologically

pertinent evidence as possible in one central location.
The following discussion of speech-in-character begins with an investigation
of two rhetorical handbooks, ad Herennium and Quintilian’s Institutio Oratoria.
I begin with ad Herennium because it is the earliest extant and complete treat-
ment of speech-in-character among all types of sources. Thereafter, generic
similarity is prioritized (Quintilian). Corroborating comments on speech-in-
character from other ancient rhetoricians will be included in the footnotes
as appropriate. After seeking to understand each text individually, their pre-
sentations of speech-in-character are compared and contrasted (Chapter 2).
The works of the progymnasmatic genre, Theon and [Hermogenes], follow,
and each similarly receives a close reading individually before being com-
pared and contrasted (Chapter 3). Here, too, chronological priority (Theon
before [Hermogenes]) sets the order in which texts are discussed. Afterwards,
findings from Chapters 2 and 3 are brought together in order to synthesize
and finalize my presentation of the central as well as secondary elements of
speech-in-character (Chapter 4). Finally, these findings are employed to ana-
lyze select practical examples of speech-in-character in the New Testament,
simultaneously documenting Paul’s awareness of and aptitude for this literary
and rhetorical technique (Chapter 5).
 chapter 2

Speech-in-Character in the Rhetorical Handbooks

Pseudo-Cicero: Rhetorica ad Herennium1

Introduction
Though other Latin treatments of rhetoric preceded the composition of
Rhetorica ad Herennium, the latter rhetorical handbook represents the old-
est extant treatment of rhetoric in Latin, much less preserved in its entirety.2
Though ad Herennium was attributed to Cicero from before Jerome’s ca-
reer through the late fifteenth century, the general scholarly consensus is
that Ciceronian authorship is erroneous.3 Instead, it is best to attribute ad
Herennium to an unknown author.4 Concerning ad Herennium’s date of com-
position, internal evidence establishes a terminus post quem in 86 BCE and a
terminus ante quem c. 80 BCE.5 Thus, ad Herennium constitutes an anonymous
treatment of rhetorical theory from the early first century BCE . Ad Herennium
addresses the practice of attributing speech to other characters under two
headings—sermocinatio and conformatio.

1  For Rhetorica ad Herennium’s treatment of speech-in-character, see Harry Caplan’s Latin

Loeb text and translation. [Cicero], Rhetorica ad Herennium (trans. Harry Caplan; LCL 403;
Cambridge: Harvard University Press, 1954, reprinted 2004), 366–69, 395–401. References to
Ad Herennium will follow the formula of [book number.section number]. I will not indicate
the alternative Roman numeral section numbers in my analysis.
Whenever appropriate, additional evidence from other rhetorical theorists who only
comment on speech-in-character briefly (for example, Cicero, Pseudo-Demetrius) are pro-
vided in the footnotes.
2  Caplan, ed., ad Herennium, vii; Kennedy, New History, 121.
3  Cicero never mentions ad Herennium, many elements in ad Herennium are at odds with
Cicero’s De Inventione, and Quintilian does not seem to be acquainted with a work by Cicero
of this type. Caplan, ad Herennium, vii–ix; Kennedy, New History, 121–22.
4  Caplan, ad Herennium, ix, xiv. See also his argument problematizing the identification of
Cornificius as the author. Ibid., ix–xiv. Cf. Kennedy, New History, 121.
5  Establishing the terminus post quem, the most recent historical references in ad Herennium
are the death of Sulpicius (88 BCE) and Marius’ seventh consulship (86 BCE); for the rough
terminus ante quem, the political interests shown in the treatise and the lack of conditions
produced under Sulla suggest a date circa 80 BCE. Ad Herennium, 1.25; 4.68; cf. 4.47. Caplan,
ad Herennium, vii, xxvi; Kennedy, New History, 122.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_004

20 chapter 2

Analysis: Sermocinatio
As one step in the task of embellishment (expolitio; 4.54), ad Herennium’s treat-
ment of speech-in-character preempts its longer treatment with a shorter dis-
cussion of sermocinatio, which it defines as follows:

Sermocinatio est … in qua constituetur alicuius personae oratio adcom-

modata ad dignitatem.

[Sermocinatio] … consists in putting in the mouth of some person

[personae] language [oratio] in keeping [adcommodata] with his charac-
ter [dignitatem]. (4.55 [Caplan, LCL])

Ad Herennium then provides a concrete example of sermocinatio, in which

each aspect of the definition is represented. In this example, a wise person
(i.e., “some person”) is characterized as one who prioritizes the welfare of the
community over his or her own self and, when in danger, is willing to suffer if
it protects the best interests of the state (i.e., “character”). Such a wise person
would say, “Not for self alone was I born, but also, and much more, for the
fatherland. Above all, let me spend my life, which I owe to fate, for the salva-
tion of my country” (i.e., language attributed in keeping with one’s character)
(4.55 [Caplan, LCL]). In this way, ad Herennium both defines and demonstrates
sermocinatio as one way rhetorically to embellish an idea.
As suggested in its shorter discussion (4.55), after addressing some addi-
tional figures, ad Herennium promptly returns to its treatment of sermocina-
tio (4.65). Again, ad Herennium begins by defining sermocinatio in a way that
wholly coheres with the previous definition:

Sermocinatio est cum alicui personae sermo adtribuitur et is exponitur

cum ratione dignitatis.

[Sermocinatio] consists in assigning [adtribuitur] to some person

[personae] language [sermo] which as set forth conforms [ratione] with
his (or her) character [dignitatis]. (4.65 [Caplan, LCL])

As before, ad Herennium follows this definition with an example. In this narra-

tival example, three “persons” to whom language is assigned appear in conver-
sation with one another: an armed enemy soldier, a wife, and a husband. When
the enemy soldier breaks into the house and threatens the husband, despite
the wife’s cries for mercy and pleas for her husband to submit, the husband
Speech-in-Character in the Rhetorical Handbooks 21

proudly criticizes the enemy. After repeating his threats, and as the husband
“began to say something or other, worthy, I am sure, of his manliness” (4.65
[Caplan, LCL]), however, the enemy killed him. Thus, as the husband’s forth-
coming words were to be “worthy … of his manliness,” ad Herennium establish-
es the requirement for attributed speech to be in accord with one’s character
both in its definitions and in the example itself. Ad Herennium repeats this
requirement a third time, concluding the example with the significant autho-
rial commentary that

Puto in hoc exemplo datos esse uni cuique sermones ad dignitatem

adcommodatos; id quod oportet in hoc genere conservare.

I think that in this example the language [sermones] assigned to each

person was appropriate [adcommodatos] to his (or her) character
[dignitatem]—a precaution necessary to maintain in Dialogue [genere].
(4.65 [Caplan, LCL])

Once more, the definition coheres with the example on all accounts.
Following this narratival example, ad Herennium notes one final aspect of
sermocinatio. Namely, sermocinatio does not have to depict an actual dialogue;
one can also create hypothetical dialogues (sermocinationes consequentes;
4.65). In sermocinationes consequentes, the scripted dialogue does not actu-
ally occur, but it is imaginary or hypothetical. In this hypothetical class of
sermocinatio, the dialogue is set forth as what some person(s) might hypotheti-
cally say given a set of circumstances, such as “what do we think those people
will say if you have passed this judgment” (4.65 [Caplan, LCL])? Additionally,
though the previous exemplary depictions of sermocinatio employed specific
persons (i.e., the “wise man” in 4.55), the example ad Herennium provides for
sermocinationes consequentes suggests that speakers may be unspecified as
well (4.65). Accordingly, the person(s) to whom speech is attributed can be
general and unidentified, such as “those people” or “every one” (4.65).
Absent from ad Herennium’s treatment, however, are any overt formal re-
quirements or suggestions for the proper structure or composition of sermoci-
natio. Structural cues, however, may be deduced implicitly, as two of the three
examples adhere to a similar sequence. The examples of the wise man and of
the enemy soldier, the wife, and the husband both adhere to a similar struc-
ture; each example (1) begins with an identification of the speaker, (2) pres-
ents a characterization of that speaker, and (3) concludes with the attributed
speech. The example of unspecified persons in sermocinationes consequentes,
22 chapter 2

however, is unfinished, as it omits the characterization and the actual words

attributed to the speaker(s). If ad Herennium completed the example, presum-
ably it would follow a similar pattern as the others.
Thus, in ad Herennium’s treatment of sermocinatio, several conventions for
how to use or compose a sermocinatio arise:

(1) the speaker or author shapes and assigns speech;

(2) the words can be attributed to:
a. a specific person or group of people,
b. a specific type of person, or
c. an unspecified person;
(3) the attributed speech can be real or imaginary;
(4) the language must agree with the character of the person in whose mouth
it is scripted; and
(5) there is an implicit pattern of identification of the speaker → character-
ization of the speaker → sermocinatio / attributed speech.

Analysis: Conformatio
Following its treatment of sermocinatio, ad Herennium engages speech-in-
character from a second perspective, namely, conformatio. Following the
same pattern of definition and example, ad Herennium defines conformatio
as follows:

Conformatio est cum aliqua quae non adest persona confingitur quasi
adsit, aut cum res muta aut informis fit eloquens, et forma ei et oratio
adtribuitur ad dignitatem adcommodata aut actio quaedam.

[Conformatio] consists in representing an absent person [non adest per-

sona] as present [adsit], or in making a mute thing or one lacking form
articulate, and attributing to it a definite form [forma] and a language
[oratio] or a certain behavior [actio] appropriate [adcommodata] to its
character [dignitatem]. (4.66 [Caplan, LCL])

At the end of the section, ad Herennium further notes that

Haec conformatio licet in plures res, in mutas atque inanimas transferatur.

[Conformatio] may be applied to a variety of things, mute and inanimate.

(4.66 [Caplan, LCL])
Speech-in-Character in the Rhetorical Handbooks 23

Two examples depict the use of conformatio. First, to a mute and inanimate
city, ad Herennium attributes speech that agrees with the character of that city.
Namely, the invincible city6 (i.e., a “mute and inanimate” thing), characterized
by numerous trophies, unconditional triumphs, opulence, and a longstanding
inability to be conquered (i.e., “character”), is cast as petitioning its tumultu-
ous citizens, “Do you now suffer to be trod upon and trampled underfoot by
worthless weaklings?” (i.e., words attributed in keeping with its character; 4.66
[Caplan, LCL]). Second, ad Herennium scripts the dead Lucius Brutus as re-
turning to life, appearing, and chastising the people (4.66). Here, Lucius Brutus
(i.e., “an absent person”), who defeated kings, created liberty, and put his own
life at risk to free the fatherland (i.e., “character”), rebukes those who do the
exact opposite by bringing in tyrants, failing to preserve liberty, and abandon-
ing freedom (i.e., “language appropriate to his character;” 4.66). In both ex-
amples, the language assigned is appropriate to the entity in whose mouth that
language is scripted. Unfortunately, ad Herennium neither provides an example
of conformatio applied to a thing “lacking form” (that is, an abstract idea) such
as Wisdom, Fate, or Love,7 or applying a behavior (actio) in lieu of speech.8
Again, ad Herennium offers no explicit advice for how to structure a confor-
matio, but the examples follow the same pattern as the first two sermocinatio
examples. Both examples of conformatio (1) identify a subject, (2) characterize
that subject, and (3) conclude by attributing speech to that subject.
Thus, ad Herennium offers four conceptual conventions for the implemen-
tation of conformatio. Namely, conformatio:

(1) attributes language (or behavior) to a person or thing;

(2) the subject to which one attributes speech can be:
a. an absent person by imagining them as present,
b. the dead,
c. a mute and inanimate thing, or
d. an abstract idea;

6  Rome, of course.
7  Quintilian’s Institutio Oratoria offers Vergil’s creation of Rumor, Prodicus’ creation of
Pleasure and Virtue, and Ennius’ creation of Death and Life, and the Progymnasmata of
[Hermogenes] and Apthonius offer Menander’s creation of the character Elenchos (refuta-
tion) as examples of speech-in-character applied to abstract ideas or things. Quintilian, Inst.,
9.2.36; Hermogenes, Prog., 20.10; Aphthonius, Prog., 34.15.
8  Since conformatio can apply behavior to a character, it may seem odd to use the term speech-
in-character. Because, in the larger picture of this project, I am only interested in attributions
of speech to imagined speakers, the technical term “speech-in-character” remains quite use-
ful and none the worse for wear.
24 chapter 2

(3) requires the attributed language or behavior to be appropriate to the

character of that person or thing; and
(4) follows an implicit pattern of identification of the speaker → character-
ization of the speaker → conformatio / attributed speech.

Synthesis: Speech-in-Character
From the above analyses, though sermocinatio and conformatio overlap to a
considerable degree in ad Herennium’s presentation of speech-in-character,
three differences emerge. First, both sermocinatio and conformatio involve
the attribution of language, but only conformatio allows the attribution of be-
havior. Second, both sermocinatio and conformatio can attribute speech to a
person, but conformatio further stipulates that this person, though absent, is
imagined as if he or she were present. Third, only conformatio envisions attrib-
uting speech or behavior to non-person, inanimate, mute, or abstract things.
Said otherwise, every element of sermocinatio can appear in conformatio, but
conformatio can include elements that lie outside the scope of sermocinatio. As
such, the elements of sermocinatio comprise one tool that can be employed in
the broader and more complex practice of conformatio. This observation cor-
roborates with the general trend for treatments of rhetorical theory to progress
from simpler to more complex skills;9 ad Herennium addresses sermocinatio
first because it amounts to one piece that speakers and writers can use to as-
semble the more complex conformatio puzzle.
Being able to recognize the differences between sermocinatio and conforma-
tio allows one to see that one may use sermocinatio by itself or conformatio by
itself, but one can never use both simultaneously, even though elements from
sermocinatio can contribute to conformatio. For instance, attributing speech to
persons who are present in a narrative qualifies as sermocinatio but not con-
formatio, since the latter requires persons to whom speech is attributed to be
absent. Additionally, speech attributed to an inanimate object is by definition
conformatio but not sermocinatio, since sermocinatio does not involve the at-
tribution of speech to inanimate objects. So, it is never possible for an assigned
speech to be both sermocinatio and conformatio in ad Herennium’s terms.
Nevertheless, notice that sermocinatio and conformatio accomplish the same
goal in ad Herennium’s treatment (i.e., attribute speech to another character);
the sole distinction ad Herennium makes regards the identity of the subject(s)

9  As Theon writes, “Easier exercises come before more difficult exercises [πρότερα γὰρ τὰ ῥᾶστα
τῶν δυσχερεστέρων].” Prog. 65.10–11.
Speech-in-Character in the Rhetorical Handbooks 25

in whose mouth words are scripted (i.e., a person, an inanimate object, or an

abstract idea).
Structurally speaking, though ad Herennium does not explicitly require
any formal structural patterns, four of its five examples adhere to a general
tripartite pattern of (1) identification, (2) characterization, and (3) attribution
of speech. The sole exception is the unfinished example of sermocinationes
consequentes.
One final note about the presentation of speech-in-character is appropriate.
When discussing the proper use of the voice, ad Herennium writes:

Si qua inciderint in narrationem dicta, rogata, responsa, si quae admira-

tiones de quibus nos narrabimus, diligenter animum advertemus ut om-
nium personarum sensus atque animos voce exprimamus.

If in the Statement of Facts there occur any declarations, demands, re-

plies, or exclamations of astonishment concerning the facts we are nar-
rating, we shall give careful attention to expressing with the voice the
feelings and thoughts of each personage. (3.14 [Caplan, LCL])

Though ad Herennium does not identify these occurrences specifically as

sermocinatio or conformatio, these concepts seem to be in mind. That is, the
supposed interjections do not physically come from the mouth of an au-
dience member. Rather, when introducing “outside” interjections into the
main speech, ad Herennium specifically requires the main speaker to tend
carefully and accurately to the character of each person presumably speak-
ing, and he or she achieves this through appropriate inflections of the voice.
Thus, ad Herennium not only assumes that speeches-in-character will be
evident to the reader or speaker, but voice inflection provides one addition-
al means by which to inform the audience that there has been a change in
speakers.
To synthesize the treatment of sermocinatio and conformatio as speech-in-
character, ad Herennium espouses the following conventions:10

10  Given that ad Herennium does not envision the simultaneous use of sermocinatio and
conformatio, how is a synthesis of these two mutually exclusive practices justified? First,
the basic element of each practice is identical—the attribution of appropriate speech to
an imagined speaker. Second, Quintilian will note how certain writers have divided vari-
ous aspects within speech-in-character into limited categories much like ad Herennium
has done (Inst. 9.2.31). Third, it is clear from Quintilian (Inst. 9.2.32) and [Hermogenes]
26 chapter 2

(1) a speaker or author attributes speech to another subject;

(2) the subject to which one applies speech can be a person, an inanimate
thing, or an abstract concept,
a. if the subject is a person,
i. it can either be a specific person, a specific type of person, or an
unspecified person,
ii. the person can be someone living or dead,
iii. if absent, the person is envisioned as present;
b. if the subject is mute and inanimate or abstract, it is depicted as pos-
sessing the ability to speak or act;
(3) the attributed speech must match the character of the subject to which it
is applied;
(4) there is an implicit pattern which progresses from identification of the
speaker → characterization of the speaker → attributed speech; and
(5) during performance, a speaker should inflect the voice only moderately
to capture the thoughts and feelings of an imaginary speaker.

Though ad Herennium employs different terms than we find in some other

writers who address speech-in-character,11 it will become clear from the fol-
lowing discussions that all of these treatments have the same concept in mind.
Indeed, Quintilian explicitly makes this point by identifying sermocinatio as
προσωποποιία (Inst. 9.2.31).

Quintilian: Institutio Oratoria12

Introduction
Appointed by Vespasian in 71 CE, Quintilian taught rhetoric in his state-
sponsored office until he retired circa 91–92 CE . Soon thereafter, Quintilian
began working on his rhetorical treatise Institutio Oratoria, which he completed

(Prog. 20.7–18) that the broader picture of speech-in-character can include both of the
aspects of what ad Herennium respectively calls sermocinatio and conformatio.
11  On the one hand, this is due in large part to the fact that ad Herennium is composed in
Latin, while most of the other texts are composed in Greek. On the other hand, at least
one important Latin writer, Quintilian, is familiar with and prioritizes the Greek termi-
nology (Inst., 9.2.29, 31, etc.).
12  For the Latin text of Quintilian’s treatment of speech-in-character, see Donald A. Russell’s
Latin Loeb edition and translation. Quintilian, The Orator’s Education (Institutio Oratoria):
Books 9–10 (ed. and trans. Donald A. Russell; LCL 127; Cambridge: Harvard University
Press, 2001), 51–55.
Speech-in-Character in the Rhetorical Handbooks 27

in 95 or 96 CE .13 In Institutio Oratoria, Quintilian aimed to account for the life-

long education of the orator, beginning with childhood, moving through adult-
hood, and idealistically arriving at a full, mature, and perfect orator.14 As such,
with the transition to Quintilian, this investigation jumps forward between
one hundred seventy-five and one hundred eighty years to examine a treat-
ment of rhetorical speech-in-character at the end of the first century CE .

Analysis: Fictiones Personarum / Προσωποποιία

As part of his treatment of figures of thought (Inst. 9.2), Quintilian addresses
speech-in-character as one such rhetorical technique that “[forms] a depar-
ture from simple ways of making a statement” and “[varies] and [animates]
a speech to a remarkable degree” (9.2.1, 29 [Russell, LCL]). Though Quintilian
does not supply a formal definition of speech-in-character, he does discuss the
terminological differences involved in describing the technique, various func-
tions of the technique, and a litany of conventions for how to use the figure
successfully.
From the very beginning of Quintilian’s discussion, the terminological prob-
lems (introduced in the previous discussion of ad Herennium) involved in dis-
cussing speech-in-character come to the fore.15 First, Quintilian provides both
Latin and Greek technical terms for the rhetorical technique under consid-
eration—fictiones personarum and προσωποποιίαι (9.2.29)—neither of which
ad Herennium employs, which uses sermocinatio and conformatio.16 Second,
Quintilian, though writing in Latin, favors the Greek term προσωποποιία and
uses it instead of the Latin term throughout the rest of the passage (9.2.29,
31, 37). Third, Quintilian is aware that some writers use different terms to dis-
cuss similar concepts. Quintilian notes that some writers limit the scope of
προσωποποιία to instances in which the speaker or writer invents both the per-
son (corpora) speaking and the reported speech (verba; 9.2.31). Such writers
then define cases in which the characters involved in conversation are real
persons as διαλόγους or sermocinationem; in this type, because the character

13  Kennedy, New History, 177–80; Russell, ed., The Orator’s Education (Institutio Oratoria):
Books 1–2 (ed. and trans. Donald A. Russell; LCL 124; Cambridge: Harvard University Press,
2001), 1–3.
14  Kennedy, New History, 181–82.
15  For a list of the various terms used for speech-in-character by ancient rhetoricians, see my
discussion above and/or Heusch, “Die Ethopoiie,” 13–14.
16  For instances in which Quintilian records the Greek προσωποποιία(ι), Russell switches
without explanation between the use of the English transliteration, prosopopoiia (9.2.29),
and the Latin loan word, prosopopoeia (9.2.31, 37; 1.8.3). For consistency, I have adjusted
each occurrence of προσωποποιία to “prosopopoiia” throughout the translations.
28 chapter 2

is a real person, the speaker or writer must only create the imagined speech
(9.2.31). This reiterates that some writers not only use various terms to discuss
a given figure, but in doing so they make terminological distinctions within the
practice of speech-in-character based on which elements a speaker or author
must invent in the process of composing an imagined conversation.
Quintilian himself, however, does not follow this tradition of dividing the
various aspects of speech-in-character; he opts to “follow the now established
usage in calling them both by the same name,” προσωποποιία (9.2.32 [Russell,
LCL]). Thus, for Quintilian, whether the speaker or writer simply invents
the imagined speech or more elaborately creates both the speech and the
character speaking, the same term applies.17 What is more, Quintilian believes
that the practice of calling both techniques by a single term has become largely
established by the end of the first century CE (9.2.32).
Concerning the functions of προσωποποιία, in addition to creating variety
and liveliness in a speech (9.2.29),18 Quintilian suggests that this figure is
particularly useful for three purposes (9.2.30). Quintilian writes:

17  Quintilian, however, throws something of a wrench in his terminological simplicity,

as a few sections later he introduces what he calls ἠθοποιία (9.2.58). Quintilian writes,
imitatio morum alienorum, quae ἠθοποιία vel, ut alli malunt, μίμησις dicitur, iam inter
leniores adfectus numerari potest: est enim posita fere in eludendo. Sed versatur et in factis et
in dictis … That is, Quintilian identifies the representation (imitatio) of the character
(morum) of another person through attributed speech (dictis) or actions (factis) as
ἠθοποιία or μίμησις. The hair Quintilian must split in order to maintain a distinction be-
tween these two terms is incredibly thin, as the only differences between Quintilian’s
presentation of προσωποποιία and ἠθοποιία seem to be that ἠθοποιία addresses gentler
“emotions” (9.2.58 [Russell, LCL]) and can include the attribution of behavior. Otherwise,
attributing words that represent the character of another speaker coheres seamlessly
with and adds nothing to Quintilian’s treatment of speech-in-character (9.2.30–37).
Though Lausberg fails to discuss the nigh contradiction in Quintilian’s treatment, see his
discussion in Lausberg, Handbook, §824.
18  See also Inst. 6.1–5, where Quintilian uses prosopopoiiai as a way to enliven and diversify a
speech in order not to be off-putting due to a straightforward repetition of facts. Similarly,
in their treatments of style, [Demetrius] and Cicero note the usefulness of προσωποποιία
to color a speech. [Demetrius] writes that προσωποποιία is a “figure of thought producing
forcefulness” (σχῆμα διανοίας πρὸς δεινότητα). For, the speech “is made to appear much
more lively (πολὺ γὰρ ἐνεργέστερα) and forceful (δεινότερα) by the characterizations
(προσώπων); rather, it actually becomes a drama (δράματα).” [Demetrius], On Style, 265–
266. Cicero writes that “impersonation of people (personarum ficta)” is “an extremely
brilliant method of amplification” (De or. 3.205 [Rackam, LCL]). Similarly, Cicero
writes in de Inventione that a speaker can produce variety by changing one’s method of
presentation; “At times you can sum up in your own person … but at other times you can
Speech-in-Character in the Rhetorical Handbooks 29

His et adversariorum cogitationes velut secum loquentium protrahimus

(qui tamen ita demum a fide non abhorrent si ea locutos finxerimus quae
cogitasse eos non sit absurdum), et nostros cum aliis sermones et alio-
rum inter se credibiliter introducimus, et suadendo, obiurgando, queren-
do, laudando, miserando personas idoneas damus.

We use them (1) to display the inner thoughts of our opponents as though
they were talking to themselves (but they are credible only if we imagine
them saying what it is not absurd for them to have thought [cogitasse
eos non sit absurdum]!), (2) to introduce conversations [sermones] be-
tween ourselves and others, or of others among themselves, in a cred-
ible manner [credibiliter], and (3) to provide appropriate characters
[personas] for words of advice, reproach, complaint, praise, or pity. (9.2.30
[Russell, LCL])

Regarding Quintilian’s first point, προσωποποιία is an effective way to disclose

the thoughts or views of one’s opponent(s) by imagining an opponent talking
to him or herself. In order for this feature of προσωποποιία to be effective,
however, the speaker or writer must script words for the opponent that are
credible for the imagined opponent to have spoken or thought. Said otherwise,
the words assigned to the opponent must accurately match the character of
the opponent in whose mouth they are placed. Second, προσωποποιία can
introduce conversations into a rhetorical context, either between oneself and
others, or simply between other characters.19 In order to be persuasive, these

bring on the stage (inducere) some person or thing (personaram aut rem aliquam) and let
this actor sum up the whole argument” (1.99 [Hubbell, LCL]).
19  On this point, I differ from Stowers in two ways. First, Stowers seems to read Quintilian’s
comment that προσωποποιία can “introduce (introducimus) conversations” as a reference
to something of a formal “introduction” or “introductory passage” preceding a subsequent
conversation, such as may occur in an apostrophe (especially in Rom 2:1–5, 17–29). Stowers,
Rereading, 100–2, 144–49; idem., “Apostrophe,” 358; cf. R. Dean Anderson Jr., Ancient
Rhetorical Theory and Paul (rev.; Contributions to Biblical Exegesis and Theology 18;
Leuven: Peeters, 1999), 201–3, cf. 203n.34. For Stowers, this means that the introduction
is itself part of the προσωποποιία. Stowers is followed by Bates, Hermeneutics, 196–97,
196n.52–58. This, however, is not the only way to read Quintilian. Quintilian does not
mean that προσωποποιία provides the formal introduction of an imaginary speaker for
an upcoming prosopopoetic speech; Quintilian intends that the use of προσωποποιία
submits into, includes, brings in, puts forth, appropriates, delivers or otherwise inserts a
conversation into a specific rhetorical context. What Quintilian suggests when he writes
that προσωποποιία can “introduce conversations” has nothing to do with apostrophe or
30 chapter 2

conversations must also be credible, such that the imagined characters could
reasonably have said the words scripted in their respective voices. Third,
προσωποποιία can serve to “provide appropriate characters for words of advice,
reproach, complaint, praise, or pity” (9.2.30 [Russell, LCL]). Essentially, through
προσωποποιία, a speaker or writer may introduce characters possessing specific
character traits in order to reveal and comment on some element pertinent
to any given rhetorical situation. In this way, a speaker or writer selects
particularly relevant characters in order to advance the plot or argument in a
specifically measured way.20
To begin discussing the conventions for composing a speech-in-character,
Quintilian notes that attributed speech must be “appropriate” to the imagined

the composing of a formal introductory passage, but with the presentation of the actual
prosopopoetic conversation for one’s audience, however it may be comprised. Such
a reading of Quintilian is defensible from the fact that Quintilian nowhere else in the
pericopae on speech-in-character discusses a formal introduction to a prosopopoetic
speech (much less an introduction as προσωποποιία), and, furthermore, from the fact
that Quintilian’s sole obsession with προσωποποιία is the attribution of words to an
imaginary speaker in one form or another. Second, Stowers later writes that the introduc-
tion of conversations “obviously involves addressing [a] person who is imagined to be
present and is, therefore, akin to apostrophe,” in which the actual speaker “participates
by addressing an imagined person.” Stowers, “Apostrophe,” 361. Again, Stowers misses the
mark. On the one hand, apostrophe only involves the actual speaker turning to speak
to an imagined person but not engaging in any back-and-forth conversation. On the
other hand, because speech-in-character attributes speech to another speaker, the only
portions of an imaginary conversation that qualify as speech-in-character are the lines
spoken in the voice of one’s imaginary dialogue partner. Thus, for Quintilian (and all of
the rhetoricians examined here), προσωποποιία or speech-in-character technically has to
do with the attribution and presentation of speech in the voice of another person and not
with the preceding introduction of the speech, though prosopopoetic speeches usually
do have some type of introduction. So, while I agree with Stowers that apostrophe can
work alongside speech-in-character to “introduce” and begin to characterize an imagined
speaker (as in Rom 2), by definition apostrophe is distinct from προσωποποιία. In fact,
Stowers himself begins to equivocate at times, noting that apostrophe “makes present
an imaginary person,” but speech-in-character “has an imaginary person speaking with
someone.” Stowers even submits, “Little is at stake in deciding if apostrophe to imaginary
persons might be called προσωποποιία.” Ibid.
20  In Book 3.8.54, Quintilian provides an example of Cicero setting up characters for par-
ticular rhetorical functions. Namely, “Cicero in the Pro Caelio makes Appius Caecus and
Clodius, her brother, address Clodia, the one to rebuke her vices, the other to encourage
them” ([Russell, LCL]).
Speech-in-Character in the Rhetorical Handbooks 31

speaker’s character in order to be most effective.21 Illustrating this need for ap-
propriate words, Quintilian writes elsewhere that because they each have a dif-
ferent character, “Caesar, Cicero, and Cato will all have to be assigned different
ways of giving the same advice (namque idem illud aliter Caesar, aliter Cicero,
aliter Cato suadere debebit)” (Inst. 3.8.49 [Russell, LCL]). Indeed, to ensure that
one attributes words that are appropriate to a given character, the speaker or
writer must not only consider the character of the imagined speaker, but also
his or her “fortune, position, and career,” as well as the subject about which the
speech is being made (3.8.50–51 [Russell, LCL]).
Additionally, Quintilian allows speeches to be scripted for several viable
candidates.22 One may script speech for a human, but one may also assign
words to the gods, the dead,23 or inanimate cities or nations (9.2.31). Quintilian
notes, however, the delicacy required to employ these latter subjects effectively,
since doing so “transcend[s] the bounds of nature” (9.2.32 [Russell, LCL]). To
model acceptable attributions of speech to mute and inanimate characters,
Quintilian includes two examples from Cicero’s first oration against Catiline.24
The first example imagines if Cicero’s “country, which is dearer to [him]
than [his] life, if all Italy, if the whole commonwealth, were to say to [him],
‘Marcus Tullius, what are you doing?’” (9.2.32 [Russell, LCL]; Cicero, Cat. 1.27).
Quintilian’s second example scripts the country pleading with Catiline and
“somehow, without uttering a word, [crying], ‘For some years past, no crime
has been committed except by your doing’” (Quintilian, Inst. 9.2.32 [Russell,
LCL]; Cicero, Cat. 1.17–18). By imagining the speech as hypothetical (“if”)
and highlighting the collective (“all”, “whole”) quality of the country in order
to align it with speaking individuals, and by noting how the mute country
paradoxically cries “without uttering a word” (Quintilian, Inst. 9.2.32 [Russell,
LCL]), respectively, Quintilian soothes the unnaturalness inherent in assigning
words to a character that cannot naturally speak.
Beyond inanimate and mute objects, it is also possible to personify and
imagine abstract concepts in dialogue. Thus, “Vergil invented Rumour, Prodicus
(according to Xenophon’s report) (invented) Pleasure and Virtue, and Ennius
(invented) Death and Life” (9.2.36 [Russell, LCL]). The orator can even imagine

21  See also Inst 6.1.25–27 and 11.1.39–41 on the effectiveness of appropriate speeches-in-
character in appealing to one’s emotions.
22  See also Inst. 11.1.41.
23  See also Quintilian, Inst. 4.1.28, where he allows that one may “raise the dead from the
grave (defunctos excitare)” (Russell, LCL).
24  Cicero, In Catilinam (trans. C. MacDonald; LCL 324; Cambridge: Harvard University Press,
1977.
32 chapter 2

an unspecified speaker (incerta persona), introduced with phrases such as,

“at this point someone says,” or “someone may say” (9.2.36 [Russell, LCL]).25
Finally, Quintilian allows that speech can occur in προσωποποιία without any
formal identification of the speaker, such as Vergil models in the Aeneid. 2.29.
Quintilian comments and quotes,

Est et iactus sine persona sermo:

‘hic Dolopum manus, hic saevus tendebat Achilles.’

One can even have speech without any person:

“Here camped the Dolopes, fierce Achilles here.” (9.2.37 [Russell,
LCL])

In this case, Quintilian writes that subtraction (detractionem)26 is combined

with προσωποποιία by omitting the speaker’s identity (9.2.37). This example,
however, is problematic on two accounts. First, though Quintilian suggests
that Vergil provides no indication of who is speaking, what is actually missing
is any overt indication that any other character is necessarily speaking at all,
as Vergil includes no verb of speech to signal the introduction of speech-in-
character. Second, allowing that this example is speech-in-character based on
Quintilian’s identification of it as such, in its original context, Vergil in fact
identifies the supposed speakers through Aeneas’ recollection of what “we”
(nos; Aen. 2.25) Trojans did when exploring the presumably abandoned Greek
camps. Though these words are not placed in the mouth of specific persons, the
imagined speakers must be the unspecified Trojans who are characteristically

25  On this point, Stowers is correct that the general “introductory phrases tell the reader
that someone is about to speak, but they do not identify who this speaker is. The reader
must infer an identity from the words of the imaginary speaker themselves in light of
clues from the preceding discourse.” Stowers, “Apostrophe,” 356. To Stowers, I simply add
the clarification that the preceding discourse usually in fact provides enough context to
allow an audience to infer sufficiently the character type or the identity of the imagined
speaker. Though it is not a perfect parallel because there is not an indication that any new
speaker has entered the scene, even if general and unspecific, see my following discussion
of Quintilian’s example from Vergil’s Aeneid 2.29 and the pertinent footnotes.
26  For Quintilian, “subtraction” (detractionem) is a category of figures that intentionally
omits certain elements in the interest of brevity, novelty, or rhetorical force, such as
what should be evident from the context (συνεκδοχῇ), conjunctions (asyndeton), or verbs
(ἐπεζευγμένον; Inst. 9.3.58–64).
Speech-in-Character in the Rhetorical Handbooks 33

overjoyed at the thought that the Greeks have sailed home (Aen. 2.25–28).27
Consequently, though Quintilian claims that speech-in-character can
occur without any distinct identification of the speaker, his example fails to
demonstrate the rule. Instead, the example suggests that speech-in-character
includes an identifcation of the speaker (i.e., unspecified Trojoans), but it
may omit a clear indication that another speaker is in fact speaking at a
particular time.28 As a result, though one may maintain despite Quintilian’s
failed example that speech-in-character can occur without any identification
of the speaker’s identity because that is what Quintilian states, one should be
twice as cautious when applying this category to concrete examples of speech-
in-character. Indeed, in such cases, one must look at the broader context for
whatever clues are present as a way, first, to recognize the mere presence of
speech-in-character and, second, to attempt to identify the supposed speaker.
Furthermore, προσωποποιία does not have to represent an “actual” speech,
but it can suggest imaginary or hypothetical speech (ficta oratio) as well

27  Coming to this reading of Vergil independently, see Will Timmins, “Romans 7 and Speech-
in-Character: A Critical Review of Stowers’ Hypothesis,” ZNW 107.1 (2016): 98–100. I agree
with Timmins at several points in his critique of Stowers, but additional evidence and
better argumentation (than Stowers) nevertheless convinces me that much of Rom 7
constitutes speech-in-character. Though this project does not aim to address Rom 7
completely, for more on this, see Part 3 below. See also Justin King, “Rhetorical Chain-
Link Construction and the Relationship between Romans 7:1–6 and 7:7–8:39: Additional
Evidence for Assessing the Argument of Romans 7–8 and the Identity of the Infamous
‘I,’” JSNT 39.3 (2017): 258–78; Bruce Longenecker, Rhetoric at the Boundaries: The Art and
Theology of New Testament Chain-Link Transitions (Waco: Baylor University Press, 2005),
88–93.
28  Without recognizing the problematic example, scholars routinely take Quintilian simply
at his word. See, for instance, Tobin, Paul’s Rhetoric, 227; Stowers, “Romans 7:7–25,” 187;
idem., “Apostrophe,” 256–57; idem., Rereading, 20; Campbell, Deliverance, 533; Lausberg,
Handbook, §824.1. The closest exception to this rule is R. Dean Anderson Jr., who recog-
nizes that, though Quintilian allows that a “speech may be inserted without indication
of the person,” in this example from Virgil “there is little room for misunderstanding that
προσωποποιία is being used,” and in both rhetorical and philosophical texts with speech-
in-character “a formal introduction and identification of the speaker is inevitably pres-
ent.” Ancient Rhetorical Theory, 202–3. See also Anderson, Glossary, 106. Anderson does
not, however, actually demonstrate how this works with respect to Quintilian’s example,
nor does he recognize that the example nevertheless excludes any specific indication that
a different speaker has even come on the scene. Indeed, this latter failure will prove quite
significant in analyzing the various arguments for speech-in-character in Rom 3 (Part 3).
34 chapter 2

(9.2.36).29 In such fictive speeches, the writer or speaker imagines what a

speaker would hypothetically say in a given situation, such as “someone may
say” (9.2.36 [Russell, LCL]).
Similarly, προσωποποιία is not limited to use in speeches, but it is also
applicable to written documents (9.2.34) and narratives (speciem narrandi;
9.2.37). To demonstrate the use of speech-in-character in a written document,
Quintilian relates the account of Asinius; when an opponent put forth a will
leaving his possessions to his benefactor, Asinius critiques the opponent’s
character by imitating it in a will imaginatively composed in the voice of the
opponent that excludes his mother from any inheritance (9.2.34–35). As a
narratival example, Quintilian quotes Livy’s History of Rome, who casts envoys
sent by Romulus to Rome’s neighboring states in order to procur marriage
rights declaring, “Cities, like other things, sprang from humble beginnings;
then, if helped by their own valour and by the gods, they made great wealth
and a great name for themselves” (Quintilian, Inst. 9.2.37 [Russell, LCL]; Livy,
Hist. 1.9.3).
Structurally speaking, however, Quintilian gives no formal requirements
concerning the arrangement of speech-in-character. Nevertheless, by
investigating the fuller context of the sources from which Quintilian draws
his examples, a consistent pattern becomes apparent. For instance, in in
Catilinam 1, Cicero first identifies the upcoming speaker as “your country”
(te patria), then characterizes the country as an authoritative and powerful
judge who hates and dreads Catiline and is fully convinced that Catiline was
seeking to destroy her, and finally attributes words to the country that indict
Catiline (1.17–18). In Cat. 1.27, Cicero again introduces this country (patriae)
which he has previously characterized as altogether opposed to Catiline
and imaginatively scripts the country asking Cicero whether he plans to
sit idly by or do something to impede Catiline. Similarly, in Quintilian’s
(misrepresented) example from Vergil’s Aeneid 2.1–29, Aeneas imagines the
experience of the Trojans (i.e., “speakers”), who, overjoyed by the thought that
the Greeks had apparently sailed home (i.e., “character”), went out to explore
the Greeks’ abandoned camps and marvelled the notion that Achilles had
once camped there (i.e., “credible speech”). In the case of Asinius, Quintilian
explains that an opponent (i.e., “speaker”) put forth a will that would leave
his estate to his benefactor (i.e., “character;” Inst. 9.2.35). Then, in an invented

29  See also Inst. 4.1.28, where Quintilian writes that, in an epilogue, one can “put imagi-
nary speeches into the mouths of [one’s] characters (fictam orationem induere personis)”
(Russell, LCL).
Speech-in-Character in the Rhetorical Handbooks 35

will, Asinius parodies the opponent’s character by carrying it out to its logi-
cal extremes, suggesting that the opponent would abandon his own mother
(i.e., “credible words;” 9.2.34). Finally, in History of Rome 1.9.1–3, Livy depicts
how in Rome’s early years she lacked a sufficient population of women to
populate the city. So, in need of women (i.e., “characterization”), the senate
and Romulus sent envoys (i.e., “speakers”) to neighboring states to petition for
the right of intermarriage (i.e, “credible words”). Consequently, every example
Quintilian lifts from his sources evinces three elements in its original context:
(1) an identification of the speaker, (2) a characterization of the speaker, and
(3) a prosopopoetic speech that is credble given the characterization of the
speaker.
Finally, in book 1 of Institutio Oratoria, Quintilian addresses the proper use
of one’s voice. In doing so, Quintilian briefly advises how one should perform
a speech-in-character so as to be most rhetorically effective. Quintilian writes:

Nec prosopopoeias, ut quibusdam placet, ad comicum morem pronuntiari

velim, esse tamen flexum quendam quo distinguantur ab iis in quibus
poeta persona sua utetur.

Nor do I think that prosopopoiiai, as some advise, should be pronounced

in the manner of the comic stage, though there should be some inflection
of the voice to distinguish them from passages in which the poet speaks
in his own person. (1.8.3 [Russell, LCL])

Here, Quintilian directs the reader to nuance one’s voice when speaking in the
voice of another character, as this signals for one’s auditors the presence of
different speakers. In his treatment of προσωποποιία (Inst. 9.2.29–37), however,
Quintilian never indicates any stipulations regarding how to demarcate or
identify the speakers when composing speech-in-character. Nevertheless,
Quintilian supposes that such shifts in speakers will be apparent enough,30
so that the only advice he is able to offer students is to “let [them] understand”
(intellegat) (1.8.2 [Russell, LCL]).31

30  Similarly, Stowers writes, “My point is that people with some education in Paul’s world
were trained to ‘read’ for—meaning to listen for—speech according to character, and
they composed their writings accordingly … [Quintilian] thus assumes that the reader
understood when to modulate his voice according to the ‘person speaking,’ whether or
not there was a formal introduction.” Stowers, “Apostrophe,” 354.
31  To “let him understand,” Russell inferentially adds, “his text.”
36 chapter 2

In summary, Quintilian teaches the concept of speech-in-character

according to the following conventions:

(1) a speaker or writer invents and assigns words to a character;

(2) the character to whom words are attributed can be real or invented,
including:
a. specified or unspecified humans,
b. gods,
c. the dead,
d. inanimate and mute objects,
e. abstract concepts, and
f. unidentified speakers;
(3) the assigned speech can represent actual or hypothetical conversations;
(4) the assigned words must be credible and not absurd for that character to
have thought or spoke, agreeing with the character, fortune, position, and
career of the imagined speaker, as well as the subject of the speech;
(5) speech-in-character may be effectively employed in speeches, narratives,
and written documents;
(6) an implicit pattern includes identification of the speaker → characteriza-
tion of the speaker → attributed prosopopoetic speech;
(7) speech-in-character may or may not explicitly indicate when and where
another speaker enters the conversation;
(8) when reading a speech-in-character for an audience, one should inflect
his or her voice in order to signal a change of speaker; and
(9) speech-in-character serves particular functions:
a. to disclose an opponent’s thoughts,
b. to introduce conversations, either
i. between oneself and others, or
ii. between others with themselves, and
c. to provide specific characters for various rhetorical purposes.

Summary: Ad Herennium and Quintilian on Speech-in-Character

This chapter has analyzed the presentation of speech-in-character in ad

Herennium and Quintilian, respectively. In doing so, though a relatively
harmonious understanding of speech-in-character has emerged, a few
differences between the two must be noted. First, ad Herennium and Quintilian
represent two distinct ways of classifying speech-in-character. As discussed
above, Quintilian uses the term προσωποποιία to cover the whole range of
Speech-in-Character in the Rhetorical Handbooks 37

speech-in-character.32 Quintilian is aware, however, that some writers make

categorical distinctions contingent on whether a speaker or writer must invent
the words and the person of the speaker (προσωποποιία), or simply the words
(sermocinatio). Interestingly, ad Herennium models just such a practice, as it
divides speech-in-character into the categories of sermocinatio, which only
invents the words, and conformatio, which invents both the words and the
character of the speaker. Second and third, ad Herennium does not comment
specifically on the function(s) of speech-in-character or on its use in narrative
or written documents, whereas Quintilian does. Fourth, only Quintilian seems
to allow that speech-in-character may be present even when it is not clearly
marked.
Nevertheless, ad Herennium and Quintilian assemble a remarkably united
front with respect to the core elements in the practice of speech-in-character.
The following conventions are present in both texts:

(1) a speaker or writer crafts speech and attributes it to an imaginary speaker;

(2) the character to whom speech is attributed can be:
a. a person, a type of person, or a group of people,
b. a dead person,
c. an unspecified person,
d. an inanimate object, or
e. an abstract concept;
(3) the attributed speech can be real or hypothetical;
(4) the attributed speech must align with the character of the supposed
speaker;
(5) one should inflect the voice when performing a speech-in-character; and
(6) the examples used to depict speech-in-character follow a common
pattern:
a. identification of a speaker,
b. characterization of the speaker, and
c. the delivery of an appropriate speech-in-character.

Though momentarily setting the rhetorical handbooks aside at this point,

I return to them in this section’s final assessment of speech-in-character.
Required next, however, is engagement with another type of source that
discusses the attribution of speech to other characters—the Progymnasmata.

32  Cf. the footnote on Quintilian’s terminological hair splitting, and Inst. 9.2.58.
 chapter 3

Speech-in-Character in the Progymnasmata

Theon: Προγυμνάσματα1,2

Introduction
Identifying the historical Theon—and thereby dating the Progymnasmata at-
tributed to him—is tricky. The tenth-century Suda records an entry on Aelius
Theon of Alexandria, who composed a work on progymnasmata (Θ 206). Is
this Aelius Theon the author of the Progymnasmata attributed to Theon, and,
if so, when did he live? Despite attempts to attribute this Progymnasmata to
another “Theon” in the fifth century CE,3 the scholarly consensus is that Aelius

1  For the Greek texts and my translations of Theon, Hermogenes, Aphthonius, and Nicolaus,
refer to Appendix A in my dissertation, “Speech-in-Character, Diatribe, and Romans 3:1–9:
Who’s Speaking When and Why It Matters” (Ph.D. diss, Baylor University, 2016), 415–35. In
this project, though the differences in Aphthonius and Nicolaus are discussed in the foot-
notes, they are excluded from comprehensive engagements due to their fourth- and fifth-
century compositions.
2  For the Greek text of Theon’s treatment of speech-in-character, see Michel Patillon and
Gioncarlo Bolognesi, eds., Aelieus Theon: Progymnasmata (Budé; Paris: Les Belles Lettres,
2002), 70–73. See also Leonardus Spengel, ed., Rhetores Graeci (vol. 2; Leipzig: Teubner, 1854),
115–18; and Butts, Theon, 444–64. All Greek translations are my own. For additional transla-
tions of Theon, see Kennedy, Progymnasmata, 47–49; Butts, Theon, 444–64.
 Scholars have traditionally cited references to Theon’s Progymnasmata by noting the page
and line numbers in Spengel’s edition. Patillon’s text, however, disrupts the line divisions of
Spengel’s text without providing an alternative reference system. Consequently, Spengel’s
reference system remains the most accessible, but Patillon’s text does not fit the mold. In
Appendix A in my dissertation, I have readjusted the line divisions of Patillon’s text to cohere
with Spengel’s divisions; the bracketed numbers refer to the text’s page and line numbers in
Spengel, and [P #] indicates the page number in Patillon.
3  Malcom Heath, “Theon and the History of the Progymnasmata,” GRBS 43 (2002/3), 129,
141–58. On the face of things, Heath puts forth a seemingly strong case. When one digs a
little deeper, however, it becomes clear that Heath’s presentation is not fair to the complete
body of evidence. Heath selects and chooses only the elements that best support his case.
To be sure, part of Heath’s selectivity is due to the fact that he cannot analyze all of Theon’s
Progymnasmata in a single essay. But it is precisely in some of the sections where Heath does
not spend much time that questions arise and problematize his argument. For example, with
respect to speech-in-character, there are remarkable similarities between Quintilian and
Theon that Heath does not discuss. Both Quintilian and Theon refer to the whole scope of
speech-in-character as προσωποποιία (Quintilian, Inst. 9.2.32; Theon, Prog. 115.12). If, as Heath

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_005

Speech-in-Character in the Progymnasmata 39

Theon of Alexandria indeed composed this Progymnasmata in the first cen-

tury CE .4 First, because Theon references Theodorus of Gadara and Dionysius
of Halicarnassus and thus establishes a terminus post quem, he could not have
written this Progymnasmata before the end of the first century BCE. Second,
Theon’s similarities with Quintilian on the classification of prosopopoiia per-
mit Theon a place squarely in the middle of the general consensus of the
mid- to late-first century CE, which later progymnasmatic writers uniformly
do not follow. Third, Theon’s placement of the chreia exercise first in his list

argues, Theon had known and used the text of [Hermogenes], which, like ad Herennium,
distinguishes between types of speech-in-character based on the subject in whose voice
speech is scripted, and if Theon was more characteristic of the later Progymnasmatic writers
who also distinguish between types of speech-in-character, should one not expect Theon
also to distinguish between the types of speech-in-character? On this score, however, Theon
disappoints. Indeed, Theon’s sole point of connection is with Quintilian in the first century
CE. What is more, Quintilian even remarks that such a presentation of speech-in-character
had become the general consensus in the first century (9.2.32). Thus, Theon’s classifica-
tion of speech-in-character is right at home in the first century CE but would be altogether
out of place among the fourth- and fifth-century Progymnasmatists. Additionally, it is only
Quintilian and Theon that expand so broadly with respect to the aspects of a given speaker’s
character and the rhetorical context that one must keep in mind when composing appropri-
ate words for that imagined speaker. None of the later Progymnasmatic writers go into such
detail as Quintilian and Theon, who specify that one must keep in mind the character, the
fortune, the social status, and the vocation of the speaker, as well as the subject about which a
speech is being composed (Quintilian, Inst. 3.8.50–51; Theon, Prog. 115.22–116.22). As it stands,
due to Theon’s similarities to Quintilian and differences from the later Progymnasmatic writ-
ers on the topic of speech-in-character, there are strong reasons to locate Theon in the first
century CE.
4  G.  Reichel, Questiones Progymnasmaticae (Ph.D. diss., Leipzig, 1909), 30, 115–27; Willy
Stegemann, “Theon (5),” PW 5A (1934), 2037–54, especially 2037–39; Butts, Theon, 1–5;
Ronald F. Hock and Edward N. O’Neil, The Chreia in Ancient Rhetoric: Volume 1: The
Progymnasmata (Society of Biblical Literature Texts and Translations Series 27; Graeco-
Roman Religion Series 9; Atlanta: Scholars Press, 1986), 63–64; Patillon and Bolognesi, Aelieus
Theon, viii–xvi; Anderson, Ancient Rhetorical Theory, 72–77; Stowers, “Romans 7:7–25,” 180;
Hock, “The Rhetoric of Romance,” 454. George Kennedy originally assumed a first-century
CE date for Theon, but he has since broadened his view to allow anytime between “the
Augustan period and the flowering of the Second Sophistic in the second century after
Christ.” George A. Kennedy, The Art of Rhetoric in the Roman World (Princeton: Princeton
University Press, 1972), 616; idem., Progymnasmata: Greek Textbooks of Prose Composition
and Rhetoric (Society of Biblical Literature Writings from the Greco-Roman World 10; Leiden:
Brill, 2003), 1. Nevertheless, Kennedy affirms, “It is the consensus of scholarly opinion that
[this Progymnasmata] is, in any event, the earliest surviving work on exercises in composi-
tion.” Ibid.
40 chapter 3

of progymnasmata,5 a feature unique among the Progymnasmata, parallels

Suetonius’ first-century list (On Grammarians and Rhetors 25.4).6 Fourth, the
cognomen “Aelius” fits well into the Alexandrian context surrounding the first
century, as it could have come to Theon or one of his ancestors from the Roman
prefect of Egypt, Aelius Gallus (26–24 BCE), or from Hadrian (Publius Aelius
Hadrianus Augustus) when he visited Alexandria in 130 CE .7 For these reasons,
I support the consensus dating of Theon early rather than late.

Analysis: Προσωποποιία
Theon treats προσωποποιία sixth in his collection of elementary exercises
(προγυμνάσματα) in Greek rhetoric and composition.8 Theon begins by
defining προσωποποιία as

Προσωποποιΐα ἐστὶ προσώπου παρεισαγωγὴ διατιθεμένου λόγους οἰκείους

ἑαυτῷ τε καὶ τοῖς ὑποκειμένοις πράγμασιν ἀναμφισβητήτως.

Prosopopoiia [προσωποποιία] is the introduction of a person [προσώπου]

who non-controversially [ἀναμφισβητήτως] sets forth words that are
appropriate [οἰκείους] both to the person himself and to the subjects
[πράγμασιν]9 being set forth. (115.12–14; P 70)

5  Butts has persuasively argued for the original order of Theon’s Progymnasmata, the first exer-
cise being chreia. Butts, Theon, 8–22.
6  Patillon and Bolognesi, Aelius Theon, xii–xvi.
7  Patillon and Bolognesi prefer to argue that the cognomen was given to one of Theon’s an-
cestors in the first century BCE and passed down to Theon. Ibid., xvi. Though Heath knows
Patillon and Bolognesi, he altogether ignores the former option. Heath, Theon, 142. If the cog-
nomen stems from Hadrian, two options allow Theon’s Progymnasmata to remain a method-
ologically sound source for the study of Paul: (1) it is entirely possible that Theon could have
composed this Progymnasmata some time before receiving the cognomen, which was only
later connected to the Progymnasmata in the Suda’s sources, or (2) though written a few de-
cades later, the contents depict and genuinely fit within the first-century rhetorical millieu,
as documented by Quintilian and Suetonius. For a scholar referencing Patillon and dating
Theon to the first half of the second century CE, see Heusch, “Die Ethopoiie,” 14, 14n.22.
8  See Butts, Theon, 8–22.
9  Here, πράγμασιν may be more exhaustive than the specific translation “subjects” suggests.
For Theon, πράγμασιν certainly includes the subject to be discussed, but it could also apply to
the circumstantial or contextual details set forth in which a speech is to take place. Thus, the
translation of πράγμασιν more generally as “things,” “holisitc circumstances,” or something
similar may be appropriate. See also Butts, Theon, 460n.3.
Speech-in-Character in the Progymnasmata 41

To supplement his definition, Theon provides several examples of contexts

for which a student would be expected to invent speech.10 Theon divides
his examples into two types, those with unspecified speakers and those with
specified speakers (ὡρισμένων … προσώπων; 16–17). For examples of προ-
σωποποιία with an unspecified speaker, Theon suggests as contexts “what
words a husband would say to his wife when he is about to depart” (14–15),
or “what a general would say to his soldiers when encountering dangers” (15–
16). In both examples, the speaker represents a type of person rather than a
specific husband or general. For προσωποποιίαι that imagine specific speakers,
however, Theon puts forth as exemplary contexts “what words Cyrus would
say while marching against the Massagetae” (17–18), or “what Datis would say
while conversing with the king after the battle of Marathon” (18–19). Thus,
when engaging in προσωποποιία, one may attribute suitable words either to
unspecified or specified persons.11

10  In the strictest sense, Theon does not provide any examples of προσωποποιία, since he
never actually records the words that would be attributed to a supposed speaker. The
first progymnasmatic writer to include an actual speech-in-character is Aphthonius,
concerning “What words Niobe would say while her children lie dead.” Prog. 35.15–36.20.
11  Gibson reads Theon’s (and Nicolaus’) treatment of speech-in-character quite uniquely.
Gibson argues, Theon and Nicolaus “seem to regard prosopopoiia as a tool for an author
or speaker’s self-presentation, rather than as a means of attributing speech to character’s
other than one’s self. So Theon suggests that the exercise is useful in speeches in which
we exhort, console, demand, dissuade, or seek forgiveness … and that it has a practical
application in letter writing.” Gibson, “Prosopopoeia,” 9–10. Gibson’s reading, however, is
untenable. First, Theon himself defines speech-in-character as the “introduction” or “en-
trance” (παρεισαγωγή) of a person to whom appropriate words are attributed. Obviously,
if Theon intended to present his own person through prosopopoiia, there would be no
need to introduce another speaker. Second, Theon’s specific examples prove that he has
a different speaker in mind than himself; Cyrus speaks, Datis speaks, and Herodotus—a
Greek—attributes speech to non-Greeks. Third, though Theon writes that prosopopoiia
is useful when “we” are trying to achieve certain rhetorical purposes (to exhort, dissuade,
console, and so forth), this in no way demands that there are contradictions in Theon’s
treatment. One should read Theon’s discussion on the functions of speech-in-character
in light of Theon’s own definition and use of speech-in-character. When reading Theon
in light of Theon, he must mean that a speaker or author attributes words to an addi-
tional speaker for a given rhetorical purpose. Though the speaker or author invariably
controls the imagined speaker and his or her speech for a particular rhetorical goal in
the speech or narrative as a whole, there is nevertheless speech attributed to another
speaker that coheres with the character of the imagined speaker. This understanding of
the usefulness of speech-in-character for any particular rhetorical goal is precisely one
aspect of what Gibson identifies as the broader result of progymnasmatic education in
42 chapter 3

Another word, however, is necessary about Theon’s specific example of Datis;

the example of Datis’ words with king Darius after Marathon seems thoroughly
hypothetical, as no record of such a speech seems to have existed in antiquity.12
At minimum, no such speech remains extant in any of the primary sources
Theon might have known.13 It appears that Theon employs this example in
order to guide his students to compose an imaginative speech based on what
they already know about Datis, his defeat in the battle of Marathon, and King
Darius, even though there is no record that such a speech actually occured
in antiquity. From this, it is clear that Theon does not limit προσωποποιία to
attributions of words actually spoken; he also allows the exercise to portray
imaginary or hypothetical speech.
Of primary importance to Theon in the creation of προσωποποιία, however,
is the requirement that the attributed speech be suitable to the rhetorical
species of the speech and its rhetorical situation. With resepct to rhetorical
species, Theon contends

Ὑπὸ δὲ τοῦτο τὸ γένος τῆς γυμνασίας πίπτει καὶ τὸ τῶν παρηγορικῶν λόγων
εἶδος, καὶ τὸ τῶν προτρεπτικῶν, καὶ τὸ τῶν ἐπιστολικῶν.

Under this genus [γένος] of exercises [γυμνασίας] fall the species [εἶδος]
of consolation [παρηγορικῶν], exhortation [προτρεπτικῶν], and epistolary
[ἐπιστολικῶν] speeches. (115.20–22; P 70)

Within the genus (γένος) of prosopopoetic exercises, therefore, Theon notes that
there are three distinct species (εἶδος) of speeches: consolatory (παρηγορικῶν),
exhortatory (προτρεπτικῶν), and epistolary (ἐπιστολικῶν; 20–22).14 To assist and

speech-in-character, and with which I agree entirely, that “students take away from it a
skill to be applied elsewhere.” Gibson, “Prosopopoeia,” 9.
12  Ctesias even holds that Datis died during the battle of Marathon. Ctesias, History of Persia
13.22, 25.
13  The following is a non-exhaustive but demonstrative list. Ctesias, History of Persia 13.22,
25; Herodotus, The Persian Wars, 6.111–120; Cornelius Nepos, Militiades 5.4; Diodorus
Siculus, Library of History, 10.27.1–3; 11.2.2; Plutarch, Aristides 5.1.1; Demosthenes, Against
Neaera 94; Plato, Laws 3.698.c.
14  Because Paul’s literature falls under the epistolary genre, one might wonder whether
Paul’s implementation of speech-in-character falls within Theon’s proposed category
of “epistolary” speech-in-character. The answer is no. Paul’s authentic letters attribute
speech-in-character to other imagined speakers (see Chapter 5). In epistolary speech-
in-character, the whole letter is itself a speech-in-character. An epistolary speech-in-
character is a letter written in the name of someone who did not in fact write it; epistolary
Speech-in-Character in the Progymnasmata 43

ensure that his students compose speeches within the parameters of a given
species and rhetorical context, Theon directs his students to understand and
hold in proper balance the personalities (το … πρόσωπον) of the speaker and
of the addressee (τὸ [πρόσωπον] πρὸς ὃν ὁ λόγος), the speaker’s age (ἡλικίαν),
the occasion (καιρόν) of the speech, the place (τόπον) in which the speech
takes place, the social status (τύχην) of the speaker, and the subject matter
(ὕλην) about which someone is going to make a speech (22–27). Indeed, Theon
continues,

Ἔπειτα δὲ ἤδη πειρᾶσθαι λόγους ἁρμόττοντας εἰπεῖν.

Only after making these considerations should one attempt to speak suit-
able [ἁρμόττοντος] words. (115.27–28; P 70)

For, a single speech is not appropriate to any and every context. Different
speeches are appropriate for different speakers, contexts, and subjects;
suitable speech is contingent on a particular speakers age, gender, social
status, vocation, character, and race, as well as on the place in, the occasion on,
the audience to, and the subject about which a particular speech is delivered
(115.28–116.22). Said otherwise, in order for προσωποποιία to be effective, it must
exhaustively account for the world in which the speech hypothetically takes
place.15
But Theon does not expect his students to work through the vast diversity of
persons and subjects entirely on their own (22–24). To help students navigate
through the complexity of persons and subjects, Theon outlines a number of
appropriate starting places (ἀφορμάς) for given subjects. For example, if the goal
is to exhort (προτρέποντες; 27–28) or request something (117.4–5), the student

speech-in-character applies to the practice of pseudepigraphy. Thus, if any of the letters

in the New Testament are deemed deutero-Pauline or otherwise pseudonymous, these
examples might fall under the category of epistolary speech-in-character. Non-Christian
examples of epistolary speech-in-character include but are by no means limited to the
letters of Alciphron, Aelian, and Philostratus. Harry O. Maier, Picturing Paul in Empire:
Imperial Image, Text and Persuasion in Colossians, Ephesians and the Pastoral Epistles
(London: Bloomsbury, 2013), 31–35; Patricia A. Rosenmeyer, Ancient Epistolary Fictions:
The Letter in Greek Literature (Cambridge: Cambridge University Press, 2001), 193–233,
255–338, esp. 259–62; and Carol Poster, “A Conversation Halved: Epistolary Theory in
Greco-Roman Antiquity,” in Letter Writing Manuals and Instruction from Antiquity to the
Present (Carol Poster and Linda C. Mitchell, eds.; Historical and Bibliographic Studies;
Columbia: University of South Carolina Press, 2007), 28, 43n.14.
15  Similarly, see Aristotle, Ars Rhetorica 3.7.1–7.
44 chapter 3

should script the speaker saying certain kinds of things, such as that what is
being exhorted or requested is “easy, good, and fitting; that it is beneficial, just,
reverent, pleasant” (116.28–32), and so forth (116.32–117.4). If the intention is
the opposite of exhortation, however—that is, to dissuade (ἀποτρέποντες)—
then the speaker should say precisely the opposite (117.5–6). For consolation
(παρηγορῶμεν; 6–24) or to request forgiveness (24–28), Theon suggests different
stock starting places. In every case, however, “One must argue from whichever
common-places are admissible, for all common-places are not suitable for all
prosopopoiiai which are under the same species” (28–30).
Finally, though he leaves this thought undeveloped, Theon notes that προ-
σωποποιία is especially helpful for depictions of character types (ἠθῶν) and
emotions (παθῶν; 117.30–32).
Structurally, Theon gives no advice with respect to the formal arrangement
of προσωποποιία. What is more, Theon’s “examples” leave much to be desired
in terms of creating some kind of consistent pattern. First, none of Theon’s “ex-
amples” present the supposed speech. Second, Theon’s example of προσωπο-
ποιία using Datis as a specific speaker is a completely hypothetical scenario, so
there is no concrete example to examine in his source text(s).16 Third, the only
example Theon provides for which an extant narrative exists in the primary
literature is that of Cyrus marching against the Massagetae, narrated in
Herodotus’ Persian Wars 1.201–16. Unfortunately it is unclear for which context
in the narrative Theon intends for his students to compose a speech. Does Theon
envision his students composing an imaginary speech concerning what Cyrus
would have said to the Queen of the Massagetae, Tomyris, when attempting
to form a marriage alliance before any battle began, for which Herodotus only
includes indirect speech (1.205)? For Cyrus’s words to his councilors when
deliberating how to respond to Queen Tomyris’s requests, to which Herodotus
simply alludes (1.206)? For Cyrus’s reported rejection of Queen Tomyris’s
appeals, which Herodotus again places in indirect speech (1.208)? Might Theon
expect his students to compose what Cyrus said concerning his dream about
Darius, even though that bears little more relationship to the battle with the
Massagetae than as a retrospective foreshadowing (1.209–210)? Or does Theon
have in mind the whole narrative, with all of its various possibilities for speech?
All that is to say, precisely what context Theon has in mind with respect to
Cyrus’s march against the Massagetae is quite unclear. For any of the options
mentioned above, however, three elements would invariably be present in light
of the narrative as a whole: (1) an identification of Cyrus as the speaker (1.201,

16  Theon seems to rely on Herodotus’ Persian Wars 1.201–16 and 6.111–20 for his two exam-
ples of προσωποποιία with specific speakers.
Speech-in-Character in the Progymnasmata 45

205, 206, etc.), (2) a characterization of Cyrus and the Massagetae (1.201–205),
and (3) the attributed prosopopoetic speech. What is additionally clear from
Theon’s examples is that every context for which Theon imagines invented
speech assumes a clear indication that someone other than the writer gives
voice to the scripted words, whether that is a specific individual (i.e., Datis) or
an unspecified person (i.e., a husband).
Therefore, when outlining the various elements of Theon’s presentation of
προσωποποιία, the following conventions take shape:

(1) προσωποποιία involves crafting and attributing speech to a person;

(2) the person for whom speech is scripted may be a specific person or a
general type of person;
(3) the attributed speech may be actual or imaginary;
(4) the attributed speech must be appropriate for:
a. the imagined speaker’s character, age, race, social status, gender, and
vocation,
b. the addressee’s character,
c. the occasion and place a speech is delivered,
d. the subject engaged in the speech;
(5) the example of Cyrus allows for a pattern of identification of speaker →
characterization of that speaker → attributed speech-in-character;
(6) there is a clear indication that someone else is speaking;
(7) one may use stock starting places appropriate for given subjects;
(8) the three primary types of speeches composed under προσωποποιία
are:
a. consolation,
b. exhortation,
c. epistolary;
(9) προσωποποιία can emphasize the character or the emotion of the
speaker.

[Hermogenes]: Προγυμνάσματα17

Introduction
The progymnasmatic treatise attributed to the second-century rhetorician
Hermogenes has also been attributed to Libanius, and Priscian’s translation

17  For the Greek text of Hermogenes’ Προγυμνάσματα, see Hugo Rabe, ed., Hermogenis
Opera (Rhetores Graeci VI; Leipzig: Teubner, 1913), 20–22. All Greek translations are my
46 chapter 3

even attributes it to either author.18 When the diverse authorial attributions

are coupled with the fact that the compiler of the Hermogenic corpus includ-
ed in it the Progymnasmata of Aphthonius rather than the one attributed to
Hermogenes, it seems quite clear that Hermogenes’s authentic authorship is
doubtful.19 What is more, scholars date the treatise quite broadly, ranging from
the second to the fourth century CE. The sole historical marker serving as the
general terminus post quem is the text’s reference to Aelius Aristides, who lived
during the second century from 117–181 CE . Thus, the treatise could have been
composed at any point after Aristides’s rise to popularity, but scholars gener-
ally agree that it was composed before the Progymnasmata of Aphthonius and
Nicolaus.20 Given the post-Pauline date of this Progymnasmata, its inclusion
serves three goals: (1) to illustrate that certain diverse features of speech-in-
character witnessed in the first century BCE persist into later rhetorical theory;
(2) to provide a measure by which to compare Theon, and (3) to demonstrate
the ways in which the two later Progymnasmata of Aphthonius and Nicolaus
address speech-in-character in distinction from the first-century CE treat-
ments of Quintilian and Theon.

Analysis: ἠθοποιία, προσωποποιία, εἰδωλοποιία

Hermogenes’s treatment of speech-in-character comes ninth in its arrangement
of progymnasmata, and it begins by distinguishing and defining three types of
speeches-in-character. First, Hermogenes defines ἠθοποιία as follows:

Ἠθοποιία ἐστὶ μίμησις ἤθους ὑποκειμένου προσώπου.

Ethopoiia [ἠθοποιία] is an imitation [μίμησις] of the character [ἤθους] of

an imagined [ὑποκειμένου] person [προσώπου]. (20.7–8; S 15)

Furthermore, the immediately following example concerning which words

Andromache would say to Hector clarifies that such “imitation” has to do

own. For an additional translation, see Kennedy, Progymnasmata, 84–85. References to

Hermogenes will indicate the page and line number(s) of a given text in Rabe’s edition.
[S #] refers to the respective page number in Spengel’s edition.
18  Kennedy, Progymnasmata, 73; Heath, “Theon,” 150.
19  Kennedy, Progymnasmata, 73; Heath, “Theon,” 158–59. For an overview of the arguments
and the maintenance of Hermogenic authorship, see Hock and O’Neil, Progymnasmata,
158–60. Recognizing that Hermogenic authorship is doubtful, from here on I nevertheless
refer to the author of this treatise as “Hermogenes” for the sake of simplicity.
20  Kennedy, Progymnasmata, 73; Hock and O’Neil, 159; Heath, “Theon,” 159.
Speech-in-Character in the Progymnasmata 47

specifically with attributions of speech, since the dependent variable in the

example is limited to Andromache’s words (λόγους; 8–9).21 In fact, Hermogenes
affirms this reading in his explanation of the differences between ἠθοποιία and
προσωποποιία when he writes,

ἡ δὲ διαφορὰ δήλη· ἐκεῖ μὲν γὰρ ὄντος προσώπου λόγους πλάττομεν, ἐνταῦθα
δὲ οὐκ ὂν πρόσωπον πλάττομεν.

The difference is clear: for, there (ethopoiia), we invent [πλάττομεν]

speeches for a person that exists [ὄντος προσώπου], but here (proso-
popoiia), we invent a person that does not exist [οὐκ ὂν πρόσωπον].
(20.12–14; S 15)

Thus, for Hermogenes, ἠθοποιία is an imitation of the character of an imagined

speaker by assigning speech that accurately models the character of that imag-
ined speaker. Second, Hermogenes describes προσωποποιία as

προσωποποιία δέ, ὅταν πράγματι περιτιθῶμεν πρόσωπον.

Prosopopoiia [προσωποποιία], however, is when we apply a personality

[πρόσωπον] to a thing [πράγματι]. (20.9–10; S 15)

Additionally, as quoted above, in προσωποποιία “we invent a person that does

not exist (οὐκ ὂν πρόσωπον)” (14). Again, Hermogenes’s examples indicate that
speech is involved in προσωποποιία, as Elenchos speaks in Menander (frag. 545),
and the Sea specifically “makes speeches” (ποιεῖται τοὺς λόγους) in Aristides
(Hermogenes, Prog. 20.10–12).22 Consequently, προσωποποιία involves both
inventing a “person” and crafting speech for that “person.” Finally, Hermogenes
defines εἰδωλοποιία as follows:

εἰδωλοποιίαν δέ φασιν ἐκεῖνο, ὅταν τοῖς τεθνεῶσι λόγους περιάπτωμεν.

And, they say it is eidolopoiia [εἰδωλοποιίαν] whenever we attribute

speeches to the dead. (20.14–16; S 15)

21  Similarly, Stowers writes, “In all of these writers, ἦθος means using words to portray a per-
son’s character.” Stowers, “Rom 7:7–25,” 181, emphasis mine.
22  Aristides’s speech of the Sea to the Athenians seems to be no longer extant. Kennedy,
Progymnasmata, 84n.43.
48 chapter 3

As an example of scripting speech for a dead person (εἰδωλοποιία), Hermogenes

refers to Aristides, who assigns speech to dead people in his speech, Against
Plato on Behalf of the Four (In Defense of Oratory 319–343; Hermogenes, Prog.
20.14–18). If one reduces each of these three types of speech-in-character,
however, a lowest common denominator emerges; for Hermogenes, speech-
in-character involves crafting and assigning speech to a supposed speaker, the
only terminological distinction being the character to which those speeches
are attributed.
So, to whom may a writer attribute speech when engaging in speech-
in-character? The examples discussed above provide some answers. One
may attribute speech to known and definite (ὡρισμένων) humans, such as
Andromache or Achilles (8, 19, 21–23), to indefinite (ἀορίστων) persons, such as
“someone” who is about to depart (19–21), to the dead, as Hermogenes indicates
with Aristides’ attributions of speech to dead people (In Defence of Oratory 319–
343; Hermogenes, Prog. 16–18), to personified abstract ideas, such as Elenchos
in Menander (10), or even to inanimate objects, like the Sea in Aristides (11–12).
In addition to distinguishing between speeches-in-character based on
the identity of the supposed speaker, Hermogenes also divides speech-in-
character into the categories of “single” and “double” (20.24–21.2). Technically,
Hermogenes is speaking about ἠθοποιία (20.24), but there is no reason to sup-
pose that these categories would not similarly extend to προσωποποιία and
εἰδωλοποιία, since an invented personage or a dead person must inevitably
speak alone or to others. In any case, by “single,” Hermogenes means that the
supposed speaker does not address other characters but delivers a speech
by him or herself, such as a monologue (21.1–4). For example, what a general
would say when he returns from victory is a “single” speech-in-character (2–4).
A “double” speech-in-character occurs when the supposed speaker addresses
other characters on the scene, such as what that general would say to his army
after a victory (1, 4–5).
Hermogenes also notes that a speech-in-character may be ethical (ἠθικαί),
pathetical (παθητικαί), or mixed (μικταί; 10–11). Said otherwise, speeches-in-
character emphasize either the character of a speaker (“ethical”), the emotion
of a speaker (“pathetical”), or a mixture of both (“mixed;” 11–18).
Regardless of any categorical distinction, however, when someone assigns
words to an imaginary speaker, the writer must attribute appropriate words to
the imagined speaker and context. Hermogenes writes,

Πανταχοῦ δὲ σώσεις τὸ οἰκεῖον πρέπον τοῖς ὑποκειμένοις προσώποις τε καὶ

καιροῖς.
Speech-in-Character in the Progymnasmata 49

Throughout, you will preserve what is appropriately fitting [τὸ οἰκεῖον

πρέπον] for the imagined [ὑποκειμένοις] persons [προσώποις] and occa-
sions [καιροῖς]. (21.6–7; S 15)

This even includes using figures and making precise diction choices23 that
represent the character of the imagined speaker appropriately (22.4–5). For,
Hermogenes informs, people who are young or rejoicing speak differently than
people who are old or grieving, respectively (21.7–9).
Finally, in the composition of a speech-in-character, Hermogenes suggests
that the elaboration (ἐργασία) should advance through a temporal progression.

Ἡ δὲ ἐργασία κατὰ τοὺς τρεῖς χρόνους πρόεισι· καὶ ἄρξῃ γε ἀπὸ τῶν παρόντων,
ὅτι χαλεπά· εἶτα ἀναδραμῇ πρὸς τὰ πρότερα, ὅτι πολλῆς εὐδαιμονίας μετέχοντα·
εἶτα ἐπὶ τὰ μέλλοντα μετάβηθι, ὅτι πολλῷ δεινότερα τὰ καταληψόμενα.

The elaboration [ἐργασία] advances according to the three times

[χρόνους]; begin with the present [τῶν παρόντων], because it is difficult;
then run back to the past [τὰ πρότερα], because it shares in much happi-
ness; then, jump forward to the future [τὰ μέλλοντα], because the things
that are going to happen are much worse.24 (21.19–22.3; S 16)

To structure an extended speech-in-character effectively, therefore,

Hermogenes suggests that the temporal progression should begin by discuss-
ing the circumstances of the present (τῶν παρόντων), then turn back to consider
the way things were in the past (τὰ πρότερα), and finally conclude by jumping
forward to imagine what will happen in the future (τὰ μέλλοντα; 21.19–22.3).
With the exception of the temporal arrangement of the attributed speech,
Hermogenes provides no other formal requirements concerning the structure
of a speech-in-character as a whole. Though Hermogenes does not provide

23  Aphthonius and Nicolaus offer additional stylistic concerns in their Progymnasmata.
Aphthonius writes, “You will elaborate the ethopoiia in a style that is clear, concise, bril-
liant, unconstrained, and freed from any inversion or figure.” Prog. 35.11–13. Similarly,
Nicolaus writes, “The reported speech should be composed with rather short and natural
phrases, but not with long periods, for to be concerned with one’s way of speaking is for-
eign to emotion, and it is characteristic for those who lament to bring forth concisely and
briefly one thing after another.” Prog. 66.9–15.
24  The note that things in the future are “much worse” is in relation to a specific context.
Depending on the context for any given speech-in-character, the events of any temporal
period will be nuanced accordingly.
50 chapter 3

any complete examples of speech-in-character,25 in their original contexts, his

concrete examples betray a pattern. For instance, the first example Hermogenes
employs considers what Andromache would say about Hector (20.8–9; 21.14).
On the one hand, the context seems quite vague in Hermogenes’s text, as
Andromache doubtlessly could say many things about Hector in many different
contexts. Nicolaus the Sophist also uses this example in his Progymnasmata,
though he specifies that the context for Andromache’s supposed speech was
“when Hector fell” (πεσόντος Ἕκτορος; Prog. 64.12). On the other hand, when
Nicolaus’s use of Andromache as an example is combined with the fact that
Homer indeed records a speech from Andromache to the dead Hector when
she learned of his death, it becomes quite likely that Hermogenes had the
same context in mind.26 As such, when one examines the Homeric context
from which Hermogenes draws the example, a simple structure is visible.
Homer first identifies the forthcoming speaker as the wife of Hector (Il. 22.437),
then characterizes Andromache in the midst of her despair (22.437–476), and
concludes with speech-in-character which imitates Andromache’s severe grief
over the loss of Hector (477–514). Such is also the case with Hermogenes’s
other Homeric example of speech-in-character, “What words Achilles would
say about Patroclus” when he died (Prog. 21.15–18). Again, Homer first iden-
tifies a potential speaker, Achilles (Il. 18.2), then characterizes that speaker’s
grief (22–35), and concludes by composing speech-in-character that matches
Achilles’ intense distress over the loss of Patroclus and his plans for battle
(78–93, 97–126). Finally, the same pattern emerges in Hermogenes’s example
from Aristides’s In Defence of Oratory, in which Aristides introduces Miltiades,
Themistocles, Cimon, and Pericles as potential speakers (2.319), characterizes
these men as injustly accused by Plato (2.319–20), and finally “resurrects” and
attributes speech to them in an imaginary defence against Plato (2.321ff).
In these ways, Hermogenes’s examples reveal a pattern of (1) identification,
(2) characterization, and (3) attribution of speech-in-character.

25  Like Theon, Hermogenes does not record any attributed speeches. Hermogenes’s “exam-
ples” uniformly put forth hypothetical contexts for which his students might compose
speeches.
26  Regardless of whether Andromache’s speech to Hector when she learned of his death is
the exact context Hermogenes had in mind, this particular speech does nevertheless rep-
resent one example of “what words Andromache would say about Hector.” Thus, whether
this example fits Hermogenes’s truncated exemplary context intentionally or accidentally
remains impossible to verify, but, in either case, it nevertheless fits.
Speech-in-Character in the Progymnasmata 51

Condensing Hermogenes’s treatment of ἠθοποιία, προσωποποιία, and εἰδω-

λοποιία as speech-in-character into a brief outline, therefore, produces the fol-
lowing results. In speech-in-character,

(1) a writer crafts speech and attributes it to an imagined speaker;

(2) the subject to whom speech is attributed may be
a. a known person or type of person,
b. an unspecified person,
c. abstract concepts,
d. inanimate objects, or
e. the dead;
(3) the attributed speech must be appropriate
a. to the characters involved in conversation, and
b. to the contexts;
(4) the attributed speech should map chronologically from the present, back
to the past, and then forward to the future;
(5) a character may speak “singly” by oneself or “doubly” to others;
(6) a speech-in-character may emphasize character, emotion, or both; and
(7) the implied structure follows a pattern of identification of the speaker →
characterization of the speaker → attributing speech appropriate to the
character of that speaker.

Summary: Theon and [Hermogenes] on Speech-in-Character

Categorically speaking, Theon and Hermogenes approach speech-in-character

differently. Functionally, however, there is a thoroughgoing commonality.
When juxtaposed, the elements common to Theon and Hermogenes include:

(1) crafting and attributing speech to an imagined speaker;

(2) the imagined speaker may be a specific person or a general type of person;
(3) the attributed speech must be appropriate to
a. the character of the imagined speaker, and
b. the occasion of the speech;
(4) the attributed speech may be actual or imagined;
(5) speech-in-character can emphasize either
a. the character of the speaker (ethical), or
b. the emotion of the speaker (pathetical); and
(6) the implied structure follows a pattern of identification of the speaker →
characterization of the speaker → the attribution of speech-in-character.
52 chapter 3

Theon and Hermogenes also evince several interesting differences. To begin,

Theon’s attention to the context for which assigned words must be appropriate
is much more robust than that of Hermogenes, as Theon aims for his students
to hold assigned speech in balance with elements of character, audience, age,
occasion, place, social status, and the subject of the speech, while Hermogenes
only specifies attention to character and occasion. Additionally, Theon
highlights the functions and subjects speech-in-character might address—
consolation, exhortation, and epistolary—and he provides detailed notes
concerning what kinds of things to say in each kind of speech.27 Hermogenes,
however, advances no further in this respect than to allow that speeches may
emphasize character, emotion, or both, which, with the exception of the mixed
category, is present in Theon as well.28
Though the text attributed to Hermogenes is the shortest of the four extant
Progymnasmata, it similarly includes material not present in Theon and
makes up for its brevity with novelty. Hermogenes’s text introduces into the
progymnasmatic corpus several features that differ markedly from Theon’s
treatment and that preempt by at least one century the later progymnasmatic
writings of Aphthonius and Nicolaus. Based on the textual evidence, Theon for
all intents and purposes seems to be unaware of the option to divide speech-
in-character into the categories of ἠθοποιία, προσωποποιία, and εἰδωλοποιία with
respect to the identity of the imagined speaker. Indeed, without noting the
possibility that others might reserve the term προσωποποιία to attributions of
speech to inanimate or abstract characters, Theon uses προσωποποιία solely
to speak of assigning speech to humans, which Hermogenes accounts for
under ἠθοποιία.29 What is more, if somewhat lackluster, Hermogenes adds
the categories of “single” and “double” speeches-in-character. And, more

27  Nicolaus also considers the rhetorical functions to which one might place speech-in-
character, but he casts the discussion in terms of the three species of rhetoric: encomium,
judicial, and deliberative, with an added nod towards the exercise’s applicability in writ-
ing letters. Prog. 66:16–67:9.
28  Regarding the function of speech-in-character in ethical, pathetical, or mixed categories,
Nicolaus writes, “Someone either focuses on general principles (i.e., character) or on what
has happened from a given circumstance (i.e., emotion), for in this way character differs
from emotion.” Said otherwise, ethical speeches highlight what character traits belong to
certain character types (such as, to a coward), but pathetical speeches emphasize what
certain characters might do in a specific, usually highly charged, context. Prog. 64.5–13.
29  Aphthonius similarly divides speech-in-character into the three categories of ἠθοποιία,
προσωποποιία, and εἰδωλοποιία based on the identity of the imagined speaker. Aphthonius,
Prog. 34.2–18. Nicolaus only mentions two categories, ἠθοποιία and προσωποποιία. For
Speech-in-Character in the Progymnasmata 53

constructively, Hermogenes is the first to give any explicit advice concerning

the general arrangement of the attributed speech according to present, past,
and future times.30
Thus, as Chapter 2 compared and contrasted the portrayals of speech-
in-character in ad Herennium and Quintilian, Chapter 3 has demonstrated
the similarities and differences between the Progymnasmata of Theon and
Hermogenes, respectively. It is time, therefore, to conclude this consideration
of the theoretical aspects of speech-in-character as it both would and could
have been relevant to the apostle Paul. Thus, I turn now to Chapter 4, in which
I first gather together the core elements pertinent to speech-in-character and
common to all of the writers currently under investigation and, second, isolate
the features unique to each writer to which Paul might have had access in his
first-century context.

Nicolaus, ἠθοποιία begins with defined persons and only creates speech, but προσωποποιία
requires the invention of the person and the applied words. Prog. 64:1–3; 64:20–65:10.
30  Aphthonius also advises dividing speech-in-character in accordance with the three times.
Aphthonius, Prog. 35.13–14. Nicolaus similarly advises dividing speeches-in-character into
the three times, but he arranges the times slightly differently—one should begin in the
present, move back to the past, return to the present, and finally contemplate what might
happen in the future. Nicolaus, Prog. 65.11–66.8.
 Because Gibson emphasizes Hermogenes and Aphthonius, he includes this division
into the three times as one of the central characteristics of progymnasmatic speech-in-
character. What is more, by examining the “themes” of speech-in-character, Gibson writes
that one common component of speech-in-character is that “the speech is most often
imagined as taking place after the death of a loved one, or when a stereotyped character
is confronted with a situation antithetical to a component of that stereotype.” Because
of these underlying structures and themes, Gibson is unable to find any instances of for-
mal speech-in-character in the New Testament. “Prosopopoiia,” 3–4. Contrary to Gibson,
because I date Theon early and lessen the influence of Aphthonius and Nicolaus on
methodological grounds, the temporal division of speech-in-character is an element
unique to Hermogenes and therefore only a subsidiary or optional element in my analy-
sis. Additionally, Theon’s engagement with the function of speech-in-character to exhort,
dissuade, console, and so forth broadens the potential “themes” of speech-in-character
exponentially wider than Gibson’s limited treatment. As a result, when it is recognized
that ad Herennium, Quintilian, and Theon neglect to require such temporal divisions, and
when the potential or likely “themes” of speeches-in-character are reconsidered in light of
Theon’s broader agenda, the New Testament is opened up to prosopopoetic analysis that
would not be possible in Gibson’s model, as Chapters 5 and 11 demonstrate.
 chapter 4

Speech-in-Character: A Synthesis

Though any writer of any time period could theoretically employ rhetorical
tools that never existed or are no longer extant in the historical evidence,
this chapter firmly establishes the elements of speech-in-character that were
doubtlessly in use during the first century CE . Because of the prevalence of
these core concepts of speech-in-character, one should expect most if not all
of these elements to be present in any appropriation of the exercise, including
the writings of Paul and other early Christians. The presence or absence of
these unique elements will depend, first, on precisely what the author under
consideration “knows” with respect to the breadth and depth of the practice of
speech-in-character and, second, on the artistic license of that writer.

Core Conventions of Speech-in-Character

The quartet or primary sources analyzed in Chapters 2 and 3 shares a re-

markable number of similarities. To be sure, the list of comprehensive simi-
larities will be limited by the lowest common denominator for any given
aspect. Nevertheless, the core concepts of speech-in-character shared by ad
Herennium, Quintilian, Theon, and Hermogenes unite to construct a coherent
and functional rhetorical tool. The four qualities each of these writers manifest
include:

(1) a writer or speaker crafts and assigns speech to an imaginary speaker;

(2) the assigned speech must appropriately model the character of the
speaker;
(3) the assigned speech may be actual (i.e., “someone says”) or hypothetical
(i.e., “what would someone say,” or “someone may say”); and
(4) there is an implicit tendency to include three elements according to
the following progression: an identification of a supposed speaker →
a characterization of that speaker → an attribution of appropriate
speech-in-character.

The consistent depiction of these core elements portrays speech-in-character

in its simplest form as the crafting and applying to some character actual or

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_006

Speech-in-Character: A Synthesis 55

imaginary speech that agrees with the character of the imagined speaker.1
Furthermore, this designation of speech should identify the imaginary speak-
er, present the character of that speaker, and then attribute words appropriate
for that speaker.
If Theon is removed momentarily, the three remaining treatments of
speech-in-character reflect additional similarities. Ad Herennium, Quintilian,
and Hermogenes include that

(5) in addition to scripting speech for specified humans, the speaker may be:
a. an unspecified person (i.e., “someone”),
b. a dead person (i.e., “Lucius Brutus returning from the grave”),
c. a personified inanimate object (i.e., “the Sea”),
d. a personified abstract idea (i.e., “Wisdom”).

Though not exhaustive, the presence of these similarities across three writers
demonstrates that one should not be surprised to find them in concrete exam-
ples of speech-in-character.2 In fact, Quintilian’s justification for his use of the
single term προσωποποιία instead of dividing speech-in-character into three
categories serves as something of an accidental apology for Theon’s omission
of potential imaginary speakers that are not human persons. Quintilian writes,
“For we cannot of course imagine a speech except as the speech of a person”
(Inst. 9.2.32 [Russell, LCL]).

Unique Features of Speech-in-Character

Each rhetorical theorist evinces features of speech-in-character that are

unique. Though these elements do not represent the core concepts of speech-
in-character, their presence in the primary literature proves their potential
relevance in the first- and second-century context. This means the writers

1  Aristotle speaks remarkably similarly, arguing that style and subject matter must be “analo-
gous,” or agree appropriately. Ars Rhetorica 3.7.1–7.
2  Similarly, in the first century BCE, Cicero allows that (1) “even dumb (muta) objects
must speak” (Part. Or. 55 [Rackham, LCL]); (2) that speech may be attributed to humans,
inanimate things (i.e., “a place, a city, or a monument”), or abstract ideas (i.e., “a law”;
de Inventione 1.99–100 [Hubbell, LCL]); and (3) that “orators and philosophers have license to
cause dumb (muta) things to talk, to call on the dead to rise from the world below, to tell of
something which could not possibly happen” (Topica 45 [Hubbell, LCL]).
56 chapter 4

who composed speeches-in-character during that time could have known and
used none, one, or several of the variously nuanced approaches to speech-in-
character. Methodologically, these unique nuances represent potential or op-
tional elements that some writers may or may not have drawn on to style their
speeches-in-character.
To begin, in distinction from the progymnasmatic writers, ad Herennium and
Quintilian comment on the need to inflect one’s voice suitably when speaking
in the voice of another presumed speaker (Rhet. Her. 3,14; Quintilian, Inst. 1.8.3).
This difference is due at least in part to the various authors’ intended goal; ad
Herennium and Quintilian emphasize oratory, whereas the Progymnasmata
emphasize composition. Both composition and oratory relate to Paul and his
letters, however. Though the issue is surely more complex, compositional em-
phases would be necessary in letter writing, and oratorical emphases would be
important in delivering the letters.3
Quintilian provides several other unique elements. First, speech-in-character
serves a number of general functions, and it will be important to consider
whether concrete examples of speech-in-character fit into one of these catego-
ries (Inst. 9.2.30):

(1) to create variety and liveliness in a speech;

(2) to display the inner thoughts of opponents;
(3) to introduce conversations; and
(4) to provide appropriate characters for specific rhetorical situations.

Second, Quintilian allows that one may attribute words to the gods (9.2.31),
and, third, that the medium through which a speech-in-character is presented
may be speech, narrative, or a written document (9.2.34, 37). Finally, Quintilian
is the only writer who seems to allow that speech-in-character may be present
even when it is not overtly indicated that certain lines are in fact spoken in the
voice of another character, and this will prove quite significant when seeking
to explain the abrupt transitions in Rom 3:1–9.

3  Discovering an autograph with something like Phoebe’s marginalia concerning how and
where to inflect her voice would truly be a boon of a discovery for understanding Paul’s dia-
logical sections in Rom 3, 7, and elsewhere. Unfortunately, such information is lost in the
sands of time, and there is no methodologically sound way to recover the vocal inflections
as performed by Phoebe in the earliest readings of the letter. As argued below, one can, how-
ever, allow the core conventions for speech-in-character to be a guide. In any case, Quintilian
assumes that readers will be able to identify speech-in-character and inflect the voice(s)
accordingly.
Speech-in-Character: A Synthesis 57

Despite his omission of allowing for several standard potential speakers,

Theon’s treatment of speech-in-character offers several unique elements as
well. For instance, Theon remarks that speech-in-character can be most ef-
fective in (1) consolation speeches, (2) exhortation speeches, or (3) epistolary
documents (Prog. 115.20–22; 116.27–117.4). What is more, Theon’s exhaustive at-
tention to characterization and the rhetorical context is unsurpassed, as he
considers elements of the speaker’s character, race, age, social status, gender,
vocation, of the audience’s character, of the occasion and place of the speech,
and of the subject of the speech (115.22–116.22). Finally, Theon’s “starting plac-
es” offer a rudimentary but novel “how to” approach to speech-in-character
(116.22–117.30). It will be important to consider whether any of these elements
improve our reading of concrete speeches-in-character in Chapters 5, 7, and 11.
Finally, though Theon and Hermogenes join together to report that speech-
in-character can emphasize the character or the emotion of the speaker (or
both, for Hermogenes; Theon, Prog. 117.30–32; Hermogenes, Prog. 21.10–18),
Hermogenes offers the only explicit advice concerning how to structure attrib-
uted speech. Hermogenes suggests that the most effective attributed speech
will begin in the present time, fall back to the past, and then jump forward
to the future (Prog. 21.19–22.3). Though other Pauline speeches-in-character
(Rom 7:7b–12, 13b–24, 25b?) might resemble this structure, most occurrenc-
es are quite brief and do not. That, however, remains to be demonstrated in
Chapter 5.
 chapter 5

Examples of Speech-in-Character in Paul

The three preceding chapters examined the four primary voices that serve as the
foundation for identifying the conventions for the rhetorical figure of speech-
in-character through the second century CE: ad Herennium, Quintilian, Theon,
and Hermogenes. In these examinations, the central elements espoused by all
or most of the texts rose to the fore. These core features were five in number:
(1) the writer or speaker crafts speech and assigns it to an imaginary speaker;
(2) the attributed speech must be appropriate to the characterization of the
speaker; (3) the speech may be portrayed as actual or hypothetical speech;
(4) three elements tend to accompany speech-in-character—an identification
of the speaker or an indication that someone else is speaking, a characterization
of the speaker, and the attributed speech-in-character; and (5) speech may be
assigned to specific humans, unspecified humans, abstract ideas, inanimate ob-
jects, or the dead. When examining concrete examples of speech-in-character,
one should expect all or most of these features to be present.1
The comparative analyses, however, also revealed the unique, or less at-
tested, elements represented in each text. For instance, Quintilian suggests
three common functions for speech-in-character that prove to be important
in the discussions below, as well as indicating that speech-in-character can be
assigned to the gods, or that the figure can even be present without any overt
shift to another speaker at all. Theon also offers a list of standard functions
for speech-in-character. Theon’s breadth of attributes to keep in mind with
respect to the characterization of the speaker and the situation in which the
speech occurs is second to none, as is his corpus of suggested starting places
that provide cues for how to begin and proceed through a speech-in-character
in order to achieve the intended result. Finally, Hermogenes was the only text
that suggested anything of a standardized way to structure speech-in-character
chronologically. Though these unique elements are less attested than the core
features, they are nevertheless attested in the relevant literature, so it is at least
possible that writers of the same time period could have been aware of and uti-
lized them. As such, these unique elements function in this analysis as second-
ary or optional features for speech-in-character that a writer, such as Paul, may

1  By “concrete examples,” I mean the use of speech-in-character in practice, or the use of
speech-in-character that is not confined to the treatments of the figure in the handbooks
and Progymnasmata.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_007

Examples of Speech-in-Character in Paul 59

or may not have been aware of and may or may not have employed in order to
style an example of speech-in-character in a particularly measured way. In the
following discussions, it will become clear that Paul indeed manifests several
of these features in his implementation of speech-in-character.
The works on rhetorical theory analyzed in the preceding chapters, how-
ever, cannot be assumed to be a complete representation of the practice of
speech-in-character. These treatments must be the sure-footed starting point
from which to begin studying the rhetorical skill, but they cannot constitute
one’s whole understanding of the exercise. All four of the treatments exam-
ined above are elementary in nature and intended as aides to educate begin-
ners in the proper use of rhetoric in general and, for the present purposes, of
speech-in-character in particular. For instance, ad Herennium is an education-
al text that provides an introduction to the theory of rhetoric; it is not a docu-
ment for master rhetoricians.2 Similarly, Quintilian’s Institutio Oratoria is a
text aimed at developing an individual beginning with youth throughout the
course of his or her rhetorical education into a mature rhetor.3 The two treat-
ments of Progymnasmata, Theon and Hermogenes (as well as Aphthonius and
Nicolaus), also aim at the elementary, or “preliminary,” education of students
in rhetoric and composition.4
In more advanced rhetorical texts, it would be of little surprise to find ap-
propriations of speech-in-character that deviate from or alter the theoretical
conventions for the exercise. In fact, it was expected that skilled rhetoricians
would amend and conform rhetorical theory to fit the needs of their practical
rhetorical situations. On this score, Gibson writes,

2  For instance, ad Herennium is addressed to an unidentifiable Gaius Herennius, who wishes

to learn rhetoric, and who is advised that “theory without continuous practice in speaking is
of little avail; from this you may understand that the precepts of theory here offered ought
to be applied in practice” (ad Herennium 1.1 [Caplan, LCL]). Whether Gaius Herennius or
students in general actually represent the audience of ad Herennium, the introductory and
instructional motivations remain the same.
3  In the prooemium to Book 1, Quintilian notes his hopes that Institutio Oratoria “will be use-
ful for the education of [Marcus Vitorius’ (to whom Quintilian dedicates Inst.)] son Geta.”
Inst. 1.6. Moreover, Quintilian’s treatment begins with the very birth of the child and his
elementary education. Inst. 1.1.1.
4  For example, Theon’s Progymnasmata is “addressed to teachers, not to students,” to help
teachers best educate youth in the preliminary stages of composition and rhetoric. Kennedy,
Progymnasmata, 2.
60 chapter 5

It is clear, then, that although prosopopoiia begins as an exercise with

certain formal characteristics and no real rhetorical context, students
take away from it a skill to be applied elsewhere … In the hands of accom-
plished writers, both ancient and modern, simpler compositional forms
either become identifiable component parts of more advanced forms, or
are abandoned as forms, leaving behind only particular skills … The origi-
nal exercise is intended to teach the student how to convey ethos and
pathos through attributed speech, and that’s all.5

This leads Gibson to conclude,

I would like to suggest that it may be better to view prosopopoiia, not as

a stable building block of discourse learned in school and plopped into
texts, but rather as one ethopoetic technique among many that accom-
plished writers used in order to advance their rhetorical goals.6

Consequently, though one must pay close attention to the general rules rep-
resented by ad Herennium, Quintilian, Theon, and Hermogenes, it is equally
necessary to notice alterations and emendations of those most elementary
patterns in practical implementations of speech-in-character, as they may be
quite indicative of the author’s intended goal in a particular rhetorical con-
text. Indeed, such sensitivity will prove quite fruitful when engaging certain
examples of speech-in-character in Paul’s letters.
The primary goals of this chapter are four in number. The first task of this
chapter is to demonstrate that the Pauline examples of attributed speech
largely agree with the conventions of speech-in-character. This will help to
ground the discussion of Rom 3:1–9 in light of diatribe and speech-in-character
in Part 3.
Second, as the previous three chapters examined the theoretical aspects of
speech-in-character and several of the concrete examples of the exercise of-
fered or referenced by them as models, this chapter expands on the examina-
tion of concrete examples by incorporating Paul’s authentic letters into the
discussion.7 In this way, the examination of speech-in-character presented
here is triangulated by theory, actual rhetorical texts, and Paul’s letters.
Third, as argued above, it is not sufficient simply to recognize the presence
of speech-in-character in a given text. Attention must also be given to the

5  Gibson, “Prosopopoeia,” 9, emphasis original.

6  Ibid., 10.
7  With one exception, I reserve discussion of Romans until Part 3.
Examples of Speech-in-Character in Paul 61

degree to and ways in which an author amends the basic conventions. Each
discussion below documents that Paul has employed speech-in-character and
analyzes Paul’s appropriation of the figure to suit the particular rhetorical con-
text being addressed in that letter or pericope.
Fourth, looking at the scope of Paul’s examples of speech-in-character al-
lows for the development of a preliminary snapshot of his use and/or aware-
ness of the central and secondary features of the exercise.
Note, however, that the goal of this chapter is not to answer each and
every exegetical oddity that arises in the texts under discussion. Such engage-
ment lies outside the scope of this project. Instead, only those points of ex-
egesis that significantly influence the shape of the speech-in-character under
examination—for instance, concerning the identity of the speaker, the ex-
tent of the attributed speech, or the rhetorical context in which the speech
occurs—will be engaged directly. Most other issues of exegetical quandary, no
matter how interesting or perplexing, must be temporarily omitted.
Finally, a preliminary note seems necessary with respect to what might be
perceived as a notable difference between the theoretical treatments of speech-
in-character and the examples that follow—size. It is true that many surviv-
ing examples of speech-in-character from Aphthonius (Prog. 35.15–36.20) and
Libanius,8 amongst others, are longer than any of the Pauline texts discussed
in this chapter. Indeed, with the exception of Rom 7 and 2 Cor 11–12 (not dis-
cussed at length in this project), the Pauline texts tend to script between one
and three lines of speech in the voice of the imagined speaker(s). Based on
Hermogenes’s prescripted chronological progression for speech-in-character,
however, Gibson suggests “a minimum requirement of three sentences.”9
But does the length of the attributed speech really matter? Ultimately, evi-
dence and argument demonstrate that the length of the attributed speech
is not a significant concern for determining whether the rhetorical figure of
speech-in-character is in play. First, none of the ancient theorists offer any spe-
cific length requirements. Even Hermogenes’s (or Aphthonius’s or Nicolaus’s)
suggested chronological progression, on which Gibson grounds his “mini-
malist” approach, fails to stipulate how long to dwell on each time period.
Plus, such a progression can hardly be taken as thoroughly characteristic of
speech-in-character, as none of the three earlier theorists show any awareness
of it. What is more, rather than composing speeches-in-character in periodic

8  Libanius’s Progymnasmata: Model Exercises in Greek Prose Composition and Rhetoric

(Craig A. Gibson, trans. and ed.; Writings from the Greco-Roman World 27; Atlanta: Society
of Biblical Literature, 2008), 355–425.
9  Gibson, “Prosopopoeia,” 7.
62 chapter 5

style, the later progymnasmatists Aphthonius (Prog. 35.11–13) and Nicolaus

(Prog. 66.9–15) even comment on the conciseness and brevity with which each
line of a speech-in-character should be composed. Quintilian even offers as
a paradigmatic example of speech-in-character the altogether brief example
of speech attributed to the amazed Trojans in Vergil’s Aeneid 2.29 (Quintilian,
Inst. 9.2.37).
Second, the quote from Gibson above already demonstrates the expecta-
tion for practical implementations of speech-in-character to be adapted and
take on new forms when placed in real (rather than fabricated for scholastic
purposes) rhetorical contexts. In such practical implementation, there is no
reason a given writer or speaker could not or would not attribute fewer words
in some instances than in others, and vice versa. Said otherwise, the proso-
popoetic skill might very well be put to use in any length of attributed speech.
Third, it is demonstrable that the tools required to compose many of the fol-
lowing Pauline examples most closely align with the skills taught for speech-
in-character rather than transferring from some other exercise. For instance,
one might consider verbal chreia to be quite applicable, especially given the
chreia’s general brevity. Chreiai, however, are always attributed to people,
whether specific (i.e., Diogenes) or unspecific (a Laconian). There are no sug-
gestions among the rhetorical theoreticians that chreiai can be attributed to
inanimate objects or abstract ideas, nor do any actual examples exist.10 On the
other hand, as highlighted in Chapters 2 through 4, such attributions of speech
are altogether par for the speech-in-character course. Thus, if Paul attributes
speech to an abstract concept (δικαιοσύνη, for instance) or an inanimate thing
(γραφή, or a body part), it is more reasonable to identify the necessary skills as
transferring from speech-in-character rather than from chreia.
Because of how consistently Paul’s writings cohere with the conventions
for speech-in-character (see below), it seems almost certain that speech-in-
character is the operative figure behind these texts and that Paul was direct-
ly or indirectly familiar with the conventions for the skill as a whole. Unless
Paul provides some clear reason to look elsewhere, the best starting place is to
consider Paul’s attributions of speech to imaginary speakers as implementa-
tions of speech-in-character, even though they are usually shorter than model

10  Theon’s comment that one can attribute chreiai to “something analogous to a person”
seems to account for the attribution of chreiai to unspecified persons as opposed to speci-
fied persons. Theon, Prog. 96.18–22. Hock and O’Neil, Chreia, 109n.2.
Examples of Speech-in-Character in Paul 63

examples of the exercise. As such, I offer for consideration the following ex-
amples of Pauline speech-in-character.11

Examples of Speech-in-Character in Paul

1 Thessalonians 5:3
The first example of speech-in-character in the Pauline corpus is present in
the earliest extant piece of Christian literature, 1 Thessalonians. After intro-
ducing the eschatological topic of “the times and the seasons,” Paul reminds
the Thessalonian Christians, “you yourselves know accurately that the day of
the Lord is coming as a thief in the night” (1 Thess 5:1–2). In order to depict the
sudden and unexpected nature with which the day of the Lord will arrive for
those who are not prepared, Paul writes in 5:3:

ὅταν λέγωσιν· εἰρήνη καὶ ἀσφάλεια, τότε αἰφνίδιος αὐτοίς ἐφίσταται ὄλεθρος
ὥσπερ ἡ ὠδὶν τῇ ἐν γαστρὶ ἐχούσῃ, καὶ οὐ μὴ ἐκφύγωσιν.

Whenever they say, “Peace and security,” then sudden destruction comes
upon them like birth pangs in a pregnant woman, and they will not
escape.

Paul then returns his focus to the Thessalonian congregation and explains how
they differ from those depicted in 5:3. The Thessalonian Christians (ὑμεῖς) “are
not in darkness” but are “children of light and of the day” (5:4–5), they should
be “alert and sober” (5:6–8), and they should be clothed with “the breastplate
of faith and love and the helmet, the hope of salvation” (5:8). Additionally,
“God has not appointed [the Thessalonian congregation] for wrath but for
the acceptance of salvation through our Lord Jesus Christ” (5:9–10). For these
reasons, in contrast to those depicted in 5:3, the Thessalonian congregation
should be prepared for the day of the Lord at all times, so that it will not
surprise them like a thief in the night (5:4).12 Paul concludes by exhorting the
congregation to encourage and build up one another (5:11).

11  For additional examples, see Appendix B in my dissertation, which examines 1 Cor 10:28;
15:35; 2 Cor 10:10. See also 1 Cor 11:23–26. King, “Speech-in-Character,” 436–48.
12  For καταλάβῃ as “surprise,” see BDAG, καταλαμβάνω 3a-b; Abraham J. Malherbe, The
Letters to the Thessalonians: A New Translation with Introduction and Commentary (AB
32B; New York: Doubleday, 2000), 294.
64 chapter 5

The shift in speakers is clearly marked. Paul speaks in his own voice to the
Thessalonians as “you” in 1 Thess 5:1–2. In 5:3a, however, the third person plural
λέγωσιν introduces a new speaker (i.e., “they”) followed by attributed speech
(5:3b), and scholars regularly note that the scripted speech is not spoken in
Paul’s own voice.13 The τότε of the temporal construction, ὅταν λέγωσιν … τότε,
signifies the return to Paul’s voice, at which time he suggests that “sudden
destruction comes upon them … and they will not escape” (5:3c). Refocusing
his attention on the Thessalonian congregation, Paul again refers to the
Thessalonians as “you” (5:4–5, 11) and “we” (5:5–6, 8–10) in distinction from the
“they” of 5:3.
But who are “they” who proclaim “peace and security” with the result that
inescapable destruction will be heaped upon them (5:3)? Paul does not provide
a formal identification, but he does suggest several characteristics. In fact, the
strong contrastives ὑμεῖς δέ (5:4) and ἡμεῖς δέ (5:8) situate the Thessalonian
congregation (with Paul) in opposition to the speakers of 5:3.14 Therefore, the
speakers of 5:3 are characterized as existing in darkness and surprised by the
day of the Lord (5:4), as children of night and darkness (5:5), as asleep and
drunk (5:6–7), and as unclothed with the “breastplate of faith and love and the
helmet, the hope of salvation” (5:8). As mirror opposites of the Thessalonian
Christians who should always be prepared, the non-Christian speakers of 5:3
are thoroughly unprepared for the day of the Lord.15
This characterization, however, does little to help identify the unspecified
speakers of 5:3 with much precision. Without identifying this as speech-in-
character, scholars tend to focus on the attributed speech of 5:3 as a means to
identify the unspecified speakers. For example, Abraham J. Malherbe writes,

13  For example, see Jeffrey A. D. Weima, “‘Peace and Security’ (1 Thess 5:3): Prophetic
Warning or Political Propaganda?” NTS 58.3 (2012), 331, who writes, “All translations place
the brief phrase … in quotation marks because it is clear from the introductory formula …
that the apostle here is not creating but citing these words.” Below I problematize Weima’s
view that Paul “is not creating but citing these words,” but Weima clearly indicates the
sweeping consensus that another speaker enters the scene at 1 Thess 5:3.
14  Gordon D. Fee argues similarly that 1 Thess 5:3 functions as a contrast to the Thessalonian
Christians, writing, 1 Thess 5:3 “allows Paul yet another opportunity to contrast the believ-
ers in Thessalonica with those in their city who are causing their present grief.” The First
and Second Letters to the Thessalonians (NICNT; Grand Rapids: Eerdmans, 2009), 190.
15  Others who identify the speakers in 5:3 as non-believers in contrast to the Thessalonian
Christians include Beverly Roberts Gaventa, First and Second Thessalonians
(Interpretation; Louisville: John Knox, 1989), 70–71; Abraham Smith, The First Letter to
the Thessalonians (NIB 11; Nashville: Abingdon, 2000), 726; Ben Witherington III, 1 and 2
Thessalonians: A Socio-Rhetorical Commentary (Grand Rapids: Eerdmans, 2006), 148.
Examples of Speech-in-Character in Paul 65

“Paul does not have to identify these people, for his readers know of whom
he is writing … The content of their teaching helps to identify them.”16 For
Malherbe, the phrase εἰρήνη καὶ ἀσφάλεια “is Paul’s own ironic formulation to
describe the teaching of false teachers” in Thessalonica, and it is a reworking
of the false prophets’ cries of “peace, peace,” in the Hebrew Bible.17 In order
to combat an attitude among some Thessalonians who were “not sufficiently
tak[ing] into consideration the eschatological dimension of their existence,”
Paul replaced the second occurrence of εἰρήνη with the Epicurean term
ἀσφάλεια, which emphasizes the Epicurean fixation “on life with friends in the
here and now.”18 Thus, Malherbe would identify those who proclaim “peace
and security” as Thessalonian false teachers who promoted a worldview that
diminished or omitted eschatological existence.
Scholars such as Peter Oakes and Jeffrey A. D. Weima, however, interpret
the speech of 5:3 in light of the Pax Romana and the Roman values of peace
(pax) and security (securitas). Oakes argues that the phrase εἰρήνη καὶ ἀσφάλεια
“seems to evoke a slogan of a current ‘Golden Age,’” since the terms represent
“a very powerful evocation of the central ideology of the new age brought in
by Augustus.”19 Similarly, Weima contends the phrase recalls “a popular theme
or slogan of the imperial Roman propaganda machine.”20 These readings,
therefore, identify the speakers of 5:3 as those advocating hope in the political
εἰρήνη καὶ ἀσφάλεια that the Roman Empire has proclaimedly achieved in
ignorance of the eschatological age yet to come.
Both readings evince problems. On the one hand, Malherbe dismisses the
Greco-Roman imperial evidence too quickly and fails to recognize that those
proclaiming the Roman values of εἰρήνη καὶ ἀσφάλεια in the present correlate
(as a type of false teacher) extremely well with his depiction of false teachers
who neglect the eschatological dimension of existence. On the other hand, the
allusions in 1 Thess 5:3 to Jer 6:14, 24 (cf. Ezek 13:10, 15) seem quite pronounced,
but Oakes and Weima largely ignore them in their respective emphases on the
Greco-Roman evidence. Moreover, Joel R. White has recently demonstrated
that there is no clinching evidence that εἰρήνη καὶ ἀσφάλεια / pax et securitas
circulated as a “slogan” for imperial propaganda from Augustus’ reign through

16  Malherbe, Thessalonians, 291.

17  Ibid., 291–304.
18  Ibid., 304–5.
19  Peter Oakes, “Re-mapping the Universe: Paul and the Emperor in 1 Thessalonians and
Philippians,” JSNT 27.3 (2005), 317.
20  Weima, “Peace and Security,” 332. Similarly, Smith, The First Letter, 675–78, 726.
66 chapter 5

Paul’s life.21 Thus, White concludes that, though either approach reasonably
accounts for the presence of εἰρήνη, neither sufficiently explains Paul’s use of
ἀσφάλεια.22 But White’s argument is problematic as well; he seems to equate
the lack of concrete evidence for a slogan with the conclusion that Paul could
not still refer to the Roman concept of ἀσφάλεια / securitas, which Oakes,
Weima, and White all document as a central Roman value, even if it is less
pronounced than that of εἰρήνη / pax.23
Combining the strengths of each approach, however, produces an interesting
explanation for the identity of the speakers in 1 Thess 5:3.24 First, the allusions
to Jer 6 (cf. Ezek 13:10, 15) seem quite pronounced, as it contains the context of
false prophets / teachers, their cries of “peace” when there actually is no peace
(Jer 6:14), and a metaphor of birth pangs (6:24), each of which is present in
1 Thess 5:3. Second, the plethora of evidence indentifying εἰρήνη / pax and
ἀσφάλεια / securitas as central (though not exclusive) Roman values best explains
the combination of these words in 1 Thess 5:3, even if it does not represent a
fixed slogan.25 It seems Paul has recast the context of Jer 6 for his Thessalonian
audience. The first way Paul adjusts Jer 6 is by replacing Jeremiah’s “εἰρήνη
εἰρήνη” with “εἰρήνη καὶ ἀσφάλεια.” Because no evidence confirms that εἰρήνη
καὶ ἀσφάλεια existed as a formal slogan from Augustus’ reign through Paul’s life,
it is best to view the phrase as Paul’s composition that would be characteristic
of the speakers of 1 Thess 5:3.26 Since εἰρήνη and ἀσφάλεια comprise two of
Rome’s core values, these speakers are best identified as those promoting the

21  Joel R. White, “‘Peace and Security’ (1 Thessalonians 5:3): Is it Really a Roman Slogan?”
NTS 59.3 (2013): 382–95. If White is correct, then Weima’s view that Paul “is not creating
but citing these words” must be readjusted to allow for Paul’s creativity in summariz-
ing these speakers’ message with the phrase εἰρήνη καὶ ἀσφάλεια. Cf. Weima, “Peace and
Security,” 331.
22  White, “Peace and Security,” 395.
23  Oakes, “Re-mapping the Universe,” 317–18; Weima, “Peace and Security,” 333–55; White,
“Peace and Security,” 384–92.
24  Without so much detail, Fee also seems to follow a similar approach. Thessalonians,
188–89.
25  Oakes, “Re-mapping the Universe,” 317–18; Weima, “Peace and Security,” 333–55; White,
“Peace and Security,” 384–92.
26  Consequently, contra Weima, Paul does have a creative role in composing the speech
of 1 Thess 5:3. Cf. Weima, “Peace and Security,” 331. Furthermore, Malherbe correctly
notes that “Paul does not have to identify these people, for his readers know of whom
he is writing … The content of their teaching helps to identify them.” Thessalonians, 291.
Necessarily, in order for the Thessalonian congregation to identify these speakers cor-
rectly, the speech Paul scripts in their voice must resemble their characterization and the
kind of things they actually could have said.
Examples of Speech-in-Character in Paul 67

present reality of Roman peace and security. Thus, as Jeremiah’s false prophets
were claiming peace in its absence, so also those proclaiming the Roman values
of εἰρήνη καὶ ἀσφάλεια were advocating the present reality of Roman peace and
security in ignorance of the eschatological peace that Paul held to be superior.27
Analogously, as the Israelites were not to hope in the false prophets’ claims of
peace despite the Babylonian threat, neither were the Thessalonians to place
hope in the peace and security advocated by Rome.
Paul also amends Jeremiah by shifting the focus of the birth pangs. Jeremiah
depicts Israel as experiencing birth pangs (6:24), but Paul realigns the metaphor
as a way to express the suddenness and unexpectedness with which the day of
the Lord would appear to those unprepared. In this reading, Paul’s agenda is
not to critique the Roman Empire directly (though it does implicitly) but to
provide a depiction of the suddenness with which the day of the Lord would
come and a contrast of the preparedness Paul exhorts for the Thessalonian
Christians, as evidenced in 5:4–10.28
How, then, does 1 Thess 5:3 cohere with the conventions for speech-in-
character? It coheres quite well indeed, as all of the central elements of
speech-in-character are present: (1) Paul crafts speech and places it in the
mouth of another speaker (i.e., “they;” 5:3a–b). (2) The assigned speech is
appropriate to the character of the imagined speakers. Paul characterizes
the speakers of 5:3 as non-Christians who are unprepared for the day of the
Lord (5:4–8); those who place hope in present Roman peace and security in
ignorance of the eschatological peace and salvation in which Christians hope
fit the charcterization quite nicely. (3) The assigned speech is imagined as
actual speech. (4) A new speaker is introduced (i.e., “they;” 5:3a), characterized
(5:4–8), and attributed an appropriate speech-in-character (5:3b). The small
difference between Paul’s composition of this speech-in-character and the
core conventions is that the reader or hearer must derive the characterization
of the speakers from Paul’s comparison of them to the Thessalonian Christians.

27  Similarly, Weima contends, “For the apostle, peace and security belong only to those who
instead trust in God, who ‘did not destine us for wrath but for the obtaining of salvation
through our Lord Jesus Christ’ (5.9).” Ibid., 359.
28  Oakes similarly writes, “In 5.3, Paul is emphasizing the unexpectedness of the Day of the
Lord. He is not specifically attacking the Roman Empire. However, he seems deliberate-
ly to be denying the central assertion of Roman imperial ideology. He asserts that the
Empire cannot guarantee ‘peace and safety’. Its claim to do so will be disproved by the
arrival of Christ … This is Christianity against Rome. However, it is neither Christianity
seeking Rome’s overthrow nor Christianity arguing against participation in the imperial
cult. It is Christian hope being asserted to be superior to Roman hope.” “Re-mapping the
Universe,” 318.
68 chapter 5

(5) The convention shared by ad Herennium, Quintilian, and Hermogenes that

the speaker(s) may be unspecified accounts for the unspecified identity of the
speakers in 1 Thess 5:3.
With respect to the more narrowly attested elements of speech-in-
character, a few additional points of connection appear between 1 Thess 5:3
and the works of the rhetorical theorists. First, 1 Thess 5:3 coheres with two of
Quintilian’s proposed functions of speech-in-character (Inst. 9.2.30). I argued
above that Paul introduces speech-in-character in 5:3 in order to highlight
the sudden and unexpected coming of the day of the Lord and to provide a
contrast to the preparedness Paul prescribed for the Thessalonian Christians.
Thus, by employing speech-in-character, Paul provides appropriate characters
for a specific rhetorical situation in his exhortation to the Thessalonians.
Additionally, as Paul situated himself and the Thessalonians as mirror opposites
of the speakers of 5:3, the implementation of speech-in-character displays the
inner thoughts of their opponents. 1 Thess 5:3 also fits within Theon’s use of
speech-in-character as a means of exhortation (Prog. 115.20–22; 116.27–117.4);
the unpreparedness for the day of the Lord by the speakers of 1 Thess 5:3, their
hope in Roman “peace and security,” and their impending destruction serve as
an example of what the Thessalonians patently should not imitate.29 Instead,
the Thessalonians must be prepared at all times and place their hope in the
salvation that comes through Christ.
Consequently, Paul’s earliest extant letter—the earliest extant piece
of Christian literature—employs the practice of speech-in-character in
a manner that is wholly within the parameters set forth in the primary
treatments of the exercise. Furthermore, Paul’s implied characterization of
the speakers in 1 Thess 5:3 suggests that Paul’s rhetorical sensibility (to one
degree or another) exceeds that expected by the elementary treatments. Paul’s
implicit and comparative (synkristic) characterization of the speakers vis-à-
vis the Thessalonian Christians satisfies speech-in-character’s requirement of
characterization, yet it does so in a way that evinces fluidity, complexity, and
freedom of expression, since the primary treatments of rhetorical theory only
model direct characterization.

Galatians 3:8
After posing a rhetorical question to the Galatians concerning whether they
received the Spirit ἐξ ἔργων νόμου or ἐξ ἀκοῆς πίστεως (3:5), Paul begins to de-
fend or prove the obvious answer to that rhetorical question—the Galatians

29  In Theon’s words, the goal of 1 Thess 5:3 is a means of dissuasion, or ἀποτρέποντες.
Prog. 117.4.
Examples of Speech-in-Character in Paul 69

received the Spirit ἐξ ἀκοῆς πίστεως. Probably due to the use of Abraham and
Gen 17 by others in Galatia, Paul’s proof begins with Abraham, for whom πίστις,
according to Gen 15:6, culminated in δικαιοσύνη (Gal 3:6). Based on Abraham’s
πίστις-grounded δικαιοσύνη, Paul argues that Abraham’s heirs are therefore
characterized by πίστις (Gal 3:7) and not by ἔργα νόμου (3:10–22). In fact, scrip-
ture itself already knew this to be the case; law was only applicable to Jews, but
Abraham’s heirs would come from, and in him blessing would extend to, all
nations (πάντα τὰ ἔθνη; 3:8).30 Paul writes,

προϊδοῦσα δὲ ἡ γραφὴ ὅτι ἐκ πίστεως δικαιοῖ τὰ ἔθνη ὁ θεός, προευηγγελίσατο

τῷ Ἀβραὰμ ὅτι ἐνευλογηθήσονται ἐν σοὶ πάντα τὰ ἔθνη.

But scripture, foreseeing that God would justify the gentiles ἐκ πίστεως,
proclaimed the gospel in advance to Abraham, “In you, all the nations
will be blessed.”

The take away for Paul is that those characterized as ἐκ πίστεως are blessed
with Abraham (3:9) because of their participation in Abraham’s seed, Jesus
(3:14–16), and their reception of the Spirit (cf. 3:2–5). Those characterized as
ἐξ ἔργων νόμου, however, experience nothing but curse (3:10–14). Blessing and
πίστις are intrinsically intertwined.
The shift out of Paul’s voice is plainly marked; it is scripture (ἡ γραφή) in the
nominative case that proclaims the gospel in advance to Abraham (3:8).31 That
scripture’s speech constitutes direct discourse is established by the second-
person pronoun, σοί, as indirect discourse would require a third-person
pronoun, αὐτῷ. To compose scripture’s speech as recorded in Gal 3:8, Paul com-
bines elements from LXX Gen 12:3 and 18:18.32 That is, Paul chose the use of σοί

30  On the complexity of translating ἔθνη in Galatians, see the discussion in footnote 34.
31  Scholars have recognized Paul’s personification of scripture, but none, to my knowledge,
have analyzed it in light of the conventions of speech-in-character. Bates identifies Gal 3:8
as “a lightweight example of … prosopopoeia,” but he does not comment any further on
this issue. Bates, Hermeneutics, 121n38. See also Richard N. Longenecker, Galatians (WBC
41; Nashville: Thomas Nelson, 1990), 115; J. Louis Martyn, Galatians (AB 33A; New York:
Doubleday, 1997), 300; Martinus C. de Boer, Galatians: A Commentary (NTL; Louisville:
Westminster John Knox, 2011), 194.
32  Gen 12:3 LXX reads, “In you (ἐν σοί), all the tribes (φυλαί) of the earth will be blessed.”
Gen 18:18 LXX reads, “In him (ἐν αὐτῷ), all the nations (ἔθνη) of the earth will be blessed.”
It is possible that, due to scribal harmonization tendencies or diction choices, the LXX
versions of Gen 12:3 and/or 18:18 that Paul was familiar with contained the reading as
he recorded it. Diminishing the likelihood of this possibility, however, is the absence of
70 chapter 5

from 12:3 over the use of αὐτῷ in 18:18, and he replaced 12:3’s φυλαί with 18:18’s
ἔθνη. By crafting scripture’s speech in this way, Paul not only brings scripture’s
declaration into conformity with the way he speaks about τὰ ἔθνη elsewhere
in Galatians,33 but he also clarifies without remainder the relevance of the
scriptural citation for the present rhetorical context, namely, what was spoken
to Abraham had direct relevance not only to Jews but to πάντα τὰ ἔθνη, all the
nations, as well.34
Paul characterizes ἡ γραφή as possessing the foresight that God planned to
justify the gentiles ἐκ πίστεως (Gal 3:8). Because, according to Paul’s argument,
Abraham’s children are οἱ ἐκ πίστεως (3:7), scripture is able to include gentiles
as Abraham’s children and to proclaim in advance the logical declaration that
gentiles too will be blessed in Abraham (3:8), which occurs by participation in
Abraham’s seed, Christ (3:14–16). For these reasons, the speech Paul attributes
to ἡ γραφή is wholly appropriate with the characterization Paul assigns to it,
thereby satisfying speech-in-character’s convention of appropriateness.

any supporting textual witnesses, as well as the fact that 12:3’s use of φυλαί and 18:18’s use
of ἔθνη translate two different Hebrew words in the MT, ‫ משפחת‬and ‫גויי‬, respectively.
Combined with Paul’s interest in τὰ ἔθνη in Galatians, which constitutes an obvious rea-
son for Paul to substitute 12:3’s φυλαί with 18:18’s ἔθνη, the chance that Paul’s LXX Vorlage
witnessed the reading as Paul recorded it seems unlikely.
33  Bates, Hermeneutics, 121 n.39; Christopher D. Stanley, Paul and the Language of Scripture:
Citation Technique in the Pauline Epistles and Contemporary Literature (SNTSMS 74;
Cambridge: Cambridge University Press, 1992), 236–38.
34  R. Longenecker argues that Paul draws a distinction between the two uses of τὰ ἔθνη in
Gal 3:8. Longenecker writes, “The expression τὰ ἔθνη [in 3:8a] means ‘the Gentiles’ whose
righteousness is under question, with πάντα τὰ ἔθνη [τῆς γῆς] to be read more inclusively
as ‘all the nations [of the earth],’” thereby allowing for the inclusion of Jews, and not sim-
ply gentiles, within the scope of τὰ ἔθνη in Gal 3:8b. Longenecker, Galatians, 115. The view
represented by Longenecker that Paul envisions the blessing of 3:8b to include gentiles
and Jews is probable given the discussion about law that follows and, especially, given
its seamless agreement with Paul’s theology of equality among all people, regardless of
race, before God. Though the use in 3:8a may also be inclusive (contra Longenecker),
Longenecker is correct that the rhetorical emphasis in 3:8a weighs on the gentiles. As
Longenecker notes, it is the Galatian gentiles “whose righteousness is under question.”
Ibid. This emphasis is strengthened when one considers the way Paul uses the term, ἔθνη,
up to this point in Galatians; the use in 1:16 is potentially ambiguous, but every other
instance distinguishes the gentiles as separate from Jews (2:2, 8, 9, 12, 14, 15). It is only in
3:8 and 14 that ambiguity/inclusivity reappears. Thus, given Paul’s usage of the term as a
reference strictly to gentiles throughout Gal 2, the substitution from Gen 12:3’s φυλαί to
18:18’s ἔθνη likely emphasizes Paul’s view that gentiles too are included in Abraham’s heirs
and God’s people solely on the basis of πίστις.
Examples of Speech-in-Character in Paul 71

In fact, Gal 3:8 satisfies all five of the primary conventions for speech-in-
character. (1) Paul nuances and attributes speech to another speaker. (2) The
assigned speech is appropriate to the character of the imagined speaker.
(3) The assigned speech is edited from actual speeches that are recorded
within scripture itself. (4) All three accompanying elements are present; Paul
first identifies a new speaker, characterizes the newly introduced speaker, and
assigns appropriate speech-in-character. Finally, (5) Paul’s attribution of speech
to an inanimate object, ἡ γραφή, is sufficiently supported in the discussions of
the rhetorical figure in the primary literature.
In consideration of the secondary elements of speech-in-character, Gal 3:8
most closely aligns with Quintilian’s view that speech-in-character can provide
appropriate characters for specific rhetorical situations. The rhetorical context
of Gal 3:6–9 involves a discussion of Abraham, his heirs, blessing, and πίστις as
a way to address the rhetorical questions posed in 3:1–5. Scripture, therefore,
which preserves the very stories about Abraham Paul references, constitutes
perhaps the most appropriate character imaginable, as (Paul’s presentation of)
ἡ γραφή represents the ultimate authority on all things Abraham.

Galatians 4:6
To stay with Galatians, in his continued explanation of Abraham’s heirs and
who belongs to God’s family, Paul writes, “All of you are children of God διὰ τῆς
πίστεως in Christ Jesus, for as many of you as were baptized into Christ have put
on Christ” (3:26–27). The Galatians can be God’s children because participa-
tion “in Christ” universally “obliterates” ethnic, social, and gender boundaries
(3:28).35 Furthermore, because Christ is Abraham’s seed (3:16), participation in
Christ renders one an heir of Abraham (3:29).
Prior to participation in Christ, however, humanity was enslaved (4:3), but
God “sent his Son” (4:4) on a redemptive mission so that humanity “might re-
ceive adoption” (4:5). Having already demonstrated that the Galatians received
this adoption through participation in Christ (3:26–29), Paul writes in 4:6:

Ὅτι δέ ἐστε υἱοί, ἐξαπέστειλεν ὁ θεὸς τὸ πνεῦμα τοῦ υἱου αὐτοῦ εἰς τὰς καρδίας
ἡμῶν κρᾶζον· αββα ὁ πατήρ.

Because you are children, God sent the Spirit of his Son into our hearts,
crying, “Abba, Father.”36

35  Beverly Roberts Gaventa, “Is Pauline Theology Just a ‘Guy Thing’?” in Our Mother St. Paul
(Louisville: Westminster John Knox, 2007), 68.
36  The logic of Gal 4:6 is not one of dependency in which adoption is a prerequisite for
the sending of the Spirit. To paraphrase, the logic is more to the effect that “you are
72 chapter 5

In sum, God has made the Galatians “no longer slaves but children” (4:7).37
In Gal 4:6, Paul indicates to his audience a shift out of his voice with the ac-
cusative neuter participle κρᾶζον, which can only refer to the neuter πνεῦμα, the
Spirit.38 As such, it is the Spirit that cries, “Abba, Father,” and Paul returns to his
own voice in 4:7 as he again addresses the Galatians (εἶ).39 Paul characterizes
the Spirit as the Spirit that one receives ἐξ ἀκοῆς πίστεως (3:2, 5) and διὰ τῆς
πίστεως (3:14), as the Spirit τοῦ υἱου αὐτοῦ (“of [God’s] Son”),40 and as sent by
God (4:6). In the attributed speech, “αββα” is the Greek transliteration of the
Aramaic ‫אבא‬, and “ὁ πατήρ” is its Greek translation.41 Bruce W. Longenecker
argues that the use of “Father” in prayer to God echoes tradition about Jesus’
own use of the term in prayer,42 and scholars often note that the retention
of the bilingual terms probably stems from its early use among bilingual
communities.43 Thus, Paul is possibly citing a formula common among some
early Christians, but he assigns its declaration to its ultimate, rather than
human, source.
In this sense, what is most interesting is not how Paul shapes the attributed
speech but its appropriateness in light of his characterization of the Spirit.
Paul characterizes the Spirit as received through πίστις (3:2, 5, 14), as sent by
God (4:6), and as the Spirit of God’s Son (4:6). In every way, Paul character-
izes the Spirit in connection with the Son. First, adoption in God’s family is
mediated διὰ τῆς πίστεως ἐν Χριστῷ Ἰησοῦ (3:26). Similarly, the sending of the

God’s children because God sent the Spirit of his Son into our hearts.” Martyn argues
similarly, “For Paul there is no chronological order between adoption into God’s
family and receipt of the Spirit.” Martyn, Galatians, 391n.11. So also Hans Dieter Betz,
Galatians: A Commentary (Hermeneia; Philadelphia: Fortress Press, 1979), 209–10; de Boer,
Galatians, 265; Ben Witherington III, Grace in Galatia: A Commentary on Paul’s Letter
to the Galatians (Grand Rapids: Eerdmans, 1998), 290; James D. G. Dunn, The Epistle
to the Galatians (BNTC; Peabody: Hendrickson, 1993), 219; Richard B. Hays, The Letter to
the Galatians: Introduction, Commentary, and Reflections (NIB 11; Nashville: Abingdon,
2000), 285.
37  Martyn similarly emphasizes God’s role. Galatians, 392.
38  Bruce W. Longenecker, The Triumph of Abraham’s God: The Transformation of Identity in
Galatians (Edinburgh: T & T Clark, 1998), 61; Betz, Galatians, 210; de Boer, Galatians, 266.
39  Rom 8:15 portrays Christians who have received the “Spirit of adoption” uttering this cry.
40  So also de Boer, Galatians, 266.
41  Betz, Galatians, 210–11.
42  Longenecker, Triumph, 61–62. So also Dunn, Galatians, 221–22; Longenecker, Galatians,
174–75; F. F. Bruce, Commentary on Galatians (NIGTC; Grand Rapids: Eerdmans, 1982), 199.
43  Longenecker, Triumph, 62; Betz, Galatians, 211; Witherington, Grace, 291; de Boer,
Galatians, 266; Sigve Tonstad, “The Revisionary Potential of ‘Abba! Father!’ in the Letters
of Paul,” Andrews University Seminary Studies 45.1 (2007), 8–12; Dunn, Galatians, 221.
Examples of Speech-in-Character in Paul 73

Spirit who is recieved διὰ πίστεως (3:14) is coterminous with and affirms one’s
adoption (4:6; cf. Rom 8:15–16). Second, the sending of the Spirit (ἐξαπέστειλεν
ὁ θεὸς τὸ πνεῦμα; Gal 4:6) is grammatically parallel with the sending of the Son
(ἐξαπέστειλεν ὁ θεὸς τὸν υἱὸν αὐτοῦ; 4:4).44 Third, and most significant, Paul char-
acterizes the Spirit specifically as the Spirit τοῦ υἱου αὐτοῦ (4:6). Having been
characterized in connection to the Son at every turn, the Spirit’s cry, “Abba,
Father,” is completely appropriate, as it is the Son who would most naturally
appeal to God as “Father.”45
Though Paul perhaps cites a traditional prayer formula, his use of the
phrase in the Spirit’s voice otherwise adheres to the theoretical conventions
for speech-in-character. To begin, (1) Paul scripts the attributed speech into the
mouth of an imaginary speaker, the Spirit. Additionally, (2) the assigned speech
appropriately models the characterization of the Spirit intimately connected
to the Son. (3) The assigned speech is actual speech. Moreover, (4) all three
structural elements are present; Paul identifies (3:2, 5, 14; 4:6), characterizes
(3:2, 5, 14; 4:4, 6), and attributes appropriate speech-in-character to the Spirit
(4:6). Finally, (5) Paul has not placed the speech into the mouth of a person,
per se, but into the mouth of a personified abstract concept, the divine Spirit.
By attributing speech to an abstract concept, the Spirit, Paul demonstrates a
fuller awareness of speech-in-character such that it is not limited to human
speakers, but that it may also be scripted in the voice of non-human entities.
If one disagrees with the view of the Spirit as a personified abstract concept,
then Quintilian’s treatment of speech-in-character explains Paul’s attribution
of speech to the divine Spirit just as well, since Quintilian allows that one may
attribute speech to the gods (Inst. 9.2.31). Quintilian’s description of the func-
tions of speech-in-character (9.2.30) may also be helpful. It is not difficult to
suppose that the Spirit’s outburst, “Abba, Father,” could have created variety
and especially liveliness in Paul’s epistolary discourse. Also, the introduction of
the Spirit’s cry fills a specific rhetorical need, since it functions as proof of the
Galatians’ adoption into God’s family. Therefore, there is nothing particularly
surprising between the general conventions of speech-in-character and Paul’s
use of the exercise in Gal 4:6.

44  Longenecker, Triumph, 60–61.

45  Dunn writes, “[The experiences] … attributable to the divine Spirit could now be rec-
ognized by the fact that this was the Spirit of the Son, the Spirit whose character was
attested by the character of Jesus … That is to say, the character of Jesus’ sonship pro-
vided the parameters for the experiences which could be attributed to the Spirit.”
Galatians, 220.
74 chapter 5

1 Corinthians 1:12; 3:4

Moving to 1 Corinthians, immediately after the epistolary prescript (1:1–3) and
thanksgiving (1:4–9), Paul begins to exhort the Corinthian Christians to be in
unity with one another (1:10). Paul wants the Corinthians to “say the same thing”
(τὸ αὐτὸ λέγητε), to be free from “divisions” (σχίσματα), and to be “restored in
the same mind and the same opinion” (κατηρτισμένοι ἐν τῷ αὐτῷ νοῒ καὶ ἐν τῇ
αὐτῇ γνῶμῃ). Paul offers such exhortation because a group of “Chloe’s people”
reported to him that there were “contentions” (ἔριδες) among the Corinthians
(1:11). Paul then explains what these divisions and contentions look like on the
ground in Corinth (1:12); Paul writes:

λέγω δὲ τοῦτο ὅτι ἕκαστος ὑμῶν λέγει· ἐγὼ μέν εἰμι Παύλου, ἐγὼ δὲ Ἀπολλῶ,
ἐγὼ δὲ Κηφᾶ, ἐγὼ δὲ Χριστοῦ.

What I am referring to is that each of you is saying, “I am of Paul,” “I am of

Apollos,” or “I am of Cephas.” I, however, am of Christ.

Margaret M. Mitchell also identifies Paul’s implementation of speech-in-

character in 1 Cor 1:12; she writes, “I understand 1:12 as the rhetorical figure
προσωποποιΐα,” citing Quintilian (Inst. 9.2.30, 37), Ad Herennium (4.53, 66),
and Demetrius (Eloc. 5.265–66).46 To be sure, the shift out of Paul’s voice and
into the voice of other speakers is clear, as Paul specifically writes, “Each of
you is saying” (1:12). Thus, the speakers of 1:12b are various members of the
Corinthian congregation to whom Paul is writing and addressing in the second
person plural (ὑμῶν). Paul characterizes these speakers as experiencing “divi-
sions” (σχίσματα; 1:10), “contentions” (ἔριδες; 1:11), and in need of restoration
(κατηρτισμένοι; 1:10).
With respect to the attributed speech, two issues arise. First, does the final
phrase, “I am of Christ,” refer to a fourth group among the Corinthians, or is
it Paul’s prescriptive remedy for the Corinthians’ divisions? The question is
significant because it determines whether the phrase should be read in the
Corinthians’ voice or as the return to Paul’s.47 The parallel structure of all

46  Margaret M. Mitchell, Paul and the Rhetoric of Reconciliation: An Exegetical Investigation

of the Language and Composition of 1 Corinthians (Louisville: Westminster / John Knox
Press, 1992), 86. Mitchell is followed by David E. Garland, 1 Corinthians (BECNT; Grand
Rapids: Baker Academic, 2003), 48. Craig S. Keener also classifies 1:12 as sermocinatio.
1–2 Corinthians (NCBC; Cambridge: Cambridge University Press, 2005), 24–25.
47  Thus, 1 Cor 1:12 constitutes an instance in which the conventions for speech-in-character
can be employed as a way to determine the boundaries of the attributed speech.
Examples of Speech-in-Character in Paul 75

four phrases leads many scholars to continue reading the fourth phrase in the
Corinthians’ voice on the assumption that Paul would not have used the same
grammatical structure if he were interjecting.48 Other scholars like Mitchell,
however, argue that the phrase does not represent a fourth party but Paul’s
commentary, since: (1) Paul does not name the “Christ group” when he names
the others elsewhere (3:4, 22); (2) he uses the phrase “you are of Christ” (ὑμεῖς
δὲ Χριστοῦ; 3:23) as a corrective to the three parties listed in 3:22; and (3) the
challenge of 1:13 (“Has Christ been divided?”) makes little sense if some of
the Corinthians claimed to be “of Christ.”49 Furthermore, the line of exege-
sis which reads the fourth phrase as Paul’s commentary extends back in early
Christian literature at least to Chrysostom (hom. In 1 Cor 3.2 [PG 61.24]).50 For
Mitchell, then, the attributed speech makes best sense if it is analyzed in light
of its larger epistolary context.
In addition to Mitchell’s arguments, one can add that the phrase, “I am of
Christ,” does not appropriately fit Paul’s characterization of the Corinthian
speakers. For Paul, to be in Christ is to be unified and free from divisions
(1 Cor 3:22–23; 12:13–14; Gal 3:26–29).51 The Corinthians, however, are patent-
ly not unified in terms of Paul’s characterization of them (1:10–11). Assuming
Paul’s adherence to the convention for speech-in-character to be appropriate
to the characterization, the fourth phrase’s out-of-character quality serves as
an additional signifier that the Corinthian factions are no longer speaking.52
Thus, despite the grammatical parallelism between the four phrases and the

48  The grammatical pattern is: (1) pronoun, (2) conjunction, (3) actual or implied εἰμι,
and (4) proper noun in the genitive case. For examples of scholars who make this ar-
gument, see C. K. Barrett, The First Epistle to the Corinthians (HNTC; New York: Harper
& Row, 1968), 44–45; Leon Morris, The First Epistle of Paul to the Corinthians (TNTC 7;
Grand Rapids: Eerdmans, 1980), 40. Cf. Hans Conzelmann, 1 Corinthians (Hermeneia;
Philadelphia: Fortress Press, 1975), 33–34; Anthony C. Thiselton, The First Epistle to the
Corinthians: A Commentary on the Greek Text (NIGTC; Grand Rapids: Eerdmans, 2000),
129–33; Charles H. Talbert, Reading Corinthians: A Literary and Theological Commentary
(Reading the New Testament; rev.; Macon: Smyth & Helwys, 2002), 16.
49  Mitchell, Rhetoric of Reconciliation, 82n.101. Other scholars who argue that the fourth
phrase does not refer to a Christ group include Garland, 1 Corinthians, 48–49; Keener,
1–2 Corinthians, 25; J. Paul Sampley, The First Letter to the Corinthians: Introduction,
Commentary, and Reflections (NIB 10; Nashville: Abingdon, 2002), 801, 804.
50  Mitchell, Rhetoric of Reconciliation, 82n.101.
51  Similarly, Sampley writes, “Divisiveness among those who are in Christ is simply unthink-
able for Paul.” The First Letter, 807.
52  Therefore, contra Gordon D. Fee, it is not the case that there is “no signal that there is a
break with the fourth member.” Rather, the signal is only recognized when one under-
stands the common convention of appropriateness for speech-in-character and considers
76 chapter 5

absence of any overt switch back into Paul’s voice (for example, a verb of
speech or strong adversative), evidence and argument favor the fourth phrase
to be spoken in Paul’s voice and not representative of a fourth party.53
The second issue concerns whether Paul quotes actual “slogans” employed by
the various Corinthian parties or invents the phrases himself. Larry L. Welborn
has argued that the phrases represent “a common formula of political self-
identification in antiquity,” which suggests that Paul was quoting slogans used
by the Corinthians.54 Again, Mitchell’s argument is to be preferred. Mitchell
demonstrates that no examples of ancient political slogans cohere with the
structure of the phrases in 1 Cor 1:12.55 Instead, Mitchell identifies the proper
names in the genitive case as “genitives of possession or belonging” or “rela-
tionship,” such that the Corinthian parties are owned or possessed as children
or slaves by the noun in the genitive.56 This reading and the probability that
Paul has crafted these phrases are then confirmed by his further development

the speech in view of its larger epistolary context. Fee, The First Epistle to the Corinthians
(NICNT; Grand Rapids: Eerdmans, 1987), 58–59, 58n.54.
53  Scholars have often commented that if the phrase, “But I am of Christ,” is to be heard
in Paul’s voice, there is “no link between 1:12 and 1:13.” Conzelmann, 1 Corinthians, 33;
Stephen M. Pogoloff, Logos and Sophia: The Rhetorical Situation of 1 Corinthians (SBLDS
134; Atlanta: Scholars Press, 1992), 179n.25. Such a view is hardly necessary; both the
phrase, “but I am of Christ,” (1:12) and the question, “Has Christ been divided?” (1:13) can
serve as a joint response to the factionalism expressed in the first three phrases of 1:12. The
question, “Has Christ been divided,” therefore, assumes that existence in Christ means to
be unified, which Paul proclaims as the ideal, and responds to the factionalism expressed
in the first three phrases of 1:12.
54  L. L. Welborn, “On the Discord in Corinth: 1 Corinthians 1–4 and Ancient Politics,” JBL 106
(1987), 90–93. The quote comes from Mitchell, Rhetoric of Reconciliation, 83. Also arguing
for political resonances in 1 Cor 1:12 is Pheme Perkins, First Corinthians (Paideia; Grand
Rapids: Baker Academic, 2012), 50–52, 54–55.
55  Ibid., 84. Thiselton, The First, 122, follows Mitchell.
56  Mitchell, Rhetoric of Reconciliaion, 85. For genitives of possession or belonging, see
Herbert Weir Smyth, Greek Grammar (Cambridge: Harvard University Press, 1984), §1301.
For genitives of relationship, see BDF §162.7. Anthony C. Thiselton misreads Mitchell; he
writes, “Mitchell insists that this is not a genitive of possession,” and “Mitchell convinc-
ingly argues that the genitives cannot plausibly be construed as genitives of possession.”
Rather, “Mitchell proposes a genitive of relationship.” Thiselton, The First, 121–22, emphasis
original. Quite the contrary, Mitchell argues, “The proper names in the genitive case …
[are] a ‘genitive of possession or belonging,’ defined as follows, ‘With persons the genitive
may denote the relations of child to parent, wife to husband, and of inferior to superior.’”
Mitchell then notes that BDF §162.7 identifies 1 Cor 1:12 and 3:4 as a “genitive of relation-
ship.” Mitchell, Rhetoric of Reconciliation, 85, italics mine. In this discussion based on
Smyth and BDF, Mitchell does not discuss different functions of the genitive but different
Examples of Speech-in-Character in Paul 77

and interpretation of these themes for the Corinthian congregation in

1 Cor 3:1–4; 6:19–20; and 7:23.57
In 1 Cor 3:4, Paul again attributes speech-in-character to the Corinthians.
After characterizing the Corinthians as not “spiritual” (πνευματικοῖς; 3:1), but
as “fleshly” (σαρκίνοις, σαρκικοί; 3:1, 3), as “infants” (νηπίοις; 3:1), as “walking hu-
manly” (κατὰ ἄνθρωπον περιπατεῖτε; 3:3), and as engaged in “zeal and conten-
tion” (ζῆλος καὶ ἔρις; 3:3), Paul attributes to them two of the same phrases from
1:12, writing (3:4):

ὅταν γὰρ λέγῃ τις· ἐγὼ μέν εἰμι Παύλου, ἕτερος δέ· ἐγὼ Ἀπολλῶ, οὐκ ἄνθρωποί
ἐστε;

For, whenever someone says, “I am of Paul,” and another [says], “I am of

Apollos,” are you not acting according to human standards?

Though Paul does not specifically identify the speakers in 3:4, it is clear that
they are the same Corinthian groups from 1:10–12. Here, Paul’s premise is
that, by adhering to specific parties, the Corinthians are acting like “humans”
(ἄνθρωποί; 3:4), which he has previously combined with being childish and not
spiritual (3:1–3).58 Thus, from Paul’s point of view, the phrases he attributes to
the Corinthian speakers perfectly match his characterization of them.
Therefore, in both 1 Cor 1:12 and 3:4, Paul again appropriates the practice
of speech-in-character wholly in keeping with the established conventions.
(1) Paul crafts speech and scripts it in the Corinthians’ voices. (2) The attrib-
uted speech appropriately models the way Paul has characterized the divided
Corinthians. (3) The speech is imagined as actual. (4) All three primary ele-
ments are present—Paul identifies the speakers as “you” Corinthians, charac-
terizes the speakers, and attributes to them appropriate speech-in-character.
Of the secondary elements of speech-in-character, 1 Cor 1:12 and 3:4 most
clearly parallel Quintilian’s proposed function for speech-in-character to dis-
play the inner thoughts of one’s opponents (Inst. 9.2.30), since Paul provides
the views of those against whom he subsequently argues.59

classifications used by different grammars to explain a single function of proper names

in the genitive case.
57  Mitchell, Rhetoric of Reconciliation, 85.
58  Ibid., 96–97.
59  So also Mitchell. Ibid., 86.
78 chapter 5

1 Corinthians 12:3
In 1 Cor 12:1, Paul turns to the topic of spiritual people (and/or gifts), περὶ δὲ
τῶν πνευματικῶν.60 In the course of arguing that the Spirit works through all
Christians, Paul utilizes speech-in-character in 12:3.61 Paul writes:

διὸ γνωρίζω ὑμῖν ὅτι οὐδεὶς ἐν πνεύματι θεοῦ λαλῶν λέγει· Ἀνάθεμα Ἰησοῦς,
καὶ οὐδεῖς δύναται εἰπεῖν· Κύριος Ἰησοῦς, εἰ μὴ ἐν πνεύματι ἁγίῳ.

Therefore, I am making known to you that no one speaking by the Spirit

of God says, “Jesus is accursed,” and no one is able to say, “Jesus is Lord,”
except by the Holy Spirit.

Because it is impossible for someone speaking by the Spirit to confess, “Jesus is

accursed” or “Jesus is a curse,” and because it is only by the Spirit that one can
confess, “Jesus is Lord,” it is the Spirit-enabled confession, “Jesus is Lord,” that
confirms for Paul the Spirit is at work in a person.62

60  That πνευματικῶν may be read as neuter, “things” or “gifts,” is supported by: (1) the neu-
ter τὰ πνευματικά in 14:1, (2) the interchange of πνευματικῶν with χαρισμάτων in 12:4,
which occurs in the neuter in 9:11 and 15:46, and (3) the discussion of charismatic gifts
that runs from 12:4–14:40. Conzelmann, 1 Corinthians, 204. If the masculine, “people,” is
to be preferred, which 1 Cor 12:2–3 perhaps suggests, it is nevertheless the case that the
arena in which the conversation about “spiritual people” occurs is that of the work of the
Spirit and the distribution of spiritual gifts (12:4–11). The discussion of people in 12:2–3 is
Paul’s demonstration that all Christians are spiritual because of participation with the
Spirit (12:3), even if there are a variety of charismatic gifts (12:4). Mitchell, Rhetoric of
Reconciliation, 267.
61  Also noting Paul’s emphasis that the Spirit works in all Christians are Sampley,
1 Corinthians, 941; Perkins, First Corinthians, 147; and Garland, 1 Corinthians, 567, who cor-
rectly writes, “[Paul] is not providing criteria for discerning authentic inspiration … The
confession is a validating sign that one is a Christian inspired by the Spirit, not a touch-
stone to gauge authentic prophetic speech.”
62  Bruce W. Winter treats the phrase ἀνάθεμα Ἰησοῦς quite differently. Rather than seeing
Jesus as the object of the curse, Winter argues that Jesus was “seen as a God who could be
invoked to deliver a curse against particular persons in Corinth” and that some Corinthian
Christians were cursing other members of the community in continuity with their previ-
ous pagan practices. Winter supports his thesis on a number of premises: (1) several an-
cient curse tablets were discovered in Corinth; (2) these curses invoke the name of a deity
to distribute a curse on one’s adversary; (3) three curses omit the verb of cursing; (4) evi-
dence demonstrates that later Christians in Corinth and elsewhere invoke God in curses;
(5) because Jesus conquered the underworld, it would have been easy for a former pagan
to substitute Jesus for the gods traditionally invoked in curses; and (6) such a reading
Examples of Speech-in-Character in Paul 79

The transitions from Paul’s voice into that of another speaker are clearly
marked by οὐδεὶς … λέγει and οὐδεῖς δύναται εἰπεῖν. Paul does not specifically
identify the speaker beyond the general recognition, “no one.” Again, however,
the characterization helps to identify the speaker a bit more fully. The
characterization of the imaginary speaker is someone ἐν πνεύματι θεοῦ and ἐν
πνεύματι ἁγίῳ (12:3). Because Paul considers being in the Spirit to be a universal
reality for Christians (1 Cor 3:16; 6:11, 19; 12:4, 7–13; Gal 3:2–5; 4:6; Rom 5:5; 7:6;
8:4–6, 9–16), the imaginary speaker must be a Christian.63
How, though, do Paul’s attributed speech-confessions measure up in light
of such a pneumatic, Christian characterization? In the first instance, “Jesus
is accursed,” Paul charts a course quite different than any of his previous

makes sense in light of the disunity of the Corinthian church, in that certain members
were vying for priority and power over others. Bruce W. Winter, After Paul Left Corinth:
The Influence of Secular Ethics and Social Change (Grand Rapids: Eerdmans, 2001), 164–83.
 Winter’s reading requires the nominative Ἰησοῦς to be the subject of an implied verb
and the neuter ἀνάθεμα to be an accusative direct object, whereas the traditional reading
understands the neuter ἀνάθεμα as a predicate nominative modifying Ἰησοῦς. Speaking
strictly with respect to grammar, Winter’s reading raises no concerns. Other issues, how-
ever, problematize Winter’s view. First, it is hardly clear that Paul’s intention in 12:3a is
to motivate the Corinthian Christians to stop cursing others. Elsewhere in 1 Corinthians,
Paul speaks quite bluntly about practices he thinks the Corinthians should cease (for
example, 4:14–21; 5:1–2, 6–11; 6:1–11). In 12:3a, however, Paul simply declares that
Christians categorically do not make a particular claim. Second, if 1 Cor 12:3a involves
invoking Jesus’ name to grant a curse, then Paul contradicts and condemns his own prac-
tice, as Paul himself (presumably relying on God’s power) issues anathemas on his op-
ponents, even in 1 Corinthians (16:22; Gal 1:8–9), which Winter notes. Thus, the problem
for Paul is not that a curse is made or the power by which it is made, but the entity that
receives the curse—Jesus. Third, contrary to Winter’s argument, a traditional reading of
12:3 can make fine sense of the διό in relation to the whole of 12:1–3; the progression
hinges on the concept of knowledge. Paul does not want the Corinthians to be ignorant
about spiritual people (12:1). The Corinthians know that they all had “spiritual” experi-
ences in their previous religious systems (12:2). Therefore, Paul is now making known
to them that all Christians are spiritual (12:3) and are vital components in the life of the
community (12:4–31). Furthermore, and fourth, such a reading fits seamlessly within the
context of a contentious Corinthian assembly in which some members are judging the
spiritual gifts and qualities of others, holding some in high esteem but downgrading oth-
ers. Paul’s point is that every Christian participates in the Spirit and is spiritual; none can
be excluded or exalted; all are equal, important, and integral members of the Christian
community.
63  Such evidence undercuts the view that the phrase, “Jesus is accursed,” is intended to be
understood as a non-Christian, Jewish proclamation, since the characterization precludes
the possibility altogether. Cf. Garland, 1 Corinthians, 571.
80 chapter 5

examples and from anything discussed in the primary treatments of the figure.
The theoretical treatments stress over and over that speech-in-character must
be appropriate to the characterization of the speaker, and the previous Pauline
examples meet this criterion. In 1 Cor 12:3a, however, the imaginary speech
poignantly fails to cohere with the characterization of the speaker—how
could a Christian confess that Jesus is anathema? The attributed speech is
altogether inappropriate for the characterization. But Paul has not lost track
of his rhetorical sensibilities. Paul knows full well that the attributed speech
would be absurd for a Christian to proclaim, as evidenced by Paul’s argument
that no Christian (οὐδεὶς ἐν πνεύματι θεοῦ) could ever make such a declaration
(12:3a). By negating what would be absurd for a Christian to proclaim, Paul
effectively confirms the convention that speech-in-character must cohere with
the characterization of the speaker. Though the speech-in-character appears
misguided at first glance, the overall effect tightly conforms to the conventions
and displays Paul’s ability to utilize speech-in-character at (at least) some
level beyond that of the general treatments in the primary sources for the
exercise.
The second attribution of speech-in-character, “Jesus is Lord,” is more to the
point. In Romans, Paul defines the confession, “Jesus is Lord,” as indicative of
Christian existence (10:9). The same is true in 1 Cor 12:3; it is only the Christian,
characterized as ἐν πνεύματι ἁγίῳ, who is able to make such a profession. Thus,
the attributed speech is altogether appropriate for Paul’s characterization of
the Christian speaker.
In these very brief attributions of speech-in-character, Paul adheres to
the conventions for speech-in-character quite well, with one interesting
alteration. (1) Paul crafts and attributes speeches to an imaginary speaker.
(2) Though the second attributed speech appropriately coheres with Paul’s
characterization of the supposed speaker in form and content, the first
example displays a (slightly) heightened level of rhetorical awareness. Paul
could have simply said, “Those without the Spirit say (or might say), ‘Jesus is
accursed,’” or he could have ommitted 12:3a altogether. Instead, Paul sets up an
absurdity and negates it, so that the construction differs from anything found
in the theoretical treatments but nevertheless coheres with and confirms the
convention of appropriateness set forth within them. (3) The first attributed
speech is impossible and therefore hypothetical, as is the second speech since
it is construed as what someone would or would not be able to say. (4) All three
typical structural elements are present; Paul indicates that another person is
speaking, he characterizes the speaker, and he assigns speech to the imaginary
speaker. Finally, (5) though exegesis allows one to identify the speaker with
more precision than Paul concretely indicates, Paul’s use of an unspecified
Examples of Speech-in-Character in Paul 81

speaker agrees seamlessly with the common allowance for speech to be placed
in the mouth of unidentified speakers.
Of the less attested features of speech-in-character, Quintilian’s proposal
that speech-in-character is well suited for introducing characters for a
specific rhetorical context best fits 1 Cor 12:3 (Inst. 9.2.30). Paul introduces
these speeches-in-character in order to address the conflict in the Corinthian
church regarding which members have better or more spiritual experiences.64
Paul’s solution to this problem is to demonstrate that the Spirit is at work in all
Christians (12:3) and that every member of the community serves a roll, even
if the Spirit works in different ways in each member (12:4–11).65 It is, therefore,
through the use of speech-in-character that Paul once again works to reconcile
and unify the Corinthian church. Quintilian’s view that speech-in-character
can also serve to create vividness also applies, as many scholars understand the
phrase, “Jesus is accursed,” to be something of a hypothetical shock treatment
situated as an antithesis to the positive confession, “Jesus is Lord.”66

1 Corinthians 12:15–16, 21
In 1 Cor 12:15–16 and 21, Paul again demonstrates knowledge of speech-in-
character’s allowance for one to assign speech to mute and inanimate objects.67
Following his explanation that the same Spirit, Lord, and God work within
all Christians to distribute gifts for them to use for the community (12:4–11),
Paul illustrates how the differently but inevitably gifted Christians constitute
a cohesive whole by way of an analogy to the body. Just as a physical body is
a single entity with multiple parts that fill diverse roles (12:12, 17–20), so also
individual Christians have been baptized “into one body” (εἰς ἕν σῶμα) and fill
different roles (12:4–11, 13, 27–31). This body of believers, though it is comprised
of many different individuals, should be a unified whole (12:14).68 In order to
depict the absurdity of a divided body, Paul places speech in the “mouths” of
various body parts.69 Paul writes:

64  For the situation in Corinth, see Sampley, 1 Corinthians, 943.

65  Mitchell, Rhetoric of Reconciliation, 267–68.
66  Fee, The First Epistle, 579–81; Sampley, 1 Corinthians, 941; Conzelmann, 1 Corinthians, 204;
Keener, 1–2 Corinthians, 100.
67  Mitchell also recognizes the use of speech-in-character, which she calls personification,
in 12:15–16, 21. Mitchell, Rhetoric of Reconciliation, 159. So also Bultmann, Der Stil, 87.
68  Mitchell, Rhetoric of Reconciliation, 268–69.
69  Keener also notes Paul’s use of speech-in-character; “Paul employs the rhetorical device
prosopopoiia to generate speaking body parts (as if each has its own mouth!) in 12:15–
16 (where they devalue themselves) and 12:21 (where they devalue others).” Keener, 1–2
82 chapter 5

(12:15) ἐὰν εἴπῃ ὁ πούς· ὅτι οὐκ εἰμὶ χείρ, οὐκ εἰμὶ ἐκ τοῦ σώματος, οὐ παρὰ τοῦτο
οὐκ ἔστιν ἐκ τοῦ σώματος; (16) καὶ ἐὰν εἴπῃ τὸ οὖς· ὅτι οὐκ εἰμὶ ὀφθαλμός, οὐκ
εἰμὶ ἐκ τοῦ σώματος, οὐ παρὰ τοῦτο οὐκ ἔστιν ἐκ τοῦ σώματος;

(21) ού δύναται δὲ ὁ ὀφθαλμὸς εἰπεῖν τῇ χειρί· χρείαν σου οὐκ ἔχω, ἢ πάλιν ἡ
κεφαλὴ τοῖς ποσίν· χρείαν ὑμῶν οὐκ ἔχω.

(15) If the foot says, “Because I am not a hand, I do not belong to the body,”
it does not for this reason cease to belong to the body. (16) And if the ear
says, “Because I am not an eye, I do not belong to the body,” it does not for
this reason cease to belong to the body.

(21) The eye cannot say to the hand, “I do not have any need for you,” nor
can the head say to the feet, “I do not have any need for you.”

In each instance, Paul clearly notes the identity of the imaginary speaker; the
foot speaks first (12:15), then the ear (12:16), the eye (12:21a), and finally the head
(12:21b). In keeping with the concept of a unified body, the characterization
of these four speakers is one and the same. Each body part is simultaneously
an individual part and a member of the whole, so that even though there are
many parts, the body is unified (12:12). Inversely, the body does not exist in
and of itself but is constituted of many parts (12:14). Thus, the body is both
greater than the sum of its parts and in need of each and every part (12:17, 19).
Moreover, God has carefully arranged each and every member just as he de-
sired (12:18). The body and the individual parts are simultaneously, and para-
doxically, inseparable and distinct.
Analysis of the attributed speeches-in-character, however, immediately
raises questions. Based on every treatment of rhetorical theory, attributed
speech must agree with or be appropriate to the characterization of the imagi-
nary speaker. Here, however, Paul for the second time (see 12:3a) charts a dif-
ferent course, as he scripts each of the attributed speeches markedly out of
character. How can the various body parts, characterized as inherently unified
in one body, claim that the body is divided or that some members are more
valuable than others (12:15–16, 21)? Such claims would be absurd70 in light of
such characterization. Paul, however, specifically plays on this absurdity as a
comparison to the conflicting Corinthian congregation. Paul knows that the

Corinthians, 103. Beyond this, Keener goes into no further detail. Fee recognizes 12:15–16,
21 as “personification,” 610, 612.
70  So also, Perkins, 1 Corinthians, 149; Fee, The First Epistle, 610.
Examples of Speech-in-Character in Paul 83

interconnected parts of a unified body should never suggest that some parts
do not belong or that some parts are more important than others (12:14, 17–20,
22–26). Nevertheless, this seems to be precisely what is happening among the
Corinthian Christians as it pertains to spiritual gifts.71 Thus, mapping the char-
acter of the divided Corinthians onto the individual body parts, Paul poses
speeches in the body parts’ voices entirely out of character for a unified body
in order to depict vividly and combat the Corinthians’ divisive attitude. Even if
some body part should actually suggest that it or another part does not belong
to the body (12:15–16)72—i.e., even if some Corinthian Christian should think
that she or another member of the congregation did not belong because of
his or her gifts—even then such a claim would never result in that member
ceasing to be a member of the unified community.73 Paul, however, negates
the incongruous declarations that some members have no need of others by
indicating that such a claim would be entirely ludicrous (οὐ δύναται; 12:21).
Consequently, Paul has not created inappropriate speeches-in-character due

71  Conzelmann, 1 Corinthians, 211–16; Barrett, First Epistle, 287–97; Morris, First Epistle, 174–
80; Mitchell, Rhetoric of Reconciliation, 267–70.
72  For Smyth, the construction ἐάν + subjunctive in the protasis with the present indicative
in the apodosis is a present general condition. “The if clause has the force of if ever (when-
ever), the conclusion expresses a repeated or habitual action or a general truth.” Greek
Grammar, §2297, 2335–2337. For Wallace, the construction would either be a third or fifth
class condition; as a third class condition, the sense of the protasis would be merely hypo-
thetical, whereas the sense of a fifth class condition would be a present general condition
in which the condition itself “gives no indication about the likelihood of its fulfillment.”
Daniel B. Wallace, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New
Testament (Grand Rapids: Zondervan, 1996), 689, 689n.9, 696–97.
 Thiselton misrepresents the condition. Thiselton identifies the first protasis simply as
ἐὰν εἴπῃ ὁ πούς and the second as ἐὰν εἴπῃ τὸ οὖς. He then identifies as the first apodosis the
imaginatively spoken phrase, ὅτι οὐκ εἰμὶ χείρ, οὐκ εἰμὶ ἐκ τοῦ σώματος, and the second as ὅτι
οὐκ εἰμὶ ὀφθαλμός, οὐκ εἰμὶ ἐκ τοῦ σώματος. Thiselton, The First Epistle, 1002–3. Thiselton’s
identifications are inaccurate and make little sense, as there is no “then” statement as
he has structured the conditions. The two phrases Thiselton identifies as an apodosis
(12:15b, 16b) are direct speech and are inseparable from the statements that introduce
them (each of which he correctly identifies as (part of) a protasis; 12:15a, 16a). Each
protasis, therefore, includes the introductory statement (12:15a, 16a) and the content
of the attributed speech (12:15b, 16b). Paul’s commentary (12:15c, 16c) constitutes the
apodosis to each condition; the phrase, οὐ παρὰ τοῦτο (“not for this reason;” 12:15c, 16c),
marks the outcome that results from the hypothetical protasis and therefore represents
the “then” statement of the condition.
73  Garland agrees, writing, “No matter what ears and feet might say if they could talk, they
are integrally part of the body.” Garland, 1 Corinthians, 594.
84 chapter 5

to a lack of rhetorical acumen. Rather, in order to meet the needs of the spe-
cific rhetorical context (i.e., Corinthian disunity), Paul tailors the general con-
ventions for appropriate speech-in-character and uses out-of-character speech
as a way to illustrate the absurdity of the Corinthians’ actions and to work to-
wards reconciling the community.74
When measured against the basic conventions, the speeches-in-character
in 1 Cor 12:15–16, 21 reflect nicely the spirit of the rhetorical figure, even though
Paul has flipped the convention for appropriateness completely on its head in
order to suit his rhetorical need. (1) Paul crafts and assigns speeches to imagi-
nary speakers. (2) Though the attributed speeches do not match the charac-
terization of the speaking body parts, Paul recognizes the absurdity of these
contradictory speeches and uses them as an analog for the divided Corinthian
community. In doing so, Paul negates the absurd speeches and thereby dem-
onstrates his maintenance of the convention for appropriateness. (3) The use
of ἐάν in 12:15–16 and the utter impossibility (ού δύναται) of the claim in 12:21
indicate that these speeches are hypothetical or purely imaginative. (4) The
three primary contents are present; Paul introduces the speaking body parts,
characterizes them, and attributes speech to them. Last, (5) Paul draws on the
convention that it is acceptable to attribute speech to inanimate objects.
From the discussion above, it is easy to see that 1 Cor 12:15–16 and 21
fit Quintilian’s allowance that speech-in-character can be quite useful for
introducing characters for a specific rhetorical context (Inst. 9.2.30), as Paul
uses these speakers in such a way that only makes sense in light of the ensuing
Corinthian dispute about spiritual gifts and status within the community. An
additional connection to Quintilian’s treatment is the manner in which Paul
attributes speech to inanimate objects. Quintilian writes, “When we transcend
the bounds of nature, however, the figure can be made less harsh like this: ‘If
(si) my country … were to say to me …’” (Inst. 9.2.32 [LCL, Russell]). One way,
therefore, Quintilian makes an “unnatural” speech-in-character sound less
harsh is to pose it as a hypothetical scenario, “if.” Paul also poses his attribution
of speech to inanimate body parts (i.e., unnatural speakers) as hypothetical in
12:15–16 through the ἐάν + subjunctive conditional construction, presumably
making it a less harsh and more natural attribution of hypothetical speech.
Furthermore, Theon’s view that speech-in-character is effective for exhor-
tation / dissuasion is particularly relevant (Prog. 115.20–22; 116.27–117.4). In
1 Cor 12:15–16, 21, Paul’s goal is to reveal to the Corinthians the absurdity of

74  Though he does not discuss 12:15–16, 21 in terms of speech-in-character, Garland under-
stands the gist of the passage, writing, “It is obvious in a body that no part is autonomous,
but Paul uses the body analogy to turn self-centered vanity upside-down.” Ibid., 595.
Examples of Speech-in-Character in Paul 85

their quarreling over spiritual gifts and to change their attitude and behavior.
That is, Paul is exhorting the Corinthians towards a particular course of action,
namely, not to act like the contradictory and absurd body parts (which they
are currently doing) but to embrace fully the useful and God-designed unified-
diversity among the Corinthian Christians.

2 Corinthians 12:9
2 Corinthians 12:9 contains another example of speech-in-character. In 12:1,
Paul continues the theme of boasting that runs through 2 Cor 10–12 (see also
1:12, 14; 5:12; 7:4, 14; 8:24; 9:2–3). After discussing a vision “someone” experienced
(12:2–4), Paul writes that he will only boast of his weakness (12:5; cf. 11:30), even
though he could boast otherwise (12:6). In fact, Paul writes that he received in
his flesh a “thorn” (σκόλοψ), a tormenting messenger of Satan, so that he would
not be exalted (ὑπεραίρωμαι) by his possible boasts (12:7). Paul petitioned the
Lord three times to remove the thorn (12:8), but the Lord did not respond quite
as Paul hoped. Paul records the Lord’s response in 12:9:

καὶ εἴρηκέν μοι· ἀρκεῖ σοι ἡ χάρις μου, ἡ γὰρ δύναμις ἐν ἀσθενείᾳ τελεῖται.
Ἥδιστα οὖν μᾶλλον καυχήσομαι ἐν ταῖς ἀσθενείαις μου, ἵνα ἐπισκηνώσῃ ἐπ᾽ἐμὲ
ἡ δύναμις τοῦ Χριστοῦ.

And he said to me, “My grace is sufficient for you, for power is perfected
in weakness.” Therefore, I will boast all the more gladly in my weaknesses,
so that the power of Christ might dwell in me.

With the third person singular verb, εἴρηκεν, Paul indicates that another char-
acter, the Lord, has entered the scene, and Paul attributes to him direct speech
(12:8–9). Assessing the appropriateness of the attributed speech, however, is
more difficult because of the complicated composition history of 2 Corinthians.
Scholars continue to debate whether 2 Corinthians is a unified letter or a com-
posite of multiple letters that an editor compiled.75 The issue is significant,
because decisions about the (dis)unity of 2 Corinthians establish specific

75  For arguments in either direction, see Thrall, 2 Corinthians, 5–20, 595; Martin, Second
Epistle, 298–99; Barrett, Second Epistle, 243–44; Hans Dieter Betz, 2 Corinthians 8 and
9 (Hermeneia; Philadelphia: Fortress Press, 1985), 3–36; Roetzel, 2 Corinthians (ANTC;
Nashville: Abingdon, 2007), 24–35; Margaret M. Mitchell, “Paul’s Letters to Corinth: The
Interpretive Intertwining of Literary and Historical Reconstruction,” in Urban Religion in
Roman Corinth, Interdisciplinary Approaches (eds. Daniel Showalter and Steven J. Friesen;
Cambridge: Harvard University Press, 2005), 307–38; Günther Bornkamm, Die
86 chapter 5

boundaries within which one should expect to find the characterization of

an imaginary speaker. The potential partition of interest in this study is that
which separates 2 Cor 10–13 from 1–9. If one considers 2 Cor 10–13 to represent
a letter separate from 1–9, then he or she must limit the examination of the
imagined speaker’s characterization to those chapters. If one accepts the unity
of 2 Corinthians, however, the search for the characterization can involve the
whole letter.
I do not intend to take on or solve the debate concerning 2 Corinthians’
compositional history at this time.76 Though I tend to favor some partition
theory, both views will receive attention here. Confining the search for
Paul’s characterization of Jesus / Christ / the Lord to 2 Cor 10–13, the reader
learns the following details leading up to the attributed speech-in-character:
(1) Christ is gentle and kind (10:1); (2) Paul belongs to Christ (10:7); (3) the Lord
gave Paul authority for building up and not for tearing down (10:8); (4) there
is a gospel about Christ (10:14); (5) the Lord “approves” people (10:18); and
(6) the Lord Jesus is God’s Son (11:31). Despite this handful of information about
the Lord, none of the examples are particularly helpful in terms of grounding
or explaining the words Paul attributes to Jesus in 12:9. In the preceding two
chapters, there is no mention of the Lord’s (or God’s) χάρις or its effects, nor
is there any description of the Lord in terms of antitheses (such as power and
weakness). The antithesis between weakness and power does return in 13:3–4
in reference to Jesus’ crucifixion in weakness, his resurrection by God’s power,
and Christian participation in Christ. To be sure, in light of 13:3–4, one could
retrospectively understand Jesus’ words in 12:9 to mean something like “God’s
perfect power is exercised in instances of weakness,” whereby Jesus’ words
would be quite appropriate to his characterization as one in whose weak-
ness God’s power was effective (13:4).77 The location of 13:3–4 in relation to
12:9 (i.e., following at some distance) would not be the most natural place for

Vorgeschichte des sogenannten Zweiten Korintherbriefes (Heidelberg: Winter, 1961), 162–94;

Talbert, Reading Corinthians, 6–9.
76  In order to assess Paul’s use of speech-in-character in 2 Cor 12:9 completely, such a task
would be required. Given the numerous examples discussed in this chapter, it already
seems evident that Paul was aware of and used the rhetorical figure of speech-in-character.
It has also been demonstrated that Paul was skilled enough to tailor the conventions
of speech-in-character for his rhetorical context and nevertheless maintain its essence.
Thus, if final conclusions regarding all the details of 2 Cor 12:9’s speech-in-character are
slightly out of reach, the plethora of other examples more than sufficiently achieves the
goals set forth in this chapter.
77  One problem with such a reading, however, is that Paul identifies the power of 12:9a not
as God’s power, per se, but as the power of Christ in 12:9b.
Examples of Speech-in-Character in Paul 87

characterization to occur, but it is not impossible when compared to other ex-

amples (i.e., Teles, Fragment I, where the characterization comes at the end
of the dialogue, or Paul’s implicit characterization in 1 Thess 5:3), and it would
certainly be easier to catch on subsequent hearings. It is also possible that Paul
expects his readers to accept his speech-in-character in light of previous com-
munications and without any specific characterization at this time, and that he
only later expands on the relationship between weakness and power in 13:3–4.78
If one accepts the unity of 2 Corinthians, an earlier verse proves interesting,
but it’s distance from 12:9 would raise questions as well. In 2 Cor 8:9, Paul writes,
“For, with respect to the grace (χάριν) of our Lord Jesus Christ (τοῦ κυρίου ἡμῶν
Ἰησοῦ Χριστοῦ), you know that, although he was rich, he became poor for your
sake, so that you might become wealthy by his poverty.” Here, Paul not only
comments about the nature of the Lord’s grace, but he does so specifically by
way of an antithesis between wealth and poverty. What the Corinthians should
know about the Lord’s grace is that it is self-sacrificial and salvific; though
Jesus was rich, he became poor in the incarnation,79 so that the Corinthians
might become rich by receiving salvation (cf. Phil 2:6–11).80 Thus, in a unified
2 Corinthians, the reader has been primed by 8:9 to understand the Lord’s χάρις
in a specific, salvific way, and he or she has been introduced to a Jesus who un-
derstood and experienced antithetical realities. Thusly primed, the attributed
speech in 12:9 could be altogether appropriate; though Paul was experiencing
a tormenting thorn in this life, the Lord’s salvific grace and the hope instilled
by it was all Paul needed, because the Lord’s power was at work in the midst of
Paul’s weakness.
Despite what decision one makes regarding the (dis)unity of 2 Corinthians,
12:9 meets the conventions for speech-in-character, allowing for some flex-
ibility in the characterization. (1) Paul attributes speech to an actual, speci-
fied speaker. (2) In a unified 2 Corinthians, the reader has been prepared for
12:9 by the characterization of 8:9. In a partition letter, the characterization is
either less helpful, assumed to be part of the Corinthians’ prior knowledge,
has been excised in the editorial process, or, more likely, follows the attributed
speech. (3) The attributed speech is portrayed as having actually occurred.
(4) The common elements of an identification of the speaker and an attribution

78  Unfortunately impossible to determine fully, there is also the question of what might have
been lost in the editing and splicing of 2 Cor 10–13 with 1–9.
79  Thrall, Second Epistle, 533–45; Barrett, Second Epistle, 223; Martin, 2 Corinthians, 263; Betz,
2 Corinthians, 62.
80  Thrall, Second Epistle, 534; Betz, 2 Corinthians, 61, 63.
88 chapter 5

of speech-in-character are both present. The details of characterization, how-

ever, are contingent on one’s view of 2 Corinthians’ composition history.81
Concerning the uniquely attested elements of speech-in-character, the
Lord’s response represents the introduction of speech-in-character for a par-
ticular rhetorical context, as it supports Paul’s defense of weakness (Quintilian,
Inst. 9.2.30). Furthermore, Theon suggests that speech-in-character is useful
for consolation (τὸ τῶν παρηγορικῶν λόγων εἶδος; Prog. 115.20–21), and this ap-
pears quite pertinent to the current situation as well. Paul has asked the Lord
for relief from the tormenting thorn (12:7–8). Though the Lord does not re-
move the thorn, he does indicate to Paul an ultimate remedy—salvific grace
and the power that works in weakness (12:9). Thus, the Lord consoles Paul in
his present torment by referring to the hope he has in God’s eschatological
activity. Theon also suggests “appropriate starting places” (ἀφορμὰς οἰκείας)
for the use of speech-in-character in consolation (Prog. 116.27). The Lord’s re-
sponse to Paul’s prayer (2 Cor 12:9) does not suggest any element of necessity
or intentionality,82 nor is there any comment that a greater evil or torment ex-
ists (cf. Theon, Prog. 117.6–13). But the Lord’s response does strike at the import
of Theon’s claim that “one should say that it was beneficial and that nothing
advantageous comes from grief over what has already happened” (Prog. 117.15–
16). That is, though Paul is presently experiencing a tormenting thorn, the
Lord’s response redirects Paul’s concern to the Lord’s future eschatological act
of grace and away from his present grief. As it stands, 2 Cor 12:9 constitutes the
first parallel between Paul and Theon’s use of speech-in-character for consola-
tion, as well as the first parallel with one of Theon’s suggested starting places.

Romans 10:6–8
The final example of speech-in-character to discuss at this time (see also
Chapters 7 and 11) involves Rom 10:6–8. Paul rounds out Rom 9’s discussion of

81  To be sure, Paul’s interactions with the Corinthians while he was physically present in
Corinth could have (probably?) involved discussing material that could be relevant to es-
tablishing a particular characterization of Jesus that would inform our understanding of
the speech-in-character in 2 Cor 12:9. Unfortunately, there is no verifiable way to confirm
what this material was or whether this was in fact the case. As such, the rhetorical context
must be confined to the limits of the text, our only solid evidence.
82  Paul, in his own voice, however, leaves room for both. First, the thorn “was given” (ἐδόθη)
to Paul, suggesting intentionality (12:7). Second, the thorn was given for a specific purpose
(ἵνα), which, though it does not require strict necessity, allows for it (12:7). Frank J. Matera,
II Corinthians: A Commentary (NTL; Louisville: Westminster John Knox, 2003), 282, 284;
Talbert, Reading Corinthians, 152; Keener, 1–2 Corinthians, 240; Sampley, 2 Corinthians,
164–66.
Examples of Speech-in-Character in Paul 89

God’s creation of Israel by noting that (non-Christian) Israel’s failure to obtain

δικαιοσύνη ἐκ πίστεως was not ultimately Israel’s fault. In the race to δικαιοσύνη
between Israel and the nations, God rigged the outcome by placing a stum-
bling block in Israel’s path; God tripped Israel (9:30–33).83 Paul then begins
Rom 10 by reiterating his deep desire for Israel’s salvation (10:1), and he explains
that the problem separating Israel from salvation is that they have zeal for God
but not “according to knowledge” (κατ᾽ ἐπίγνωσιν; 10:2). Rom 10:3 expands on
what this means; Israel was ignorant (ἀγνοοῦντες) of “the δικαιοσύνη of God”
(τἠν τοῦ θεοῦ δικαιοσύνην) and sought to establish its own δικαιοσύνη through
law.84 Again, however, Paul mitigates the blame attributed to Israel by noting
God’s role in Israel’s experience; Israel was not subjected to God’s δικαιοσύνη.85
Romans 10:4, despite its deep and perennial exegetical “ruts,” then highlights

83  Beverly R. Gaventa, “Questions about Nomos, Answers about Christos: Romans 10:4 in
Context,” forthcoming.
84  Rom 10:3 does not specifically state that Israel’s sought after righteousness was through
law, but the following discussion of νόμος in 10:4–5 indicates that this is the case.
85  Paul refocuses the discussion once more, noting that the ultimate problem with Israel is
that “they were not subjected to the δικαιοσύνη of God” (τῇ δικαιοσύνῃ τοῦ θεοῦ οὐχ ὑπετά-
γησαν; 10:3). Jewett’s view of 10:3 is common. Jewett writes, “Ὑποτάσσω in the middle voice
used here implies voluntary submission or subordination of oneself to a superior.” Jewett,
Romans, 618. See also Dunn, Romans, 588–89; Keck, Romans, 247–48; Joseph A. Fitzmyer,
Romans: A New Translation with Introduction and Commentary (AB 33; New Haven: Yale
University Press, 1993; reprint 2008), 583–84; C. E. B. Cranfield, Romans: A Critical and
Exegetical Commentary on the Epistle to the Romans (ICC; 2 vols.; London: T & T Clark,
1975–79), 515; N. T. Wright, Romans, 655; Moo, Romans, 636; Witherington and Hyatt,
Romans, 260. The verb, however, is not technically in the middle voice but the passive,
ὑπετάγησαν. BDAG (ὑποτάσσω 1.b.β) allows the passive verb in Rom 10:3 to carry the pos-
sible middle meaning of “subject oneself,” but it also allows for the full passive sense, “be
subjected or subordinated.” In light of God’s active role in regards to Israel in particular
(Rom 9:6–10:3) and the world as a whole throughout Romans (especially 1, 5, 8, 11:25–32),
the passive sense, as a divine passive, seems most appropriate in 10:3. First, a divine pas-
sive in 10:3 makes better sense of God’s role in tripping Israel in 9:30–33; as Israel’s stum-
bling was caused by God’s jutted out stone (9:30–33), so also Israel’s ignorance was due
to God not subjecting them to his δικαιοσύνη (10:3). Second, such an idea is not foreign
to Romans or Paul. For instance, creation is subjected in Rom 8:20, and the Son will be
subjected to “the one who subjects all things” in 1 Cor 15:28. In this passive sense, Israel
not being subjected to God’s δικαιοσύνη is similar to the notion of God’s “handing people
over” (Rom 1:24, 26, 28). On this treatment of παραδίδωμι, see Beverly Roberts Gaventa,
“God Handed Them Over,” in Our Mother Saint Paul (Louisville: Westminster John Knox,
2007), 113–23.
 In my reading, the participles ἀγνοοῦντες and ζητοῦντες can be taken as adverbial par-
ticiples indicating result. God did not subject Israel to his δικαιοσύνη; as a result, Israel
90 chapter 5

the dividing line between the law and Christ as the means of δικαιοσύνη.86
As such, Paul signifies that Israel was chasing after δικαιοσύνη in all the wrong
places. As a proof for 10:4,87 Rom 10:5–8 suggests that it was the simple act of
chasing that was problematic. Rom 10:5 cites scripture (Lev 18:5) as evidence
that the “law righteousness” Israel was chasing after is a matter of “doing”
(ποιήσας). Rom 10:6–8 contrastingly argues that righteousness based on πίστις
has nothing to do with “doing” by “ascending” or “descending” (i.e., “chasing”);
instead, the proclaimed word about πίστις is already nearby in one’s mouth
and heart and requires no one to retrieve it.
In the course of making his argument, Paul employs speech-in-character.88
In Rom 10:6–8, Paul writes:

(10:6) ἡ δὲ ἐκ πίστεως δικαιοσύνη οὕτως λέγει· μὴ εἴπῃς ἐν τῇ καρδίᾳ σου·

τίς ἀναβήσεται εἰς τὸν οὐρανόν; τοῦτ᾽ ἔστιν Χριστὸν καταγαγεῖν·(7) ἤ· τίς
καταβήσεται εἰς τὴν ἄβυσσον; τοῦτ᾽ ἔστιν Χριστὸν ἐκ νεκρῶν ἀναγαγεῖν.
(8) ἀλλὰ τί λέγει; ἐγγύς σου τὸ ῥῆμά ἐστιν ἐν τῷ στόματί σου καὶ ἐν τῇ καρδίᾳ
σου, τοῦτ᾽ ἔστιν τὸ ῥῆμα τῆς πίστεως ὅ κηρύσσομεν.

(10:6) But, the δικαιοσύνη that is based on πίστις speaks in this way, “Do not
say in your heart, Who will ascend into Heaven?” This means, in order to
bring Christ down. (7) “And do not say, Who will descend into the abyss?”
This means, in order to bring Christ up from the dead. (8) But what does
it say? “The word is near to you in your mouth and in your heart.” This is
the word about πίστις which we preach.89

was (understandably) ignorant of God’s δικαιοσύνη and sought to establish their own
through law.
86  Gaventa, “Questions about Nomos.”
87  Ernst Käsemann, Commentary on Romans (trans. Geoffrey W. Bromiley; Grand Rapids:
Eerdmans, 1980), 284; Tobin, Paul’s Rhetoric, 342.
88  Scholars who recognize 10:6–8 as speech-in-character (or a similar term) include Jewett,
Romans, 625–29; Keck, Romans, 252; Arland J. Hultgren, Paul’s Letter to the Romans (Grand
Rapids: Eerdmans, 2011), 386; Frank J. Matera, Romans (Paideia Commentaries on the
New Testament; Grand Rapids: Baker, 2010), 246–47; Cranfield, Romans, 522; Käsemann,
Romans, 284; Bultman, Der Stil, 87–88; Bates, Hermeneutics, 225–40; Anderson, Ancient
Rhetorical Theory, 236; Stowers, Rereading, 309–10; Tobin, Paul’s Rhetoric, 343–47. Others
who generally recognize the introduction of a second speaker include Gaventa, “Questions
about Nomos,” forthcoming; Luke Timothy Johnson, Reading Romans: A Literary and
Theological Commentary (Reading the New Testament; New York: Crossroad, 1997), 160.
89  Bates argues for a more complex script in 10:6–8. Bates argues, “Not only does Paul make
the Righteousness by Faith the speaker of Deuteronomy 9:4 and 30:14 in Romans 10:6–8,
Examples of Speech-in-Character in Paul 91

The nominative clause, ἡ δὲ ἐκ πίστεως δικαιοσύνη, identifies the subject of the

third-person verb of speech, λέγει, as “the δικαιοσύνη based on πίστις” (10:6).90
The quality of πίστις is Paul’s overriding characterization of δικαιοσύνη, and
he has been discussing (i.e., characterizing) the concept of δικαιοσύνη on the
basis of πίστις throughout the course of Romans, beginning in 1:17.91 Of the
several nuances Paul applies to δικαιοσύνη and πίστις in Romans, one aspect

Paul also assigns another prosopon to the ‘you’ whom the Righteousness by Faith ad-
dresses, and it is this ‘you’ whom Paul understands to be the speaker of Deuteronomy
30:12–13.” Bates defines this “you” as “Paul’s unbelieving compatriots—addressed via a
collective singular—who have been unflatteringly assimilated by Paul to the presumptu-
ous Israelites described in Deuteronomy 9:4.” As such, Rom 10:6–8 is “a speech that it-
self contains reported speech. The Righteousness by Faith is reporting the direct speech
of the addressee, the Presumptuous Person, and it is the Presumptuous Person whom
Paul assigns by way of prosopological exegesis as the speaker of Deuteronomy 30:12–13!”
Consequently, for Bates, Paul speaks Rom 10:6a and 6d, 7c, 8a and 8c, the Righteousness
by Faith speaks 10:6b, 7a, 8b, and the Presumptuous Person poses the questions of 10:6c,
7b. Bates, Hermeneutics, 226, 230, 232, 233–38, emphasis original.
 To clarify, though the contents of ἡ ἐκ πίστεως δικαιοσύνη’s speech in 6c and 7b re-
semble the sentiments of those non-Christians about whom Paul is speaking, there is
little reason to suppose that Paul intends his Roman audience to hear 6c and 7b as a
second layer of speech-in-character. First, Paul specifically states that ἡ ἐκ πίστεως δικαι-
οσύνη is speaking (λέγει), but he never suggests that a new speaker enters the scene at
6c or 7b. Second, without the questions of 6c and 7b, it would be unclear what ἡ ἐκ πί-
στεως δικαιοσύνη wishes to instruct its audience regarding δικαιοσύνη, πίστις, and νόμος.
The contents of 6c and 7b are the necessary object clauses that complete ἡ ἐκ πίστεως
δικαιοσύνη’s μή + subjunctive prohibition. Third, the argument in Rom 10:4–8 is not
about non-Christian Israel per se, though it is of course connected. The thrust of 10:4–8
is more narrowly to demonstrate how the τέλος νόμου is Χριστός, resulting in δικαιοσύνη
for everyone characterized by πίστις (10:4). In this argument, Paul cites Lev 18:5 to estab-
lish that law-righteousness depends on “doing” (Rom 10:5). Paul then places other texts
from the law, Deut 9:4 (8:17) and 30:12–14, into the mouth of ἡ ἐκ πίστεως δικαιοσύνη
to demonstrate that the law itself also affirms δικαιοσύνη on the basis of πίστις (10:6–8).
So, Rom 10:6–8 has less to do directly with Israel and much more with the question of
what νόμος endorses as a mediator of δικαιοσύνη. In this way, Rom 10:6–8 supports Paul’s
claim in 10:4 by indicating how νόμος itself allows that δικαιοσύνη is based on πίστις (and
not “doing”). Therefore, the questions in Rom 10:6 and 7 make little sense if they are dia-
logically scripted in the voice of Paul’s contemporary, non-Christian compatriots. Instead,
ἡ ἐκ πίστεως δικαιοσύνη speaks these words as a scriptural proof from the law to support
Paul’s thesis in 10:4.
90  Similarly, Jewett argues, “In view of the speech-in-character that Paul employs here,
the δέ that opens verse 6 indicates a change of speaker from Moses to the personified
Righteousness by Faith.” Jewett, Romans, 625.
91  Ibid.
92 chapter 5

stands above the rest as particularly relevant to Rom 10:6–8; πίστις-based

δικαιοσύνη has nothing to do with works or “doing.”92 In 3:21–22, God’s δικαιοσύνη
is made manifest not by νόμος (i.e., “doing;” cf. 10:5), but through Jesus’ πίστις.
Romans 4:3–6 stipulate that it is πίστις that results in δικαιοσύνη, and 4:9–10
reject the idea that Abraham’s circumcision functioned as a work producing
δικαιοσύνη. Rom 4:11 even informs that Abraham’s circumcision was simply a
“sign” or “seal” of the πίστις-based δικαιοσύνη he already possessed. Romans 4:13
then distinguishes between νόμος and δικαιοσύνη πίστεως as the means
through which God’s promise to Abraham was effected. Additionally, Rom 5:17
identifies δικαιοσύνη as a gift (τὴν περισσείαν τῆς χάριτος καὶ τῆς δωρεᾶς τῆς
δικαιοσύνης) and thus not a self-attainable goal. Interestingly, δικαιοσύνη and
πίστις language is wholly absent in Rom 7:7–12, 13b–24, 25b’s discourse of the
agonized “I” who “does” what he or she does not want to do but cannot “do”
what he or she wants (7:15–21).93 Finally, πίστις based δικαιοσύνη is that which
the nations obtained without even pursuing it (9:30; see also my discussion of
2:12–16, 25–29 in Chapter 10). Thus, the attribution of speech which rejects any
requirement of “doing” and affirms “πίστις” as the mediator of δικαιοσύνη is
altogether appropriate to the characterization of δικαιοσύνη πίστεως Paul de-
velops throughout Romans.
The attributed speech is not entirely a Pauline creation but an adaptation.
The attributed speech in Rom 10:6–8 is Paul’s recasting of Deut 9:4 (cf. 8:17) and
30:11–14. Rom 10:6’s introductory line, μὴ εἴπῃς ἐν τῇ καρδίᾳ σου, is verbatim with
LXX Deut 9:4a (cf. 8:17a). Jumping to LXX Deut 30, Moses exhorts the Hebrews
that the commandment (ἐντολή) is neither difficult or distant (μακρὰν ἀπὸ σοῦ;
30:11). The commandment is not in heaven, so that you Israelites say, “Who
(τίς) among us will ascend (ἀναβήσεται) into heaven (εἰς τὸν οὐρανόν) and bring
it to us, and when we hear it we will do it?” (30:12)? Nor is the commandment
on the other side of the sea, so that you say, “Who among us will cross over
to the other side of the sea and bring it back for us and make it heard among
us, and we will do it?” (30:13). Instead, “the word (τὸ ῥῆμά) is very near to you
(ἐγγύς σου) in your mouth (ἐν τῷ στόματί σου) and in your heart (ἐν τῇ καρδίᾳ
σου) and in your hands (ἐν ταῖς χερσίν σου) in order to do it” (αὐτὸ ποιεῖν; 30:14).94

92  Gaventa, “Questions about Nomos,” forthcoming. Similarly, Dunn writes, “What Paul is
objecting to throughout this letter is not the law or the commandment as such, but the
law and the commandment understood in terms of works” (i.e., doing). Dunn, Romans, 613.
93  In Rom 7, the only cognate, δικαία, refers to the law and not the “I” (7:12).
94  Deuteronomy’s text, therefore, is an example of speech-in-character inside another
speech-in-character. The first degree of speech-in-character is the attribution of speech
to Moses (Deut 29:2–30:20). The second-level of speech-in-character is that which Moses
Examples of Speech-in-Character in Paul 93

Paul, however, significantly reivses Deuteronomy for the present context. To

begin, Israel no longer speaks; it is ἡ ἐκ πίστεως δικαιοσύνη that speaks in Rom
10:6–8, which Paul sets in juxtaposition with the way Moses characterizes τὴν
δικαιοσύνην τὴν ἐκ τοῦ νομοῦ as a matter of “doing” (10:5).95 Paul also adjusts the
depiction of the second phrase from the imagery of crossing the sea to that of
descending into the abyss (10:7; cf. LXX Ps 106:26). In this way, Paul creates a
vertical rather than horizontal depiction of the “doer’s” attempt at δικαιοσύνη,
which he describes as an attempt to bring Christ down or raise Christ up from
the dead. Such “doing” is altogether misguided; Christ has already come down
and been raised up (cf. Phil 2:6–11), and “the word of πίστις” is already present
(10:8).96 Additionally, Paul omits the references in Deut 30:12, 13, and 14 to the
concept of “doing” (ποιέω; Rom 10:6, 7, and 8, respectively). Such omissions
play an integral role in Paul’s argument, since he is specifically addressing
the opposition between πίστις and human “doing” as a route to δικαιοσύνη.97

ascribes to Israel. The nominative masculine participle (λέγων) and the first person plu-
ral pronouns (ἡμῖν) require the subject to be Israel or, more generally, “you” and not the
nominative feminine commandment (ἐντολή).
95  On the question of whether the relationship between Rom 10:5 and 10:6 is connective or
adversative, see Jewett, Romans, 625; Dunn, Romans, 613; Thomas R. Schreiner, Romans
(BECNT 6; Grand Rapids: Baker Academic, 1998), 551–54; Moo, Romans, 645–46; Bates,
Hermeneutics, 226–29; Richard B. Hays, Echoes of Scripture in the Letters of Paul (New
Haven: Yale University Press, 1989), 73–83; Stowers, Rereading, 308–10; Elliott, Rhetoric,
266–67, 267n.2; Tobin, Paul’s Rhetoric, 343–47; Francis Watson, Paul and the Hermeneutics
of Faith: London: T & T Clark, 2004), 330–41.
96  Jewett, Romans, 625–28; Keck, Romans, 251–53; Barrett, Romans, 185–86; Moo, Romans,
655–56.
97  Contra Bates, who argues that “the argument that Paul was theologically motivated to
omit this phrase is cogent only with regard to popular English translations that cannot
mark the difference between the feminine and the neuter, not on the basis of a careful
inspection of the Greek text … in Deuteronomy 30:14 the ‘in order to perform it (αὐτό)’
is in the neuter, and refers to the ‘the utterance’ (τὸ ῥῆμα) of 30:14, which Paul goes on to
gloss as ‘the utterance of faith’ (τὸ ῥῆμα τῆς πίστεως). Paul’s omissions of the performance
theme throughout Romans 10:6–8 do not appear to be theologically motivated, since
it would be unproblematic for the Presumptuous Person to favor doing the command-
ment and equally unobjectionable for the Righteousness by Faith to favor enacting the
utterance from within the dramatic world.” Bates, Hermeneutics, 237–38, see also 234n.31.
Bates’s argument contains several weaknesses. First, Bates makes much ado about
Deut 30:11–14’s references to the feminine “commandment” (ἐντολή, αὐτήν; 30:11–13) and
the neuter “utterance” (τὸ ῥῆμα; αὐτό; 30:14). In context, though, τὸ ῥῆμα is a synonymous
reference to ἡ ἐντολή. Though Bates’s argument could stand in relation to Rom 10:6–7,
if Paul retained the phrase “and in your hands in order to do it” from Deut 30:14 in
Rom 10:8b, his argument would crumble. To retain the phrase would contradictorily
94 chapter 5

Furthermore, Paul offers alternative commentary on each phrase of the

attributed speech.98 The metaphorical travels are no longer conceived of as a
way to grasp the commandment but Christ (10:6–7). Finally, 10:8 reiterates that
it is not the commandment that is near in one’s heart and mouth but “the word
of πίστις.” In these ways, Paul does not simply repeat the words of Deuteronomy
in another speaker’s voice; he reshapes those words to fit his rhetorical needs
and the voice of the imagined speaker, ἡ … ἐκ πίστεως δικαιοσύνη.
In Rom 10:6–8, therefore, all primary expectations for speech-in-charac-
ter are present. (1) Paul crafts speech and assigns it to an imaginary speaker.
(2) The attributed speech appropriately matches the character of the speaker.
(3) The attributed speech is actual speech; it is what δικαιοσύνη is imagined as

affirm the notion that ἡ ἐκ πίστεως δικαιοσύνη is correlative with human “doing,” the very
thing Paul has been denying throughout Romans. Second, even if one grants Bates’s argu-
ment that “Righteousness by Faith” attributes speech-in-character to “the Presumptuous
Person” in 10:6c and 7b, that certain characters could have reasonably affirmed what Paul
has omitted about the commandment and the utterance does not necessitate the con-
clusion that Paul’s omissions were not theologically motivated. One could argue all the
more that though certain characters could have affirmed certain claims about doing the
commandment or utterance, the fact that they do not make such affirmations is quite sug-
gestive of Paul’s agenda. What Bates has momentarily missed is the substance of Paul’s
juxtaposition in these verses. The juxtaposition is not between νόμος (or ἐντολή) and πίστις
but between πίστις and “doing” (ποιέω), and Paul is working to obliterate completely the
concept of human “doing” as a mediator of δικαιοσύνη. Rhetorically, one might very well
expect Paul to work towards this theological goal by removing completely the affirmation
of “doing” from any and all participants in the dialogue, which sets 10:6–8 over against the
“doing” of 10:5. Third, Paul’s argument about δικαιοσύνη is categorically and unavoidably
theological. The omission (three times) of the concept of “doing” is best understood as
Paul’s attempt to dissociate δικαιοσύνη from human “doing” as much as possible.
98  On this point, Jewett is inconsistent. At first, Jewett agrees that 10:6c, 7b, and 8c represent
“Paul’s remarks in clarification” of δικαιοσύνη’s words. Later, however, Jewett argues, “In the
style of Jewish pesharim, the character called Righteousness by Faith comments on each
phrase of the Deuteronomic citation, beginning with the traditional formula τοὺτ᾽ ἔστιν,”
i.e., 10:6c, 7b, and 8c. Jewett, Romans, 625, 626. Obviously, the speaker cannot be both Paul
and δικαιοσύνη. Jewett overlooks two details that demonstrate the whole of 10:6b–8 does
not constitute the speech of δικαιοσύνη. First, 10:8a, ἀλλὰ τί λέγει, indicates an undeniable
return to Paul’s voice as it is clearly a reference to δικαιοσύνη (and, therefore, cannot be
spoken by δικαιοσύνη). Second, 10:8c, τοῦτ᾽ ἔστιν τὸ ῥῆμα τῆς πίστεως ὅ κηρύσσομεν, cannot
be spoken in the voice of the singular δικαιοσύνη; 10:8c must be spoken in the voice of Paul
and his cohort, as evidenced by the 1st plural κηρύσσομεν. Given that the τοῦτ᾽ ἔστιν phrase
in 10:8c must be spoken in Paul’s voice, I likewise take the τοῦτ᾽ ἔστιν phrases in 10:6c and
7b to be spoken in Paul’s voice. So also Bates, Hermeneutics, 234; Watson, Hermeneutics of
Faith, 336–37; Campbell, Deliverance, 798–97.
Examples of Speech-in-Character in Paul 95

actually saying. (4) The three major components are present: identification,
characterization, attribution of speech-in-character. Finally, (5) for the second
time,99 Paul places speech on the lips of an abstract concept, δικαιοσύνη. Of the
secondary aspects of speech-in-character, Rom 10:6–8 most closely resonates
with the use of the figure to introduce characters for a specific rhetorical
context (Quintilian, Inst. 9.2.30). Paul is constructing an argument about
the opposition between πίστις and human “doing” as a means to δικαιοσύνη.
Introducing ἡ δὲ ἐκ πίστεως δικαιοσύνη and presenting its rejection of human
achievement as a valid avenue to δικαιοσύνη, even arguing from the law itself,
emphasizes Paul’s claim all the more.

Conclusion

This chapter analyzed select texts that demonstrate Paul’s use of speech-in-
character. The previous examples do not comprise an exhaustive list of attrib-
uted speech in Paul’s letters but are offered as representative. Through engaging
these texts, a number of conclusions are deduced. First, five of the seven undis-
puted Pauline letters contain speech-in-character (1 Thessalonians, Galatians,
1 Corinthians, 2 Corinthians, and Romans).
Second, Paul is altogether able to match the conventions for speech-in-
character set forth in the treatments of rhetorical theory. Though most of Paul’s
examples cohere in one way or another to the conventions set forth, a num-
ber of them monotonously agree with the elementary conventions. The in-
stances that agree with the conventions at the most basic level are: Gal 3:8; 4:6;
1 Cor 1:12; 3:4; 12:3c; and Rom 10:6–8 (see also 1 Cor 10:28 and 15:35).100
Third, Paul was also capable of transcending the basic conventions and tai-
loring the form of his speeches-in-character for specific rhetorical goals, all the
while maintaining the central essence of the figure. To begin, Paul can apply
speech to specific people (1 Cor 1:12; 3:4), types of people (see 2 Cor 10:10), or
unspecified people (1 Thess 5:3; 1 Cor 10:28; 12:3; 15:35). Paul, however, can also
attribute speech to abstract concepts (Gal 4:6; Rom 10:6–8), inanimate objects
(Gal 3:8; 1 Cor 12:15–16, 21; see also Rom 9:20), or the gods (perhaps Gal 4:6;
2 Cor 12:9). Furthermore, Paul can also forego any direct characterization of the
imaginary speaker, supplying instead implicit characterizations for the reader
or auditor to deduce accordingly (1 Thess 5:3; 2 Cor 10:10). In terms of appropri-
ateness, on three occasions Paul scripts speeches altogether out of character

99  See the section on Gal 4:6.

100  For discussions of 1 Cor 10:28; 15:35; 2 Cor 10:10, see King, “Speech-in-Character,” 436–48.
96 chapter 5

(1 Cor 12:3b; 12:15–16, 21; cf. Rom 9:20). On each of these occasions, however,
Paul negates the out-of-character speeches and, as a result, brings these exam-
ples of speech-in-character back into conformity with the criterion of appro-
priateness. Paul’s ability to manipulate and tailor speeches-in-character to suit
his rhetorical needs in each of these ways strongly demonstrates Paul’s comfort
and tact with the figure to be advanced beyond that of an elementary level.
Fourth, Paul’s use of speech-in-character parallels many of the secondary
elements proposed in the theoretical treatments. Most notably, Quintilian’s
three suggested functions of speech-in-character seem quite relevant to
Paul’s use of the figure. To begin, on numerous occasions, it was easy to rec-
ognize how Paul introduced speech-in-character in order to address specific
needs within the rhetorical context (1 Thess 5:3; Gal 4:6; 1 Cor 12:3; 12:15–16, 21;
2 Cor 12:9; Rom 10:6–8; see also 1 Cor 10:28; 15:35; 2 Cor 10:10; Rom 9:20).101
Quintilian’s allowance that speech-in-character is useful for illustrating an op-
ponent’s thoughts also finds affirmation in Paul’s letters (1 Thess 5:3; 1 Cor 1:12;
3:4; see also 1 Cor 15:35; 2 Cor 10:10). And, though one might argue that every
example of speech-in-character creates vividness and variety, such a function
was easily recognizable in texts like Gal 4:6; 1 Cor 12:3; and 12:15–16, 21. The final
point of connection between Paul’s use of speech-in-character discussed so
far and Quintilian involves Paul’s attempt to express the unnatural speech of
inanimate body parts as naturally as possible through the use of hypothetical
conditions (1 Cor 12:15–16, 21).
Elements from Theon’s Progymnasmata appear in Paul’s attributions of
speech, too. Theon informs that speech-in-character is useful for exhortation
or dissuasion, and these uses mirror several of Paul’s examples (1 Thess 5:3;
1 Cor 12:15–16, 21; see also 1 Cor 10:28). Additionally, Paul even approaches one
of Theon’s suggested starting places for the use of speech-in-character as a
means of consolation (2 Cor 12:9).
The primary unique element from Hermogenes concerning the chronologi-
cal structure of speech-in-character, however, does not appear in any of the
discussed examples. This is no doubt due, at least in part, to the very short
nature of most of the examples.

101  The generous degree to which Paul’s use of speech-in-character overlaps with Quintilian’s
suggestion that speech-in-character can be useful for introducing characters into a spe-
cific rhetorical context cannot be taken as weighty proof that Paul was intimately aware
of Quintilian. The very general nature of Quintilian’s comment allows that speech-in-
character can be used in any number of rhetorical contexts as needed. Additionally, Paul’s
letters, understood as circumstantial letters addressing particular historical situations,
cannot help but to use speech-in-character in precisely this way.
Examples of Speech-in-Character in Paul 97

In conclusion, it is altogether clear that Paul knows how and attributes

speech to other speakers in his letters. Moreover, Paul’s thorough agreement
with the conventions for speech-in-character (especially when he manipulates
them for his rhetorical needs) is quite suggestive not only that Paul was well
aware of them in one form or another, but also that his tact with the figure was
at least somewhat advanced when compared to the elementary treatments.
Paul’s agreement with the rhetorical conventions stands firm, regardless of the
degree to which Paul was or was not formally trained in Greco-Roman rhet-
oric.102 This sentiment is further supported by Paul’s usage of the secondary
aspects of speech-in-character, of which Paul employed many, and the exami-
nation of Rom 3:1–9 in Part 3 continues to build on this presentation of Paul’s
abilities with the figure.
At this point, it is time to shift away from speech-in-character in its own
right and towards the concept of diatribe and its dialogical elements. After
Chapter 6’s discussion of secondary literature on diatribe, Chapter 7 brings
speech-in-character back into the discussion as an additional means of ana-
lyzing diatribe’s dialogical sections. The combination of diatribe and speech-
in-character will, in Part 3, be particularly insightful for determining the script
of the dialogue that takes place in Rom 3:1–9.

102  Though it diverges from this project’s sentiments, for an argument suggesting Paul was
not formally trained in rhetoric and for relevant bibliography, see Ryan S. Schellenberg,
Rethinking Paul’s Rhetorical Education: Comparative Rhetoric and 2 Corinthians 10–13
(Society of Biblical Literature: Early Christianity and Its Literature 10; Atlanta; Society of
Biblical Literature, 2013). See also Anderson, Ancient Rhetorical Theory and Paul.
part 2
Diatribe

∵
Introduction to Part 2

More than a century ago, Rudolf Bultmann directed the attention of Pauline
scholars to elements of Greco-Roman diatribe in Paul’s letters, most notably
1 Corinthians and Romans.1 Almost seven decades later, Stanley K. Stowers
reassessed this issue, emphasizing the dialogical portions of diatribal and
Pauline texts.2 Part 2 of this project, “Diatribe,” advances this trajectory.
Chapter 6, “Portrayals of Diatribe,” examines (particularly New Testament)
scholarly engagement(s) with diatribe beginning with Bultmann. The result is
a summation of the current state of the topic on diatribe and Paul among New
Testament scholars that analyzes strengths, weaknesses, and developments
along the way. Generally speaking, and though other voices must be heard,
the consensus is that Stowers’s presentation of diatribe supersedes Bultmann’s
earlier analysis. The nuanced portrayal of diatribe that emerges from this sur-
vey may then be used as a hermeneutical lens through which to view Paul’s
writings that evince diatribal features, culminating (in Part 3) with a detailed
engagement with Rom 3:1–9.
Chapter 7, “Examples of Diatribal Dialogue,” serves two primary roles.
It explores the use of dialogue by two writers other than Paul that scholars
generally agree employ diatribe and provide the foundation for our under-
standing of the practice in antiquity. It also analyzes each of these diatribal
dialogues in light of the conventions for appropriate speech-in-character.
In this way, the hermeneutical lens of Greco-Roman diatribe is largely
employed to shape the meaning and import of a given text, while the rhetori-
cal conventions for speech-in-character carry the weight in terms of defining
the boundaries of the various dialogical exchanges within the script of that
text. Said otherwise, speech-in-character’s conventions determine who
speaks which lines in a dialogue, and the characteristic tones, goals, and tra-
jectories of diatribe indicate how that dialogue should be understood within
its larger argumentative context. By blending these two features, this proj-
ect offers methodological progress as it allows both to exercise collaborative
exegetical influence.

1  Bultmann, Der Stil.

2  Stowers, Diatribe.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_008

 chapter 6

Portrayals of Diatribe

As indicated above, Bultmann constitutes the starting place for consideration

of Greco-Roman diatribe and its relation to the Pauline literature, and his para-
digm held sway for nearly seventy years.1 With the work of Stowers, however,
many scholars began to reconfigure their understandings of diatribe and to
align more closely with Stowers’s model. In the aftermath of Stowers’s work on
diatribe, although many simply adopt Stowers at face value, there have been
a number of publications that have continued to challenge or at least nuance
views of diatribe. These scholars include Abraham J. Malherbe,2 Thomas
Schmeller,3 Runar M. Thorsteinsson,4 Thomas H. Tobin,5 and Changwon
Song.6 The present task is to survey and analyze the ways these scholars por-
tray diatribe, taking them up in chronological order. How these scholars (and
others) interpret Paul vis-à-vis their respective understandings of diatribe are
discussed in Part 3.
Though Bultmann was the first to contribute an in-depth study of diatribe
and its relevance for understanding Paul’s letters, he was obviously not the first
to think about diatribe. Many of Bultmann’s views were influenced by the pre-
vious one hundred thirty years of scholarship on the subject.7 Bultmann8 set
out to demonstrate that Paul’s letters shared a particular affinity with the well-
known Gattung of the diatribe, though this was not due to any conscious effort

1  Bultmann, Der Stil.

2  Malherbe, “Μη Γενοιτο.”
3  Thomas Schmeller, Paulus und die “Diatribe:” Eine vergleichende Stilinterpretation (NTAbdh
19; Münster: Aschendorff, 1987).
4  Runer M. Thorsteinsson, Paul’s Interlocutor in Romans 2: Function and Identity in the Context
of Ancient Epistolography (ConBNTS 40; Stockholm: Almqvist & Wiksell, 2003).
5  Tobin, Paul’s Rhetoric.
6  Song, Reading Romans.
7  For analyses of these earlier works, see King, “Speech-in-Character,” 126–29; Stowers, Diatribe,
7–17.
8  Stowers’s survey of Bultmann’s treatment of the dialogical character of the diatribe is helpful
and similar to what I offer here. This survey also covers Bultmann’s treatment of the tone and
mood of diatribe. Stowers, Diatribe, 17–25.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_009

104 chapter 6

on Paul’s part.9 This connection was significant for Bultmann for two reasons.
Bultmann affirmed the earlier consensus that the diatribe was a form of oral
street preaching.10 Bultmann was also convinced that Paul’s letters were repre-
sentative of his preaching style. By showing the connections between diatribe
and Paul’s letters, Bultmann thought he could simultaneously form a better
understanding of Paul’s preaching style.11
Bultmann’s sources for diatribe were: (1) Bion, (2) Teles, (3) Horace,
(4) Seneca, (5) Musonius Rufus, (6) Epictetus, (7) Dio Chrysostom and
(8) Plutarch.12 Like some earlier scholars, Bultmann considered Bion to be
the earliest source for diatribe, and he held that Teles followed Bion’s model.13
Epictetus later recapitulated the early Bionean model, for which reason
Bultmann prioritized Bion and Epictetus as the principal sources for diatribe.14
In Bultmann’s model, diatribe took on “der Form von Rede und Gegenrede.”15
Such speech and reply were often created by the direct interjection of a fic-
titious opponent who objected to something the primary speaker had just
stated. The objecting interlocutor usually posed a question introduced with
ἀλλά, and the interlocutor’s direct speech was systematically, “in der Regel,”
introduced by a short introductory formula (φησί or inquit, for example).16
Bultmann noted that the speaker could respond to the objector in a number
of ways, such as by issuing a counter-question or a “blow by blow” string of

9  Bultmann writes, “Die vorliegende Arbeit sieht ihre Aufgabe also in dem Nachweis, daß die
paulinischen Briefe Verwandtschaft mit einer bestimmten literarischen Gattung zeigen. Es
ist dies, wie man schon seit langem erkannt hat, die Gattung der Diatribe.” Bultmann, Der
Stil, 2–3.
10  Amongst others, cf. Ulrich von Wilamowitz-Moellendorff, “Excurs 3: Der kynische
Prediger Teles,” in Antigonos von Karystos (Philologische Untersuchungen IV; Berlin:
Weidmannsche Buchhandlung, 1881), 292–319.
11  Bultmann, Der Stil, 3.
12  Ibid., 6–9. Stowers inadvertently omits Seneca from Bultmann’s list. Stowers, Diatribe, 19.
13  Cf. Herman Usener, Epicurea (Berlin: Teubner, 1887); Otto Hense, Teletis reliquiae
(Tübingen: Teubner, 1889; 2nd ed., 1909).
14  Bultmann, Der Stil, 6–9. For Bultmann, the high degree to which the other sources were
steeped in rhetoric or had been altered in the course of transmission resulted in the view
that, while they were still valuable within reason, they were less important than Bion and
Epictetus.
15  Ibid., 10. Stowers correctly notes that Bultmann over exaggerates when he limits diatribe
so narrowly to cycles of speech and reply. Stowers, Diatribe, 20n.105.
16  Bultmann, Der Stil, 10–11. Cf. Rudolf Hirzel, Der Dialog: Ein literarhistorischer Versuch
(2 vols.; Leipzig: S. Hirzel, 1895), 1.371, see also 2.250. Bultmann was mistaken concerning
the degree to which the use of such an introductory formula can be considered a Regel.
Rom 3:1–9 represents a case in point, as it contains no such phrases.
Portrayals of Diatribe 105

questions, or by entering into dialogue with the opponent.17 Additionally,

it was not the case that the imaginary interlocutor must initiate the dialogue
by objecting to the speaker, but the speaker could also initiate dialogue by
addressing the interlocutor.18
According to Bultmann, the identity of the interlocutor usually remained
unspecified. The interlocutor was often classified as an ἰδιώτης and consid-
ered a representative of the general audience (or communis opinio) or even a
member of an opposing philosophical school.19 Personifications (such as Law,
Nature, the Fatherland, and the Virtues) appeared and conversed with the
speaker or the audience, often as allies rather than opponents, and characters
from legend and poetry could appear as allies with the speaker or as a proof
for the woes of ideals falsely believed to be heroic.20 The interlocutor might
even be an entirely fictive individual created by the speaker.21 Additionally, the
interlocutor need not address the speaker or the audience directly; instead,
a group of imaginary speakers could appear and speak among themselves.22
What is more, the speaker often used the interlocutor simply as a rhetorical
device in order to emphasize and clarify his or her own thoughts rather than
intentionally presenting the views of an opponent or the audience.23
In describing the tone and mood of diatribal dialogue, Bultmann argued
that a primary characteristic was vividness and movement. Bultmann wrote,
“Schon nach wenigen Sätzen befindet sich der Redner im Feuer.”24 Such fiery dis-
course was achieved through the use of

lively questions, energetic imperatives, the back-and-forth of questions

and answers, objections and rejections, in the change of mood between
jest and earnest, in the change of tone between encouragement and
enthusiasm, insult and humiliation.25

17  Bultmann, Der Stil, 11. Cf. Hirzel, Der Dialog, 1.372.
18  Bultmann, Der Stil, 11.
19  Ibid., 11–12.
20  Ibid., 12.
21  Ibid., 13.
22  Ibid., 12–13.
23  Ibid., 12.
24  “After only a few sentences, the speaker finds himself in the fire.” Ibid., 58.
25  Bultmann writes, “… in den lebhaften Fragen, in den energischen Imperativen, in dem
Hinundher von Fragen und Antworten, von Einwänden und Zurückweisungen, in dem
Wechsel der Stimmung zwischen Scherz und Ernst, in dem Wechsel des Tones zwischen
Ermahnung und Begeisterung, Scheltwort und Beschämung.” Ibid.
106 chapter 6

For this reason, the tone and mood of diatribe could be quite diverse, the fast
movement of the diatribe was visible in its ability to switch from a light to a
heavy mood and back again, and to do so quickly.
Bultmann also described the diatribe as vivid, which was best seen in the
direct, concrete, and practical nature of diatribe. The speaker did not spend a
great amount of time addressing theoretical philosophical aspects but painted
a picture of the way things were; for example, instead of speaking theoretically
about some philosophical concept, a personification of the theory was intro-
duced in order to speak on its own behalf.26
In Bultmann’s analysis, diatribe’s vividness was also present in its purpose.
The speaker’s words were not pointless musings. They were a personal appeal
to elicit a desired effect in one’s audience.27 The speaker’s personal appeal to
the audience took on various moods, but the chief component was that of
refutation (ἐλέγχειν).28 The speaker refuted (censored) problematic popular
opinions, false preconceptions and dogmas, pleasure, pain, indecision, and
weakness.29 For a practical example, teachers could refute foolish students
with such phrases as ὦ ταλαίπωρε, τάλας, μωρέ, ὦ πονηρέ, infelix, miser and
stulte.30 There was also a positive side to diatribe, exhortation (νουθετεῖν). The
boundary between the two was blurred, as comparison (Vergleich) frequently
found a place in exhortation and inherently made demands on the audience
vis-à-vis the thing to which they were compared.31 Thus, both in its primary
polemical mood and its muddied exhortative-protreptic mood, Bultmann’s
model for diatribe displayed an agenda to generate a calculated change in the
audience for the better, at least as the speaker envisioned it.
No doubt in part creating room for his own contributions, concerning the
period of research on diatribe following Bultmann, Stowers remarks, “It is
clear that there has been no fundamental advance in understanding the dia-
tribe or its dialogical style among New Testament scholars since Bultmann.”32
Nevertheless, critique of the earlier works on diatribe eventually led to posi-
tive refinements during the period between Bultmann and Stowers. For
example, Otto Halbauer and Tadeusz Sinko rejected the view that diatribe was

26  Ibid., 60.

27  Ibid., 61.
28  Ibid., 62.
29  Ibid.
30  Ibid., 14.
31  Ibid., 63.
32  Stowers, Diatribe, 45. For a fuller discussion of this period, see King, “Speech-in-Character,”
134–38; Stowers, Diatribe, 41–48.
Portrayals of Diatribe 107

a literary Gattung,33 Hermann Throm attempted to locate diatribe within the

genre of the rhetorical θέσις,34 and Helmut Rahn simply opted to discuss liter-
ary works as “Diatribenartiges”—diatribal or diatribe-like.35 Wilhelm Capelle
and Henri-Irénée Marrou, however, argued that, though diatribe did not begin
as a literary genre, it eventually became a literary Gattung characterized by
personifications and dialogues, amongst other features.36 George L. Kustas
contended similarly, “In the course of time the looser structure of the early
pieces of diatribe would have been tightened and readily identifiable rhe-
torical parts established.”37 Hildegard Cancik, on the other hand, argued that
diatribe represented a “subliterary” form that existed in oral form apart from
carefully composed literary works.38
Perhaps the most significant development in the period of research be-
tween Bultmann and Stowers was the emphasis on the social setting or func-
tion of the diatribe. Abraham J. Malherbe, for instance, argued that

evidence suggests that there is a correlation between the style of the dia-
tribes and the social setting in which they were delivered … We shall have
to take more seriously the possibility that the discernible differences in
form and style of what are known as diatribes are related to their socio-
logical functions.39

33  Otto Halbauer, De Diatribis Epicteti (Leipzig: Robert Noske Bornen, 1911), 1–7; Tadeusz
Sinko, “On the So-called Cynic-Stoic Diatribe” (in Polish), Eos 21 (1916): 21–63. See also
Barbara P. Wallach, “A History of the Diatribe from its Origin up to the First Century
B.C. and a Study of the Influence of the Genre upon Lucretius” (Ph.D. diss., University of
Illinois, 1974).
34  Hermann Throm, Die Thesis (Rhetorical Studies 17; Paderborn: F. Schöningh, 1932).
35  Helmut Rahn, Morphologie der antike Literatur: Eine Einführung (Darmstadt:
Wissenschaftliche Buch-Gesellschaft, 1969), 156.
36  Wilhelm Capelle and Henri-Irénée Marrou, “Diatribe,” in Reallexicon für Antike und
Christentum: Sachworterbuch zur Auseinandersetzung des Christentums mit der antiken
Welt (Theodor Klauser, ed.; vol. 3; Stuttgart: Anton Hiersemann, 1957), 990–92.
37  George L. Kustas, Diatribe in Ancient Rhetorical Theory (Protocol of the colloquy of the
Center for Hermeneutical Studies in Hellenistic and Modern Culture 22; Berkeley: Center
for Hermeneutical Studies, 1976), 6.
38  Hildegard Cancik, Untersuchungen zu Senecas epistulae morales (Spudasmata 18;
Hildesheim: Georg Olms, 1976), 47–48.
39  Abraham J. Malherbe, Social Aspects of Early Christianity (Baton Rouge: Louisiana State
University Press, 1977), 50. Quoted in Stowers, Diatribe, 42–43. See also Mario Puelma
Piwonka, Lucilius und Kallimachos (Frankfurt: V. Klostermann, 1949); E. G. Schmidt,
“Diatribe and Satire,” in Wissenschaftliche Zeitschrift der Universität Rostock 15 (1966):
507–15.
108 chapter 6

In the pre-Bultmannian tradition extending back to Wilamowitz-Moellendorf,

the social setting of diatribe was understood as the orations of the wandering
street preacher. Wilamowitz-Moellendorf and Hirzel, however, had noted the
scholastic setting of diatribe, even if they did not assign it any real import.40
The emphasis on diatribe as the sermons of the Cynic-Stoic street preacher,
therefore, resulted in little interest being paid to the significance of a potential
scholastic setting for diatribe. This would be a primary agenda of Stowers.41
Stowers begins by identifying eight primary sources for diatribe: (1) Teles
(Bion); (2) Lucius (Musonius Rufus); (3) Arrian (Epictetus); (4) Dio Chrysostom;
(5) Plutarch; (6) Maximus of Tyre; (7) Seneca; and (8) Philo of Alexandria.42
Stowers notes that the criteria for categorizing these eight sources as repre-
sentative of diatribe are primarily twofold. First, these sources share a “com-
mon appropriation of a certain body of popular philosophical traditions”
and, second, combine a “dialogical style together with certain other stylistic
or rhetorical features.”43 In addition to these attributes, Stowers adds a third
criterion—each of these sources are dependent on or evince affinities for a
“scholastic social setting.”44 This latter addition distances the social setting
of diatribe from philosophical Cynic-Stoic street preaching for the masses,
locates it in the school setting, and proves most influential for Stowers’s inter-
pretation of diatribal texts, including Paul.
Finally, before engaging the primary sources, Stowers addresses the ques-
tion of whether or not diatribe constitutes a Gattung or genre. Stowers allows
for the critique of scholars such as Halbauer and Cancik that diatribe does
not represent a literary genre “in the sense of a family of writings which con-
sciously reflect back on and follow a literary tradition with common literary
form.” Stowers, however, maintains that denying diatribe is a literary genre does
not exclude diatribe from belonging to rhetorical or oral speech categories.45

40  Wilamowitz-Moellendorff, “Excurs 3,” 301, 306; Hirzel, Der Dialog, I.368–74; 2.117.
41  The following discussion refers to Stowers’s The Diatribe and Paul’s Letter to the Romans.
Stowers has addressed diatribe in a number of other places, including “The Diatribe,” in
Greco-Roman Literature and the New Testament: Selected Forms and Genres (SBLSBS 21;
David E. Aune, ed.; Atlanta: Scholars Press, 1998), 71–83; idem., “Diatribe,” in The Anchor
Bible Dictionary (vol. 2; David Noel Freedman, ed.; New York: Doubleday, 1992), 190–93.
42  Stowers, Diatribe, 48. Stowers admits that Philo presents some unique problems for being
included among other diatribal authors, but he nonetheless allows Philo’s voice to be
heard.
43  Ibid., 48–49.
44  Ibid., 49.
45  Ibid.
Portrayals of Diatribe 109

Because the noted diatribal sources share a number of traits, he holds that they
either belong to or represent characteristic features of a type or genre, even if it
is not a genre in the literary sense.46
Turning to the primary literature, Stowers prioritizes Teles over Bion as
the earliest source for diatribe, though Bion preceded Teles.47 Additionally,
Stowers contends that the stylistic traits characteristic of diatribe belong
(almost) exclusively to Teles and not to Bion.48 Consequently, Stowers not only
prioritizes Teles as a chief representative of diatribe chronologically, but he
also emphasizes Teles qualitatively, as it is Teles and not Bion who evinces the
features most characteristic of diatribe.
Finally, Stowers reconsiders the social setting of Teles’s discourses.
Wilamowitz-Moellendorff classified Teles as a wandering Cynic preacher,
whereby the social setting for Teles’s discourses was believed to be public
lectures delivered in the streets and marketplaces. Most scholars followed
Wilamowitz-Moellendorff in this respect despite the absence of evidence
that Teles was ever a wandering Cynic preacher, and despite Wilamowitz-
Moellendorff’s own recognition of evidence identifying Teles as a teacher
(παιδαγωγός; III.60). Stowers allows Teles’s self-designation as a teacher to
come to the fore and to help explain certain features of his style. These features
include the use of asides referring to previous conversations, the use of the

46  Ibid.
47  Teles overtly refers to Bion’s words seven times, four times in fragment II, one time in
fragment III, and two times in fragment IVA. Cf. Stowers, Diatribe, 50. Hense’s attempt
to define the limits of the Bionean materials in Teles’s fragments had focused scholarly
attention on Bion rather than Teles as the earliest representative of diatribe. Hense, Teletis
reliquiae. Stowers, however, follows Jan Frederik Kindstrand, who concludes, “While there
is a strong probability that Teles contains more of Bion than meets the eye, I think the
question is better left open, as we have no possibility of reaching a definite answer.” Jan
Frederick Kindstrand, Bion of Borysthenes: A Collection of the Fragments with Introduction
and Commentary (Studia Graeca Upsaliensia 11; Stockholm: Almquist & Wiksell, 1976), 85;
Stowers, Diatribe, 50–51.
48  For example, with the exception of the speech of Poverty (II.19–44), the attributions to
Bion lack any trace of the dialogical element. On the contrary, dialogical exchanges and
the use of objections characterize Teles’s own material, which “warrants the conclusion
that the dialogical element was an important feature of the Telean diatribe.” Plus, the
brevity of the Bionean fragments disallows the possibility of constructing something of
a complete Bionean discourse, since the evidence is simply lacking. References to Teles’s
fragments refer to Edward O’Neil, Teles (The Cynic Teacher) (SBLTT 11; Missoula: Scholars
Press, 1977). For Stowers’s argument, see Diatribe, 51.
110 chapter 6

first person plural, and, most notably, the extensive use of the dialogical style.
These dialogical elements include the use of unnamed and fictitious inter-
locutors, objections, responses, and rhetorical questions.49 In Stowers’s view,
Teles’s use of such stylistic elements (both dialogical and otherwise) is best
explained “as instances of Teles speaking personally to his class.”50 Thus, these
features are best understood in light of their scholastic context, and Stowers
forms and nuances his understanding of these elements as he examines the
additional diatribal sources.
Moving to Epictetus, Stowers stresses the fact that Epictetus “taught in a
school,” whereby his writings should be examined in light of this scholastic
social setting.51 In fact, Epictetus’s discourses only exist today because one of
his students, Arrian, stenographically recorded them.52 In Epictetus’s school,
the diatribes are the “more practical lectures and conversations, which prob-
ably followed [the technical lessons],” during which time the teacher could
address student questions and problems.53 The diatribes may address the stu-
dents as a group, or they could be “occasional responses addressed to specific
problems, situations, or individuals.”54 Furthermore, the diatribes could be
unsolicited or initiated by student questions.55
As with Teles, one of the most characteristic features of Epictetus’s diatribes
is their dialogical elements. Unnamed and/or fictitious interlocutors interject
isolated comments or objections, or they engage in sustained dialogue with
the primary speaker, Epictetus. At other times, Epictetus initiates dialogue
by addressing or questioning the audience or an imaginary interlocutor. As
Kindstrand demonstrated the tenuous nature of determining the extent of
Bion’s voice in the Teles fragments, Stowers indicates it is often similarly diffi-
cult to distinguish the extent of the speech of Epictetus’s interlocutors. Stowers
writes, “Often these dialogical flourishes are obscure and difficult to punctu-
ate and edit since much which was communicated by voice intonation is now

49  Dialogical elements appear in six of the eight fragments. Fragments V and VI do not con-
tain dialogical elements. Ibid., 52–53.
50  Ibid., 52.
51  Ibid., 54.
52  Ibid.; W. A. Oldfather, Epictetus: Discourses I–II (LCL 131; Cambridge: Harvard University
Press, 2000), xii–xiii.
53  Stowers, Diatribe, 54–55.
54  Ibid.
55  Ibid., 54–55.
Portrayals of Diatribe 111

lost.”56 In other words, it is frequently the case that no overt textual markers
exist to assist the reader to define the precise limits of each speaker’s words.57
The tone of Epictetus’s diatribes varies; it can be polemical or reprimanding,
such as in Discourses 1.23, 2.20, and 2.23, but it is usually simply educational.58
In most cases, Epictetus employs elements of censure (or indictment, refuta-
tion; ἐλεγκτικός) and protreptic (προτρεπτικός) in keeping with the Socratic
method of question and answer, which Stowers takes to be indicative of
Epictetus’s pedagogical method.59 With censure, Epictetus exposes one’s inner
contradictions or errors. With protreptic, Epictetus makes clear the correct
course of thought or action and encourages his students to appropriate it.60
Thus, Stowers argues, the purpose of diatribe “is to point out error, to convince
and to convict and then to lead one to the truth, to a right way of life.”61 The
dialogical elements support this purpose, as they allow Epictetus to challenge
his audiences’ views and to convey his own (correct) thoughts through the use
of question and answer.62
Stowers also identifies Epictetus’ teacher, Musonius Rufus, as significant for
diatribe.63 This is particularly because, in fragment VI, Musonius uses διατριβή
as a reference to what takes place in the philosophical school.64 Like Teles and
Epictetus, Musonius’ conversational expressions and addresses in the second
person singular amount for Stowers to “[clear] examples of the dialogical style

56  Ibid., 55.

57  This makes perfect sense, assuming diatribe’s oral rather than literary generic qualities.
Instead, the primary speaker could indicate that an imaginary interlocutor was entering
or exiting the conversation simply through changes in his or her voice, which would have
been perceptible to the original audience but lost in Arrian’s written records.
58  Ibid.
59  Ibid., 55–58. See Epictetus, Disc. 2.26.4.
60  Stowers, Diatribe, 55–58.
61  Ibid., 58.
62  Ibid.
63  Ibid., 58–59. One of Musonius’s students, a certain Lucius, preserved twenty-one of
Musonius’s diatribes (with significant editing). Whereas Arrian produced a stenographic
record of Epictetus’s diatribes, however, Lucius did not. Lucius reveals that he often
approximates or paraphrases Musonius’s words. As a result, Lucius’s writings are not
verbatim reports but are “semi-literary re-writings of Musonius’s diatribes.” See also Otto
Hense, ed., C. Musonii Rufi reliquiae (Leipzig: Teubner, 1910), XIV; Cora E. Lutz, Musonius
Rufus: “The Roman Socrates” (Yale Classical Studies 10:3–147; New Haven: Yale University
Press, 1947), 6–7.
64  Stowers, Diatribe, 58. The text reads, ὅσοι γε φιλοσόφου διατριβῆς μετεσχήκαμεν, ὡς οὔτε
πόνος οὔτε θάνατος οὔτε πενία κακὸν οὐδαμῶς ἐστιν οὐδ᾽ἄλλο τι τῶν κακίας ἀπηλλαγμένων.
112 chapter 6

of the diatribe, the addressing of an interlocutor.”65 Additionally, Musonius

quotes and directly responds to Euripedes in discourse IX, and discourse XV
includes a small dialogue with an objector.66 Epictetus even cites Musonius as
an example of a teacher who employed censure and protreptic in his pedagogi-
cal method (Discourses 3.23.28–29).67
With respect to Dio of Prusa (Dio Chrysostom), who was exiled by Domitian
in 82 CE, Stowers notes that Dio “took up the life of a wandering Cynic,” and
that Dio does not evince an overt scholastic social setting.68 Stowers admits,
“Dio does not seem ever to have had a formal school.”69 Nevertheless, Stowers
argues, “The social context of Dio’s diatribes … in some ways approximates
that of a school, and Stowers argues similarly regarding Maximus of Tyre,
who uses διατριβή to refer to the educational activity in schools as well as an
“equivalent” for the schools themselves.70 Stowers supports this view by argu-
ing that, though Dio delivered public orations at times, he also worked more
directly with small groups for periods of time.71 For instance, Dio can refer to
details from previous lessons, for which reason Stowers argues that “[Dio’s]
diatribes reflect a limited but established relationship with his audience and
especially with his discussion partners.”72 So, though Dio can go so far as to
express Socratic disdain for established schools, Stowers avers, “in his true
diatribes he did establish a student-teacher relationship.”73
Plutarch’s discussions significantly shape Stowers’s narrative of the goal of
diatribe, which Stowers again sets within “a specific type of discourse grow-
ing out of a certain social setting,” the scholastic setting.74 Stowers buttresses
his view by noting Plutarch’s use of σχολή and διατριβή not only to refer to the
school itself but also as a partial indication of the form of the instructional
activity that took place in the school.75

65  Ibid., 58.

66  Ibid., 58–59.
67  Ibid., 59–60.
68  Ibid., 60–61.
69  Ibid., 61.
70  Ibid., 61, 67–68.
71  Ibid., 61.
72  Ibid., 61–62.
73  Ibid., 62, 207n.340.
74  Contra Fritz Krauss, Die rhetorischen Schriften Plutarchs und ihre Stellung im Plutarchischen
Schriftkorpus (Nürnberg: J. L. Stitch, 1912); Stowers, Diatribe, 63. Stowers, Diatribe, 63–64.
75  Ibid., 63.
Portrayals of Diatribe 113

Plutarch’s works show a significant degree of diversity.76 Plutarch’s diatribal

writings also display dialogical elements, albeit less prominently. Imaginary
interlocutor’s issue objections, and Plutarch occasionally initiates conversa-
tion by addressing his interlocutors.77
What Plutarch lacks in dialogical prominence, however, he makes up for
with what he suggests about the tone of diatribe and the censure-protreptic
pedagogical process. Stowers notes that Plutarch “displays the same language
of indictment or censure as seen in the other diatribal authors.”78 Plutarch also
explains in De recta ratione audiendi how an audience member should react
to this censure-protreptic method. According to Plutarch, when an audience
member is censured, he or she should not respond as if it were nothing seri-
ous; the censure should weigh sufficiently in order to achieve its desired effect
(47A). The initial censure, however, must not be the end goal. The censured
audience member should wait in hope for the ensuing “sweet and bright” pro-
treptic element that follows (47A).79 Said otherwise, in the censure-protreptic
progression, harsh censure should give way to a great cure of the censured
party’s perceived illness(es).80 As Epictetus and Musonius, Plutarch invokes
Socrates as the exemplary model for this method and connects it to diatribe,
which Plutarch notes is a “type of discourse in the philosophical school”
(43E).81 Therefore, Plutarch’s diatribes are significant not only because they
demonstrate how oral diatribes could be shaped into literary forms,82 but also
because they provide a picture of the optimal audience response to the meth-
od of censure and protreptic which Stowers views to be so central to diatribe.
Philo’s literature, however, presents certain problems for Stowers. There is
ambiguity concerning the date and social setting of Philo’s works, and, Stowers
admits, “there is almost nothing to indicate that any of them grew out of a
school situation.”83 Additionally, Philo only infrequently utilizes dialogical ele-
ments and, even when he does, he uses them quite dully.84 Paul Wendland’s
earlier examination of Philo in light of diatribe, however, not only placed Philo
squarely among those thinkers influenced by diatribe, but it also opened the

76  Ibid., 63–65.

77  Ibid.
78  Ibid., 65–66.
79  Plutarch writes, … ἐλπίζειν τι γλυκὺ καὶ λαμπρλον.
80  Stowers, Diatribe, 66.
81  Ibid., 67.
82  Ibid., 63–65.
83  Ibid., 68.
84  Ibid., 69.
114 chapter 6

door for scholars to begin consideration of other Jewish writers with possible
diatribal influences, namely Paul.85
With Seneca’s epistulae morales, however, Stowers finds a generic analog
for Paul’s writings, given that the moral epistles and Paul’s letter to the
Romans combine the style of diatribe with the form of a letter.86 Seneca’s liter-
ary epistles obviously differ in form from formal philosophical teaching in a
school setting. Stowers insists that a student-teacher relationship nevertheless
exists between Seneca and the recipient, Lucilius. Following Cancik, Stowers
affirms that ancient letters acted “as a surrogate for being together” and were a
“method of self-revelation.”87 As such, the letters create a literary environment
in which it was as if Seneca was physically present before Lucilius’s eyes, and
vice versa.88 In this way, “Seneca provides Lucilius with a type of epistolary
paraenesis where he acts as spiritual guide and presents Lucilius with [himself
as an] example.”89
Another characteristic of the letter form, which overlaps with diatribe, is its
dialogical quality. Stowers questions how to distinguish between the dialogical
quality of letters and that of diatribe.90 Stowers asserts that the dialogical ele-
ment of the letter primarily appears in its “philophronetic,” friendly character,
which manifests itself “in the tone of two friends holding a conversation.”91 The
dialogical elements of diatribe, however, are less personal, employ the meth-
ods of censure and protreptic, and tend to have a moral-philosophical subject
matter. Stowers argues that Seneca regularly shifts back-and-forth between
the friendly epistolary tone and the didactic, diatribal style. It is when Seneca
dons the didactic, diatribal style with its method of censure and protreptic
that Stowers thinks “[Seneca] presents himself as the popular philosophical
teacher.”92
Concerning the dialogical element in Seneca’s epistles, the use of an
imaginary interlocutor who objects and questions the teacher is common,
as has been seen in other sources. Stowers divides Seneca’s interlocutor’s

85  See Paul Wendland and Otto Kern, Beiträge zur Geschichte der griechischen Philosophie
und Religion (Berlin: Georg Reimer, 1895).
86  Ibid., 70.
87  Ibid.; see also Stanley K. Stowers, Letter Writing in Greco-Roman Antiquity (Library of Early
Christianity 5; Philadelphia: Westminster Press, 1989), 23, 38–39.
88  Stowers, Diatribe, 70
89  Ibid. For additional discussion regarding the use of letters for various rhetorical contexts/
purposes, see also Rosenmeyer, Ancient Epistolary Fictions, esp. 45–60.
90  Stowers, Diatribe, 71.
91  Ibid.
92  Ibid.
Portrayals of Diatribe 115

responses into three types.93 First, the interlocutor’s responses may represent
the response of Lucilius to whatever Seneca is discussing. These responses are
present primarily in the friendly and conversational tone of the letter, or in the
more epistolary sections of the letter. Often Seneca introduces the responses
with a general verb of speech, but at times he uses no introductory formula at
all. The second group of responses contains those instances when the inter-
locutor’s identity is less certain but is probably Lucilius. These responses are
found in the primarily didactic portions of the letter and “appear to function
in the argumentation like the objections of the fictitious interlocutor in the
diatribe.”94 Occasions when the imaginary interlocutor interjects objections
in a standard diatribal way belong to Stowers’s third category, in which “the
interlocutor is typically anonymous and his views usually represent ideas or
types of behavior which Seneca wants to censure.”95
Furthermore, Stowers argues that, because all three categories of interlocu-
tor responses frequently occur in a single letter, ambiguity exists with respect
to “just who is objecting and to whom Seneca is responding.”96 Stowers’s expla-
nation of the ambiguity is worth quoting at length. In this epistolary context,

Seneca has skillfully created the same type of situation found in the
diatribe where the philosopher moves back and forth between contact
with his audience and the interlocutor, producing a calculated ambiguity
about precisely who is being responded to or addressed. When Seneca
censures the interlocutor, one often senses that Lucilius is the real target
of the indictment. It is as if Lucilius were present in the school of Seneca.97

Thus, for Stowers, the epistles create space for Seneca to teach Lucilius in
much the same way as would have occurred if the two men were together in a
schoolroom setting.
In light of his analysis of these primary diatribal sources, Stowers draws a
number of conclusions. First, all of the sources display dialogical elements,
some more prominently than others, which manifest primarily as address to
or objections from the interlocutor.98 Second, with the possible exception of

93  Ibid., 72.

94  Ibid. Stowers adds, these responses “are more like the questions he might ask in a class-
room than in a personal conversation.”
95  Ibid., 73.
96  Ibid.
97  Ibid., 75.
98  Ibid., 76–77.
116 chapter 6

Philo, each of the sources belong to a philosophical scholastic social setting.

This is in distinction from earlier works on diatribe, which imagined the social
setting as the public, propagandistic lectures of the wandering Cynic street
preacher. Teles, Epictetus, Musonius, and Plutarch headed formal schools; Dio
and Maximus created discourses for a social setting approximating that of a
school; Seneca approximated the context of a school and its dialogical context
through the use of his epistles to Lucilius.99 Third, diatribe is not the techni-
cal instruction on various topics, but the conversations that follow the formal
lectures.100 Fourth, in this scholastic setting and through the use of diatribe
and its dialogical elements, the pedagogical method of censure and protreptic
aims to point out errors and inconsistencies (censure) and to correct them by
explaining the correct course of action and encouraging students to adopt it
(protreptic).101 The audience should take this process seriously, but it is not
to be considered destructive or overly onerous. Instead, the audience should
endure the censure and wait in hope for the sweetly bright protreptic to fol-
low. Nevertheless, diatribe may still appropriate polemic against its interlocu-
tor or audience, as represented by Epictetus’s Discourses 1.23, 2.20, and 2.23.102
Finally, fifth, the diversity among the diatribal sources is due to the personal
background of each author, the varying degree to which the diatribes have
been adapted and prepared for composition, and the varying relationships of
each source’s discourses to the philosophical scholastic context.103
These characteristics account for the significant similarities among these
sources. These similarities also shape the way these sources should be consid-
ered as a group—as a genre. To be sure, it is incorrect to speak of diatribe as
a literary Gattung or genre. In its primary form, diatribe was an oral endeavor.
These core characteristics, however, attest that these works belong to a specific
rhetorical type, for which “diatribe” is an appropriate and useful term.104 Said
otherwise, Stowers argues, diatribe and diatribal features would have been
recognizable, and one can expect Paul’s audiences in Rome to have recognized
its influence on his letter to them in a similar fashion.
The publication of Stowers’s dissertation represented a watershed moment
in the history of research on diatribe. Before Stowers, Bultmann’s synthe-
sis of the early work on diatribe held sway with the vast majority. Stowers’s

99  Ibid., 76.

100  Ibid.
101  Ibid., 76–77.
102  Ibid., 55.
103  Ibid., 76.
104  Ibid., 76–78.
Portrayals of Diatribe 117

reassessment, however, opened a new path.105 Since Stowers’s dissertation, no

significant reassessment of diatribe has been published. Scholars have opted
to prioritize one or another of the two prominent views, Bultmann or Stowers,
and Stowers receives the vast majority of endorsements.106 To be sure, most
of the offerings discussed below nuance points in Bultmann or Stowers, but
the holistic presentations of the diatribe remain the same. Is diatribe the pro-
pagandistic street preaching of the wandering philosopher addressed to the
masses (Bultmann)? Or, is diatribe best understood as a type of scholastic
discourse, through which a teacher uses censure and protreptic as a way to
point out and correct contradictions and errors among his or her students as a
means of transformation?
In Paulus und die “Diatribe,” Thomas Schmeller attempts to reshape how
scholars define and apply diatribe. Schmeller proposes a new method for
how to define what is and what is not diatribal, which relies on what he calls
the Strukturprinzip (“principal of structure”). In this method, Schmeller limits
his engagement with diatribal texts to a very narrow selection, which he con-
firms against a second even narrower selection of texts. Ultimately, I suggest
that Schmeller’s project falls short in both of these regards.
Schmeller begins with a discussion of “Die Problematik der ‘Diatribe,’”107
where he questions whether any concept of diatribe existed in antiquity or
whether it is a modern scholarly construct. Schmeller argues that the ancients
never used the term diatribe to refer to a specific genre.108 For Schmeller, it
was modern scholars, beginning with Usener, who began to use the term in a
sense different from that of the ancient sources.109 Consequently, representa-
tive texts should not be considered to belong to some specific genre known as
“diatribe,” though they do share identifiable literary, stylistic, and philosophical

105  Technically, Stowers did not reveal a “new” path; rather, Stowers pressed a point recog-
nized but not pursued by several previous scholars to its logical extreme, i.e., the scholas-
tic social setting.
106  The following works are analyzed so as to answer the following questions: (1) Does this
scholar ultimately endorse Bultmann’s or Stowers’s presentation of diatribe? (2) In what
ways does this work nuance its preferred depiction of diatribe? (3) What strengths or
weaknesses in this work should influence one’s understanding of diatribe at this point
in time, over a century past Bultmann and almost four decades after Stowers’s initial
contributions?
107  Schmeller, Diatribe, 1–54.
108  Relying on the work of Halbauer, Schmeller thinks the ancients would have called this
dialexis. Ibid., 17–20. See Stowers’s critique of Halbauer. Stowers, Diatribe, 28–29.
109  Ibid., 1, 13–14.
118 chapter 6

features.110 To answer the question of whether diatribe is a modern fiction or

not, Schmeller concludes, “ja und nein.”111 Thus, with respect to the question of
genre, Schmeller disagrees with Bultmann that diatribe is a literary Gattung,
and he disagrees with Stowers that diatribe is a legitimate oral / rhetorical
type or genre. Instead, Schmeller situates himself among the early critics of
Bultmann’s paradigm, the chief of which is Halbauer.
Schmeller chooses his diatribal camp, however, when he discusses the ques-
tion of the social context for diatribe.112 Schmeller’s model for diatribe fits
three main contexts: (1) mass propaganda, (2) use in philosophical schools, and
(3) literary production.113 Though diverse, these three contexts share a com-
mon element. In each context, diatribe proclaims popular philosophy aimed
at the general public.114 Essentially, it is philosophy for everyman; Schmeller
argues, “eine ‘Diatribe’ ist eine persönliche und existentielle Anrede an jedermann
mit ethischer Prägung.”115 Schmeller, therefore, ultimately sides with the view
of diatribe represented by Bultmann rather than by Stowers.
Schmeller’s new method for deciphering what qualifies as genuine diatribe
is a search for what he calls the Strukturprinzip.116 Schmeller’s goal is not to
identify which individual traits, when combined together in specific ways, are
able to prove that a text is or is not a diatribe. For Schmeller, these features

110  Ibid., 20, 33–54; see also 205–7, 215–16, 219.

111  Ibid., 20. Schmeller writes, “Eine Fiktion ist die Bezeichnung an sich: Sie wird seit Ende des
letzten Jahrhunderts in einem anderen Sinn als in der Antike gebraucht und soll nun eine Art
sprachlicher Äußerungen benennen, die zur Zeit ihrer Entstehung anders benannt wurde.
Keine Fiktion ist dagegen das, was mit dieser Bezeichnung belegt wird. Es gibt eine Reihe
antiker Texte, die solche Gemeinsamkeiten in Form und Inhalt aufweisen, daß sie zu Recht
unter einem gemeinsamen Namen zusammengefaßt werden.”
112  Ibid., 51. Schmeller agrees with Stowers that Teles, Musonius, and Epictetus might legiti-
mately be read in the context of a philosophical school. Schmeller contends that Stowers’s
thesis falls short, however, with respect to Maximus, Plutarch, Philo, Seneca, and Dio.
But Schmeller’s dismissal of Maximus, Plutarch, Philo, Seneca and Dio as indebted to
the scholastic social setting is overly simplistic (whether he is right or wrong). A meager
assertion that these sources do not belong in the scholastic social setting does not an
argument make. Though it is correct that the scholastic context is more readily identifi-
able for Teles, Musonius and Epictetus, Schmeller offers little that is constructive to refute
Stowers’s arguments with respect to the other diatribal sources.
113  Ibid., 52.
114  Schmeller writes, “Was das Publikum betrifft, besteht demnach die Gemeinsamkeit inner-
halb der ‘Diatribe’ lediglich in einer populären Tendenz, in einer Ausrichtung auf Laien statt
Fachphilosophen.” Ibid., 52.
115  Ibid., 99.
116  Ibid., 98–99.
Portrayals of Diatribe 119

are merely products of the Strukturprinzip, which may or may not utilize
other elements.117 The Strukturprinzip is the underlying essence that holds
all of these individual features together and integrates them into a tex-
tual unit. To borrow Stowers’s translation of Schmeller’s definition, “The
[Strukturprinzip] for the diatribe is ‘the transformation of intellectual content
into existential encouragement with ethical delivery.’”118 Perhaps stated overly
simplistically, Schmeller’s Strukturprinzip is reducible to ethical exhortation.
Schmeller’s selection of sources, which he uses to identify the Struktur­
prinzip, is interesting. Schmeller opts for “comprehensive analysis and
interpretation” of a very select number of texts from three sources, which he
then compares to an even narrower selection from three different sources.
Schmeller’s primary selection of texts includes: (1) Fragments 16A, 17, 21, and
68 from Bion; (2) Fragment XVII from Musonius Rufus; and (3) Diss. I.12.8–35
from Epictetus. Schmeller then “confirms” his reading of these texts against:
(1) Teles’s second diatribe, Περὶ αὐταρκείας; (2) Dio’s Or. 16, Περὶ λύπης; and
(3) Plutarch’s Περὶ ἀρετῆς καὶ κακίας.119 Schmeller recognizes that he is utilizing
a very limited number of texts but insists that such selectivity can still produce
positive results.120
Schmeller’s project is problematic on both methodological fronts. Given
the amount of diatribal literature, Schmeller’s highly selective selection of
texts can hardly be considered representative. Plus, as Stowers also notes,
Schmeller is searching for something of an Aristotelian form or essence for
diatribe. Unfortunately, Schmeller defines this form, the Strukturprinzip,
so broadly that it can encompass almost any if not all hortatory literature.121
Diatribe is but one mode of exhortation; not all exhortation is diatribe.122
Schmeller’s overly broad definition of the Strukturprinzip is perhaps con-
nected to his limited concept of genre. By denying that diatribe might be a
rhetorical or oral (rather than literary) genre, one wonders whether Schmeller
inevitably diminishes the significance of the similarities these sources share,
which results in an overgeneralization of his proposed underlying principle.

117  Ibid., 99.

118  Stanley K. Stowers, Review of Thomas Schmeller, Paulus und die ‘Diatribe’: Eine verglei-
chende Stilinterpretation, JBL 108.3 (1989): 540.
119  Schmeller, Diatribe, 203–24.
120  Ibid., 99.
121  As Tobin writes, “Definitions [of genres] often have to be so general that they prove vir-
tually useless in understanding the conventions and expectations of ancient works of
literature or rhetoric.” Tobin, Paul’s Rhetoric, 90.
122  Stowers, Review of Schmeller, 539–40.
120 chapter 6

In any case, Schmeller’s suggested methodology contains severe issues and

ultimately fails to convince.123
As Schmeller aligns more with Bultmann, Thorsteinsson endorses
Stowers on the characteristics of diatribe.124 The reason Thorsteinsson is
important for the current project is because his work is an attempt to answer
the question, “How can one identify an imaginary interlocutor, especially
in Paul’s epistles?” The correlative questions, of course, are, “How can one
deduce when the interlocutor is or is not speaking?” and “Why does it
matter?” which are some of the same overarching questions this project aims
to resolve. Thorsteinsson’s method is quite different, however. Rather than
corroborating diatribal evidence in tandem with the rhetorical exercise
of speech-in-character, which transcends generic categories, Thorsteinsson
attempts to bring in evidence only from epistolary materials.
Thorsteinsson identifies many similarities between the diatribal and epis-
tolary uses of dialogical elements, but he also finds differences. Thorsteinsson
observes that one of the principal dialogical features in epistles is the use of an
interlocutor, who is envisioned as entering into conversation in order to ques-
tion, object, or be questioned by the author. When the interlocutor’s speech is
explicitly marked, it is marked with a verb of speech.125 If it is unmarked, cer-
tain clues often hint to the reader the presence of an additional voice, such as
interrogative phrases (i.e., τί οὖν), adversative or inferential conjunctions, or an
author’s immediate response to a statement.126 In most cases, the immediate
context is also helpful for determining whether an interlocutor has entered the
scene.127
Diverging from diatribe’s broader use of interlocutors, Thorsteinsson argues
that epistolary interlocutors typically represent the letter’s recipient, whereby
the interlocutor fills the dialogical roll of the absent party.128 When epistolary

123  Since Schmeller’s publication, to the best of my knowledge, no work on diatribe adopts his
views on these methodological points. At most, Schmeller is generally listed as an alterna-
tive source for diatribe in addition to those of Bultmann and Stowers, primarily. For a rep-
resentative selection, see Elliott, Rhetoric, 120 n.2; Song, Reading Romans, 11n.8; Douglas A.
Campbell, Deliverance, 1078n.29; and Tobin, Paul’s Rhetoric, 91n.35. Thorsteinsson plainly
suggests that Schmeller’s project fails in comparison to Stowers’s. Thorsteinsson, Paul’s
Interlocutor, 124n.6.
124  Thorsteinsson, Paul’s Interlocutor, 123–30.
125  Ibid., 135–36.
126  Ibid., 137–39.
127  Ibid., 140.
128  Ibid., 136–41.
Portrayals of Diatribe 121

interlocutors do not represent the letter’s recipient(s), Thorsteinsson suggests

it is usually clearly indicated, often with a verb of saying and an indefinite
pronoun.129 Also, unless otherwise noted, the identity of the interlocutor
generally remains the same throughout the course of a letter.130
Finally, Thorsteinsson identifies exceptions to these epistolary norms,
similar to diatribal style, whereby “the epistolary interlocutor may represent
different groups and views of people … specific types of persons, or common
opinion.”131 Also similar to the diatribe’s use of interlocutors, the epistolary
interlocutor functions as a device to develop the argument as desired, and as a
means of engaging the audience.132
Despite Thorsteinsson’s helpful analysis of epistolary evidence, several
questions remain unanswered. For instance, how many exceptions to a rule are
required to problematize said rule? Seneca’s epistles contain numerous exam-
ples in which the interlocutor does not represent the recipient, as Stowers and
Thorsteinsson have documented. At what point does the general rule for the
interlocutor to represent the recipient become less significant? Additionally,
what if an epistle displays heavy diatribal features? Should epistolary norms or
diatribal norms (when they differ) receive priority in these instances? Is it not
possible that Paul, like Epictetus, moves in and out of more diatribal sections
in Romans, as argued by Tobin,133 so that it would make more sense to evalu-
ate these diatribal pericopae in light of diatribe rather than epistolography?
Or, could diatribe’s allowance for the author to initiate conversation with an
interlocutor mitigate what Thorsteinsson finds to be the epistolary tendency
to introduce interlocutors with the interrogative τί οὖν?
Tobin’s work also endorses Stowers’s model, as far as diatribe is concerned.
Diatribe belongs to the philosophical, scholastic social setting, and it “usually
[has] an ethical-religious nature.”134 Additionally, diatribes were not the tech-
nical instruction but were “discourses or instructions of a more popular sort
in which there was a strong dialogical or Socratic component.”135 The goal of
diatribe was to transform, “to point out error and cure it,” that is, Socratic cen-
sure and protreptic.136

129  Ibid., 141.

130  Ibid., 144.
131  Ibid., 141.
132  Ibid., 140–41.
133  Tobin, Paul’s Rhetoric, 84–88.
134  Ibid., 91.
135  Ibid.
136  Ibid., 91, 93.
122 chapter 6

Tobin’s contribution proceeds from his conviction that genre matters sig-
nificantly in terms of how an author intends his composition to be understood,
as well as with respect to how an audience would have in fact read or heard the
composition. Tobin writes,

a genre [is] the fairly stable clustering of different conventions such

that they formed a commonly recognized pattern. These patterns then
informed the ways in which authors composed, and readers and hearers
understood, these compositions.137

For Tobin, the body of Romans belongs to the genre of diatribe, and it would
have readily been recognizable as such.138
When Tobin looks at the macrostructure of select diatribes from Epictetus,
he observes that some of Epictetus’s diatribes contain passages of a more
expository nature. For Tobin, these passages represent “less controversial
explanation[s] and foundation[s] for the other arguments.”139 In the ensuing
more lively and diatribal sections, Epictetus applies these less controversial
explanations and foundations in more controversial ways to how he thinks
humans should conduct their lives.140 Thus, some of Epictetus’s diatribes dem-
onstrate a pattern in which more expository passages are further expounded
in passages with a heightened diatribal style, and Tobin contends that Romans
follows this pattern too.141
Song represents another scholar who adopts Stowers’s presentation of the
diatribe basically in a holistic fashion. Song notes that Stowers “convincingly
demonstrates” diatribe’s implementation of a Socratic censure and protreptic
method, so that the dialogical elements are used for a pedagogical rather than

137  Ibid., 90.

138  Ibid., 89. Tobin’s identification of the body of Romans as diatribe will be discussed in
more depth in Part 3. Tobin of course recognizes that, technically, Romans is an epis-
tle. Correctly, Tobin also notes that it is perfectly permissible and common for the body
of an epistle to vary in and employ diverse generic categories, which problematizes
Thorsteinsson’s narrow epistolary approach. Said otherwise, there is no reason, based
simply in terms of genre, that Romans cannot contain a diatribal body situated within
an epistolary frame. Ibid., 96. See also David E. Aune, “Romans as a Logos Protreptikos,”
in The Romans Debate (Karl P. Donfried, ed.; rev.; Grand Rapids: Baker Academic, 1991),
278–96; Rodríguez, If You Call, 37n.40.
139  Ibid., 94.
140  Ibid., 95.
141  Ibid., 84–88, 95–98.
Portrayals of Diatribe 123

polemical purpose.142 Additionally, diatribe is not a form of mass propaganda;

“its form and function presuppose a student-teacher relationship.”143 As such,
Song does not have much to offer in terms of any overarching understanding
of diatribe. Song’s contribution rests in his attempt to identify stylistic features,
which he calls diatribe markers, that concretely identify diatribe as diatribe
and not some other rhetorical or literary category.144
By “diatribe markers,” Song means those elements of diatribe that are
unique to diatribe and do not overlap with other rhetorical categories.145 Such
“markers” differ from general traits or characteristics that are shared by other
generic categories. These “diatribe markers” include: (1) vivid dialogues, espe-
cially with fictitious interlocutors, (2) the emergence of an imaginary second-
person singular, (3) characteristic rejection phrases, such as μἠ γένοιτο,146 and
(4) characteristic apostrophic vocatives.147 Theoretically, these markers belong
to diatribe and diatribe alone. Thus, for Song, if a text has these categories, it is
either a diatribe or diatribal.
Song’s contribution is insightful, but it requires two qualifications.
First, Song leaves himself open to criticism due to a lack of specificity. Allowing
Song a generous reading, what he seems to mean is that each of these “mark-
ers” come from the primary speaker / author. For instance, one can conceive
of vivid dialogue between fictive characters in a non-diatribal context, such
as the example of speech-in-character discussed in Part 1 from ad Herennium
4.65.148 Second, it should at least be noted that all of Song’s markers belong to
the dialogical aspect of diatribe and do not provide a holistic system by which
to identify diatribe. The significance of Song’s reading of Romans, as with all of
the above sources, of course remains to be seen until Part 3.

142  Song, Reading Romans, 2–3. Song fails to notice Stowers’s own concession that some of
Epictetus’s diatribes contain polemic. Stowers, Diatribe, 55.
143  Song, Reading Romans, 2–3.
144  Ibid., 6.
145  Ibid., 6, 16.
146  Song is heavily indebted to Malherbe, who argues that μὴ γένοιτο, “as it appears as a re-
sponse in a dialogue without being part of a larger sentence, is unique to Epictetus and
Paul.” Malherbe, “Μὴ γένοιτο,” 231–40, quote from 232.
147  Song, Reading Romans, 16.
148  Witherington and Hyatt actually seem to identify ad Herennium 4.65 with diatribe.
Ad Herennium 4.65 represents speech-in-character, not diatribe. Ad Herennium 4.65
depicts fictional characters speaking with one another; there is no dialogue between
the primary speaker / narrator and the hypothetical speakers, which would be character-
istic for diatribe. Witherington and Hyatt, Romans, 74–75. This is, of course, not to say that
the conventions for speech-in-character cannot be used to create and fashion the words
of the diatribal interlocutor, but more on this in Chapter 7.
124 chapter 6

Conclusion

This survey and developing analysis demonstrates that Stowers’s model has
superseded that of Bultmann and continues to represent the general scholarly
consensus.149 This does not mean that Stowers is completely divorced from his
predecessors. The difference is that Stowers puts the pieces together in a way
at which had previously only been hinted.
Concerning the works on diatribe chronologically following Stowers,
Schmeller amounts to something of a methodological detour, while
Thorssteinsson, Tobin, and Song each endorse Stowers as far as the gen-
eral depiction of diatribe is concerned. This is not to say that these scholars
have not made helpful contributions or nuances on various points, but nei-
ther have they successfully recast the nature of diatribe studies for future New
Testament research. Three of them never even aimed to do so. This is also not
to say that Stowers is correct in every single respect, either concerning diatribe
or his application of it to Paul. To be sure, these scholars and others will be
assessed with respect to their application of diatribe to Pauline studies in the
following chapters. Consequently, though it will be prudent to listen to other
scholars at various points in the discussion of diatribal and Pauline texts, the
presentation of diatribe followed in this project is heavily indebted to Stowers.
What, therefore, are the characteristics of diatribe which will be employed
in the remainder of this project? First, the social function of diatribe did not
primarily belong to the Cynic-Stoic propagandistic street preaching for the
masses. Diatribe was most at home in the setting of the philosophical school,
in the teacher-student relationship. In this setting, diatribe was not the formal
instruction or lecture about a given topic; diatribe was the conversation that
occurred after the formal instruction, during which the teacher could clarify
his thoughts and/or correct his students.
Second, one of the most prominent characteristics of diatribe is its imple-
mentation of a dialogical element. The primary speaker often introduces a

149  In addition to the scholars discussed above, the following represents a list of scholars
who adopt Stowers’s presentation of diatribe and do not express any specific interest
in adding to or nuancing Stowers’s presentation. John L. White, The Apostle of God: Paul
and the Promise of Abraham (Peabody: Hendrickson, 1999), 76–77; Rodríguez, If You Call,
36–37; Fitzmyer, Romans, 91; Elliott, Rhetoric, 120n.2; David E. Aune, The New Testament
in its Literary Environment (LEC 8; Philadelphia: Westminster Press, 1987), 200–2; idem.,
The Westminster Dictionary, 127–29; Witherington and Hyatt, Romans, 74–75; Abraham J.
Malherbe, “Hellenistic Moralists and the New Testament,” Principat 26.1 (1992): 313–20,
esp. 313n.222.
Portrayals of Diatribe 125

fictitious interlocutor on to the scene and assigns speech in his, her, or its voice.
The primary speaker may initiate conversation with the interlocutor through
direct address, or the interlocutor can interrupt and address the primary
speaker. This interlocutor can represent a general type of person, a specific per-
son, or a personified object or abstract idea. Often the interlocutor is unidenti-
fied and/or simply a rhetorical device by which the primary speaker advances
his own argument in a particularly measured rhetorical manner. Finally, some-
times the interlocutor’s words are introduced with a basic verb of speech, but
frequently they are not introduced at all. Thus, it is often difficult to determine
when an interlocutor’s speech begins or ends.
Third, Socratic censure and protreptic are significant aspects of the teach-
er’s pedagogical method. Through censure, the teacher is able to point out
contradictions or errors among his students. Through protreptic, the teacher
provides the student(s) with the correct view or course of action and exhorts
them to adopt it. Extending back to Socrates, this method capitalizes on the
use of question and answer as a means of guiding one’s conversation partner
in a particular direction.
Fourth, the tone of diatribe varies. Early work on diatribe considered the
tone to be heavily polemical. Stowers’s reassessment swung the tonal pendu-
lum in quite the opposite direction, so that the tone of diatribe became pri-
marily that of collaboration and education. To this end, Plutarch is particularly
significant in Stowers’s argument. Plutarch informs that, though the censure
might be difficult, the audience should await the pleasant exhortation that
follows.150 Stowers does, however, continue to allow for polemic in diatribe, cit-
ing three of Epictetus’s Discourses as cases in point. But “polemic” in Stowers’s
presentation takes on a very narrow sense; Stowers’s identification of polemic
in Epictetus is limited to instances in which Epictetus specifically targets and
attacks other philosophical movements.151 For instance, in Disc. 1.23 and 2.23
Epictetus attacks Epicurus, and in 2.20 he targets Epicureans and Academics.
But to say that most diatribes are not “polemical” in this minimalistic sense

150  Plutarch represents the ideal student response from the teacher’s perspective. Both
confirming and problematizing Plutarch’s presentation is Epictetus’s Discourse 2.21,
Περὶ ἀνομολογίας. Here, Epictetus implicitly expects students to undergo transformation
through attending his school, but he also documents the reality that not all students are
actually prepared (or interested) to allow themselves to change and be transformed. As
a result, Epictetus critiques such students who are unwilling to question or lay aside the
preconceptions about various topics that they held upon entering his school. See espe-
cially Disc. 2.21.15–22.
151  Stowers, Diatribe, 55n.289.
126 chapter 6

does not demand that their scholastic settings are all fun and collegial games.
Nothing necessitates that a teacher must be “nice” to his or her students. The
scholastic environment can be just as hot and heated as those diatribes Stowers
identifies as “polemical” in his limited sense. As corroborating evidence, one
needs simply to consider Illustrations 1 and 2 below, a painting from the house
of Julia Felix in Pompeii of a student being punished. In this painting, the
student is depicted as being stripped, restrained by two other students, and
whipped by the schoolmaster. In fact, Teles comments to this end as well. Teles,
in Fragment V, writes that if a child survives and escapes his nurse, a litany
of teachers are the next in line to get their hands on him, and “by all of these
he is beaten, scrupulously observed, and forcibly manhandled by the neck”
(ὑπὸ πάντων τούτων μαστιγοῦται, παρατηρεῖται, τραχηλίζεται).
Though the social setting for diatribe might be established as a scholastic
environment, the tone within that environment is certainly malleable to the
specific rhetorical context. To exclude the possibility of more general polemi-
cal tones from diatribe is a confusion of terms likely stemming from Stowers’s
narrow presentation of “polemical diatribes.” This means that in the normal

illustration 1 Painting of a schoolboy being punished from the house of Julia Felix in
Pompeii. This painting is held in the National Archaeological Museum of
Naples, inv. 9066.
Photo taken by Bruce Longenecker.
Portrayals of Diatribe 127

illustration 2 A drawing of the painting of a schoolboy being punished from the house of
Julia Felix in Pompeii, from Delle antichità di Ercolano, volume 3 (Napoli:
Regia Stamperia, 1767), pp. 213.

censure-protreptic progression of scholastically situated diatribe, one must be

open to finding both more negative / pejorative examples as well as more posi-
tive / optimistic instances of diatribal discourse.
Fifth, in this early scholastic setting, diatribe did not represent a literary
genre.152 The forms of the primary sources for diatribe are too diverse to per-
mit such a claim. The sources, however, do share numerous characteristics that
group them together and suggest they belong to a rhetorical genre or category.
Indeed, it is only as a rhetorical category or genre that these texts or their char-
acteristic features would be comprehensible and able to be adapted for and
into literary forms.

152  For scholars who continue to argue that diatribe represents a genuine genre, see Stanley
E. Porter, “The Argument of Romans 5: Can a Rhetorical Question Make a Difference,” JBL
110.4 (1991): 657–59; Song, Reading Romans, 58–62.
128 chapter 6

The following chapter seeks to apply this presentation of diatribe, in com-

bination with speech-in-character, to non-Pauline diatribal texts. In this way,
Chapter 7 represents a demonstration of the method to be employed when
examining Rom 3:1–9 in Part 3. Again, as needed, diatribe will carry much
of the hermeneutical weight, whereas speech-in-character will assist in defin-
ing the limits of the various dialogical exchanges. This combination should
allow significant progress to be made on Rom 3:1–9 and its place in the larger
pictures of Romans and Paul’s thought.
 chapter 7

Examples of Diatribal Dialogue

Introduction

Having addressed speech-in-character in Part 1 and diatribe in Part 2,

Chapter 6, it is now possible to examine how these phenomena operate simul-
taneously in non-Pauline diatribal texts. Though never discussed in any great
depth to my knowledge, the observation that speech-in-character can inform
questions about diatribe, and diatribe questions about speech-in-character,
has not gone entirely unnoticed. For instance, Kustas argues that diatribe’s
“special habit” of attributing speech to a fictitious interlocutor “has achieved
a classification of its own” in the rhetorical treatises, which he identifies
as Ad Herennium’s concept of sermocinationes consequentes—i.e., speech-
in-character.1 Additionally, at times Stowers speaks primarily in terms of dia-
tribe.2 At other times, Stowers speaks more in terms of speech-in-character.3
Nonetheless, these are to be considered mutually informative categories.
Thus, there is a precedent in scholarship for associating diatribe and speech-
in-character in some way, and the present chapter aims to address this
relationship more fully.
The most obvious connection in the relationship between the rhetorical fig-
ure of speech-in-character and diatribe is the practice of crafting and scripting
(or attributing) speech in the voice of another speaker. In this way, it is specifi-
cally the dialogical aspect of diatribe that overlaps with speech-in-character.
This similarity runs deeper than surface level. For example, the identity of the
imaginary speaker varies greatly; in both diatribe and speech-in-character,
the imaginary speaker can be an actual person, a hypothetical person, an
unidentified person, an inanimate object, or an abstract idea. And, in both
diatribe and speech-in-character, the imaginary speaker’s words may or may
not be clearly marked by an introductory formula. Finally, the purpose of dia-
tribe often overlaps with that of speech-in-character. Stowers demonstrates
the basic function of diatribe to be censure and protreptic, pointing out faults
in one’s dialogue partner and/or audience and attempting to correct them.4

1  Kustas, Diatribe, 11–12. See my treatment of sermocinationes consequentes in Part 1.

2  Stowers, Diatribe.
3  Idem., “Apostrophe,” 351–69.
4  Idem., Diatribe, 76–77.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_010

130 chapter 7

Similarly, Quintilian notes that speech-in-character is particularly useful for

revealing the thoughts of an opponent or introducing appropriate characters
for specific rhetorical situations (Inst. 9.2.30), and Theon lists protreptic as one
of the primary uses of the exercise (Prog. 115.20–22). The attributions of speech
in diatribe and speech-in-character, therefore, reflect a remarkable similarity.
Indeed, each utterance of an imaginary, diatribal interlocutor represents the
use of speech-in-character, as the primary speaker or writer must craft appro-
priate speech for the dialogue partner in order to move the discourse in a mea-
sured rhetorical direction.
This is not to say that one might not uncover differences between the
dialogical sections of diatribe and speech-in-character. The very first example
discussed below (Teles, Fragment 1, “On Seeming and Being”) deviates from
the norm of speech-in-character by implementing a more complex and pluri-
form characterization than is typical for speech-in-character in the rhetorical
treatises addressed in Part 1. Moreover, it is unclear whether an interlocutor is
even on the scene until half way through the dialogue.
The biggest divergence between diatribe’s dialogues and speech-in-
character, however, is of a related but somewhat different sort. As discussed
in Part 1, speech-in-character contains several tools by which readers and
auditors can usually cue in to the presence of an imaginary speaker. More
specifically, two highly common aspects of speech-in-character can assist
readers not only to identify that speech is being attributed to an imaginary
speaker, but also to deduce the extent of the attributed speech. Thus, the reader
or auditor can best understand the precise exchanges in the script of the dia-
logue and who is to be understood as responsible for speaking them. These
two features are the tendency to provide a characterization of the imaginary
speaker and the unanimous rule that the attributed speech must be appropri-
ate vis-à-vis the characterization in order to be rhetorically effective. In theory,
therefore, even if there are no overt markers distinguishing the limits of a given
dialogical exchange (such as an introductory verb of speech), one should often
be able to define the boundaries within the script by measuring any phrase
against the characterization of the speaker(s) imaginatively present.
On the contrary, diatribe benefits from no such guidelines concerning how
to determine the extent or the voice of individual lines in a discourse, and this
is for several reasons. First, multiple treatments of rhetorical theory directly
address speech-in-character, but no primary literature systematically discusses
diatribe and unpacks its presumed or prescribed conventions. Such guidelines
simply do not exist.5 Second, because no handbooks of diatribal theory exist,

5  At least, no such evidence has been discovered.

Examples of Diatribal Dialogue 131

scholars must extrapolate diatribal tendencies from primary diatribal texts.

When scholars make these extrapolations, the conclusion is regularly that the
manners of creating dialogical exchanges in diatribe are diverse, inconsistent,
loose, unsystematic, and so forth. Of course, there are various introductory
formulas that often introduce an interlocutor’s speech (such as φησί or inquit)
and stock phrases that are frequently placed on the interlocutor’s lips (such
as τί οὖν or statements beginning with ἀλλά), but none of these tenden-
cies are employed consistently. An introductory verb of speech very well
might introduce an interlocutor’s words, but it is equally acceptable for the
interlocutor simply to enter the conversation without any indication. Similarly,
Epictetus can use identical forms (such as τί οὖν or μὴ γένοιτο) for his own words
as well as for the words of his interlocutors, which creates a “looseness and
variability of their usage.”6 Third, heretofore, speech-in-character has typically
been ignored or not afforded the appropriate weight in terms of analyzing
diatribe’s dialogical exchanges. As discussed in Part 3, Stowers’s contributions
are the most helpful in this respect, but they are not without fault and more
remains to be said.
This project narrows the gap between diatribe’s “inconsistent and loose”
dialogical pericopae and speech-in-character’s (typically) more systematic
attributions of speech by allowing the conventions set forth for speech-in-
character to offer insight into diatribe’s dialogues, specifically with respect to
the shape of the script. The merit of the method will of course be borne out
in the exegesis of primary texts. The method will doubtlessly work better for
some texts than it will for others. For instance, in Fragment III, Περὶ φυγῆς,
lines 9–31, Teles quotes a diatribal dialogue from Stilpon. In this discourse, no
characterization of the interlocutor occurs (though perhaps it would have been
present in the original context of the dialogue), and the extent of the interlocu-
tor’s comments amounts to nothing more than abrupt agreements with each
of Stilpon’s leading questions.7 As it fortuitously turns out, the script of this
particular discourse is quite clear. If, however, the script were more complex
or difficult to deduce, speech-in-character would have a troubled time assist-
ing in one’s understanding of the conversation due to the lack of characteriza-
tion and the limited amount of speech attributed to the interlocutor. When
the necessary components are present, however, the conventions governing
speech-in-character function as tools that are able to improve understandings
of diatribal discourse significantly (see further below).

6  Stowers, Diatribe, 128–29, 160. Cf. Malherbe, “Μὴ Γένοιτο;” Song, Reading Romans, 32–37.
7  The interlocutor’s respective responses are: οὐ δή, line 13; οὐδὲ τοῦτο, line 16; and καὶ μάλα,
line 21.
132 chapter 7

Under the assumption that the texts discussed below in one way or another
belong to or evince features of a diatribal rhetorical category, diatribe will be
alotted the primary hermeneutical role in terms of how these texts should
be understood. From the perspective of diatribe, four questions will be
posed to each text: (1) what is the identity of the interlocutor, (2) what is the
function of the interlocutor, (3) is the scholastic tone more positively collegial
and collaborative, or is it more negative or generally polemical, and (4) does
the argument employ censure, protreptic, or both? At the same time, the
conventions for speech-in-character—especially characterization and the ap-
propriateness of the attributed speech—will be referenced in order to define
and/or confirm as precisely as possible the various exchanges within the dia-
logical script under examination. I also consider whether speech-in-character
offers any additional information regarding the identity of the interlocutor or
the function of the attributed speech and/or dialogue. In these ways, speech-
in-character largely defines the limits of the content to be analyzed, while
diatribe generally suggests how that content is to be best interpreted. The fol-
lowing three examples of diatribal dialogue serve as models for the proposed
method.

Dialogue and Attributed Speech in Primary Sources for Diatribe

Teles, Fragment I: Περὶ τοῦ δοκεῖν καὶ τοῦ εἶναι

Overview. In “On Seeming and Being,” which is Fragment I of Teles’s extant writ-
ings, Teles engages in conversation with an imaginary interlocutor concerning
whether it is better simply to seem to have some quality or attribute or actually
to possess or embody it. Teles begins the conversation as follows:

Κρεῖττόν φασι τὸ δοκεῖν δίκαιον εἶναι τοῦ εἶναι· μὴ καὶ τὸ δοκεῖν ἀγαθὸν εἶναι
τοῦ εἶναι κρεῖττόν ἐστιν;

Some claim [φασι] that seeming [τὸ δοκεῖν] to be [εἶναι] just [δίκαιον]
is better than being [τοῦ εἶναι] just. Seeming to be good [ἀγαθὸν] is not
better than being good, is it? (I.1–3)

Introducing the subject matter for the remainder of the discourse, Teles’s
opening statement posits that some unidentified people affirm the notion that
seeming to be something is better than truly being so. Though Teles could have
easily used direct speech to record the affirmation of these claimants, his use
Examples of Diatribal Dialogue 133

of the infinitive (τὸ δοκεῖν) indicates that he is reporting indirect rather than
direct speech.8 The whole opening is to be heard in Teles’s own voice.
Following from his topical cue in the opening statement, Teles next poses
a leading question to his interlocutor, petitioning the interlocutor to affirm or
deny whether it is better to seem to be good or actually to be good (2–3). The
interlocutor is never identified in the discourse, nor do any verbs of speech
overtly distinguish the interlocutor’s words from Teles’s. At first glance, one
could be forgiven for wondering whether Teles’s question was simply rhetori-
cal, so that both the question and answer would be spoken in Teles’s own voice.
Teles, however, makes it clear elsewhere that he is in fact engaging an inter-
locutor. Teles addresses the interlocutor in the second person singular (βούλοιο,
line 19, 21; βούλει, line 32), and he scripts the interlocutor as responding in
the first person singular (βουλοίμην, line 29) and referring to himself with the
first person singular personal pronoun (τιμήσουσί με, line 33). Since it is clear
elsewhere in the discourse that an interlocutor answers Teles’s questions, it is
justifiably argued that the response to Teles’s first question is also to be heard
in the interlocutor’s voice. Thus, in this instance of diatribal dialogue, the pri-
mary speaker, Teles, poses a leading question to an unnamed, imaginary inter-
locutor. Teles then composes a response in the interlocutor’s voice.9
The interlocutor’s scripted response, ἀμέλει, rejects the possibility that seem-
ing is somehow better than being (4). As such, the interlocutor and Teles are
in agreement, and this consensus gives rise to a series of three back and forth
questions and answers. Teles ponders whether people are good actors because
they seem to act well or because they actually act well (5–6), whether people
play the cithara well because they seem to play it well or because they actu-
ally play it well (8–9), and whether it is generally the case that, with respect to
whatever people do well, they simply seem to be good or actually are good at
that particular undertaking (11–12). To each of these scenarios, the interlocutor
responds, διὰ τὸ εἶναι (7, 10, 13). Consequently, at every turn to this point in the
dialogue, the interlocutor agrees with Teles and affirms that people excel at
various skills not because they seem to be good but because they actually are
good.10 So far, seeming is not better than being.
Having agreed that being good is better than seeming to be good (14–
17), Teles proceeds by firing a litany of questions towards the interlocutor

8  Smyth, Greek Grammar, §2016–17, 2019–21, 2026–27. So also O’Neil, Teles, 3.

9  So also Stowers, “Paul’s Dialogue,” 711.
10  Stowers rightly observes, “Teles asks questions which the student or interlocutor can only
answer as Teles wishes.” Ibid.
134 chapter 7

concerning whether it is preferable to see or simply seem to see, to be well or

merely seem to be well, and so forth (17–25). Teles’s question of whether it is
preferable to be confident, fearless, and courageous (ἀνδρεῖος) or only to seem
to be so, however, brings the dialogue partners’ agreement to a halt. Given the
interlocutor’s agreement that being good is better than simply seeming to be
good, one would expect the interlocutor to respond in like kind here. Quite the
opposite, the interlocutor answers in the first person,

Ἀλλὰ καὶ ἀνδρεῖος ἂν μᾶλλον βουλοίμην δοκεῖν ἢ εἶναι.

I would prefer to seem courageous rather than to be courageous. (I.29)

Contrary to the thrust of the conversation, the interlocutor now chooses seem-
ing over being.
Naturally, Teles probes further and asks the interlocutor why he would
prefer seeming courageous to actually being courageous (31–32). The interloc-
utor responds, because “they will honor me” (33), at which point Teles takes
the opportunity to inform his imaginary conversation partner of the dire out-
comes of such a position. Teles hypothesizes that, because they (presumably
soldiers) believe the interlocutor to be courageous, they will place him in the
front of the battle formations and conscript him to fight in one-on-one combat
(34–35). Furthermore, when the interlocutor is taken captive, the enemy will
also perceive him to be courageous and able to endure considerable suffering.
They will bind him, lock him away, torture him, stretch him on the racks, and
burn him over a fire (35–45). All of these maladies will come upon the inter-
locutor because he feigned to be courageous, hiding his true character, which
Teles likens to the rhetoricians (45–47).
Diatribe. As mentioned above, the identity of the imaginary interlocutor
is entirely unspecified. Additionally, the words Teles attributes to the inter-
locutor are never introduced with verbs of speech. The clinching pieces of
evidence that an interlocutor is on the scene are Teles’s references to the in-
terlocutor in the second person and the interlocutor’s references to himself in
the first person. Furthermore, it is unclear whether the interlocutor represents
one of Teles’s students or more generally a certain type of person. Assuming
that Teles’s leading question is directed to a schoolroom setting, it is perfectly
plausible that the interlocutor’s responses could represent a specific student’s
thoughts or concerns. It is also plausible that the interlocutor’s responses sim-
ply represent a general type that Teles uses rhetorically in order to model for
his students a particular lesson about authenticity and hypocrisy. Whether
addressing a particular student or the class as a whole, the interlocutor nev-
ertheless functions to provide Teles with a character to engage with within a
Examples of Diatribal Dialogue 135

specific rhetorical situation. Through this engagement, Teles is able to demon-

strate the various ways seeming to be something is deficient when compared
to actually being or possessing some trait, and vice versa.
Moving forward, if censure and/or protreptic are the primary argumenta-
tive strategies in diatribe, how does Teles utilize them in this discourse? The
presence of censure is unmistakable. Again, censure (ἐλεγτικός, ἐλέγχειν) is
the feature by which the teacher or primary speaker exposes errors and con-
tradictions endorsed by his students or the imaginary interlocutor. Here, the
trajectory of the interlocutor’s responses models Teles’s use of censure per-
fectly. The interlocutor’s first four responses agree wholly with Teles’s view that
being is better than merely seeming to be. Through Teles’s continued probing
and questioning, however, the interlocutor reveals that he ultimately wishes
to seem courageous rather than actually to be so. Thus, Teles’s use of censure
points out a contradiction between what the interlocutor consents to early in
the discourse (i.e., being is greater than seeming) and what he later discloses
that he secretly desires for himself (i.e., merely to seem to be courageous).
Protreptic argumentation (προτρεπτικός, προτρέπω), on the other hand,
offers the correct course to pursue and encourages one’s audience to follow
it. Such protreptic argumentation is less pronounced in Fragment I. Once
the interlocutor has revealed his inner contradiction and broken consensus
with Teles’s point of view, Teles offers no positive protreptic argument to per-
suade the interlocutor to pursue some other—better—course of action. What
Teles does is to present a fully apotreptic (ἀποτρεπτικός, ἀποτρέπω) argument.
Apotreptic argumentation is characterized not by an attempt to persuade pos-
itively (i.e., protreptically) towards some goal, but by an attempt to dissuade
negatively away from something (see Theon, Prog. 116.22–117.6). In this man-
ner, Teles outlines the numerous tragedies that will befall his interlocutor due
to his desire simply to seem courageous (I.34–45). Such apotreptic argumen-
tation would, Teles would hope, challenge the interlocutor’s reasoning and
implicitly suggest that he should abandon such a view. Thus, though it is not
explicitly protreptic, technically speaking, Teles nevertheless maintains in “On
Seeming and Being” a censure-exhortation (in the form of implicit dissuasion)
progression aimed at confronting and correcting his interlocutor’s error.
Finally, what is the tone of the diatribe? Does Teles engage with his interloc-
utor in a collegial and optimistic manner, does he treat his imaginary dialogue
partner more pejoratively, or does he engage in polemic in the narrower sense
employed by Stowers? Three features of the dialogue suggest that the tone
should be heard in a more pejorative, if not narrowly polemical, sense. First,
Teles never suggests that the interlocutor actually abandons his desire to seem
courageous, comes full circle, and rejoins Teles’s point of view. This means Teles
and his interlocutor have not worked collegially or collaboratively (at least not
136 chapter 7

successfully) in order to bring the interlocutor to the presumed correct way

of thinking. As far as the dialogical evidence is concerned, the interlocutor
remains in his folly.11 Second, though it would implicitly suggest the need for
reform, the apotreptic argumentation is thoroughly negative, only highlighting
the terrible results that would follow from the interlocutor’s foolish point-of-
view.12 Third, and quite suggestive, Teles derides the interlocutor for putting
forth a false presentation of himself (I.45–47). Such hypocritical self-presenta-
tion, Teles avers, associates the interlocutor with the rhetors (ὥσπερ οἱ ῥήτορες;
I.47). Teles’s disdain for the rhetors as an opposing or anti-philosophical group,
and his association of the interlocutor with them, is similar to Epictetus’s treat-
ment of the Epicureans and Academics (Disc. 2.20; see below), which Stowers
identifies as polemic.13 Consequently, though the censure-implicit exhortation
argumentation of the scholastic setting is present, “On Seeming and Being”
reminds readers that diatribe is fully capable of donning generally negative
and even narrowly polemical moods, even within the student-teacher relation-
ship of the scholastic setting.
Speech-in-Character. But what about speech-in-character? For all intents and
purposes, the script of “On Seeming and Being” is quite straightforward,
and speech-in-character is not really needed in order to help define it any fur-
ther. For this reason, however, “On Seeming and Being” is a great entry point,
as it constitutes a relatively stable control against which to test the validity
of speech-in-character as a means of further understanding diatribe and the
degree to which the two overlap.
Because the speaker is never identified in any way, the first items to con-
sider with respect to speech-in-character are the speaker’s characterization
and the appropriateness of the attributed speech. In the entire discourse, Teles
only once comments directly about the character of the interlocutor, and that
comment comes only after the interlocutor’s final words. In the onslaught of
difficulties that Teles suggests merely seeming to be courageous will produce,
Teles writes:

11  Based on the extant evidence, this must be the conclusion. If the text were not a frag-
ment, however, it is of course possible that the dialogue continues and Teles eventually
persuades the interlocutor to abandon his desire to seem to be courageous. Until new
copies of Teles surface, however, one must do one’s best with the extant fragments.
12  The apotreptic argument would have a similar implicit effect on Teles’s students / audi-
ence. The litany of tragedies that Teles suggests would befall the hypocritical interlocutor
serve as a model aimed at dissuading others from following the interlocutor’s manner of
thinking.
13  Stowers, Diatribe, 55n.289.
Examples of Diatribal Dialogue 137

εἶτα τί οἴει πείσεσθαι δειλὸς μὲν ὤν, κινδυνεύων δέ;

Then, what do you think you will experience, because you are a coward
(δειλός) and are in danger? (I.37–38)

Teles characterizes the interlocutor ultimately as a coward. This cowardly char-

acter trait makes fine sense of the interlocutor’s preference for seeming to be
courageous rather than actually being courageous (I.29–30). For, as Teles clari-
fies, a truly courageous person is both fearless and painless (ἤ οὐχ ὁ ἀνδρεῖος
καὶ ἄφοβος καὶ ἄλυπος, οὐχ ὁ δοκῶν; I.31–32). The interlocutor cannot actually
be courageous since he would then have to be fearless as well, which is impos-
sible because he is a coward. The interlocutor’s speech in the latter half of the
discourse is entirely appropriate in light of the characterization.
More difficult is the question of how to explain the appropriateness of the
interlocutor’s speech in the first half of the dialogue. Since Teles later discloses
that the interlocutor is a coward who prefers seeming over being courageous,
how can the interlocutor’s early agreements with Teles be considered appropri-
ate? There is no explanation for why the interlocutor might agree that being is
better than seeming with respect to certain realities but not others, nor is there
any hint that the interlocutor has simply changed his mind. The best solution
seems to be that the diatribal rhetorical category has superseded the singular
characterization typical of speech-in-character.14 Said otherwise, the very na-
ture of diatribe can implicitly assume a particular characterization, namely,
inner inconsistency and contradiction in the person of the interlocutor.15 If a
particular diatribe’s primary aim is to expose and/or root out error and incon-
sistency, then, by necessity, the interlocutor must display inconsistency. Thus,
in this instance, speech-in-character, which usually operates based on a sin-
gle characterization, is put to the service of diatribe, which here requires two
characterizations, one of which is assumed (inconsistency and contradiction),
and one of which is explicit (coward). In this way, when viewed through the
overarching lens of diatribe, all of the interlocutor’s speech can be considered
appropriate as far as speech-in-character is concerned.

14  By “singular characterization,” I mean that speech-in-character typically functions based
on a single or fixed characterization of the imaginary speaker rather than multiple and/or
contradictory characterizations. For examples, see the engagement with actual speeches-
in-character in Part 1.
15  For characterization in diatribe, see also Stowers, Diatribe, 106–10.
138 chapter 7

Speech-in-character also parallels diatribe in terms of the function of the

attributed speech. In the rhetorical context of the diatribe, Teles’s interlocu-
tor provides him with a conversation partner against whom he can express
to his class his views about seeming and being. Such a use of the interlocutor
maps seamlessly with Quintilian’s allowance that speech-in-character is quite
useful for “[providing] appropriate characters for words of advice, reproach,
complaint, praise, or pity” (Inst. 9.2.30 [Russell, LCL]). In both conversations—
diatribe and speech-in-character—the interlocutor serves an identical
rhetorical function. As a result, with respect to the attributed speech, char-
acterization, appropriateness, and the function of the imaginary speaker,
speech-in-character fits quite snuggly inside the diatribal glove.16

Epictetus, Discourses 2.20: Πρὸς Ἐπικουρείους καὶ Ἀκαδημαϊκούς

Introduction. The second example of diatribal discourse to engage is Epictetus’s
“Against Epicureans and Academics” (Disc. 2.20). Here, Epictetus addresses
the basic view taught within the Middle Academy. Beginning with Arcesilaus
(3rd century BCE), Plato’s Academy entered into a period of philosophi-
cal skepticism, which denied the possibility of absolute knowledge. Instead,
the Academy’s leaders insisted on delaying cognitive judgment.17 Epictetus
strongly disagrees with these skeptical sentiments. In the interest of demon-
strating the validity of knowable, absolute truths, Epictetus begins by setting
forth his primary premise:

Τοῖς ὑγιέσι καὶ ἐναργέσιν ἐξ ἀνάγκης καὶ οἱ ἀντιλέγοντες προσχρῶνται.

Even those who oppose sound and obvious propositions use them by
necessity. (Disc. 2.20.1)

That is, even critics of absolute and knowable truth must endorse the prin-
ciple in order to reject it, and Epictetus considers this to be one of the stron-
gest proofs effectively verifying the reality of such truth (2.20.1). To support his

16  If Teles had brought the interlocutor around full circle and placed apotreptic speech on
the interlocutor’s lips rather than only in his own, another point of connection between
diatribe and speech-in-character would be Theon’s discussion of starting places for apo-
treptic, dissuasive argumentation. See Theon, Prog. 116.22–117.6.
17  Epictetus, Discourses: Books I–II (trans. W. A. Oldfather; LCL 131; Cambridge: Harvard
University Press, 2000), 360n.95.
Examples of Diatribal Dialogue 139

position, Epictetus enters into four dialogical exchanges, three of which will be
discussed at this time.18
The first dialogical exchange: Disc. 2.20.1–5. In the first dialogical exchange,
Epictetus enlists a number of interlocutors and exposes how each of their cri-
tiques of knowable truth are in fact statements proving such truth. Epictetus
introduces each new interlocutor. First, Epictetus suggests that anyone who
denies that any true universal statement exists must also affirm the opposite.
Epictetus writes,

… δῆλον ὅτι τὴν ἐναντίαν ἀπόφασιν οὗτος ὀφείλει ποιήσασθαι· οὐδέν ἐστι
καθολικὸν ἀληθές.

… it is clear that (the person who rejects the existence of knowable truth)
[οὗτος] is obligated to affirm the opposite and say [ἀπόφασιν], ‘No univer-
sal statement is true.’ (2.20.2)

Epictetus promptly opposes the interlocutor, referring to him as a captive

(ἀνδράποδον), and clarifies that such a confession ultimately degenerates into
the view that there is no absolute truth, whereby every universal statement is
necessarily false (2.20.3). Epictetus then imaginatively scripts three additional
interlocutors speaking to the same end as the first interlocutor. The second
interlocutor19 declares,

γίγνωσκε, ὅτι οὐδέν ἐστι γνωστόν, ἀλλὰ πάντα ἀτέκμαρτα.

Know that nothing is knowable; rather, all things are uncertain. (2.20.4)

The third interlocutor20 remarks,

πίστευσόν μοι καὶ ὠφεληθήσῃ·οὐδὲν δεῖ ἀνθρώπῳ πιστεύειν.

Trust me and you will be benefited; one must trust a person in no way
whatsoever. (2.20.4)

18  The fourth dialogical exchange is part of a forthcoming article. As Chapter 7 of this proj-
ect primarily aims to model the ways speech-in-character and the dialogical pericopae in
diatribe correlate, omitting this particular dialogical exchange does no disservice to the
current project.
19  Epictetus introduces the second interlocutor with the formula, ἄν τις … λέγῃ.
20  Epictetus introduces the third interlocutor with the formula, ἢ ἄλλος.
140 chapter 7

The fourth21 states,

μάθε παρ’ ἐμοῦ, ἄνθρωπε, ὅτι οὐδὲν ἐνδέχεται μαθεῖν·ἐγώ σοι λέγω τοῦτο καὶ
διδάξω σε, ἐὰν θέλῃς.

Learn from me, human, that nothing is possible to learn; I am telling you
this, and I will teach you, if you are willing. (2.20.4–5)

Epictetus then explains that there is no difference between these four interloc-
utors and a fifth, the Academics (οἱ Ἀκαδημαϊκοὺς αὑτοὺς λέγοντες), who claim,22

ὦ ἄνθρωποι, συγκατάθεσθε ὅτι οὐδεὶς συγκατατίθεται· πιστεύσατε ἡμῖν ὅτι

οὐδεὶς πιστεύει οὐδενί.

O people, agree completely (to the view) that no person can agree com-
pletely (to anything); trust us that no person can trust anyone. (2.20.5)

Consequently, in their respective attempts to deny the validity of knowable

truth, Epictetus illustrates that all five interlocutors nevertheless ironically and
hypocritically affirm the principle of knowable truth by employing it.
Transitioning to speak in terms of diatribe and the questions being posed
to these texts, the identity of the interlocutor is clear enough. Epictetus spe-
cifically identifies the fifth iteration of the interlocutor as a body of skeptical
“Academics” (2.20.5). Though Epictetus does not precisely identify the first
four interlocutors, he suggests that there is no discernible difference between
their proclamations and that of those he clearly identifies as Academics, so
it is reasonable to group all five interlocutors within an overarching skeptical
Academic umbrella. As such, the interlocutors represent Academics and the
like who are skeptical about humanity’s ability to know. Furthermore, for this
reason, they are almost certainly to be differentiated from Epictetus’s actual
classroom audience.
The function of the dialogical exchange is twofold. First, enlisting the
Academic interlocutors affords Epictetus an avenue for illustrating the ways
rejections of knowable, absolute truth nevertheless require statements of such

21  Epictetus introduces the fourth interlocutor with the formula, ἢ πάλιν ἄλλος.
22  Epictetus offers no verb of speech to introduce the Academics’ words. The textual fea-
tures that suggest the Academics speak these lines include the use of the first-person
plural personal pronoun (ἡμῖν), and the similarity between the scripted words with those
of the other interlocutors, which Epictetus suggests should have no dissimilarities.
Examples of Diatribal Dialogue 141

truth in order to make their respective arguments. As a result, the interlocu-

tors’ comments buttress Epictetus’s principal premise. Second, the use of the
interlocutors allows Epictetus to set forth the Academics’ skeptical sentiments
about knowledge and truth. This, in turn, provides Epictetus with a contrast-
ing view against which to pose his own particular views about knowledge and
truth for the consideration of his actual audience.
On the one hand, it is somewhat premature to begin to speak about the
tone of the diatribe, since much still remains to be discussed. On the other
hand, certain features need to be addressed that begin to push the discourse
in a particular direction. First, Epictetus’s engagement with an opposing anti-
philosophical group fits within Stowers’s narrow presentation of diatribal
polemic. In fact, Stowers directly cites Disc. 2.20 as a leading example of such
polemic.23 Second, Epictetus makes no collaborative attempt to engage the
interlocutors in order to help, cure, or correct what he perceives to be their
misconceptions about human knowledge. Quite the contrary, Epictetus leaves
the interlocutors in, and uses them as examples of, their folly. Consequently,
Epictetus’s tone towards the interlocutors up to this point is far from collabora-
tive and collegial and borders on straight polemic.
Implicit in the discussion about tone are observations about Epictetus’s
method of argumentation in this exchange. Again, Epictetus makes no attempt
to help or cure the Academics; up to this point, no protreptic exhortation
is present. Censure, however, is rampant at every turn. Epictetus’s primary
premise is based on the observation that those who deny knowable truth nec-
essarily contradict themselves, and each interlocutor’s statement functions as
an illustration of this self-contradiction. Thus, Epictetus points out the inter-
locutors’ inner contradictions and constructs this whole section of the diatribe
on a series of five censures.
Given the structure of 2.20.1–5, in which each interlocutor is introduced
practically one after the other, bringing in speech-in-character as a conver-
sation partner is not really necessary, though it will be quite significant in
subsequent conversations. In order to model once more the way speech-
in-character and diatribe correlate, note the following brief observations.
There is an introduction of each interlocutor, either by way of a verb of speech,
or by way of referencing the new subject of an implied verb of speech (for
instance, ἢ ἄλλος, 2.20.4; οἱ Ἀκαδημαϊκοὺς αὑτοὺς λέγοντες, 2.20.5). Additionally,
Epictetus provides a characterization that covers all five interlocutors,
namely, those who reject knowable and absolute truth but nevertheless rely
upon it (2.20.21). Furthermore, all five instances of speech attributed to the

23  Stowers, Diatribe, 55n.289.

142 chapter 7

interlocutors appropriately fit the characterization, which is to be expected,

since they are examples intended to prove the characterization. So, all of
the primary features of speech-in-character are present: (1) identification,
(2) characterization, (3) appropriate attributed speech-in-character. Finally,
as in the example from Teles’s “On Seeming and Being,” Quintilian’s allow-
ance that speech-in-character effectively portrays the thoughts of one’s
opponents and introduces characters for words of reproach or complaint
(Inst. 9.2.30) plainly overlaps with the function of the interlocutors up to this
point in Epictetus’s diatribe. Therefore, diatribe and speech-in-character
collaboratively speak to the same results.
The second dialogical exchange: Disc. 2.20.6–20. Epictetus does not stop with
the Academics. From Epictetus’s point of view, Epicurus’s methodology is simi-
larly suspect, as he also capitalizes on the very principle he intends to dismiss
(2.20.6). Though Epicurus attempts to argue that rational people cannot share
natural fellowship with one another, he expects such rational people neverthe-
less to share fellowship with himself, and, as Epictetus accuses him, his own
concern for others’ thoughts even models such fellowship. In the second dia-
logical exchange, Epictetus writes,

τί γὰρ λέγει; μὴ ἐξαπατᾶσθε, ἄνθρωποι, μηδὲ παράγεσθε μηδὲ διαπίπτετε· οὐκ

ἔστι φυσικὴ κοινωνία τοῖς λογικοῖς πρὸς ἀλλήλους· πιστεύσατέ μοι. οἱ δὲ τὰ
ἕτερα λέγοντες ἐξαπατῶσιν ὑμᾶς καὶ παραλογίζονται.

For, what does (Epicurus) say [λέγει]? ‘Do not be deceived, people, nor
mislead; do not fall away. There is no natural fellowship [φυσικὴ κοινωνία]
between rational people with one another. Believe me. Those who say
otherwise are deceiving and defrauding you.’ (2.20.7)

At once, Epictetus attacks Epicurus’s view. Epictetus asks, if there really is no

natural fellowship between rational people, why does Epicurus worry about
what others think regarding the presence or absence of natural fellowship
(2.20.8–14); why not rather maintain an Epicurean laxity and simply “eat, drink,
have intercourse, defecate, and snore” (2.20.10)? Epictetus’s answer is that it
was the strongest force within people—nature—that drew Epicurus to write.

τί γὰρ ἄλλο ἢ τὸ πάντων τῶν ἐν ἀνθρώποις ἰσχυρότατον, ἡ φύσις ἕλκουσα ἐπὶ τὸ

αὑτῆς βούλημα ἄκοντα καὶ στένοντα;

For, what (was it) other than that which is the strongest of all things
within people—nature (φύσις)—which compels the unwilling and
groaning person to do her will? (2.20.15)
Examples of Diatribal Dialogue 143

On a surface level reading, it is unclear whether Epictetus intends the immedi-

ately following words to be heard in his own voice or in the voice of personified
Nature. Epictetus writes,

ὅτι γὰρ δοκεῖ σοι ταῦτα τὰ ἀκοινώνητα, γράψον αὐτὰ καὶ ἄλλοις ἀπόλιπε καὶ
ἀγρύπνησον δι’ αὐτὰ καὶ αὐτὸς ἔργῳ κατήγορος γενοῦ τῶν σαυτοῦ δογμάτων.

For, since these anti-fellowship principles [ταῦτα τὰ ἀκοινώνητα] seem

good to you (i.e., Epicurus), write them and leave them behind for
others and lie awake thinking because of them and, in practice, you your-
self become the critic [κατήγορος] of your own dogmas. (2.20.16)

Oldfather, for instance, uses quotation marks to place these lines in the
mouth of personified Nature. Epictetus, however, does not introduce these
words with a verb of speech indicating the presence of an interlocutor, nor do
any grammatical aspects of the text strictly require Nature to be the subject
speaking. Consequently, Disc. 2.20.16 represents a prime example by which to
test the proposed method of allowing the conventions for speech-in-character
to inform the dialogical script of a diatribal text. What can be said at this point
is that Epictetus, whether in his own voice or in the voice of personified Nature,
uses these words in order to point out the contradiction in Epicurus’s logic,
so that Epicurus becomes his own critic; being concerned with the thoughts
of others essentially undermines Epicurus’s whole rejection of natural fel-
lowship. Moreover, Epictetus informs, such a natural appeal is altogether
inescapable, just as it is impossible for the Academics to remove their sensory
perceptions by which they can truly know (2.20.17–20). Both the Academics
and Epicurus inevitably affirm the issue they set out to reject, and they do so
by ignorantly employing that very principle in attempting to make their case.
There are, therefore, two occasions within this pericope that deserve
the attention of speech-in-character: Epicurus’s speech and the potential
speech of Nature. Epicurus’s attributed speech is quite easy to address. With
the exception of the different identity of the imagined interlocutor, the func-
tion, tone, and argumentation surrounding Epicurus’s speech are identical to
the conclusions deduced above with respect to the Academics. If for no other
reason, this is supported by Epictetus beginning the discussion about Epicurus
with οὕτως, suggesting a continuation of the previous argument along similar
lines (2.20.6). In this vein, the function of Epicurus as an interlocutor simply
provides yet another example of someone who employs and ignorantly affirms
the same principle he hopes to reject. Epictetus clarifies how this is the case
by petitioning Epicurus to explain how he can mentally reject natural fellow-
ship yet practically maintain a concern for what others think about natural
144 chapter 7

fellowship (2.20.7–14). Furthermore, Epictetus makes no attempt to correct

Epicurus’s logic; Epictetus opts only to censure Epicurus’s self-contradictory
viewpoint. The conventions for speech-in-character also operate on the same
lines as in the previous engagement with the Academics; characterized by con-
tradiction and utilizing the very principle he sought to reject, Epictetus places
appropriately contradictory words on Epicurus’s lips as a way to indicate the
thoughts of his opponent and to provide a character for words of reproach.
The place where speech-in-character begins to demonstrate its usefulness
as a tool for explaining scripted speech in diatribe involves the potential dis-
course of Nature (φύσις; 2.20.15–16). As above, it is unclear whether the lines
following Epictetus’s introduction of nature are best heard in Epictetus’s own
voice or in the voice of personified Nature (2.20.16). Textually, there are no
introductory verbs of speech or suggestions that another speaker has began
to speak, nor, grammatically speaking, are there any verbal forms or pronouns
(or any other indicators) within the speech that cue the audience into the use
of attributed speech.
Considering the text in terms of characterization and whether or not the
scripted speech is appropriate, however, is quite indicative of the usefulness
of the conventions for speech-in-character. When Epictetus begins to discuss
nature as the force capable of overpowering Epicurus’s mental aspirations,
he describes nature as “the strongest of all the forces within people,” which
“compels the unwilling and groaning person to do her will” (2.20.15). Assuming
for the moment that this description amounts to a characterization typical of
speech-in-character (see Part 1), it is possible to consider whether Disc. 2.20.16
would be appropriately scripted for personified Nature, or whether it would
better fit Epictetus’s voice. As it turns out, 2.20.16 is altogether better suited
for the voice of personified Nature. Epictetus implicitly characterizes himself
as one who has no power to compel Epicurus to act one way or another; this
power is largely within the sphere of Nature (2.20.15). The combination of four
imperatives addressed to Epicurus in the second person (γράψον, ἀπόλιπε,
ἀγρύπνησον, γενοῦ; 2.20.16), therefore, would have no effect on Epicurus’s deci-
sions if they are to be heard in Epictetus’s voice. Contrastingly, Nature, as the
strongest force which characteristically draws unwilling and begrudging par-
ticipants to do her will, could altogether appropriately command Epicurus to
act in this way or that. Thus, personified Nature can most appropriately speak
2.20.16, not Epictetus. This means the conventions of characterization and
appropriateness that are intrinsic to speech-in-character strongly suggest that
2.20.16 represents speech attributed to an abstract or inanimate object and, in
this way, constitute the primary evidence governing this portion of the script
Examples of Diatribal Dialogue 145

of the diatribe’s dialogical discourse. The following first person plural, φῶμεν
returns the discourse to Epictetus’s voice (2.20.17).
Overlap between Disc. 2.20.15–16 and speech-in-character, however, goes
further than characterization and the appropriateness of the attributed
speech. Recall that Theon’s elaborate treatment of appropriateness involves
not only the speaker but numerous other elements as well, including the audi-
ence of the attributed speech (Prog. 115.22–27). Personified Nature addresses
Epicurus as one who holds “anti-fellowship principles” (Disc. 2.20.16) and com-
mands him to write them down and leave them for others. This is precisely
how Epictetus depicts Epicurus (2.20.6–9). The attributed speech is not only
appropriate for the personified speaker, but it is also appropriate for the attrib-
uted speech’s target audience.
Having established the script, how then shall Epictetus’s use of per-
sonified Nature be understood in terms of diatribe? First, the function of
personified Nature as an interlocutor is entirely different from the other inter-
locutors discussed to this point. Personified Nature does not afford Epictetus
an opponent against whom to argue but rather a fellow critic. Nature further
explains and censures Epicurus’s contradictory error, even reiterating how
that contradiction ultimately dismantles Epicurus’s views about natural fel-
lowship. Second and third, though Epictetus is still pejorative and polemical
towards Epicurus, his tone with, and manner of argumentation through, the
interlocutor are quite different. Epictetus enlists Nature as a supporting rather
than opposing voice in the development of his argument. Nature collaborates
with Epictetus to expose Epicurus; the two are in complete agreement and
work towards a common, collegial goal. In this way, Epictetus’s use of Nature
to censure Epicurus represents a positive engagement with an interlocutor.
The third dialogical exchange and closing: Disc. 2.20.21–37. Epictetus considers
as altogether unfortunate (ἀτυχίας) the state of affairs in which the Academics
and Epicurus respectively fail to recognize their flawed logic (2.20.21). As far as
Epictetus is concerned, the Academics’ and the Epicureans’ views amount to
nothing more than instances in which naturally endowed humans attempt
to destroy the very natural endowments that allow them to know truth (2.20.21).
Epictetus models this phenomenon with a third dialogical pericope, this time
with a “philosopher.” Epictetus asks,

τί λέγεις, φιλόσοφε; τὸ εὐσεβὲς καὶ τὸ ὅσιον ποῖόν τί σοι φαίνεται;

What do you say, philosopher? How do devotion and sanctity appear to

you? (2.20.22)
146 chapter 7

Epictetus’s leading question gives way to a sustained dialogue with the phi-
losopher, but Epictetus never introduces the interlocutor’s speech with a verb
of speaking or any other introductory formula. Instead, the dialogue instantly
shifts back and forth between Epictetus’s voice and the interlocutor’s. Below, I
further confirm the script of this passage through the conventions of speech-
in-character. For now, however, grammatical cues will suffice to set the bound-
aries within the dialogue. To Epictetus’s leading question, the philosopher
responds in the first person,

ἂν θέλῃς, κατασκευάσω ὅτι ἀγαθόν.

If you wish, I will prove that it is good. (2.20.22)

Epictetus retorts, “Yes! Prove it, so that our citizens might be turned back and
honor the divine” (2.20.22). To this, the philosopher asks,

ἔχεις οὖν τὰς κατασκευάς;

So, do you possess the proofs? (2.20.22)

Up to this point, the philosopher seems to be working in conjunction with

Epictetus’s requests. When Epictetus responds in the affirmative (2.20.23),
however, the philosopher charts an unexpected course. Rather than proceed-
ing to prove the goodness of devotion and sanctity, the philosopher declares,

ἐπεὶ οὖν ταῦτά σοι λίαν ἀρέσκει, λάβε τὰ ἐναντία· ὅτι θεοὶ οὔτ’ εἰσίν, εἴ τε καὶ
εἰσίν, οὐκ ἐπιμελοῦνται ἀνθρώπων οὐδὲ κοινόν τι ἡμῖν ἐστι πρὸς αὐτοὺς τό τ’
εὐσεβὲς τοῦτο καὶ ὅσιον παρὰ τοῖς πολλοῖς ἀνθρώποις λαλούμενον κατάψευσμά
ἐστιν ἀλαζόνων ἀνθρώπων καὶ σοφιστῶν ἢ νὴ Δία νομοθετῶν εἰς φόβον καὶ
ἐπίσχεσιν τῶν ἀδικούντων.

Therefore, since these (proofs) are acceptable to you, receive (i.e., hear)
the opposite [τὰ ἐναντία]. Namely, the gods do not exist, and even if they
exist, they do not care about people, nor do we have anything in common
with them. This devotion and sanctity spoken about by many people is
a lie that comes from arrogant people and sophists and, by god, lawyers
intended to cause fear and to check the unjust. (2.20.23)

Far from proving the goodness of devotion and sanctity towards the
gods as proposed, in characteristic Epicurean fashion, the philosopher has
Examples of Diatribal Dialogue 147

attempted to demonstrate the opposite. No doubt sarcastically (see also

2.20.26–27), Epictetus lauds the philosopher for serving the citizens so well,
who, in Epictetus’s view, really needed to be turned towards the gods rather than
repelled from them (2.20.22, 24). The philosopher’s final response takes his
views even further; he states,

τί οὖν; οὐκ ἀρέσκει σοι ταῦτα; λάβε νῦν, πῶς ἡ δικαιοσύνη οὐδέν ἐστιν, πῶς ὁ
υἱὸς οὐδέν ἐστιν.

What then? Are these things not acceptable to you? Receive now how
righteousness is nothing and how a son is nothing. (2.20.25)

In response, Epictetus simply heaps sarcasm upon sarcasm (2.20.26–27).

Though the philosopher possesses the proofs by which to know the goodness
of devotion and sanctity, he abandons them and argues for the opposite. In this
way, the philosopher mirrors the Academics and the Epicureans by sacrificing
his faculties for knowing the truth.
Epictetus begins to draw the discourse to a close by depicting how he would
engage with such a skeptic, if he happened to be enslaved to one (2.20.28).24
Epictetus relates how he would spite his master at every turn, sarcastically
feigning an inability to know this from that, intentionally serving him vinegar
and fish sauce rather than actual food (2.29–31). Epictetus concludes with an
appeal to his audience. Epictetus challenges them to think about what is good
or evil, this way of thinking or the other (2.20.35). Moreover, Epictetus sugges-
tively wonders whether there is any use whatsoever in further engaging with
such skeptics, and he contends “a person might rather hope, by god, to convert
an altogether lewd person than such people who have become so deaf and
blind” (2.20.37).
In diatribal terms, therefore, the “philosopher” as an interlocutor func-
tions to depict even more clearly how Epictetus thinks the Academics and
Epicureans sacrifice their means of knowing truth. The philosopher progresses
Epictetus’s diatribe by serving as a model through which Epictetus further sup-
ports his argument against the Academics and Epicureans. The tone Epictetus
takes with the interlocutor is similarly pejorative-polemical, as he responds to
the interlocutor with sarcasm and makes no attempt to correct the interlocu-
tor’s mistakes. This also means that the argument is based entirely on censure
rather than protreptic; Epictetus criticizes the way the philosopher knows the

24  This would constitute the fourth dialogical exchange.

148 chapter 7

correct proofs but abandons them in order to propose the opposite and, in
Epictetus’s view, incorrect conclusions.
Incorporating the conventions for speech-in-character as a means of check-
ing the suggested script of the discourse, one finds that all three primary
features are present. Epictetus establishes a characterization of people who de-
stroy their naturally endowed ways of discovering and knowing truth (2.20.21),
introduces the “philosopher” as just such a person, and then attributes speech
to this interlocutor (2.20.22). Additionally, the sum total of the scripted lines
proves to be appropriate to the established characterization; the philosopher
initially agrees to prove the benefit of devotion and sanctity, indicates that he
possesses the proofs for doing so, and finally commences to abandon those
proofs and argue contrarily that devotion and sanctity are useless (2.20.23).
Thus, from beginning to end, the philosopher sacrifices his faculties for know-
ing truth (i.e., the proofs) and demonstrates his inner inconsistencies, just like
the Academics and Epicureans. Once more, the diatribal discourse functions
on the same principles as established for speech-in-character.
Conclusions. In “Against Epicureans and Academics” (Disc. 2.20), Epictetus
employs interlocutors of two sorts. Epictetus uses the Academics, Epicurus,
and the “philosopher” as illustrations of his primary thesis that the best argu-
ment for a principle being true is the use of that principle by those who reject it
(2.20.1). Epictetus’s attitude towards these interlocutors is thoroughly negative
and polemical. Epictetus never attempts to cure his interlocutors’ maladies; he
simply points out their errors and criticizes them. Accordingly, censure is the
dominant rhetorical trajectory in Epictetus’s engagements with these interloc-
utors. Quite differently, Epictetus enlists personified Nature as a fellow critic of
Epicurus. Nature and Epictetus work together in a collaborative and collegial
manner to explain and expose Epicurus’s flawed logic. But both interlocutors
function to teach Epictetus’ actual classroom audience a lesson. It is demon-
strated, therefore, that interlocutors can function, and primary speakers can
engage with them, in diverse ways in diatribal texts.
The conventions established for speech-in-character—especially char-
acterization and appropriateness—correlate seamlessly with each itera-
tion of the interlocutor. In every instance, the words Epictetus scripts for
his interlocutor(s) appropriately align with the ways he characterizes that
particular interlocutor. Additionally, the theoretical treatments of speech-in-
character are able to explain the function of the imaginary speakers in terms
that parallel the functions of diatribal discourse. Most importantly, however,
in the case of personified Nature, the conventions for speech-in-character
constitute the clinching evidence that Epictetus does in fact attribute speech
to Nature. Without considering characterization and the appropriateness of
Examples of Diatribal Dialogue 149

the potentially attributed speech, it is at best ambiguous in whose voice the

lines should be heard. As it stands, the conventions of characterization and
appropriateness solidly suggest that Epictetus attributes speech to personi-
fied Nature, whereby speech-in-character sets the script of diatribe’s dialogical
discourse.

Epictetus, Discourses 2.23: Περὶ τῆς τοῦ λέγειν δυνάμεως

Overview. The final representative of diatribal dialogue to discuss with an eye
towards the proposed method involves a conversation between Epictetus and
yet another Epicurean. In the discussion of Disc. 2.20, I noted how Epictetus
uses Epicurus strictly as a negative example to censure, ultimately treating
him quite pejoratively. In Disc. 2.23, however, Epictetus engages the Epicurean
in quite a different fashion. Rather than simply pointing out his flaws, Epictetus
eventually brings the Epicurean interlocutor from a point of disagreement to
a point of agreement, so that the interlocutor himself becomes the witness
proving Epictetus’s point.25
To begin the diatribe, Epictetus remarks that denying a faculty or power of
expression (δύναμις … ἀπαγγελτική) exists is impious and cowardly (2.23.2). It
is impious because it appears as though one is dishonoring the gifts graced
from god, such as vision, hearing, or speech (2.23.2). God graced these gifts for
a purpose, Epictetus argues, and people should use them accordingly (2.23.3–
4). While keeping these gifts in mind, however, people should not forget that
god also provided something better than these gifts. In fact, this greater gift is
able to use, judge, and determine the worth of the lesser gifts (2.23.6). As it is,
the lesser gifts (i.e., sight, vision, and so forth) are appointed as ministers and
slaves to serve the greater gift (2.23.7). This greater gift that controls the other
faculties is the power of will (προαιρετική; 2.23.9).26 An ear simply hears, and an
eye simply sees, but the power of will informs the person concerning what to
see or hear, how to interpret what is seen or heard, and how to respond in any
number of situations (2.23.9–15). In every way, the power of will supersedes the
other faculties, yet not in such a way as to render them unnecessary or useless.
Moreover, nothing is able to thwart the power of will (προαίρεσιν) unless that
same power of will corrupts itself. For this reason, the power of προαίρεσις ulti-
mately constitutes the only vice or the only virtue (2.23.18–19).

25  For Epictetus’s discussion of the Socratic method as a means for educational transfor-
mation, see Disc. 2.12, which loosely, and albeit briefly, discusses Plato’s Philebus with
Epictetus’s authorial commentary. For discussions of Disc. 2.12, see Stowers, Diatribe,
158–59; idem., “Paul’s Dialogue,” 712.
26  Oldfather consistently translates προαιρετική as “moral purpose.”
150 chapter 7

In the course of presenting his case for the superiority of προαίρεσις, how-
ever, the interlocutor interjects. Introduced with the common verb φησίν, the
interlocutor asks,

‘Τί οὖν,’ φησίν, ‘εἰ οὕτως τὸ πρᾶγμα ἔχει, καὶ δύναται τὸ διακονοῦν κρεῖσσον
εἶναι ἐκείνου ᾧ διακονεῖ, ὁ ἵππος τοῦ ἱππέως ἢ ὁ κύων τοῦ κυνηγοῦ ἢ τὸ ὄργανον
τοῦ κιθαριστοῦ ἢ οἱ ὑπηρέται τοῦ βασιλέως;’

‘What then,’ he says [φησίν], ‘if the situation is as follows, and the thing
that serves is able to be better than that which it serves, (for instance), the
horse over the horse rider, the dog over the hunter, the instrument over
the musician, or the attendants over the king?’ (2.23.16)

In this interjection, the interlocutor willingly concedes that some faculties

serve others, but he indicates that he is not content with Epictetus’s prioritiza-
tion of προαίρεσις. The interlocutor’s concern is whether it is perhaps possible
for one of the other faculties to be greater than the power of προαίρεσις, even
though the former serves the latter. Epictetus will have nothing of it, and he
rehashes his case once more for the supremacy of προαίρεσις (2.23.17–19).
Having established to his satisfaction that the power of will (προαίρεσις)
governs the other faculties, Epictetus turns to his Epicurean interlocutor
and addresses him directly (2.23.20). Based on the discourse up to this point,
Epictetus draws out the conclusion that only the faculty of προαίρεσις, as that
which judges and determines the value of the other faculties, could hypotheti-
cally declare the flesh to be the greatest of the faculties (2.23.20), and he press-
es Epicurus to answer which faculty is greater. Epictetus writes,

νῦν δὲ τί ἐστιν, Ἐπίκουρε, τὸ ταῦτα ἀποφαινόμενον; τὸ περὶ Τέλους συγγεγραφός,

τὸ τὰ Φυσικά, τὸ περὶ Κανόνος; τὸ τὸν πώγωνα καθεικός; τὸ γράφον, ὅτε
ἀπέθνῃσκεν, ὅτι “τὴν τελευταίαν ἄγοντες ἅμα καὶ μακαρίαν ἡμέραν;” ἡ σὰρξ
ἢ ἡ προαίρεσις; εἶτα τούτου τι κρεῖσσον ἔχειν ὁμολογεῖς καὶ οὐ μαίνῃ; οὕτως
τυφλὸς ταῖς ἀληθείαις καὶ κωφὸς εἶ;

Now, what is it, Epicurus, that declares these things? What composed the
treatise On the End, or The Physics, or On the Standard? What grew out
your beard? What wrote, (when he was dying,) ‘We are at the same time
experiencing our last and a happy day?’ Was it flesh or the will [ἡ σὰρξ ἢ ἡ
προαίρεσις]? So, do you agree [ὁμολογεῖς] that you have something greater
than (the flesh), and are you not enraged (by this admission)? Are you so
blind and dumb to the truth? (2.23.21–22)
Examples of Diatribal Dialogue 151

Epictetus would need to question his interlocutor in this fashion only if the
interlocutor does not initially share Epictetus’s views. The interlocutor’s pre-
vious interjection demonstrates that he in fact does not share Epictetus’s
hierarchical organization scheme for the various faculties (2.23.16). Based
on the dichotomy Epictetus poses between σάρξ and προαίρεσις, he seems to
presume that the Epicurean interlocutor would naturally prioritize the flesh
rather than the will (ἡ σὰρξ ἢ ἡ προαίρεσις; 2.23.22). Epictetus’s questions,
however, demand the interlocutor to answer otherwise that the faculty of
προαίρεσις is greater than that of σάρξ. As Epictetus already demonstrated,
it is the faculty of will and not of flesh that judges and assigns value to the
lesser faculties and could make a declaration concerning which is greatest.
Thus, to continue answering Epictetus’s line of questions, it is the power of
προαίρεσις that motivated Epicurus to compose his various treatises, that
moved him to become a philosopher and grow his beard (see Disc. 1.2.29), and
that guided him to interpret his last day as a “happy” day. Following Epictetus’s
discourse and questions, the interlocutor can in no reasonable way maintain
that σάρξ is greater than προαίρεσις. This is why Epictetus asks the interlocutor
whether he is enraged, blind, and dumb; how, Epictetus wonders, could the
interlocutor possibly continue to affirm σάρξ over προαίρεσις if he were in his
right mind?
In order to make sense of Epictetus’s barrage of questions in light of the con-
tinuing discourse, however, an unstated, but necessary, implication requires
discussion. Namely, Epictetus’s questioning is successful, and, as a result, the
interlocutor grants Epictetus’s prioritization of the faculty of will. Though
the text does not explicitly note the interlocutor’s submission in the form of
attributed speech, it does hint at it in Epictetus’s questions, “So, do you agree
(ὁμολογεῖς) that you have something greater than [the flesh], and are you not
(οὐ) enraged [by this admission]?” (2.23.22). If the interlocutor could disagree
and answer “no” to Epictetus’s first question, there would be no reason for the
interlocutor to be “enraged,” as he would simply continue to maintain his ini-
tial position. The only reason the interlocutor would be “enraged” is if Epictetus
has successfully persuaded him to abandon his former views about σάρξ and
to answer “yes” to Epictetus’s question of whether he “agrees” that προαίρεσις
is the greatest faculty. Such an allowance by the interlocutor is the necessary
response that allows the discourse to develop into the subsequent dialogical
exchanges.27

27  Coming to the same conclusion, in less detail, is Stowers, “Paul’s Dialogue,” 712–13.
152 chapter 7

So, having come to consensus that the faculty of the will is the greatest,
Epictetus continues to ask his interlocutor what he thinks such a conclusion
means for the lesser faculties. Epictetus asks,

Τί οὖν; ἀτιμάζει τις τὰς ἄλλας δυνάμεις;

What then? Does a person dishonor the other faculties? (2.23.23)

The interlocutor’s prompt rejection indicates he believes that one faculty

being greater results in no disservice to the lesser faculties; he simply declares,

μὴ γένοιτο.

Absolutely not! (2.23.23)

Satisfied, Epictetus poses a different, but related, question,

λέγει τις μηδεμίαν εἶναι χρείαν ἢ προαγωγὴν ἔξω τῆς προαιρετικῆς δυνάμεως;

Does a person claim there is no use or progress outside of the faculty of

the will? (2.23.23)

Epictetus himself has already addressed this very question earlier in 2.23.2–4,
where he maintains that people must continue to use the lesser faculties and
view them as god’s gracious gifts. Consequently, the leading question tests to
what degree the interlocutor has adopted Epictetus’s views about προαίρεσις
and their consequences. When the interlocutor responds to Epictetus’s second
leading question, he does more than simply reject it. This time the interlocutor
replies at length, as follows,

μὴ γένοιτο. ἀνόητον, ἀσεβές, ἀχάριστον πρὸς τὸν θεόν. ἀλλὰ τὴν ἀξίαν ἑκάστῳ
ἀποδίδωσιν. ἔστι γάρ τις καὶ ὄνου χρεία, ἀλλ’ οὐχ ἡλίκη βοός· ἔστι καὶ κυνός,
ἀλλ’ οὐχ ἡλίκη οἰκέτου· ἔστι καὶ οἰκέτου, ἀλλ’ οὐχ ἡλίκη τῶν πολιτῶν· ἔστι καὶ
τούτων, ἀλλ’ οὐχ ἡλίκη τῶν ἀρχόντων. οὐ μέντοι διὰ τὸ ἄλλα εἶναι κρείττονα
καὶ ἣν παρέχει τὰ ἕτερα χρείαν ἀτιμαστέον. ἔστι τις ἀξία καὶ τῆς φραστικῆς
δυνάμεως, ἀλλ’ οὐχ ἡλίκη τῆς προαιρετικῆς.

Absolutely not! That would be thoughtless, impious, and thankless

towards god. Rather, he gives to each its value. For, there is a use for a
donkey, but not as much as for an ox. There is also a use for a dog, but not
Examples of Diatribal Dialogue 153

as much as for a slave. There is a use for a slave, but not as much as for
the citizens. There is a use for these (citizens), but not as much as for the
rulers. Nevertheless, although some (faculties) are better, one must not
dishonor the use that the others provide. There is a value for the power
of speaking [τῆς φραστικῆς δυνάμεως], but not as much as for the power of
will [τῆς προαιρετικῆς]. (2.23.23–26)

That this material remains in the interlocutor’s voice beyond the simple rejec-
tion (μὴ γένοιτο) is supported by how the whole pericope answers Epictetus’s
leading question.28 That is, given the prominence of the faculty of προαίρεσις,
Epictetus asks whether anyone would claim that there is no use for the lesser
faculties and, therefore, no need to pursue their development. The whole of
Disc. 2.23.23–26 quoted above addresses this question; the correct response
to the lesser faculties is not to devalue and abandon them (μὴ γένοιτο; 2.23.23),
but for each person to give (ἀποδίδωσιν)29 to each faculty its proper value
(2.23.23–24), as all of the faculties have some value, even if none has more
value than the faculty of προαίρεσις (2.23.24–26). The interlocutor, therefore,
has made a full conversion. By the end of the scripted speech,30 the interlocu-
tor is confessing Epictetus’s own argument; the faculty of will is greatest, but
the others ought to be put to the correct service as well.
Diatribe. Many of the questions about the diatribal aspect of this discourse
are answered above in the Overview. Consequently, some points only need
to be reiterated briefly, and the others only require a bit more filling out of
the specifics. First, the Epicurean identity of the interlocutor is obvious, as
Epictetus specifically addresses him in the vocative as “Epicurus” (2.23.21).
Second, in the Socratic manner Epictetus outlines in Disc. 2.12, the inter-
locutor ultimately serves as Epictetus’s witness. Though they initially have
different views on the chief faculty, Epictetus’s guiding questions bring the in-
terlocutor to a position from which he can solidly affirm and declare Epictetus’s

28  Stowers also supports the suggested script and observes that the dialogue continues in
the interlocutor’s voice after the μὴ γένοιτο rejection. “Paul’s Dialogue,” 712–13.
29  No subject is supplied for the verb, ἀποδίδωσιν. I take the subject to be the τις from
Epictetus’s leading question. Thus, the interlocutor’s point is that the person (τις) does
not claim that there is no use or progress outside of the faculty of will; instead, he or she
allots the proper value to each.
30  The discourse shifts back to Epictetus’s voice immediately following the quoted material,
as evidenced by the verbal forms and pronouns (Disc. 2.23.26-29). For a detailed discus-
sion, see below in the Speech-in-character section.
154 chapter 7

own argument.31 The interlocutor’s confession is not an end in its own right,
however, as the discourse takes on a larger rhetorical function in the whole
of the diatribe. In its schoolroom setting, such a discourse would serve to
undergird for Epictetus’s students (and other auditors) the lesson Epictetus
was communicating about the supremacy of προαίρεσις.
Third, the function of the diatribe tips its hat towards the tone of the di-
atribe as well. To be sure, Epictetus spouts off some words and accusations
that could be perceived as pejorative or polemical, primarily with respect to
the question of whether the interlocutor is blind or dumb (2.23.22). The man-
ner in which the conversation unfolds, however, reveals that this is not the
case. Instead, Epictetus’s use of this Epicurean interlocutor represents a text-
book example of collaboration and collegiality between the primary diatribal
speaker and the interlocutor. As Epictetus corrects the interlocutor by guiding
him towards the better view, the interlocutor sides with Epictetus and the two
work together to present Epictetus’s argument to his actual audience.
Fourth, censure and protreptic argumentation operate to achieve Epictetus’s
goal. With respect to censure, Epictetus illustrates how the interlocutor’s
original view (i.e., that σάρξ is the primary faculty; 2.23.20) contradicts with
his impending confession that προαίρεσις surpasses all of the other faculties
(2.23.21–22). In terms of protreptic, Epictetus’s leading questions, which rely
on false conclusions based on the developing argument, successfully guide
the interlocutor to think through the stipulations and consequences of his
new position (2.23.23–25). Thus, censure and protreptic combine to reveal the
interlocutor’s inconsistencies and to propel him well into the correct view.
Speech-in-Character. Thinking in terms of identification, characterization,
and appropriate attributed speech-in-character, Disc. 2.23 presents some dif-
ficult issues. The identification of the speaker is, of course, the same Epicurean
interlocutor. With characterization, however, things become a bit more intri-
cate. To begin, there is no direct characterization of the interlocutor. Instead,
the characterization of the interlocutor must be deduced from his interjection
at 2.23.16 and the questions Epictetus poses to him leading up to the remain-
der of the dialogue. The interlocutor’s challenge against Epictetus’s prioritiza-
tion of προαίρεσις by positing that a servant might actually be superior to that
which it serves clearly reveals that the interlocutor is not yet on Epictetus’s

31  So also Stowers, who writes, “Epictetus states the questions which represent false rea-
soning or unthinkable alternatives so sharply that the interlocutor is forced to reject
the questions and, in fact, state the logical alternative toward which Epictetus is leading
him. Thus, the interlocutor himself provides the evidence or conclusion.” Stowers, “Paul’s
Dialogue,” 712–13. Stowers is followed by Elliott, Rhetoric, 136.
Examples of Diatribal Dialogue 155

side. Furthermore, the questions Epictetus directs toward the interlocutor

suggest that the interlocutor, in opposition to Epictetus, originally priori-
tized the flesh and not προαίρεσις (2.23.20–22). Therefore, especially when
the interlocutor’s challenge to Epictetus is combined with the implications of
Epictetus’s questions, auditors and readers are able to distinguish the inter-
locutor as initially at odds with Epictetus’s line of reasoning about σάρξ, the
lesser faculties, and προαίρεσις.
How, though, does this characterization avoid circular reasoning? The con-
cern is that, as presented, some of the interlocutor’s words provide elements
of characterization for the remainder of the conversation. If attributed speech
must be appropriate to the characterization, but the attributed speech itself
is used as evidence for the very characterization against which it is to be mea-
sured, circularity potentially threatens the validity of the argument. Epictetus’s
questions to the interlocutor concerning whether σάρξ or προαίρεσις is greater
(2.23.20–22), however, alleviate the risk of circularity as they implicitly supply,
and thereby confirm, the necessary characterization of the interlocutor as one
who disagrees with Epictetus by favoring σάρξ. In this way, the interlocutor’s
question about whether it is possible for a servant to be superior to that which
it renders service is scripted appropriately in light of the implications drawn
from Epictetus’s questioning, and circularity is avoided.
The next problem to consider is whether 2.23.23–26 actually contains any
discourse from the interlocutor at all. Epictetus does not use any verb of speech
or introductory formulas to demarcate that the interlocutor has returned in a
dialogical role, nor does any speaker utilize any clarifying pronouns or verb
forms in these lines. Three arguments support not only that dialogue takes
place in these lines, but also the form of the script as presented in the Overview
section. First, Epictetus has already scripted an interlocutor into the discourse
(2.23.16), and he addresses and questions this interlocutor in 2.23.21–23. It
would hardly be surprising for the answers given to Epictetus’s questions to
be spoken in the voice of the interrogated interlocutor. Second, when the dis-
course has certainly returned to Epictetus’s voice, verb forms and pronouns
clearly distinguish between Epictetus, the interlocutor, and Epictetus’s audi-
ence. In 2.23.26–27, one voice speaks in the first person singular. This same
voice speaks to its conversation partners as “you” in both the singular (πυνθάνῃ)
and the plural (ὑμᾶς). If the interlocutor were responsible for the first per-
son singular lines, it would make no grammatical sense for him to address
Epictetus with the second person plural ὑμᾶς. Contrastingly, it would be gram-
matically accurate for Epictetus to address his actual audience with the second
person plural, thereby affixing these lines in Epictetus’s voice. Third, there is
no need for Epictetus to speak the extended lines placed in the interlocutor’s
156 chapter 7

voice, beginning with the second μὴ γένοιτο (2.23.23–24). The content of the
lines amounts to a recapitulation of what Epictetus has already demonstrated;
in Epictetus’s voice, they would be redundant.32 Therefore, it seems quite likely
that Epictetus and the interlocutor engage in conversation in these lines.
Finally, granting that 2.23.23–26 documents an imaginary dialogue, how do
the lines scripted in the interlocutor’s voice cohere with the characterization
of the speaker? At first glance, they do not cohere with the characterization at
all. The only speech that appropriately fits the characterization is the inter-
locutor’s interjection at 2.23.16, which I have already discussed. By the time the
interlocutor enters into the conversation again, he has, perhaps begrudgingly
(i.e., in an enraged fashion; 2.23.22), agreed with Epictetus about the primacy
of προαίρεσις. As a result, the characterization of the interlocutor as one who
disagrees with Epictetus no longer applies. As was the case with the example
from Teles, it appears once more that the overarching diatribal agenda to trans-
form and enlist the interlocutor as Epictetus’s witness demands a double char-
acterization, one implicit and one more explicit. In this Epictetian example,
the more explicit characterization involves the interlocutor’s initial disagree-
ment with Epictetus. In order to achieve the transformation of the interlocutor
and to employ him as witness for Epictetus’s argument, however, the implicit
characterization requires that he come to agree with Epictetus’s point of view.
Put to this diatribe’s overarching goal, therefore, the interlocutor’s rejections of
Epictetus’s proposed false conclusions and his summarizing endorsement
of Epictetus’s argument appropriately fit the implicit characterization. Thus,
much like the way the faculty of προαίρεσις trumps yet uses that of σάρξ (or
vision, etc.), the faculty of diatribe has superseded but nevertheless made use
of the faculty of speech-in-character in an appropriate manner.
In terms of the functions of speech-in-character, Quintilian’s general cat-
egory that speech-in-character is useful for providing appropriate characters
for specific rhetorical situations (Inst. 9.2.30) readily parallels the discussion
of function in diatribal terms. The specific rhetorical situation is Epictetus’s
endeavor to persuade his actual audience that προαίρεσις is the superior fac-
ulty. The introduction of the interlocutor, therefore, appropriately fits the bill

32  Every element in the summary of 2.23.24–25 has a parallel in Epictetus’s previous dis-
course. Epictetus comments that neglecting the lesser faculties would be impious, cow-
ardly, and thankless towards god (2.23.5), which is reiterated in 2.23.23. Epictetus argues
that each faculty possesses a degree of value that must be determined (2.23.2–5); this is
repeated in 2.23.24. Epictetus notes that one should put the lesser faculties to their proper
use (2.23.5–7), just as in 2.23.25.
Examples of Diatribal Dialogue 157

as he partners with Epictetus to prove Epictetus’s argument. Similarly, Theon’s

note that speech-in-character is quite effective for exhortation (i.e., protreptic;
Prog. 115.20–22) also tracks with the diatribal agenda for Epictetus’s audience
to endorse his views, just like the interlocutor.

Conclusions

The preceding forays into the relationship between diatribe and speech-in-
character are quite illustrative. To be sure, these three examples are not exhaus-
tive, but they are test cases by which to examine whether the proposed method
is helpful and to what degree. That said, the examples are representative of the
commonality and diversity among diatribal texts. The examples demonstrate
that diatribal texts vary in their depictions of dialogue, such as concerning the
details they include or in whose voice certain phrases are spoken. Variety even
exists within diatribes from single figures, such as Epictetus. The examples
also illustrate commonality, such as the significance of the dialogical element.
This commonality allows for the application of a single method to each of the
dialogues, and the diversity accounts for the range within which that method
is helpful for confirming or determining the script of a dialogue. From these
analyses, a number of conclusions are possible.
The first observation is that dialogical passages in diatribal texts and
speeches-in-character overlap to a remarkable degree. In both, speech is attrib-
uted to a diverse range of imaginary speakers, including unspecified people,
specified people, abstract concepts, and inanimate objects. Moreover, diatribe
and speech-in-character have ways to discuss the functions of dialogical peri-
copae that parallel one another. Diatribal texts also often include an accompa-
nying characterization along with the imaginary speech of an interlocutor. As
in speech-in-character, this characterization can be implicit or explicit.
Second, the instances in which a diatribal script is quite certain due to ver-
bal forms, introductory formulas, and/or other grammatical features provide
a standard against which to test whether speech-in-character’s conventions of
characterization and appropriateness can reliably account for the exchanges
within the script. In many instances, speech-in-character’s conventions of
characterization and appropriateness do in fact confirm the diatribe’s script
without further qualification. These test cases indicate that a correlation often
exists between the elements included in a diatribal dialogue and the way a
reader or auditor should be able to distinguish a speech-in-character from the
voice of the primary speaker by considering the established conventions of
characterization and appropriateness.
158 chapter 7

Third, when a diatribal dialogue’s script is uncertain or altogether ambigu-

ous, there are occasions where speech-in-character’s conventions can clarify
or prove (or, at minimum, strongly suggest) that a particular text is scripted
for the voice of an imaginary interlocutor. The example of personified Nature
in Disc. 2.20 is a case in point. Examining the context for clues about charac-
terization and considering whether the potentially attributed speech would
be appropriately spoken in Nature’s voice in fact suggest that Epictetus has
placed these words on Nature’s lips. Left to its own faculties, diatribe can at
best render the material as ambiguous. Speech-in-character, however, rises to
the cause and helps to establish the script of diatribe’s dialogical discourse.
Fourth, the rhetorical category of diatribe operates at a level beyond that
of speech-in-character. At times, diatribe uses the conventions of speech-
in-character seamlessly in relation to the rhetorical handbooks and
Progymnasmata. At other times, diatribe’s own goals require speech-in-
character to be adapted for a particular purpose. This was seen in Teles,
Fragment I, and Epictetus’s Disc. 2.23, where diatribe required a more complex,
contradictory, double characterization. Thus, though there is a correlation
between diatribe and speech-in-character as evidenced by the numerous
similarities, the correlation is not always 1:1; sometimes diatribe’s rhetorical
agenda takes precedence and manipulates speech-in-character as needed.
Fifth, speaking more strictly in terms of the diatribal examples rather than
the relationship between diatribe and speech-in-character, the dialogues ex-
amined depict a high degree of diversity and fluidity. To begin, introductory
formulas and common phrases do not and cannot consistently (i.e., reliably)
indicate which speaker in a discourse is responsible for this or that line. To
repeat a point made above and illustrated in the examples, both Epictetus
and his interlocutor(s) are able to use the famous negation, μὴ γένοιτο, just as
both are capable of speaking the formula τί οὖν. The mere presence of one of
these (or other) formulas does not constitute a persuasive argument for the
identity of the person speaking the relevant lines.33 Additionally, the speak-
er can treat the interlocutor collegially, pejoratively, or narrowly polemically.
Epictetus’s interactions with Epicurean interlocutors demonstrate this nicely;
in Disc. 2.20, Epictetus is polemical towards his adversary, but, in Disc. 2.23, he
treats his Epicurean interlocutor collegially and collaboratively, both of which
presumably occur in classroom settings. As such, a speaker’s engagement with
an interlocutor cannot be assumed to project a particular mood; the tone of
each discourse must be weighed on its own. Finally, some engagements be-
tween the speaker and the interlocutor employ censure, some protreptic, and

33  Cf. Malherbe, “Μὴ γένοιτο;” Song, Diatribe, 32–37.

Examples of Diatribal Dialogue 159

many both. Said otherwise, the dialogical pericopae serve various functions.
Quite generally, some dialogues serve simply to critique an interlocutor, some
to transform the interlocutor, and some to go the whole way and enlist the
interlocutor as one’s principle witness. In all instances, engagements with the
interlocutor function on a grander level to progress the argument the primary
speaker is making to his actual audience.
In sum, although it is not absolutely perfect, the proposed method can be
effective and helpful in terms of analyzing and even establishing the dialogical
script of a diatribal discourse. Part 3 demonstrates that the method is alto-
gether useful for analyzing Paul’s discourses with his interlocutor in Romans,
to which I now turn.
 part 3
Romans 3:1–9

∵
Introduction to Part 3

Part 3, “Romans 3:1–9,” considers how the previous analyses of speech-in-

character in Part 1 and diatribe in Part 2 come to bear on Paul’s letter to the
Romans. As discussed in Chapters 8 and 9, many scholars understand Romans
to evince diatribal features, and some even consider Romans to be a bona fide
diatribe. Moreover, no matter how a scholar conceives of the relationship be-
tween Romans and diatribe, most recognize Rom 3:1–9 to represent dialogue
between Paul and a fictitious interlocutor, often consisting of a series of five
question and answer exchanges. For these reasons, Rom 3:1–9 is an excellent
entry point and test case for assessing the import of diatribe and diatribal dia-
logue in Romans.
But who is speaking which lines in 3:1–8 and 9–20? Paul offers no verbs of
speech to differentiate the interlocutor’s words from his own, nor does Paul
include any overt and/or grammatical indication of the subject of each line,
such as “the self-proclaimed Jew” (cf. 2:17) or the like. Unsurprisingly, this
renders discerning the dialogical script of Rom 3:1–9 terribly difficult. For the
most part, scholars have presented three opposing scripts of the dialogue.
First, the traditional reading of the script understands the questions to be
posed in the voice of an interlocutor who objects to Paul’s argument.1 Second,
beginning with Stowers, a few scholars begin to revise the script of the dia-
logue so that Paul asks some or, third, all of the questions for the interlocu-
tor to answer. Naturally enough, the way each scholar conceives of the script
then contributes, to some degree, to shaping his or her understanding of the
letter as a whole. To be sure, understanding who’s speaking when matters
significantly.
Though Parts 1 and 2 offer contributions in their own right, Part 3 is
the point to which this whole project has been building, as it addresses the
questions of who (Paul or his interlocutor) speaks which lines in Rom 3:1–9
and why it matters for Romans. Part 3 begins with Chapter 8, which engages
the work of scholars who discuss Rom 3:1–8/9 as dialogue and understand the
script in a traditional arrangement. Chapter 9 similarly examines works that
revise or “rescript” the dialogue in 3:1–8/9. Chapter 10 begins the actual analysis

1  I say, “for the most part,” because of minute variations in presentations of the traditional
reading, in which a scholar will attribute this or that small phrase to a different voice than is
the norm.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_011

164 part 3

of Romans and addresses the beginning of Paul’s argument and his prepara-
tion for the imaginary conversation that begins in 3:1. Chapter 11 analyzes the
script of the dialogue in light of speech-in-character and diatribal conventions,
after which it considers how a well defined dialogical script influences one’s
understanding of Romans and Paul’s thought as a whole.
 chapter 8

Traditional Readings of the Dialogue in Rom 3:1–9

and Its Role in the Letter

The vast range of generalizing comments scholars make about Rom 3:1–9 is
fascinating. For instance, Frédéric Louis Godet touts Rom 3:1–9 as “one of the
most difficult, perhaps, in the epistle.”1 William Sanday and Arthur C. Headlam
view the passage as a “digression.”2 More pessimistically, C. H. Dodd remarks
that the passage is “obscure and feeble” to the extent that the whole epistle
would make better sense if the pericope were omitted.3 More recently, N. T.
Wright argues that Rom 2:17–3:9 contains a key to understanding the whole
of Romans.4 Such a diversity of views about Rom 3:1–9 raises (at least) two
types of questions. First, what is Paul doing in the dialogue in 3:1–9? If it is such
a difficult passage, how is the passage best understood? Second, what is Paul
doing with the dialogue in 3:1–9? How does it advance the argument of Romans
as a whole? Do some readings of 3:1–9 cause tension with material elsewhere
in Romans that other readings successfully alleviate, though they may cause
others?
This chapter assesses readings that demonstrate a traditional understand-
ing of 3:1–9’s script, highlighting each approach’s engagement with Rom 3:1–9
as dialogical and/or diatribal. In order to track the developments made in the
history of scholarship on Rom 3:1–9, the examination begins with two treat-
ments that precede the publication of Bultmann’s dissertation. Then, begin-
ning with Bultmann, scholarship is addressed that sees Romans as diatribal
and affirms (some variation of) the traditional script. Afterwards, a taxonomy
is drawn of potential strengths and weaknesses for these works. Though not
exclusively, critique of individual points is withheld until Chapters 10 and 11, so
that analysis can be combined with detailed exegesis.5

1  Godet, Romans, 131.

2  William Sanday and Arthur Headlam, A Critical and Exegetical Commentary on the Epistle to
the Romans (ICC; Edinburgh: T & T Clark, 1902), 75.
3  Dodd, Romans, 46.
4  Wright, “Romans 2:17–3:9,” 1–25.
5  With a few exceptions, I reserve discussion of the various commentaries and additional sec-
ondary literature until Chapters 10 and 11. My intent in this chapter and the next is to focus

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_012

166 chapter 8

Pre-Bultmannian Readings of Romans 3:1–9

Two readings that predate Bultmann provide a brief snapshot of how (at least
some) scholarly works struggle to explicate Rom 3:1–9 without recourse to dia-
tribe or diatribal tendencies, and they suffice as a jumping off point into the
discussion of Rom 3:1–9.
Arriving at Rom 3:1–8, Godet transitions into the pericope by positing
that these verses represent Paul’s own anticipation of an objection to the argu-
ment he has just set forth in the preceding text, which Paul feels the need to
obviate. This objection revolves around the question of Jewish privilege, as set
forth in Rom 3:1.6 Rather than understand the objection as the introduction of
an opponent, Godet argues that such an introduction is altogether unneces-
sary. Godet observes that “Paul does not here make use of the formula: “But
someone will say,” which Paul utilizes elsewhere to indicate the presence of an
imaginary speaker.7 Instead, this and subsequent objections (cf. 3:3, 5) “arise of
themselves from the affirmations, and Paul puts them in a manner to his own
account.”8
Godet depicts the argument of 3:1–8 as follows. If Jews find themselves
in the same sinful state as gentiles and deserving of God’s wrath as Paul has
argued in Rom 2, what has become of Jewish privilege (3:1)? Paul’s answer is
that Jewish privilege remains in the fact that they were deemed faithful to
be “the depositories of the divine oracles,” which are the Old Testament writ-
ings, especially the messianic promises.9 This affirmation, however, gives rise
to another logical objection. Has this advantage not (μή) been abrogated by
(most of) Israel’s unbelief and rejection of Jesus, its Messiah (3:3)? Can Israel’s
unbelief void God’s faithfulness?10 The use of μή in the objection implies and
anticipates a negative response, which Paul forcibly supplies with μὴ γένοιτο; of
course Jewish unbelief will not void God’s faithfulness (3:4).11 Rather, let God
be found true and all people liars. The possibility that God’s “veracity” might
be magnified by Israel’s unbelief, however, leads to yet another objection;

by and large on those publications that are specifically interested in diatribe and dialogue
in Romans.
6  Godet, Romans, 131.
7  Ibid., 131–32, italics original. See also 133, 136. Godet buttresses this argument based on the
diction used to negate several of the questions in the pericope. See below.
8  Ibid., 132.
9  Ibid., 132–33, italics original.
10  Ibid., 133–34.
11  Ibid. For the use of negatives in interrogatives, see BDF §427.
Traditional Readings of the Dialogue in Romans 3:1–9 167

if human unrighteousness establishes God’s righteousness, is God not (μή)

unrighteous for inflicting wrath upon those who cause his own righteousness
to increase (3:5)? Though this question has “special application to the Jewish
unrighteousness which gives rise to the objection,” Paul’s inclusion of “all peo-
ple” in 3:4 makes his use of “our” in 3:5 more naturally applicable to “human,”
rather than simply Jewish, unrighteousness.12 Thus, again, the objection does
not need to arise from an imaginary opponent; quite the contrary, “it is from
the depths of the human conscience that the apostle fetches his question.”13
Furthermore, the use of μή in the interrogative, which, “as it always does,”
implies a negative answer, confirms that “it is certainly the apostle who is
speaking and not an opponent, for the objection is thus expressed in the out-
set as one resolved in the negative.”14 Paul supplies the anticipated negative μὴ
γένοιτο on the premise that no final judgment would be possible if the benefi-
cial consequences of human sin (i.e., in magnifying God’s righteousness) could
justify the sinner.15 In 3:7–8, Paul does not offer another objection but a further
confirmation of the answer given in 3:6 to the objection in 3:5. The sense is, if
God were unjust for inflicting wrath, any sinner could approach God, note the
way his or her sin increases God’s righteousness, be acquitted, and God’s judg-
ment would be brought to nothing.16 Paul takes this even further in Rom 3:8 by
positing, if God’s judgment is abrogated, why not sin all the more and provide
fodder for God to turn into his own righteousness.17 Though unstated by Paul
or Godet, if Godet’s argument is correct, the same rejection Paul voices in 3:6
applies to 3:7–8—μὴ γένοιτο.
As for the way 3:1–8 fits within the rest of Romans for Godet, as Rom 1 depicts
the reign of God’s wrath over the gentile world, 3:1–8 functions as an “appendix”
to the depiction of God’s wrath over the Jewish people in Rom 2, sweeping
away the objections that preclude God’s chosen people from becoming the
obejct of “divine animadversion.”18 Nevertheless, this passage finds its further
explanation in Rom 11, where Paul argues that God’s judgment of the Jews
“subserve[s] the salvation of mankind and that of Israel themselves.”19

12  Godet, Romans, 136, italics original.

13  Ibid.
14  Ibid. Godet’s point is that an objector would not pose an objection in such a way as to
presume Paul’s rejection of it, but in such a way as to hope Paul would have to consent
positively to it.
15  Ibid., 137.
16  Ibid., 137–38.
17  Ibid., 138.
18  Ibid., 139.
19  Ibid., 139, 391.
168 chapter 8

Concerning the structure (not a script, in this instance) of 3:1–8, Godet’s

argument outlines the passage in keeping with the way traditional dialogical
readers outline the script of the passage (see below), with one minor
exception. Romans 3:1, 3, and 5 represent (anticipatory) objections to Paul’s
arguments, and 3:2, 4, and 6 are Paul’s asnwers. The minor exception is that
though traditional readers often see 3:7–8c as a fourth objection and 3:8d as
Paul’s response, Godet takes these verses to be further affirmation of the
answer in 3:6.
Sanday and Headlam do not present the argument of Rom 3:1–8 much
differently than Godet.20 The primary difference in the argument is that
Sanday and Headlam read 3:7–8c as an objection in keeping with 3:5 rather
than as an affirmation of 3:6, but from a different perspective.21 Thus, Sanday
and Headlam read 3:1, 3, 5, and 7–8c as objections, and 3:2, 4, 6, and 8d as
responses, which is precisely how traditional dialogical readers generally
structure the discourse.
What is remarkable about Sanday and Headlam is their discussion of 3:1–8
and the question of dialogue. Sanday and Headlam do reject the presence of
real discourse in the pericope, but they seem to do so much more difficultly
than Godet. Sanday and Headlam write:

It is characteristic of this Epistle that St. Paul seems to imagine him-

self face to face with an opponent, and that he discusses and answers
arguments which an opponent might bring against him … No doubt
this is a way of presenting the dialectical process in his own mind. But
at the same time it is a way which would seem to have been suggested
by actual experience of controversy with Jews and the narrower Jewish
Christians.22

It is clear that Sanday and Headlam see where dialogue might take place in
Romans, but they are not able to explain it adequately. As with Godet, they
simply do not have the tools—diatribe—to understand how such dialogical
passages function. With Bultmann’s dissertation, however, the tide begins to
change.

20  Sanday and Headlam, Romans, 68–74.

21  Ibid., 73.
22  Ibid., 69.
Traditional Readings of the Dialogue in Romans 3:1–9 169

Diatribal Readings that Affirm the Traditional Script of

Romans 3:1–9

Bultmann’s predecessors began to draw preliminary connections between

diatribe and New Testament studies, but it is was Bultmann’s dissertation that
turned scholars to the question of diatribe and its relevance to Pauline studies,
especially Romans.23 In fact, quite recently, three scholars have advanced read-
ings of Rom 3:1–9 that, like Bultmann, emphasize the diatribal and dialogical
nature of the text and divide the script of the discourse along traditional lines.
These readings include Changwon Song24 (who relies heavily on Abraham J.
Malherbe’s assessment of μὴ γένοιτο in Epictetus and Paul25), Thomas H.
Tobin,26 and Rafael Rodríguez.27 This is not to say that these scholars under-
stand diatribe in the same way as Bultmann, as each of these scholars to one
degree or another endorses Stowers’s presentation of diatribe. But it is inter-
esting that, though they adhere to Stowers’s model, they nevertheless argue
for a different script of Rom 3:1–9 than Stowers does. What is more, there are
few exceptions to the rule that commentaries on Romans that identify dia-
logue in 3:1–9 affirm a traditional script of the passage as well.28 For instance,
three powerhouses among Romans commentators suppose or affirm a tradi-
tional script: James D. G. Dunn,29 N. T. Wright,30 and Robert Jewett,31 of whom

23  Bultmann, Der Stil. See Norden, Die Antike Kunstprosa, 2.556; Carl Friedrich Georg
Heinrici, Das erste Sendschreiben des Apostel Paul an die Korinther (Berlin: Wilhelm
Hertz, 1880); idem., Das zweite Sendschreiben des Apostel Paulus an die Korinther (Berlin:
Wilhelm Hertz, 1887); idem., Der litterarische Charakter der neutestamentlichen Schriften
(Leipzig: Durr, 1908); Johannes Weiss, “Beiträge zur paulinischen Rhetorik,” in Theologische
Studien: Bernhard Weiss ze seinem 70. Geburtstag dargebracht (Göttingen: Vandenhoeck
& Ruprecht, 1897); idem., Die Aufgaben der neutestamentlichen Wissenschaft (Göttingen:
Vandenhoeck & Ruprecht, 1908); idem., Der erste Korintherbrief (Göttingen: Vandenhoeck
& Ruprecht, 1910).
24  Song, Reading Romans.
25  Malherbe, “Μη Γενοιτο.”
26  Tobin, Paul’s Rhetoric.
27  Rodríguez, If You Call Yourself a Jew.
28  Exceptions include Witherington and Hyatt, Romans, 93–94; Keck, Romans, 89–95; Byrne,
Romans, 106–8. Witherington and Hyatt and Keck follow Stowers’s rescription; Byrne’s is
closer to Elliott’s.
29  Dunn, Romans, 128–49; idem., Theology, 45, 114–19.
30  Wright, Romans, 452–55; idem., “Romans 2:17–3:9,” 1–25; idem., Paul and the Faithfulness
of God (Christian Origins and the Question of God 4; 2 vols.; Minneapolis: Fortress Press,
2013), 836–39.
31  Jewett, Romans, 238–52.
170 chapter 8

I discuss the first two in this chapter.32 Given the array of results that stem
from a single pool of evidence, perhaps what is needed is a tool external to, yet
broadly overlapping with, diatribal dialogue that is able to refine and define
more precisely the breaks in a dialogical script, namely, speech-in-character
(but more on this below).
For all the emphasis on Bultmann’s dissertation, he comments on
Rom 3:1–8 very sparingly.33 Bultmann argues that Rom 3:1 and 3 represent
objections Paul introduces as questions without any introductory formula,
that Paul often crafts words for his opponent rather than recording their ac-
tual words, and that 3:1 (and 4:2) are the only instances in Paul where one
might glimpse the back-and-forth question and answer progression of a real
dialogue.34 Furthermore, Paul’s quick rejections (cf. 3:4, 6) of the objections
demonstrate that the objections amount to nothing more in Paul’s thought
than blatant absurdities.35 Finally, Bultmann notes that Paul’s scriptural cita-
tion in Rom 3:4 serves as evidence for, or an explanation of, his rejection of the
objection posed in 3:3.36
Based on Bultmann’s view of diatribe (see Part 2), though he does not actu-
ally discuss Rom 3:1–8 holistically in diatribal terms, one can rather confidently
extrapolate what he might say about the passage.37 Based on what Bultmann
actually says about verses within 3:1–8, however, one can deduce even further,

32  The commentaries of Jewett and Moo are also significant and deserve attention in this
project. In the interest of not making this project even longer, I address their commentar-
ies in the footnotes and conclusions, indicating where they significantly parallel or differ
from other traditional readings. They receive equal attention in Chapters 10 and 11.
33  On Bultmann’s sparse engagement with 3:1–8, Paul J. Achtemeier comments, “What is
noteworthy, I think, is the fact that Bultmann all but ignored [Rom 3:1–8]. In the whole of
his book, if I have counted correctly, he refers only three times to any of the verses com-
prising Romans 3:1–8, twice on p. 67 (vv. 1, 3, and 1–3) and once on p. 95 (v. 4). In none of
these instances is the reference more than passing, and there is no attempt to relate the
whole passage to the diatribal form.” Achtemeier, “Romans 3:1–8,” 79.
34  Bultmann, Der Stil, 67. Concerning Paul’s crafting of his opponent’s speech, Bultmann
writes, “Und er formuliert deshalb die Einwendung oft nicht mit direkten Worten des
Gegners, sondern als seine eigenen Worte, freilich im Sinne des Gegners.” And concerning
the back-and-forth, question and answer progression, Bultmann writes, “Auch zu einem
wirklichen Zwiegespräch, zu einem Hinundher von Fragen und Antworten kommt es kaum.
Rom. 3.1ff und 4.2 sind nur Ansätze dazu zu erblicken.”
35  Ibid.
36  Ibid., 95.
37  Bultmann would say something to the effect that Paul, in the form of a Cynic-Stoic street
preacher, polemically attacks the Jewish interlocutor who raises objections against his
argument.
Traditional Readings of the Dialogue in Romans 3:1–9 171

such as to how Bultmann understands the script of the dialogue. For instance,
Bultmann identifies Rom 3:1 and 3 as absurd objections, and he argues that
Paul formulates the objections for the fictitious interlocutor. Additionally, the
instances of μὴ γένοιτο represent Paul’s rejections of such absurd objections.38
Thus, the fictitious interlocutor is responsible for Rom 3:1, 3, and 5, and Paul is
responsible for 3:4 and 6.39 Based on these fixed points in the script, it is easy
to imagine that Bultmann would similarly identify 3:7–8c as the interlocutor’s
objection, and 3:2 and 8d as Paul’s responses. Thus, Bultmann, as best as one
can tell, espouses a traditional script of the dialogue: the interlocutor poses
objections to Paul in 3:1, 3, 5, (7–8c,) and Paul responds to the interlocutor in
3:2, 4, 6, (8d) in his own voice.
Malherbe40 does not intend to treat 3:1–8 as a whole, but his article is quite
influential and serves as a (corrective) bridge between Bultmann and later
scholars like Song. Malherbe’s “limited purpose” is “to examine the way μὴ
γένοιτο (‘by no means’)” functions in diatribal literature.41 Malherbe concludes
that, while Paul’s use of the phrase does not parallel diatribe in general, it does
align with one of Epictetus’s common uses of the phrase to begin a new seg-
ment in an argument.42
Malherbe sees his work as a clarifying expansion and corrective to
Bultmann’s views about the phrase. Bultmann notes that when Paul rejects an
objection with μὴ γένοιτο, he follows the phrase with an explanation that illus-
trates his own views and points out the objector’s error.43 Malherbe contends
that scholars have given “insufficient attention” to this observation, especially
since Paul is consistent in this regard.44 For instance, Rom 3:4 and 6 contain
and explain two of Paul’s uses of the phrase with constructions characteristic

38  Ibid., 67.

39  Rom 3:5 would be spoken in the voice of the interlocutor because 3:6 is spoken in Paul’s
voice as a rejection of whatever is said immediately preceding it.
40  At this point in the chapter, things become a bit chronologically messy. Because I group
traditional readings in this chapter and rescriptive readings in the next, I momentarily
pass over Stowers, who temporally precedes the following traditional works. Stowers
completed his dissertation in 1979, Malherbe published “Μη Γενοιτο” in 1980, and Stowers
published his dissertation in 1981. Consequently, all of the following scholars in this sec-
tion had access to Stowers’s dissertation and generally endorse his conclusions about
diatribe. With the exception of Malherbe, these scholars also had access to some or all of
Stowers’s later writings on diatribe and Romans.
41  Malherbe, “Μη Γενοιτο,” 231.
42  Ibid., 239.
43  Bultmann, Der Stil, 67–68.
44  Malherbe, “Μη Γενοιτο,” 236.
172 chapter 8

of diatribe; 3:4 explains μὴ γένοιτο with an imperative introduced by δέ, and 3:6
explains the phrase through rhetorical questions introduced by πῶς. More sig-
nificantly, the supporting explanations provide the theme of the discussions
that follow. Malherbe shows how the explanation in 3:4, which contains the
verb δικαιόω, gives rise to the discussion in 3:5, which contains the cognates
ἀδικία and δικαιοσύνη. Similarly, the supporting argument in 3:6 uses the verb
κρίνω, which 3:7 follows by using κρίνομαι and 3:8 κρίμα.45 In this way, Malherbe
argues that the phrase μὴ γένοιτο does not function in Paul to conclude a
conversation (as a certain reading of Bultmann allows), but to start a new line
of reasoning.46
Though Malherbe had access to Stowers’s dissertation—in fact, Malherbe
supervised Stowers’s dissertation—the publication of his article occurred just
before the onset of rescriptive readings of Romans.47 It is, therefore, unsurpris-
ing to find in Malherbe an affirmation of the traditional script of the dialogue
in 3:1–8 (for the verses Malherbe addresses). For Malherbe, “the characteris-
tic short questions … function as introductions to the objections raised by the
interlocutor,” such as τί γάρ (3:3) and τί ἐροῦμεν (3:5).48 Also, the uses of μὴ
γένοιτο and their respective explanations represent Paul’s rejections of the
interlocutor’s objections. Thus, the interlocutor speaks in 3:3 and his objection
is given in Paul’s words in 3:5, and Paul responds to the interlocutor in 3:4 and
6. Like Godet, Malherbe also seems to understand 3:7(–8?) as a further expla-
nation of 3:6 rather than as another objection.49 Malherbe does not discuss
Rom 3:1–2, but his discussion of τί οὖν and τί οὖν ἐροῦμεν, as well as his endorse-
ment of the shape of the rest of the script, point in the direction that 3:1 would
belong to the interlocutor, and 3:2 certainly belongs to Paul.
Of two works addressing Romans as diatribe and affirming the traditional
script that were published in 2004, Song relies heavily on Malherbe. Following
Stowers, Song notes how a second-person singular apostrophe begins in
Rom 2:1, and that the “theme of this chapter is an ‘apostrophe to the ignorant
and inconsistent judge.’”50 Concerning the identity of this judge, Song avers,

45  Ibid.
46  Ibid., 237.
47  This is because Stowers only sets forth his views about the shape of the script in 3:1–8 four
years later in “Paul’s Dialogue.”
48  Malherbe, “Μὴ Γένοιτο,” 233–34, but Paul’s use of the first plural indicates that he has for-
mulated the objection in his own words.
49  Ibid., 236. Malherbe groups 3:6–7 as “rhetorical questions introduced by πὼς.”
50  Song, Reading Romans, 92–93. Song notes that the title is adopted from Stowers.
Traditional Readings of the Dialogue in Romans 3:1–9 173

“it is indisputable that the second-person singular here is not an actual

person,”51 but it refers more generally to Jews.
For Song, Rom 3 transitions to the main body of the diatribe, which the
μὴ γένοιτο formula dominates, and the Jewish judge Paul addresses in Rom 2
enters into conversation with Paul. In typical, traditional script fashion, Song
outlines the script of Rom 3:1–8 as follows: the interlocutor poses objec-
tions to Paul in 3:1, 3, 5, and 7–8c; Paul, the teacher, responds in 3:2, 4, 6, 8d.52
Rom 3:9–20 continues this pattern with the interlocutor objecting in 3:9a
and Paul rejecting the objection in 3:9b–20.53
Unfortunately, Song does not comment on the meaning of the passage or
on its function in the larger argument of Romans. Song does engage with
Stowers’s rescription of the pericope (Elliott’s further rescription of 3:1–9 is
absent from Song’s project54), but he does so minimally and unpersuasively.
Song outlines Stowers’s script of the dialogue (see below) and admits, “This
analysis may be also possible.”55 Song’s lone defense for his reading is that, in
Epictetus, μὴ γένοιτο “as a statement of rejection and its supporting statements
are usually attributed to the Teacher, not to the Interlocutor.”56 Song then lists
a number of exceptions to his rule and admits to the difficulty involved in
separating the voice of the teacher from that of the interlocutor in μὴ γένοιτο
formulas.57 Song’s limited argument for the shape of the script in 3:1–9, there-
fore, rests on evidence that Song himself recognizes as unstable and capable of
being argued in either direction. More solid evidence and argumentation must
be brought to the discussion.
Also published in 2004, Tobin offers a much more robust analysis of Romans
and diatribe than does Song. Based on similarities between Epictetus’s
Discourses and the body of Romans (which he identifies as 1:16–11:36), Tobin
argues, “Paul’s Roman audience probably would have understood this sec-
tion of Romans as a diatribe and that Paul himself was intentionally using the
conventions of the diatribe as he wrote this section of the letter.”58 Tobin cor-
rectly notes how his assertion goes further than other interpreters, and he

51  Ibid., 94.

52  Ibid., 94–95. For 3:1, Song accidentally records 3:2. See also 35–36.
53  Ibid., 96.
54  Elliott, The Rhetoric of Romans, 132–41.
55  Song, Reading Romans, 112n.5, 7.
56  Ibid., 112n.5. See also Malherbe, “Μη Γενοιτο.”
57  Song, Reading Romans, 112n.5, 51n.109.
58  Tobin, Paul’s Rhetoric, 95. These similarities include: similar rhetorical devices, dialogi-
cal style, length, quotations from scripture and/or authorities, and alterations between
shorter, calmer, expository passages and longer, more argumentative passages.
174 chapter 8

offers several reasons for other scholars’ reluctance to identify Romans as

diatribe rather than diatribal.59 Tobin then discusses the multifaceted purpose
of Romans, and why Paul adapts and uses the genre of diatribe.60 Tobin argues
that Paul would have found diatribe attractive because of its broadness, espe-
cially since in Romans Paul addresses a wide range of issues in comparison
with his other letters. Additionally, the use of diatribe allows for a degree of
indirectness (such as through censuring the interlocutor rather than the
Romans directly) by which Paul is able “to place [the Roman Christians]
rhetorically on his side from the beginning” and to address “their issues and
misunderstandings of him without ever having to confront them directly.”61
Said otherwise, it is diatribe that allows Paul to confront, often critically, an
unknown and distant audience.
Tobin begins his actual discussion of the body of Romans by noting that
Paul and diatribal authors begin diatribes in a similar fashion. Tobin argues,
“Paul begins his argument in Rom 1:16–17, as the authors of most diatribes do,
by setting out the basic proposition he wants to argue in the body of the letter.”62
This proposition involves the gospel as God’s power for salvation to everyone
who has faith, the Jew first, and then the Greek, because God’s righteousness is
revealed in it through faith for faith, as Hab 2:4 records.63
Paul begins the first stage of his argument in 1:18–3:20 with a “subproposi-
tion,” which argues that God’s wrath is revealed against all human ungodli-
ness and wickedness.64 In 1:18–32, Paul presents “standard Hellenistic Jewish
apologetic motifs against Gentile religiosity and conduct.”65 Because in Tobin’s
reading most of Paul’s Roman Christian audience would have previously
been members of the Roman Jewish community or sympathizers with them,
the Roman Christians would have readily recognized and agreed with Paul’s

59  Ibid., 96–98. Tobin is not at fault for being unaware of Song’s work, which was published
in the same year. The reasons Tobin gives for the reluctance of others to identify the body
of Romans as diatribe include: the fact that Romans is a letter; the presence of expository
passages in Romans, which did not appear to be diatribal before Tobin’s comparison of
Romans with some of Epictetus’s Discourses (see Part 2); Paul’s extensive use and inter-
pretation of Jewish scripture; the level of generality in the diatribe versus that of Romans;
and the appearance of Romans in juxtaposition with other diatribes.
60  For Tobin’s discussion of the purposes of Romans, see ibid., 98–103.
61  Ibid., 103.
62  Ibid., 104. For examples of this practice, see the discussions of Teles, Fragment I, and
Epictetus, Disc. 2.20, in Part 2, Chapter 7.
63  Ibid.
64  Ibid., 108.
65  Ibid., 109, italics original. Examples include Wisdom 13–15 and Sib. Or. 3:8–45.
Traditional Readings of the Dialogue in Romans 3:1–9 175

presentation of “human godlessness” in these verses.66 This means “Rom 1:18–

32 is not so much an argument as an exposition of viewpoints Paul and the
Roman Christians shared. On the surface, at least, there was nothing contro-
versial about it.”67
Tobin argues, however, that in 1:18–32 Paul stacks the deck in preparation for
something of a bait-and-switch.68 Namely, because Paul never refers to gentiles
in 1:18–32, and because he opens the argument with the claim that God’s wrath
is revealed against “all” human ungodliness and wickedness, he stages the con-
versation for an unexpected broadening in the following sections.69 Though
the audience would have theoretically agreed with Paul’s comments in 1:18–32,
Tobin argues that 2:1–3:20 is a more argumentative and diatribal section of
the letter that presents Paul’s more controversial conclusions about the inclu-
siveness of human sinfulness. Paul does not present his views all at once or
bluntly, but “gradually and indirectly,” slowly building on the points he and his
audience share (i.e., 1:18–32).70 Paul begins by addressing a “fictitious person
who hypocritically condemns those who commit the acts described in 1:18–32
but who also commits the same acts.”71 Paul’s use of “all” (2:1) insinuates but
does not explicitly state that both Jews and gentiles fall under this category—
which his maintenance of an anonymous addressee further allows—and he
asks whether such a person thinks he or she will escape God’s judgment (2:3).
It is only in 2:9–10, however, that Paul spells out for his audience how this anon-
ymous, hypocritical addressee represents both Jews and gentiles, based on the
principle of God’s impartiality (2:11).72
In Rom 2:12–29, Paul turns to address Jews and their relation to the law.
Romans 2:12–16 functions to transition from a passage primarily dealing
with gentiles to an argument primarily addressing Jews, which Paul accom-
plishes by illustrating God’s impartiality to punish Jews who sin under the law
and gentiles who sin without the law.73 Thus, Paul reformulates the argument
of 2:6–11 in Jewish terms, showing how God judges Jews and gentiles by the
single criterion of law observance.74 Paul’s address to a Jew in 2:17–29 further

66  Ibid., 16–46, 106, 109.

67  Ibid., 109–10.
68  “Bait-and-switch” is my term, not Tobin’s.
69  Ibid., 110.
70  Ibid.
71  Ibid.
72  Ibid., 110–12.
73  Ibid., 113–14.
74  Ibid., 114–15.
176 chapter 8

establishes the parallelism between 2:1–11 and 2:17–29.75 Here, “Paul castigates
the Jew who, while claiming to know the law and to be able to instruct others
in it, does not himself observe the commandments of the law.”76 Thus, both
addressees, the gentile of 2:1–11 and the Jew of 2:17–29, are guilty of hypocrisy,
and Paul expects his readers to condemn their actions based on the principle
of divine impartiality.77 As such, Tobin concludes his discussion of Rom 2 by
stating,

On the basis of this impartiality, both Gentiles and Jews, in terms of their
conduct … are similarly liable to God’s judgment. If Gentiles’ knowledge
of what is right means that they have no excuse (2:1) for doing wrong,
then similarly Jews’ knowledge of the law is not a privilege that can be
appealed to in transgressing the law.78

As a result, Paul begins with agreed upon points of view about human sin-
fulness (1:18–32), broadens them out to include all people in light of God’s
impartiality (2:1–11), and finally illustrates how such sinfulness and culpability
applies directly to Jews (2:12–29).
Arriving at 3:1–20, having argued that Jews and gentiles are on equal footing,
Paul addresses possible objections to his (controversial) position in 2:12–29.
Tobin characterizes the passage as “objections from, and replies to, a fictitious
interlocutor,” thereby revealing his understanding of the dialogical script to be
along traditional lines.79 Tobin interacts with Stowers’s rescription of the pas-
sage in 3:3–6 (but, like Song, not Elliott’s). Tobin argues that 3:3 belongs in the
interlocutor’s voice and 3:4 in Paul’s because “throughout Romans 1–3 … Paul
appeals to Scripture, and so the scriptural references in 3:4 … are more natu-
rally placed in his mouth.”80 Concerning Rom 3:5–6, Tobin argues, again contra
Stowers, that 3:5 should belong to the interlocutor and 3:6 to Paul, “since Paul …
is defending himself and answering objections, it is more natural to place the
answers in Paul’s mouth (3:6) and the objections in the interlocutor’s. The end
of 3:[5] … is Paul’s editorial comment.”81 So, Tobin identifies 3:1, 3, 5a–b, 7–8c,
and 9a as the interlocutor, and 3:2, 4, 5c, 6, 8d, and 9b and following as Paul.

75  Ibid., 115, 117–18. Tobin now identifies the addressee of 2:1–11 specifically as a gentile.
76  Ibid., 115.
77  Ibid., 117–18.
78  Ibid., 118.
79  Ibid.
80  Ibid., 119n.43.
81  Ibid., 120n.44. Tobin erroneously records 3:6 for 3:5.
Traditional Readings of the Dialogue in Romans 3:1–9 177

Particularly interesting is Tobin’s discussion of the interlocutor. Tobin cor-

rectly observes that Paul does not overtly identify the interlocutor in 3:1–8.82
Though most scholars identify the interlocutor of 3:1–8 as the addressee in
Rom 2:17–29 (variously defined), Tobin considers this unlikely.83 Tobin charac-
terizes the addressee in 2:17–29 as “a hypocrite who claimed to know the law
but did not keep its commandments.”84 Tobin argues,

The interlocutor in 3:1–8, however, is someone quite different. He is

someone concerned over the negative implications of Paul’s views about
the equal sinfulness of Jews and Gentiles … In reality, Paul’s fictitious in-
terlocutor in 3:1–8 is the Roman Christian who is suspicious of, or hostile
to, what he sees as the implications of Paul’s controversial arguments.85

Again, the interlocutor is still anonymous, which allows Paul to be indirect

in his dealings with issues held by his Roman Christian audience concerning
God’s faithfulness or righteousness in relation to the Jews.86
Tobin depicts the argument of Rom 3:1–20, therefore, in the following way.
Paul answers the interlocutor’s objection about whether there is any value to
being a Jew or circumcised if Jews and gentiles are on equal footing before God
(3:1) in the affirmative—God has entrusted to the Jews the Jewish scriptures
(3:2). None of Paul’s Roman audience would have objected to this affirmation
of the value of the Jewish scriptures. When the interlocutor objects a second
time and argues that Jewish unfaithfulness does not nullify God’s faithfulness
(3:3), Paul readily agrees and cites scripture in support (3:4).87 Paul also agrees
with the interlocutor’s third objection concerning God’s justice in judging
sinners (3:5) and notes the absurdity of any opposing view (3:6). Tobin sug-
gests that Paul intends these three agreements with his interlocutor to lead his
audience to see him as sharing the same principles as they do and to render
them open to his final two responses.88
In the final two objections, the interlocutor wonders why he is considered
a sinner if his sin ultimately results in God’s glory; why not do evil and build
up God’s glory (3:7–8)? At the same time, Tobin maintains, “In this objection,

82  Ibid., 118.

83  Ibid., 118–19.
84  Ibid., 119.
85  Ibid.
86  Ibid.
87  Ibid.
88  Ibid., 120.
178 chapter 8

Paul explicitly mentions that there are some who blasphemously claim that
this is indeed his position,” whom he “summarily condemns” (3:8d).89 The fifth
and final objection wonders whether Paul is making excuses for human sinful-
ness (3:9a). Paul rejects the objection and reiterates that he has charged both
Jews and Greeks all to be under sin (3:9), referring to the whole argument of
1:18–2:29.90 Thus, Paul “finally states clearly and explicitly what he has only
insinuated at crucial points earlier in his argument … this ‘all’ is now seen to
include not only Gentiles but also Jews,” and 3:10–20 further substantiate this
claim.91 Paul, therefore, “crafted his rhetoric very carefully in order to persuade
his audience that his apparently controversial positions about the equal sinful-
ness of both Jews and Gentiles were really rooted in … the Jewish scriptures
and Jewish tradition,” which he and his Roman Christian audience shared.92
Tobin presents a robust case affirming a traditional script of Rom 3:1–9.
Tobin’s argument evinces many strengths, but it has many potential weak-
nesses too. Before tending to these, a few more traditional readings deserve a
hearing.
For instance, Rodríguez’s If You Call Yourself a Jew casts the traditional
reading of Romans in a completely different direction. This is primarily due
to Rodríguez’s identification of Paul’s interlocutor not as a Jew, but as a gen-
tile proselyte.93 As Rodríguez rightly notes, “The choice between an actually
Jewish interlocutor in Rom 2:17–29 and an ethnically-gentile-religiously-Jewish
interlocutor [would] prove to be the fork in the road for [an] understanding of
Romans as a whole.”94 But how does Rodríguez’s argument take shape in, and
influence readings of, Romans?
When Rodríguez takes up the audience of the letter, the primary question is
whether Paul envisions his audience in Rome to be primarily Jewish, gentile, or
mixed.95 Citing Stowers and Andrew Das in support, Rodríguez advocates “for
an exclusively gentile audience.”96 Rodríguez argues, “Paul only ever explicitly

89  Ibid.
90  Ibid., 120–21.
91  Ibid.
92  Ibid., 122.
93  This is not to say that Rodríguez is the first to offer such a reading, especially of Rom 2.
Rodríguez gives ample support to Thorsteinsson’s Paul’s Interlocutor in Romans 2, which
Rodríguez draws upon heavily in his identification of Paul’s interlocutor. I address
Thorsteinsson in Part 2, Chapter 6, and Part 3, Chapter 10.
94  Rodríguez, If You Call, 51. I would say this decision is “a,” rather than “the,” fork in the road.
95  Ibid., 7.
96  Ibid., 7–8. See Stowers, Rereading, 29–30; Andrew Das, “The Gentile-Encoded Audience of
Romans: The Church Outside the Synagogue,” in Reading Paul’s Letter to the Romans (ed.
Jerry L. Sumney; RBS 73; Atlanta: Society of Biblical Literature, 2012), 29–46.
Traditional Readings of the Dialogue in Romans 3:1–9 179

identifies the audience he imagines himself addressing as gentiles,” and, there-

fore, it is best to understand the audience of Romans as predominantly or
exclusively gentile.97 This carries over significantly in Rodríguez’s interpreta-
tion of Rom 1–3.
As Rodríguez approaches the body of Romans, he identifies 1:16–17 as the
thematic statement of the letter on the grounds that the rest of the letter
relates to it in one manner or another. Rodríguez connects God’s righteousness
to his faithfulness to the Abrahamic covenant, a feature Paul must explain in
light of Israel’s failure to maintain its part of the covenant.98 Romans 1:18–32
sets forth the problem for which the gospel is the solution.99 In 1:18–32, Paul
“employs standard Jewish rhetoric against gentile idolatry, which strongly
suggests that Paul’s critique aims at gentiles and not Jews.”100 Rodríguez
agrees that elements in 1:18–32 might suggest universalistic possibilities and
that it takes little imagination to apply this critique to humanity as a whole.
Rodríguez argues, however, “a careful reading of the text … reveals that Paul
does not apply his comments either to himself or to his gentile readers.”
Instead, “the knowledge of God is evident,” Paul denotes, “among them,” who
exchanged God’s incorruptible glory for corruptible images (1:23).101 This sug-
gests to Rodríguez that Paul is not railing against human depravity or gentile
religiosity as a whole—nor against “the universal condition of human bond-
age”—but more narrowly against “the worship of graven images patterned
after mortal creatures,” idolatry, the failure to worship God as Creator.102 At
this point, Rodríguez argues that it is best to imagine Paul’s Roman Christian
audience “nodding in agreement along with him” as he offers critiques “about
another party (debauched, idol-worshipping gentiles).”103
In Rom 2, however, Paul’s rhetoric changes suddenly as he “turns his attention
to a class of individual who, like him, condemns the very people he described
in 1:18–32.”104 Scholars often identify this individual as a Jew, but Rodríguez
argues that Paul continues to address a gentile, based on the connective

97  Rodríguez cites Rom 1:5–6, 13; 11:13; 15:15–16. Rom 2:17 does not qualify as addressing a
Jewish audience for Rodríguez, since he argues that “even here Paul is imagining himself
addressing a gentile who ‘call[s him]self a Jew.’” Rodríguez, If You Call, 10n.26, emphasis
original.
98  Ibid., 23–24.
99  Ibid., 27.
100  Ibid., 28.
101  Ibid., 28–29, emphasis original.
102  Ibid., 29–30, 33.
103  Ibid., 31–32, emphasis original.
104  Ibid., 33.
180 chapter 8

function of διό in Rom 2:1 and a linear reading of the text.105 Rodríguez identi-
fies 2:1 as the first occasion of diatribe in Romans, and he follows Stowers’s
emphasis on its function as a pedagogical rather than polemical tool.106 In
2:1, Paul “conjures up an imagined dialogue partner, a gentile moralist who,
like Paul, disapproves of those who lose control of their emotions or desires
and succumb to the power of their passions.”107 Rodríguez continues to argue
that nothing in Rom 2:1–16 indicates that Paul’s interlocutor has transitioned
to Christian existence; for this reason, the interlocutor is “without excuse” for
precisely the same reason as those described in 1:18–32, namely, the failure
to worship God as God.108 The moralist, therefore, is no better off than the
gentile in 1:18–32, for “God judges everyone who refuses to acknowledge him
as Creator, regardless of their moral status,” “for there is no favoritism with
God” (2:11).109
In 2:12–16, however, Paul draws a distinction between gentiles who do not
have the Torah yet do the things of Torah and those who have the Torah but
fail to do it. Rodríguez explains the case of gentiles who do not have the Torah
but nevertheless do the work of the Torah inscribed on their hearts as those
who worship God as creator, which, for Paul, is the summation of Torah obser-
vance. Thus, these gentiles do not circumcise, observe Sabbath, or follow food
laws, but they effectively do the Torah. On the contrary, gentiles who attempt
to practice Torah inevitably fail to observe the Torah, the reason for which Paul
(and Rodríguez) spells out later in the chapter.110
With 2:17–29, a new interlocutor comes on the scene; the addressee in 2:1–11
is a gentile moralist who does not worship Israel’s God, but the addressee in
2:17–29 is a “gentile proselyte to Judaism.”111 By taking on the name and call-
ing himself a Jew, this ethnically gentile but religiously Jewish interlocutor
has adopted the signs of the Mosaic covenant, including circumcision.112 By
adopting such an identity for the interlocutor, Rodríguez’s reading “configures
Paul’s discussion here in terms of an enthusiasm for Torah among Paul’s gentile

105  Ibid., 35–36.

106  Ibid., 36–37.
107  Ibid., 37.
108  Ibid., 38.
109  Ibid., 39–43.
110  Ibid., 44–46.
111  Ibid., 49–50. Rodríguez therefore distances himself from Thorsteinsson at this point, as
Thorsteinsson holds that epistolary interlocutors tend not to change unless specifically
indicated. Thorsteinsson, Paul’s Interlocutor, 144.
112  Rodríguez, If You Call, 50–51.
Traditional Readings of the Dialogue in Romans 3:1–9 181

audience rather than any anti-Jewish sentiment.”113 As such, Paul describes the
interlocutor in positive terms in 2:17–22. In fact, Rodríguez reads 2:21–22, which
are usually read as Paul’s critique of the interlocutor, as functioning to bolster
his character. Based on the presence of οὐ (which expects a positive answer)
rather than μή in the interrogative, Rodríguez argues that “Paul grants his in-
terlocutor the benefit of the doubt: His moral behavior is consistent with his
moral instruction.”114 That is, the interlocutor teaches himself, does not steal,
and so forth. In Rom 2:23–29, however, Paul reveals this interlocutor’s problem.
The interlocutor boasts in Torah rather than worshipping God as Creator (2:23)
and transgresses the law by improperly observing the rite of circumcision by
not completing it on the eighth day and by not being members of Abraham’s
household, as stipulated in Torah (2:25–29).115 Thus, “rather than a sign of the
gentile proselyte’s ‘ultimate commitment’ to observe Torah’s commandments
and prohibitions, circumcision has become the paradigmatic locus of the pros-
elyte’s transgression of Torah!”116
Rodríguez recognizes that rhetorical dialogue “usually involves the speaker
characterizing his dialogue partner,” and he considers 2:17–29 to fill this role.
Thus, the gentile proselyte addressed in 2:17–29 remains on the scene for
Rom 3’s diatribal dialogue.117 Indicative of his understanding of the script in
3:1–9, Rodríguez suggests that, in their discourse, Paul “[fields] questions and
[explains] why [the interlocutor’s] confidence in Israel’s Torah and its ability
to facilitate his worship of Israel’s God was misplaced.”118 Rodríguez allows one
alteration in the traditional script, 3:8. Rodríguez identifies 3:1, 3, 5a, 7, and 9a
as the interlocutor, and 3:2, 4, 5c, 6, 8 and 9b as Paul. Consequently, with the
exception of Paul’s asides in 3:5b and 8, “The interlocutor consistently identi-
fies an (erroneous) implication from Paul’s argument, and Paul responds to
correct the proselyte’s misunderstanding.”119
To move on to the dialogue, Rodríguez’s proselyte has taken on the full
yoke of the Torah and, as such, has missed what doing law really requires
of gentiles. Understandably, the interlocutor’s opening question is personal,
as he asks what advantage exists for being called a Jew and undergoing

113  Ibid., 51.

114  Ibid., 53–55, emphasis original.
115  Ibid., 56–59.
116  Ibid., 59, emphasis original.
117  Ibid., 62n.58.
118  Ibid., 62.
119  Ibid., 64–65.
182 chapter 8

circumcision (3:1).120 Paul affirms an advantage “for being—and being called—

a Jew,” as represented by the covenant with Yahweh, i.e., Torah (3:2).121 This
leads the proselyte to wonder about the status of God’s faithfulness in view of
Jewish unfaithfulness, which Paul claims the proselyte has committed through
inappropriate circumcision (3:3). Paul affirms God’s faithfulness as fixed
through and through (3:4). Noticing how unfaithfulness functions to magnify
God’s faithfulness, the proselyte then asks how God can be just for bringing
wrath on those who magnify his righteousness. Essentially, the proselyte asks,
“why would circumcision after the eighth day not be reckoned as faithful
obedience of God’s command?” (3:5, 7); Paul summarily dismisses these ques-
tions as absurd (3:6, 8).122 Consequently, Rodríguez argues, the proselyte’s
questions “assume that neither Israel’s nor his own transgressions of Torah
nullify the promises God made with his people, and he attempts to wrangle
from Paul a concession that, in fact, his circumcision brings him in the sphere
of God’s people,” which Paul emphatically denies.123 The proselyte’s final plea
again asks whether Jews have any advantage (3:9a). This time, however, Paul
rejects Jewish advantage; “despite the advantage of Israel’s election, her being
entrusted with Torah/the oracles of God, of being the covenant people, and so
on, in actual fact no advantage has attached to being a Jew … Because Jews, like
Greeks, find themselves under sin.”124 Paul’s catena in 3:10–18 further supports
this view.
Paul returns to the question of Jew-non-Jew relations before God in chap-
ters 9–11, where he argues that God deals equally with Jew and gentile. In
fact, Rodríguez argues, God brings his promises to fruition for Jews and non-
Jews according to a single set of mechanics: God meets disobedience with
divine mercy. In 11:25–32, however, Paul apparently reverses his “for the Jew
first, then for the Greek” formula, as it is non-Jews who first experience God’s
mercy, which only subsequently extends to disobedient Jews.125 Nevertheless,
Rodríguez contends, “The discrepancy hardly matters, however, for the point
of Paul’s argument, here in Romans 9–11 and throughout the entire letter, cen-
ters on the equality between Jew and gentile in God’s economy.”126

120  Ibid., 62–63, 62n.59.

121  Ibid., 63–65.
122  Ibid., 66–67.
123  Ibid., 66.
124  Ibid., 68, emphasis mine.
125  Ibid., 229.
126  Ibid.
Traditional Readings of the Dialogue in Romans 3:1–9 183

Though their attention to dialogue in 3:1–9 is significantly less pronounced,

two scholars who have each composed influential commentaries on Romans,
Dunn first, then Wright, deserve mention. The concluding taxonomies of
potential advantages and problems manifested in traditional readings illus-
trate that their presentations also prove to contain significant gaps.
According to Dunn, after the catalog of traditional Hellenistic Jewish cri-
tiques of non-Jews in 1:18–32 that ultimately applies to all humanity,127 Paul
begins to address an interlocutor. Dunn argues that though this interlocutor
is initially unidentified, the “Jewish” style of 1:19–32 indicates that the interloc-
utor is a Jew.128 In fact, the interlocutor is a Jew who agrees with the critiques
of non-Jews in the Wisdom of Solomon that are paralleled in Rom 1:19–32.129
Rom 2 builds in specificity regarding the interlocutor and Paul’s critique of
him. By the end of the chapter, it is clear that the Jewish interlocutor presumes
that his possession of νόμος and its physical manifestation in circumcision
distinguish him from non-Jews in God’s eyes. The effect of this presumption is
overconfidence in God’s favor for, and obligation to, Israel, such that God will
only meet the privileged Jew’s sin with mercy.130 Thus, the Jewish interlocutor
believes his possession of νόμος and circumcision functions as a “talisman” or
“prophylactic” at the judgment and grants him a superiority over non-Jews,
thereby demonstrating a nationalistic, or ethnocentric, exclusivity in which
God belongs solely to the Jews.131
Paul, of course, critiques this view; God does not judge Jews who do not do
νόμος any differently than gentiles who do not do νόμος (2:12–16). Thus, Dunn
claims, “God’s judgment will be completely evenhanded,”132 and simple posses-
sion of νόμος affords the Jew no advantage over non-Jews. In this way, Paul aims
to “deflate” the Jew’s presumption that God is predisposed in Israel’s favor.133
The dialogue in 3:1–8 clarifies that this is the case, as the interlocutor
objects and wonders what advantage there is to being Jewish and circumcised
(3:1).134 In his own voice, Paul responds that Jews are advantaged by being
entrusted with the λόγια (3:2), but he does not indicate how that constitutes

127  Dunn, Romans, 55–56, 70.

128  Ibid., 76–79; idem., Theology, 115.
129  Idem., Romans, 82–83, 89; idem., Theology, 116.
130  Idem., Romans, 77, 90–91, 108; idem., Theology, 116.
131  Idem., Theology, 117; idem., Romans, 110. The term “ethnocentric” comes from Bruce W.
Longenecker, Eschatology and the Covenant: A Comparison of 4 Ezra and Romans 1–11
(JSNTSup 57; Sheffield: JSOT Press, 1991), 27–31.
132  Dunn, Theology, 117.
133  Idem., Romans, 77.
134  Ibid., 129; idem., Theology, 118n.75.
184 chapter 8

an advantage.135 At this point, for Dunn, the dialogue quickly begins to

break down and gets “out of hand.”136 Dunn asserts that though “Paul at first
perseveres with the diatribe style” of Rom 2 in 3:1–8, “the debate becomes
increasingly with himself.”137 The only other comment Dunn makes relevant
to the structure of 3:1:8 as dialogue is that Paul is surely responsible for the
μὴ γένοιτο rejections in 3:4 and 6, on the grounds that Paul could not attribute
such a phrase to another speaker.138 So, to the degree that Dunn concerns him-
self with the dialogue in 3:1–8, he resembles the traditional reading at each of
these points. In any case, by the end of the pericope, Paul illustrates that God
does not only judge gentile idolatry, “but also the idolatrous misplaced con-
fidence of his own people in their own God-given religion and status before
God,” including their nationalistic pride.139 Thus, 3:1–8 draws the final connec-
tion that leads to the indictment of all in 3:9 and sets the stage for conversa-
tions to occur in the rest of the letter.140
One such conversation involves the question of Jewish advantage and
Israel’s fate. Are Jews advantaged over non-Jews (cf. Rom 2)? Do the λόγια
really constitute an advantage for Jews (cf. 3:2)? According to Dunn, Paul says
“no” to the former but “yes” to the latter. This conversation of course returns
in chapters 9–11, where Paul addresses Israel’s eschatological fate. Here, Dunn
argues, Paul affirms Israel’s call, election, and the promises, but these apply
neither on the basis of genealogy nor covenant faithfulness (9:7–13), as Paul
indicates in Rom 3–4.141 Instead, God hardens or shows mercy to whomever he
wills (9:18). God’s sovereignty over creation does not have as its primary pur-
pose the hardening of humanity, but a display of mercy. Even if part of Israel
is hardened and “fills the role of the ‘vessels of wrath’” in the present, Paul’s
concept of predestination looks forward to the “not yet.”142 At that time, the

135  Ibid., 45.

136  Idem., Romans, 145. So also Heikki Räisänen, “Zum Verständis von Röm 3,1–8,” in The Torah
and Christ: Essays in German and English on the Problem of the Law in Early Christianity
(Publications of the Finnish Exegetical Society 45; Helsinki: Finnish Exegetical Society,
1986), 185.
137  Dunn, Romans, 129. Similarly, Moo argues, “Paul is not so much reproducing for his read-
ers an argument between himself and another person as he is posing questions and
objections to himself.” Moo, Romans, 181.
138  Dunn, Theology, 118n.75. Dunn intends his comment as critique of Elliott (Rhetoric, 139–
41) and Stowers (Rereading, ch. 5), based on Malherbe’s assessment in “Μη Γενοιτο.”
139  Dunn, Theology, 118–19.
140  Idem., Romans, 77, 148–49, 156.
141  Idem., Theology, 510–11.
142  Ibid., 511–13.
Traditional Readings of the Dialogue in Romans 3:1–9 185

division within historic Israel will be healed, and the division between historic
Israel and the rest of those called by God will disappear.143 Thus, Paul “does
not weaken his commitment to the Gentiles in any degree. But he holds forth a
hope for the final salvation of Israel which is characteristically Jewish through
and through,” being based on God’s call.144 In this way, “The gifts and calling
of God are irrevocable … The God of Israel remains faithful to Israel; his righ-
teousness endures to the end.”145
Finally, for Wright, Rom 2:1–11 critiques all humans, but 2:17 turns specifi-
cally to a Jew.146 The Jew Paul begins to address in 2:17 is, at first, quite similar
to Dunn’s. This Jew is not one who attempts “to use the law as a ladder of good
works up which to climb to a moral self-righteousness,” but one who relies on
a “national righteousness.”147 That is, this Jew boasts in and relies on the law
“as the covenant badge which would keep membership within that covenant
limited to Jews and Jews only.”148 The covenant, therefore, in this Jew’s mind,
is exclusive and based on ethnicity and Torah possession.149 This view stems
from the sense that, though they have returned from Babylon, Israel remains
in exile and longs for the renewal of the covenant, which they seek to establish
on and as their own.150
The way Wright formulates Paul’s problem with this Jewish position, how-
ever, is quite different.151 Wright argues, “The presuppositions of all Paul’s
thought, as of more or less all serious Jewish thought, is that in some way or
other Israel is the solution of the creator god, [Yahweh], to the problem of
the world.”152 Said otherwise, Israel’s vocation is to be the light of the world.153
Israel, however, lost sight of that vocation and, by clinging to an exclusive
nationalism and possession of law as that which sets Israel apart from other

143  Ibid., 527.

144  Ibid., 528.
145  Ibid.
146  Wright, “The Law,” 147.
147  Ibid., 139.
148  Ibid.
149  Ibid., 142.
150  Ibid., 147. The theme of Israel’s continuing exile plays a significant role in many of Wright’s
publications.
151  Not only is Wright’s formulation different, it is also more difficult to discuss linearly.
When Wright addresses Rom 2 and 3:1–8, he tends to argue from 3:2–3 backwards into
Rom 2. This summary of Wright attempts a linear rather than retrospective presentation
of Wright’s argument.
152  Wright, “The Law,” 147.
153  Idem., “Romans 2:17–3:9,” 2.
186 chapter 8

nations, became sinful. Israel, therefore, needs a “physician’s physician, one

who could do for Israel, and hence for the world, what neither could do for
themselves or for each other.”154 In Wright’s schematic, Paul argues that this
physician had in fact come and the covenant been renewed in Christ, as is
clear from the existence of (Christian) gentiles who surprisingly and paradoxi-
cally keep the law (2:12–16; 25–29) and find themselves to be the beneficiaries
of this new covenant.155 Essentially, what Israel failed to do for the world, God
accomplished by sending Christ in its place, so that he, the creator God, might
rescue the created order that he has committed himself to rescuing in some
sense through Israel.156
For Wright, the “linchpin” for this view comes in 3:1–9, especially 3:2–3.157 If
the gospel and God’s wrath equally apply to Jew and gentile, and if Paul “really
does envisage people of any and every background being regarded as ‘circum-
cision’ and as ‘Jew’” (2:25–29), then, indeed, Paul needs to ask what advantage
there is to being Jewish (3:1).158 Paul’s answer is that Israel was “entrusted” with
the λόγια (3:2), which means to be given something which one must “take care
of and pass on to the appropriate person,” as Paul speaks concerning his own
gospel (1 Cor 9:17; Gal 2:7; 1 Thess 2:4).159 In this sense, Wright argues, God gives
the λόγια to Israel in order for Israel to pass them on to the people for whom
they were intended, namely, the gentiles.160 Israel, however, is unfaithful to this
vocation (3:3); instead of sharing the λόγια with the world, Israel boasts in its
exclusive possession of them. Accordingly, Paul does not critique Israel for a
lack of πίστις in the Christian sense, but specifically for failing to relay God’s
λόγια to the world.161 Nevertheless, despite Israel’s faithlessness to her com-
mission, God remains faithful to his promise to bless the world through Israel

154  Idem., “The Law,” 147.

155  Ibid., 147.
156  Idem., “Rom 2:17–3:9,” 2; idem., Paul, 836.
157  Idem., “Rom 2:17–3:9,” 2. Note that in Wright’s numerous publications on Rom 3, he never
addresses the question or significance of dialogue in the pericope, much less the issue of
which voice speaks which lines.
158  Idem., Paul, 836–37.
159  Ibid., 837, italics original; idem., “Rom 2:17–3:9,” 2–3.
160  So also Jewett, Romans, 243, but see 244–45 where he connects ἀπιστία with a rejection of
Jesus as Messiah.
161  Wright, “Rom 2:17–3:9,” 3–4, 7. Moo argues the opposite point. For Moo, in light of
Rom 11:17, Jewish faithlessness to the covenantal obligations (3:2–3) indeed carries the
meaning of “the Jews’ failure to embrace Jesus as the Messiah.” Moo, Romans, 184–85.
Traditional Readings of the Dialogue in Romans 3:1–9 187

(3:4), which he accomplishes through Christ, the “faithful Israelite,” “Israel’s

representative, Israel-in-person if you will” (3:21–26).162 This insures that God
cannot be found unfaithful, unjust (cf. 3:5–6), or untrue (cf. 3:7), and it (i.e.,
3:1–9) provides a “summary of the key moves that Paul will make throughout
much of the letter,” especially Rom 9–11.163
For instance, the instrumentality in God’s plan, not only of Israel but also of
the church, rises again to the fore in Rom 9–11. On the one hand, it is through
Israel’s trespass, disobedience, and casting away that God rescues the world
through Israel, “if not through their faithfulness, then through their unfaith-
fulness” and by means of the Messiah.164 On the other hand, in keeping with
Israel’s own scriptures (cf. Deut 32:21), non-Jewish Christians play the part of
making Jews jealous in order to shock Jews into faith (11:14).165 Thus, Wright ar-
gues, it is indeed through Israel that God blesses the gentiles, and it is through
the gentiles that God then works to rescue Jews. Moreover, lest it appear as
though God in fact shows any degree of national partiality for Israel as the Jew
in 2:17–29 supposes, chapter 11 concludes with a resounding note of equality.
God imprisoned Israel and non-Israel under disobedience (11:32), so that the
Messiah’s faithful and obedient death and resurrection procure salvation for
all who come to faith in Jesus as the risen Messiah and Lord. Justification and
salvation are matters of mercy not only for gentiles, but also for Jews. In this
way, God is faithful to the covenant with respect to both Jews and gentiles.166

Preliminary Conclusions

Traditional readings of the script of 3:1–8/9 tend to understand an imaginary

interlocutor posing objections to Paul in the form of questions, which Paul
responds to in his own voice. With minute variations, such readings identify
the speaker in 3:1, 3, 5a, 7–8a, c, and 9a as the interlocutor, and the respondent
in 3:2, 4, 6, 8d, and 9b as Paul. Many view 3:5b and 3:8b as authorial asides or
commentary by Paul that do not belong to the imaginary dialogue. The script
is as follows:

162  Wright, Paul, 838–39, italics original.

163  Idem., “Rom 2:17–3:9,” 5.
164  Ibid., 20–21.
165  Ibid., 21.
166  Idem., Romans, 694–95.
188 chapter 8

Interlocutor: Therefore, what advantage comes from being Jewish, or

what benefit accrues from circumcision? (3:1)
Paul’s Response: There is much [advantage] in every way! To begin,
they were entrusted with the oracles of God. (3:2)

Interlocutor: To what end? If some lacked πίστις, their lack of πίστις will
not nullify God’s πίστις, will it? (3:3)
Paul’s Response: Absolutely not! Instead, let God be true but every
human a liar, as it is written, “So that you might be justified in your
words, and you will overcome when you are judged.” (3:4)

Interlocutor: But if our unrighteousness proves God’s righteousness, what

shall we say? God is not unjust when he brings wrath, is he? (3:5a)
Paul’s Authorial Aside: (3:5b) I speak in a human way.
Paul’s Response: (3:6) Absolutely not! Otherwise, how will God judge
the world?

Interlocutor: But if God’s truthfulness is increased for his glory by my

lie, why am I still being judged as a sinner? Why not, (Paul’s Authorial
Aside?) as we are slandered and as some claim that we say, (Return to
Interlocutor?) “Let us do evil so that good might come?” (3:7–8c)
Paul’s Response: Their judgment is justly deserved. (3:8d)

Interlocutor: What then? Are we advantaged or disadvantaged? (3:9a)

Paul’s Response: By no means! For, we have charged both Jews and
Greeks all to be under Sin, as it is written … (3:9b–10)

Each of the scholars discussed above read Rom 3:1–8/9 according to a tradi-
tional script and displays areas of strength and areas in need of further clar-
ification.167 Points of obvious strength extend back to Godet, and no doubt
even further. The shining points in Godet’s treatment are his relatively simple
but significant exegetical observations about the interrogatives in 3:1–8 that
are negated with μή, which imply a negative answer, and his insistance that

167  I do not intend the following discussions of strengths and weaknesses to be exhaustive
or even determinative at this point. Rather, these taxonomies indicate points of both
potentially positive and potentially negative sorts for which a successful treatment of
Rom 3:1–8 must be able to account, in one way or another. I address the merit of these
issues and others at the exegetical level in Chapters 10 and 11
Traditional Readings of the Dialogue in Romans 3:1–9 189

Paul speaks these verses (3:3, 5, [and 8, though he does not discuss it]).168 Any
arrangement of the dialogue in 3:1–9 must account for this use of μή, and
traditional readings especially must explain why an interlocutor would pose
“objections” he or she expects Paul to dismiss from the very beginning. For
this reason, Godet’s work impacts all subsequent dialogical readings of the
pericope. Of the views discussed so far, only Tobin’s portrayal of the dialogue
accounts for this, since Paul employs this interlocutor specifically to demon-
strate where he and his Romans Christian audience are in agreement.169
Moving on, some of Bultmann’s observations are also noteworthy.
Bultmann’s recognition that Paul formulates the words of the interlocutor in
such a way as to advance his own argument is spot on, and this holds true
whether Bultmann’s arrangement of the script is correct or not.170 To be sure,
Bultmann arrives at this conclusion as a result of finding Paul’s rhetorical acu-
men to be beneath that of other Greek writers.171 One can, however, arrive at
the same conclusion by attributing to Paul a more moderate degree of rhetori-
cal sensibility. Speech-in-character teaches that it is the duty of the primary
speaker or author to craft appropriate speech for an imaginary speaker and to
do so in such a way as to meet various rhetorical goals (see Part 1). Bultmann’s
conclusion matches the conventions for speech-in-character, even though
he grounds his argument elsewhere. Also positive, especially when viewed
in retrospect, is Bultmann’s classification of the objections as “absurdities”
(Absurditäten), incorrect conclusions based on the way Paul has developed his
argument to this point.172 This, too, can hold true regardless of the structure of
the script.
Malherbe’s refinement of Bultmann concerning the function of the μὴ
γένοιτο formulas is also quite helpful. Malherbe not only provides an histori-
cal example (Epictetus) which Paul’s use of μὴ γένοιτο mirrors. Malherbe also

168  Godet, Romans, 131–39; BDF §427.

169  Tobin, therefore, implicitly identifies a hole in Godet’s otherwise astute reasoning, name-
ly, what if the imaginary interlocutor agrees with Paul on certain points? Then, such an
interlocutor could naturally pose interrogatives with μή in expectation of Paul’s negative
response. Tobin, Paul’s Rhetoric, 119–20. Moo also correctly recognizes the significance of
μή in the interrogatives, such that Paul should be the one responsible for formulating the
interlocutor’s objections in this way. Moo, Romans, 185–91, 194n.92. For Moo, however,
this becomes the undoing of any meaningful dialogue in Rom 3:1–8, as the dialogue es-
sentially dissolves into a back-and-forth monologue between Paul and himself. Ibid., 181.
170  After all, Paul is responsible for the whole letter. Bultmann, Der Stil, 67.
171  Ibid. Bultmann writes, “Die Fiktion des mitredenden Gegners hat für ihn nicht die Kraft, die
sie bei den Griechen hat.”
172  Ibid.
190 chapter 8

persuasively documents the supporting, explanatory, and introductory role of

the comments that follow the μὴ γένοιτο rejections.173 Accordingly, Malherbe is
followed by many, especially Song.174
Tobin’s work contains many positive points as well. First, Tobin’s iden-
tification of the body of Romans as diatribe, or at least heavily diatribal, is
attractive, as long as “diatribe” is defined as a rhetorical genre or category
rather than as a literary genre.175 Also, Tobin’s comparison of the structure of
the body of Romans with some of Epictetus’s diatribes that rotate cyclically
between expository and argumentative pericopae is quite strong.176 Second,
on an exegetical level, Tobin’s discussion of the anonymity and ambiguity
at several points in Rom 1–3 helpfully paves the way for how to understand
Paul’s use of “all” throughout the letter.177 Third, and finally, Tobin’s emphasis
on God’s impartiality as a, or the, driving hermeneutical key in this passage is
especially helpful.178 For, clearly, divine impartiality is a feature that ultimately
renders Paul’s inclusive gospel for the gentiles possible.
The final contribution of these “traditional” views is represented by many,
but specifically so by Dunn and Wright. This point has more to do directly with
Rom 9–11, but it relates to 3:1–9 as well. At times, Paul speaks primarily about
Jews. At other times, Paul speaks more about non-Jews. At still other times,
Paul speaks about all, Jew and non-Jew, without distinction. It is important to
keep both Jew and non-Jew in mind at all times and not to lose sight of what
Paul is saying about, and more importantly what he thinks God is doing to,
one or the other. It may very well turn out that in texts where it seems Paul
is speaking primarily about Jews or non-Jews, the other is equally in sight, if
perhaps slightly in the background. Even in Paul’s discussion of Israel in 9–11,
Dunn recognizes that Paul “does not weaken his commitment to the Gentiles
in any degree.”179 Wright also finds non-Jews to be prominent in 9–11; as God
works through Israel to procure the redemption of non-Jews, so also God will
use non-Jews to provoke Israel to jealousy and faith.180 In this sense, what
God does to Israel, God also does to the gentiles, and reading Rom 9–11 with

173  Malherbe, “Μη Γενοιτο,” 236–37.

174  Song, Reading Romans, 112n.5.
175  See the discussion of Stowers on this point in Part 2, Chapter 6.
176  Tobin, Paul’s Rhetoric, 91–98.
177  Ibid., 108–12.
178  Ibid., 104–18.
179  Dunn, Theology, 528.
180  Wright, “Rom 2:17–3:9,” 20–21.
Traditional Readings of the Dialogue in Romans 3:1–9 191

an eye only toward Israel risks missing this feature of Paul’s argument (see
Chapters 10 and 11).
Despite the contributions of these traditional readers, their approaches
also have several important problems. First, many are unable to explain the
use of μή rather than οὐ in the interrogatives of 3:3, 5, and 8 and the relevance
this discussion has for understanding the script and import of the dialogue. It
is only Tobin who provides a reasonable explanation for this phenomenon.181
Second, and a matter of fact rather than critique, many of these traditional
readings appear before discussion of diatribe and Romans come into full swing,
much less rescriptive readings of Romans. Godet and Sanday and Headlam
do not find dialogue in 3:1–8, as they precede Bultmann and are not privy
to later discussions of diatribal dialogue and Romans. Similarly, Malherbe’s
“Μη Γενοιτο” sits in limbo between traditional diatribal readings and rescriptive
readings. Malherbe is aware of Stowers’s dissertation,182 but Stowers had not
yet published “Paul’s Dialogue,” in which he argues for the first revised script
of the passage. Thus, one would altogether expect Malherbe to affirm a tradi-
tional script based on the timing of his publication.
Third, even among traditional readings that appear after the onset of rescrip-
tive readings, there is inadequate engagement with the rescriptive offerings.
For instance, several (incorrectly) conclude that Stowers altogether dismisses
polemic from diatribe in favor of collegial and educational discourse (see
Part 2),183 or they neglect to engage with relevant literature such as Elliott’s
The Rhetoric of Romans (1990, 2007) or Campbell’s The Deliverance of God
(2009).184 This latter critique is particularly pronounced, especially since
scholars like Elliott, who is followed by Campbell, use Stowers’s own terms

181  Rodríguez’s explanation of the interlocutor and the dialogue accounts for the occurrence
of μή in 3:3, but it fails to do so for 3:5. Rodríguez’s argument accounts for 3:3 because
his interlocutor actually wants Paul to reject the premise that faithlessness might nul-
lify God’s faithfulness. Rodríguez’s analysis fails with respect to 3:5, however, because,
in order for the interlocutor’s unseasonably late circumcision to be credited as faithful
obedience to the Torah, he needs Paul to affirm rather than deny that it would be unjust
for God to bring wrath on those who magnify his glory. Rodríguez, If You Call, 66–67.
182  Malherbe supervised the dissertation and cites Stowers twice. Malherbe, “Μη Γενοιτο,”
231n.2, 239n.23. At the same time, however, Stowers had read Malherbe’s unpublished
paper before the completion of his dissertation. Stowers, The Diatribe, 124.
183  Cf. Jewett, Romans, 239n.3.
184  Song engages with Stowers and confesses that his reading “may be also possible,” but there
is no awareness of Elliott (Rhetoric), and the same is true for Tobin. Similarly, Rodríguez
neglects to engage Campbell (Deliverance). Song, Reading Romans, 112n.5, 7; Tobin, Paul’s
Rhetoric; Rodríguez, If You Call.
192 chapter 8

to argue for an even more aggressive revision of 3:1–9 than Stowers imagines
(see below).
Fourth, practically all presentations of the traditional script of 3:1–9 are
based on impermissible arguments, ‘which are not really other arguments’
(cf. Gal 1:6–7) but sweeping assumptions. It is often the case that no argu-
ment whatsoever is made about the script of the dialogue, and the tradition-
al formulation is holistically appropriated.185 Bultmann, for example, gives
no voice to the possibility that the script might be altered or flipped; such a
question is altogether unconsidered at his point in time. Similarly, Rodríguez
assumes rather than argues for his presentation of the dialogue’s arrangement.
Rodríguez at least suggests that his presentation maintains consistency in that
the interlocutor asks all the questions,186 but he does not take diatribe’s high
degree of variability into account in the least (see Part 2), which limits the
import of this consistency.187
When supporting points are presented, however, even they are by and large
based on unwarranted assumptions. This is particularly true regarding the
use of μὴ γένοιτο. To begin, Stowers’s dissertation demonstrates that Epictetus
can use μὴ γένοιτο as a rejection either in his own voice or in the voice of an
interlocutor.188 Part 2 illustrates this further in its discussion of Disc. 2.23. As
far as Malherbe is concerned, however, his shape of the script in Rom 3:1–8
is a foregone conclusion; given the assumed identity of the speaker behind
occurrences of μὴ γένοιτο elsewhere in Paul, the phrase belongs irrevocably to
Paul’s voice in 3:4 and 6.189 So also Song, who actually comments on Epictetus’s
Disc. 2.23 but ignores the question of who speaks which lines. Instead, Song
simply assumes Epictetus must be responsible for the μὴ γένοιτο formula, just
like he assumes Paul must be responsible for the formula in Rom 3:4 and 6.190
This line of thinking is crystal clear in Dunn’s work as well. Dunn critiques
Elliott’s “surprising” proposal and Stowers’s “only partially more plausible read-
ing,” both of which identify the interlocutor as responsible for the rejection
formulas in 3:4 and 6 rather than Paul, simply on the grounds that the phrase
is used elsewhere in Paul.191 Matera too dismisses Stowers’s reading on the

185  See Jewett, Romans, 239–52, especially 239n.1.

186  Rodríguez, If You Call, 64–65.
187  Though this is true, note that the script I ultimately propose appeals to a consistency as
well, but as a second-order piece of evidence.
188  Stowers, The Diatribe, 128–29.
189  Malherbe, “Μη Γενοιτο,” 234–36.
190  Song, Reading Romans, 35–36, 66–67, 94–95, 112n.5, 7.
191  Dunn, Theology, 118n.75.
Traditional Readings of the Dialogue in Romans 3:1–9 193

grounds that “it attributes questions that begin with [μὴ γένοιτο] (of course
not) in 3:4 and 3:6 to Paul’s dialogical partner, whereas in the rest of Romans,
Paul reserves this expression for himself (3:31; 6:2, 15; 7:7, 13; 9:14; 11:1, 11).”192
Matera, however, makes no argument for the case that any of these examples
of the phrase actually belong to Paul (at the moment, whether they do or do
not makes no difference). The point is, Matera assumes they belong to Paul,
the case is closed, and Matera’s reading of Romans is unjustifiably limited with
respect to possible outcomes.193
The problems with this line of reasoning are pluriform. To begin, the whole
system is based on circular reasoning; because this phrase is present elsewhere
in Paul’s letters in his own voice, it is assumed it must also be spoken in Paul’s
voice in 3:4 and 6. In addition, scholars who appeal to the use of the phrase
elsewhere in Paul’s writings neglect to consider whether or not those other
instances are actually spoken in Paul’s voice. In fact, more than one scholar
has argued that the dialogue of the infamous “I” in Rom 7, which contains
two instances of the μὴ γένοιτο formula (7:7, 13), is best understood as spo-
ken in a voice other than Paul’s.194 Furthermore, the view that all μὴ γένοιτο
formulas must be spoken in Paul’s voice altogether ignores the fact that Paul
is responsible not only for his own side of the dialogue but also for crafting his
interlocutor’s speech, as Bultmann notes195 and the conventions for speech-in-
character presuppose (see Part 1). Since Paul is so comfortable with μὴ γένοιτο,
it is surely possible that Paul could attribute this language to his interlocutor
in an appropriate context, especially a diatribal context in which Epictetus
represents a precedent for the use of the phrase both on the diatribal teacher’s
lips as well as on the interlocutor’s. The point I am trying to make is not whether
μὴ γένοιτο does, does not, or cannot be spoken in Paul’s voice in 3:4 or 6; this
will be addressed in Chapter 11. The point is that arguments for the traditional
script are based on insufficient logic. If the traditional script is going to be
maintained, so that Paul speaks 3:4 and 6 and the μὴ γένοιτο formulas therein,
actual evidence must be offered, arguments weighed, and conclusions drawn.
Simple assumptions cannot suffice, especially given the presence of rescriptive
alternatives.
I argue rather bluntly here not only because the (ab)use of the μὴ γένοιτο
phrase is so problematic and entrenched among traditional readers, but also

192  Matera, Romans, 80.

193  Similarly, see Jewett, Romans, 245n.61.
194  See Longenecker, Rhetoric, 88–93; King, “Rhetorical Chain-Link Construction.” See also
Stowers, “Romans 7.7–25.” Contra Timmins, “Romans 7.”
195  Bultmann, Der Stil, 67.
194 chapter 8

because the same type of assumptions seep into other supporting claims
for the traditional reading of 3:1–9. For example, concerning 3:3–4, Tobin
argues that Rom 3:4 belongs to Paul because he cites scripture elsewhere
in chapters 1–3, and 3:4 contains a citation of scripture. Consequently, 3:3
belongs to Tobin’s interlocutor.196 Tobin’s attempt to identify Paul’s speech by
his practices elsewhere is not necessarily wrong-headed, but neither is it overly
strong in this case. By itself, it is no less circular than arguments regarding
μὴ γένοιτο. Plus, on the premise that Paul would only attribute speech to his
interlocutor that appropriately models the interlocutor’s character (see Part 1),
attributing to the interlocutor a citation of scripture would not be problem-
atic in the least for Tobin’s characterization of the interlocutor. For, in Tobin’s
own terms, Paul’s Roman Christian audience, who the interlocutor represents,
reveres the Mosaic law and Jewish scriptures. Similarly, concerning 3:5–6,
Tobin argues that Paul speaks 3:6 (and the interlocutor 3:5) because Paul is
defending himself and answering objections.197 This, too, is based altogether
on an assumption resulting from decades of the traditional script occupying
the position of the majority view. Again, the point is not that such a script
necessarily cannot be argued for; the point is that no one has actually argued
successfully for it.
Fifth, to put the last point plainly, speech-in-character and its conventions
are widely ignored among these readings, though they arguably provide the
best point of view from which to consider the attributed speech in 3:1–9’s
dialogue.198 Rarely a scholar will include a line about the importance or
applicability of speech-in-character, but the reference inevitably amounts
to nothing more than a “name drop.” For instance, Rodríguez correctly notes
the norm for characterization to precede imaginary dialogue, but he does not
provide any detailed discussion of it, nor does he model how drawing on the
characterization helps to define the divisions within the script.199 Instead,
the reference to speech-in-character remains relatively untapped and all its
possible supporting evidence stored away to continue fermentation.
Sixth, traditional readings present a dialogical script that is potentially inco-
herent. For example, the authorial asides in 3:5 and 8 are obviously disruptive
to the flow of the discourse—that is how traditional readers recognize them
as authorial asides. These awkward interruptions surely contribute to Dunn’s

196  Tobin, Paul’s Rhetoric, 119.

197  Ibid., 119–20.
198  For instance, Moo, Romans, 177–97; Dunn, Romans, 129.
199  Rodríguez, If You Call, 62n.58.
Traditional Readings of the Dialogue in Romans 3:1–9 195

view that the dialogue “gets out of hand”200 and to Wright’s decision not to
talk about dialogue in 3:1–9 at all.201 The presumed authorial asides, however,
are not only intrusive; they are also unnecessary. Tobin suggests that Paul
inserts an editorial comment in 3:5b, “lest even the mention of the possibility
of God being unjust be taken amiss.”202 Such a comment, however, ends in
redundancy. Granting momentarily that the questions in 3:1–9 represent the
interlocutor’s objections, surely Paul’s μὴ γένοιτο denial in 3:6 (emphatically
translated “hell no!”) would sufficiently stifle any potential misunderstanding
about God’s justice from the objection in 3:5. Moreover, why would Paul allow
any of these objections to have the possibility of being misunderstood by his
audience? Why does Paul feel obliged to offer supplemental authorial com-
mentary on this objection? Surely God’s πίστις (3:3) is not (μή) less significant
than his δικαιοσύνη (3:5), is it? Absolutely not (μὴ γένοιτο), but Paul apparently
feels no need to issue an authorial comment in 3:3.
Finally, connected to the previous issue of incoherency, the traditional
script seems to depict Paul both affirming (3:2) and rejecting (3:9) Jewish
advantage over non-Jews. To many, the affirmation in 3:2 is particularly sur-
prising given Paul’s universal leveling in Rom 1–2. As C. H. Dodd famously
puts it, “the logical answer” to the question of Jewish privilege “on the basis of
Paul’s argument is, ‘None whatever!’”203 The question therefore arises, if God
is impartial and ultimately treats all equally (2:9–11), how can the Jews possess
any real, substantive advantage over gentiles? How can the “oracles” (λόγια)
give the Jews a salvific edge if God himself orchestrates their stumbling over
them (Rom 9:32–33; 11:8; cf. 11:32)?204 Of course, though some scholars do not
comment at any length on this seemingly strange juxtaposition,205 others offer
possible solutions. Perhaps the point is that Jews have an advantage, but it is
not a salvific advantage.206 Perhaps Paul is here holding in tension the idea of
God’s impartiality to all along with his faithfulness to Israel.207 Perhaps a better
translation of key terms in 3:9 eases the apparent contradiction.208 Or, perhaps

200  Dunn, Romans, 129, 146. See also Moo, Romans, 181.
201  Wright, “Rom 2:17–3:9;” idem., Romans; idem., Paul.
202  Tobin, Paul’s Rhetoric, 120n.44. So also Jewett, Romans, 248.
203  Dodd, Romans, 43.
204  On this latter point emphasizing God’s tripping of Israel, see Beverly Roberts Gaventa,
“Questions about Nomos, Answers about Christos: Romans 10:4 In Context,” forthcoming.
205  For example, Bultmann, Der Stil; Malherbe, “Μη Γενοιτο;” Song, Reading Romans.
206  For example, Hultgren, Romans, 135.
207  Longenecker, Eschatology, 195–97, 251–52.
208  So Jewett, Romans, 256–57.
196 chapter 8

the emphasis rests on 3:9, and Paul altogether rejects Jewish privilege.209
Again, deciding the correct or best view is not presently the agenda; that is
the task of Chapters 10 and 11. The point is, in every case, ‘traditional readings
first and rescriptive readings,’ scholars must be able to account for Paul’s cease-
less insistence on God’s impartiality to all, on the one hand, and his discussion
of Israel, non-Israelites, and God’s faithfulness to his promises to Abraham,
on the other hand. Such arguments will doubtlessly extend well beyond
Rom 3:1–9 in both directions and address multiple issues throughout the letter,
not least Paul’s discussion of Abraham in Rom 4 and Israel in 9–11.
But addressing any of these strengths and weaknesses is contingent first and
foremost on ‘correctly dividing the words’ and lines of 3:1–9’s dialogical script.
A properly arranged script cannot simply be assumed but must be argued for
with valid evidence and critical analysis. No one, however, has in fact done this
for the traditional reading. The long-standing, majority view of the traditional
reading can no longer be assumed to be the best reading simply on face value.
The reality that scholars or the church have read 3:1–9 according to the tradi-
tional script for so long does not suffice as an acceptable reason to endorse
such a reading. The door stands open for more persuasive arguments regard-
ing the script of the diatribal dialogue Paul stages in Rom 3:1–9 and following,
whether that proves to adhere to a traditional script or some rescripted version
of the discourse. With one small interlude, therefore, this project turns now to
consider rescriptive alternatives in order to investigate what they offer as far as
readings of Rom 3:1–9 and its place in the letter are concerned.

209  Rodríguez, If You Call, 68.

 chapter 9

Rescriptive Readings of the Dialogue in

Romans 3:1–9 and Its Role in the Letter

Non-Diatribal Critique of Traditional Readings

A brief interlude between traditional and rescriptive readers takes this

conversation back in time over three decades with David R. Hall’s article,
“Romans 3.1–8 Reconsidered.” Hall represents something of an unresolved
thorn in the side of traditional readings, as Hall recapitulates many Godet-
esque critiques. Discussing Hall at this point resets the stage for the consider-
ation of rescriptive readings of Rom 3:1–9, as they begin to appear soon after
the publication of Hall’s article.1
Hall critiques readings of Rom 3:1–8 that follow “the diatribe hypothesis.”
The readings Hall specifically has in mind are those that understand the ques-
tions in 3:1–8 as objections spoken by an imaginary interlocutor, i.e., tradition-
al readings of the script. Hall first argues against traditional readings on the
ground that it is best not to read Rom 3:1–8 as diatribe in the first place. Then
Hall raises a number of exegetical observations that problematize the tradi-
tional script of the passage. Hall consequently addresses both the forest and
certain trees in the traditional reading’s standard position.
Hall pushes against reading the passage as diatribe in two ways. Hall notes
that in traditional readings of Rom 3:1–8, “the objections are stated in detail,
and Paul’s replies are brief and inadequate.” This arrangement, Hall thinks,
falls clearly outside of the norm for diatribal dialogue, which he thinks would
balance the discourse in precisely the opposite direction.2 Hall also notes that
when Paul introduces interlocutors elsewhere, he uses common introduc-
tory formulas that indicate a new speaker. Hall understands the phrase, τί
ἐροῦμεν (3:5), however, as an indicator for “internal debate rather than external
objection,” so that the “origin” for Paul’s argument “is to be found not only in
his debates with Jewish objectors, but also in the internal debate within his
own conscience.”3 For these reasons, according to Hall, it is best not to read

1  Again, with a few exceptions, I reserve discussion of the various commentaries and addi-
tional secondary literature until Chapters 10 and 11.
2  Hall, “Romans 3.1–9,” 183.
3  Ibid., 183–84.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_013

198 chapter 9

Rom 3:1–8 as diatribe but as an internal debate in which Paul explains Ps 51:4,
quoted in Rom 3:4, and defends God’s righteousness.4
For his narrower critiques of the traditional script of the passage, Hall first
draws on the same grammatical rules as Godet concerning the use of μή and
οὐ in interrogatives. Concerning Rom 3:5, Hall writes, “There is, however, a
major difficulty in the hypothesis that in this verse Paul is quoting an imagi-
nary objector. The question is introduced by the word μή, and is a ‘rhetorical
question anticipating a negative answer.’”5 Hall continues, “This is not the way
in which objectors speak,” and he quotes Godet to the end that Paul must be
the speaker of this verse.6
Then, concerning 3:7–8a, Hall argues that the position of the “diatribe
hypothesis,” which also places these verses in the mouth of an interlocutor,
“presents two minor difficulties and one major difficulty.”7 The first minor dif-
ficulty is that, if these verses are spoken in the mouth of an interlocutor, one
should favor textually the reading εἰ δέ rather than the also well attested εἰ γάρ.
The former is supported by ‫ א‬and A, but the latter is supported by B D G K L P
Ψ, amongst others on both accounts.8 The second minor difficulty involves the
parenthesis, καθὼς βλασφημούμεθα … Hall argues that the first person plural
“‘we’ refers to ‘Paul and his colleagues,’” so that the parenthesis “comes very
awkwardly in the middle of a sentence attributed to the objector.”9 The major
difficulty for the traditional reading is, again, that Paul dismisses the lengthy
objection of 3:7–8a with a terse, five-word rejection, which “conflicts with the
normal diatribe style.”10 Hall actually raises a fourth problem for the script of
3:7, arguing that the first-person phrase, κἀγὼ ὡς ἁμαρτωλὸς κρίνομαι, has as its
subject “‘I, Paul,’ as it always does in his letters.”11
Hall’s overall argument is not without fault, however, since he interacts with
an overly narrow understanding of diatribe and diatribal readings of Rom 3:1–
8. This is largely due to the timing of his publication. At the time of Hall’s pub-
lication, no rescriptive readings of 3:1–8 were in print. Stowers’s dissertation

4  Ibid., 195. For a recent publication following Hall (and Moo) and dismissing dialogue from
3:1–8, see Richard N. Longenecker, The Epistle to the Romans: A Commentary on the Greek Text
(NIGTC; Grand Rapids: Eerdmans, 2016), 329–33.
5  Hall, “Romans 3.1–8,” 190. Hall quotes Cranfield, Romans, 184.
6  Hall, “Romans 3.1–8,” 190.
7  Ibid., 192.
8  Hall nevertheless admits that both readings have sufficient textual evidence to justify the
adoption of either one. Ibid.
9  Ibid.
10  Ibid.
11  Ibid., 194. Hall is referring specifically to the phrase, κἀγὼ.
Rescriptive Readings of the Dialogue in Romans 3:1–9 199

had been published—with which Hall does not interact—but Stowers’s initial
discussion of 3:1–9 is far from an exhaustive or even clear demonstration of his
thoughts about the shape of its dialogue. As a result, Hall fails to comprehend
the diversity and variability represented in diatribe, supposing it instead to be
one monolithic practice that produces a similarly fixed interpretation of 3:1–8.
Hall does nevertheless (re-)raise multiple issues against the traditional script
of the dialogue in 3:1–8, several of which go without any persuasive rejoinder
from traditional readers who chronologically follow him (see above). How can
one make sense of the interrogatives, traditionally understood as objections
posed to Paul, which contain μή and expect to be rejected? Furthermore, if 3:8
contains a parenthesis, how would the audience know how to separate the
first-person plural references to Paul and his colleagues from those spoken
in the voice of his imaginary interlocutor(s)? Would they know how to identify
the first-person singular reference in 3:7 in relation to the various first-person
plural references elsewhere in the passage? Also, how are the brief rejections
to be explained in light of the more lengthy and detailed objections, which
Hall suggests fly in the face of diatribal norms? On this latter point, traditional
readers would rightly respond that Paul addresses these issues in more detail
later in Romans (e.g., Rom 6, 9–11),12 but this begs Hall’s very question of just
how far Romans can deviate from diatribal style and still be readable as dia-
tribe. So, can a traditional reading answer these questions? Or, can a rescriptive
reading solve some of these reoccurring concerns?

Rescriptive Readings of the Dialogue of Romans 3:1–9

As is clear by now, “rescriptive readings” refer to those engagements with

Rom 3:1–9 that revise the script of the dialogue by reassessing which lines are
most appropriately spoken in which voice, Paul’s or his interlocutor’s. Such
rescriptions of the dialogue take on two shapes. In the first place is Stowers,
who offers the first significant revision for consideration shortly after the
publication of his dissertation. The second arrangement of the revised script
belongs to Elliott, whom Campbell follows. Though Elliott and Campbell share
the same script of the pericope, this in no way results in identical understand-
ings of the dialogue in the argument of Romans.13

12  Examples are legion. For instance, Wright, “Rom 2:17–3:9,” 5; Moo, Romans, 180.
13  Note that Stowers and Elliott both have multiple relevant publications on Rom 3. I
address all of Stowers’s work together, and similarly with Elliott’s, rather than tracking the
conversation chronologically.
200 chapter 9

Stowers has four publications that are particularly relevant for the pur-
poses of this project, taken here in chronological order. In his dissertation,
The Diatribe and Paul’s Letter to the Romans, after presenting his reassessment
of the diatribal primary sources, Stowers narrows his engagement to illus-
trate three diatribal phenomena as represented in diatribal texts and Romans.
Stowers addresses Rom 3:1–9 under “Objections and False Conclusions.”
Stowers opens his discussion by defining objections and false conclusions,
which is helpful to quote in full:

An objection raises a problem, contradicts or takes exception to some-

thing in the author’s line of argument. A false conclusion is indicated
when the author himself or an interlocutor states a false inference de-
duced from the author’s position. False conclusions are usually stated
rhetorically and usually imply an objection. Objections and false conclu-
sions are often the same or very similar in form. These basic observations
also apply to Paul’s use of these devices in Romans.14

Stowers then outlines the instances in Romans where he finds objections and
false conclusions.15 Stowers suggests that two interrogative clauses in Rom 3:1
are objections that “[imply] a false conclusion from what precedes,” and 3:2
answers the questions with explanation. Rom 3:3 introduces another objec-
tion, which is a false conclusion or objection to Paul’s argument in 2:17–29,
and which is rejected with reasoning in 3:4. Rom 3:5 presents yet another pair
of objections, the latter of which is a false conclusion that expects a negative
answer; 3:6 rejects these objections and explains why. Rom 3:7–8c poses two
more objections and contains a parenthetic statement. Rom 3:8d responds to
the accusers introduced in the parenthetic statement. Finally, 3:9a presents
a false conclusion repeated from 3:1, which 3:9b rejects and explains. So, in
Stowers’s analysis, 3:1, 3, 5, 7–8c and 9a are objections or false conclusions, and
3:2, 4, 6, 8d, 9b–20 are responses to those objections / false conclusions.16
Stowers, however, makes almost no comment regarding the voice in which
each line of the discourse is spoken, and Stowers himself argues that false con-
clusions can be spoken by either voice in a dialogue.17 The sole instance where

14  Stowers, The Diatribe, 119.

15  Unless otherwise noted, the following paragraph is based on ibid., 119–20.
16  It is unclear why Stowers identifies certain verses as “objections” but describes them as
false conclusions or false interpretations. Later, Stowers writes that 3:1a contains objec-
tions stated as questions, and 3:3, 5, 7, 8, and 9 are false conclusions put forth as questions.
Ibid., 134.
17  Ibid., 119.
Rescriptive Readings of the Dialogue in Romans 3:1–9 201

Stowers might comment regarding the arrangement of 3:1–9 is with regards

to Rom 3:4. Stowers writes, “Quotations from the scriptures play an important
part in Paul’s reply to objections in 3:4,” whereby Stowers seems to suggest that
Paul is responsible for speaking this verse.18 This is markedly not what Stowers
argues in subsequent publications, however. As such, it is unclear precisely
what Stowers thinks about the script of the dialogue in 3:1–9 based solely on
his dissertation.19 Furthermore, Stowers does not attempt to situate 3:1–9 with-
in the larger argument of Romans; this too must wait for later publications.
Stowers does, however, discuss 3:1–9 in a diatribal, educational context.
Commenting against the frequency with which the passage is understood as
polemic against Jews or Judaizers, Stowers argues,

The objections in 3:1–9 and elsewhere should not be thought of as aimed

at Jews as opponents, but rather as addressed to the Roman church in
the mode of indictment and censure. Their intent is not polemical but
pedagogical.20

Thus, Stowers differentiates between the tone Paul takes with his imaginary
interlocutor and his intended historical audience. Though Stowers clearly
indicates his thoughts about Paul’s tone towards his audience, he does not
comment on Paul’s tone towards his interlocutor. Is Paul’s tone towards his
interlocutor indeed polemical (which Stowers maintains as one possible func-
tion of diatribe, as depicted in Epictetus’s Disc. 2.20 and 23), or is it similarly
collaborative and educational?
In his 1984 article, “Paul’s Dialogue with a Fellow Jew in Romans 3:1–9,”
Stowers engages Rom 3:1–9 and traditional readings more completely and
proposes the first revised script of the dialogue. Stowers begins by noting
numerous difficulties he finds with traditional readings, which he hopes to
alleviate with his revision of the script informed by diatribal practices.21

18  Ibid., 137.

19  The ambiguity in Stowers’s dissertation impacts other scholars as well. For instance,
Song attempts to support (albeit marginally) his traditional reading of the script of 3:1–9
by noting that “Stowers’ original reading of this section was very similar to mine in his
dissertation published in 1981.” Song seems to suppose that “objections” and “false con-
clusions” in Stowers’s analysis must belong to the interlocutor. Song, Reading Romans,
112n.5, 7. Similarly, Elliott supposes Stowers’s comments in his dissertation adhere to the
traditional script. Elliott, Arrogance, 205n.74.
20  Stowers, The Diatribe, 153.
21  The problems include: (1) to the degree traditional readings view the passage as a “digres-
sion” or “Paul getting ahead of himself,” they fail to account adequately for the function
of the passage in the letter as a whole; (2) traditional readings lack unity and coherence;
202 chapter 9

Regarding Rom 3:1–9, Stowers’s criteria for understanding the pericope

depend on “one of the most common methods of characterization,” namely,
“an apostrophe to the imaginary person,”22 the tendency for the diatribal teach-
er to lead the questioning after an initial interruption from an interlocutor,23
and other contextual and/or diatribal clues. Stowers begins with 2:17–29,
which he identifies as “an apostrophe to an imaginary Jewish interlocutor in
the style of the diatribe.”24 Here, “Paul characterizes the interlocutor as a per-
son who proudly claims to be a Jew and a teacher of truth, but whose behavior
is inconsistent with these claims.”25
This interlocutor’s intrusion at 3:1 then leads into a dialogue between the
interlocutor and Paul. Stowers narrates the dialogue according to the following
progression:26

Interlocutor: What then is the advantage of the Jew, or what is the value
of circumcision? (3:1)
Paul’s Response: Much in every way! To begin with, the Jews were
entrusted with the oracles of God. (3:2)

Paul’s Guiding Question: What else! If some were unfaithful, their unfaith-
fulness does not nullify the faithfulness of God, does it? (3:3)
Interlocutor’s Response: By no means! Let God be true, though every
man be false, as it is written, “That thou may be justified in thy
words, and prevail when thou art judged.”27 (3:4)

(3) traditional readings fail to make sense of the dialogical exchanges, specifically con-
cerning the identity of the interlocutor; (4) the traditional view’s accepted reading (i.e.,
the script) of the text is incoherent, especially the bewildering transitions between voices
in first-person speech; and (5, but essentially a reiteration of 1) traditional readings fail
to account “satisfactorily for the function of the passage in the rhetoric of the letter.”
Ibid., 710.
22  Ibid., 713.
23  Ibid., 714.
24  Ibid., 715; idem., The Diatribe, 79–118.
25  Idem., “Paul’s Dialogue,” 715.
26  Ibid. Stowers notes that his text, which I have quoted, follows the RSV as much as possible.
27  So, either Stowers changed his mind about the voice of 3:4 between the time of his disser-
tation and this article, or his dissertation does not clearly communicate his views about
the verse.
Rescriptive Readings of the Dialogue in Romans 3:1–9 203

Paul’s Guiding Question: But if our wickedness serves to show the righ-
teousness of God, what shall we say? That God is unjust to inflict wrath
on us? (I speak in a human way.) (3:5)
Interlocutor’s Response: By no means! For then how could God judge
the world? (3:6)

Paul’s Guiding Question: But if through my falsehood God’s truthfulness

abounds to his glory, why am I still being condemned as a sinner? And
shall we then say (as certain people also slanderously charge us with
saying), “Let us do evil, that good may come?” (3:7–8c)
Paul’s Own Response: Those who so slander us are deservedly
condemned. (3:8d)

Interlocutor’s Question: What then? Are we Jews at a disadvantage? (3:9a)

Paul’s Response: Not at all! For we have already charged that all, both
Jews and Greeks, are under sin … (3:9b)

Thus, Stowers maintains the frame of the traditional reading (3:1–2 and 9a–b)
but inverts each exchange in the middle of the discourse (3:3–8). The inter-
locutor speaks 3:1, 4, 6, 9a, and Paul speaks 3:2, 3, 5, 7–8, 9b.
According to Stowers’s reading, based on Paul’s comments in 2:17–29 about
God’s impartial judgment contingent on behavior, the Jewish interlocutor
wonders whether membership in the Jewish community provides any ben-
efit and asks the logical question about Jewish advantage (3:1). Perhaps un-
expectedly, Paul answers in the affirmative; Jews were entrusted with the
oracles of God (3:2). Rather than reading 3:3 as a return to the interlocutor’s
voice, Stowers argues that it is “more plausible … to read 3:3 as a diatribal
teacher’s leading question,” so that “Paul begins to question him and will lead
the interlocutor not only to answer his own objection, but also to an admis-
sion of the apostle’s basic theological claims.”28 Said otherwise, Paul leads the
interlocutor “to be the witness, to provide the evidence” for Paul’s own argu-
ment.29 The interlocutor, therefore, emphatically rejects Paul’s leading ques-
tions about God’s (lack of) faithfulness (3:3) and (lack of) righteousness (3:5)
with μὴ γένοιτο in 3:4 and 6, respectively.30 Then, due to the interlocutor’s belief
“in a kind of cheap grace for Israel,” Paul offers himself (hypothetically) as a
reductio ad absurdum, clearly demonstrating the absurdity involved with

28  Stowers, “Paul’s Dialogue,” 716.

29  Ibid.
30  Ibid., 716–17.
204 chapter 9

a member of the covenant community complaining that he should not be

punished for his sin since it ultimately magnifies God (3:7–8). As such, Paul
leads his interlocutor “to see that, on the one hand, God has not abandoned
his people … but, on the other hand, that God is at the same time the judge of
the world.”31 Supposing Paul to be eliminating God’s mercy toward Israel, the
interlocutor then asks whether Jews are actually at a disadvantage in relation
to gentiles (3:9a). Paul rejects the interlocutor’s concern and reiterates that all
people are on equal footing under sin (3:9b).32
Stowers suggests that his reading solves many of the problems he finds
with the traditional reading. First, he argues the apostrophe in 2:17–29 and the
dialogue in 3:1–9 form a coherent unit, so there is no break or digression in
the discussion. Second, understanding the first-person plural references as the
dialogical ‘we,’ “meaning Paul and his discussion partner … makes sense con-
sistently for the whole text.”33 Third, whereas the traditional reading is unable
to explain 3:1–9 in its larger epistolary context, Stowers argues that his read-
ing illustrates how the passage is “a continuation of the discussion of God’s
impartiality, only now by means of the dramatic fiction of a dialogue with a
Jewish interlocutor,” both in terms of God’s judgment and his righteousness
(cf. 3:21–26).34 Fourth, it resolves anti-Semitic tendencies, since the dialogue is
not a polemic against Jews or Judaism, but provides a pedagogical and protrep-
tic “model of reasonable discussion with Jews,” which Paul’s Roman audience
is allowed to overhear.35
More shall be said about Stowers below. At this point, a number of strengths
and weaknesses are notable. At minimum, it is commendable that Stowers
approaches the text with open eyes rather than blindly adhering to previous,
traditional, readings. Additionally, Stowers does offer a reading that resolves
one of the problems that perennially plague the traditional reading. Though
Stowers could make the point more clearly, his rescription of 3:3 and 5 into
Paul’s voice entirely eliminates the concern of scholars like Godet and Hall
who find the use of μή in the interrogatives to be a massive stumbling block
for the traditional script. In Stowers’s reading, the μή works wonderfully in
the interrogatives, as Paul intentionally guides his interlocutor to reject (as
μή expects) the false conclusions he raises. Furthermore, Stowers offers a

31  Ibid., 717–18.

32  Ibid., 720. Stowers takes the difficult verb προεχόμεθα as a passive, “are we Jews at a disad-
vantage.” See my discussion of this below and in Chapter 11.
33  Ibid.
34  Ibid., 721.
35  Ibid., 722.
Rescriptive Readings of the Dialogue in Romans 3:1–9 205

coherent reading of 2:17–3:9 (and following) that does not resort to labeling
3:1–9 as a digression or deserving of omission, and he correctly recognizes that
the dialogue with the interlocutor functions not as an end in its own right but
as a model for Paul’s actual Roman audience.36
Stowers does not solve all of the problems, however. In fact, Stowers creates
additional problems. Stowers believes he has resolved the tension between
Paul’s various comments about Jewish privilege in Rom 2–3 (cf. 9–11). Stowers
argues that Paul affirms Jewish advantage in 3:1–2 but rejects Jewish disadvan-
tage in 3:9. This, however, only results in 3:1–9 not holding seemingly opposing
views in tension, at least until one digs a little deeper. The tension still exists
between Paul’s comments about divine impartiality and the equality of Jew
and gentile in Rom 2, on the one hand, and those supposedly about Jewish
advantage in Rom 3, on the other hand. The tension remains as real as ever.
What is more, Stowers’s analysis of 3:9 does not actually cohere with 3:2. For
Stowers, Rom 3:2 positively affirms Jewish advantage. Rom 3:9, however, nega-
tively rejects Jewish disadvantage, only to remark once again on the equality
of Jew and gentile in the remainder of 3:9–20. Thus, in Stowers’s reading, 3:1–2
offers Jews an advantage that ultimately amounts to nothing as 3:9–20 com-
pletely obliterates it.
Also, though Stowers substantially revises the script of 3:1–9, his argumen­
tation for doing so rests on little solid evidence. For instance, Stowers’s identifi-
cation of 3:3 as a shift to Paul’s leading questions is supported merely by the not
uncommon practice for diatribal teachers to take over the questioning after an
interlocutor’s initial interjection. This suffices as evidence for Stowers, despite
the fact that he is well aware of the vast variety of expression and form among
diatribal texts. The same argument applies to Stowers’s identification of 3:5
and 7–8 as Paul’s voice, and 3:4 and 6 as the interlocutor’s. Moreover, Stowers
offers no argument for why the interlocutor resumes asking questions in 3:9;
he simply assumes it to be the case and narrates how he would understand
such a script. If, however, Paul is leading his interlocutor to be the witness and
make Paul’s case for him as Stowers suggests, would it not make all the more
sense for Paul to go all the way and have the interlocutor answer Paul’s leading
question (3:9a) in 3:9b?
Moving on, in A Rereading of Romans, Stowers’s chapter on Rom 3:1–9 in
Rereading largely repeats “Paul’s Dialogue” with a few revisions.37 Rereading
does, however, afford Stowers the space to explore other features relevant to

36  Note that more recent traditional readings like Tobin’s can accomplish this as well. Tobin,
Paul’s Rhetoric.
37  Stowers, Rereading, 159–75. For the repeating of Stowers’s script of 3:1–9, see pages 165–66.
206 chapter 9

Rom 3:1–9 more broadly, and the feature of particular interest is Stowers’s
discussion of speech-in-character as a tool for understanding diatribe.
Though Stowers emphasizes diatribe in The Diatribe and “Paul’s Dialogue,”
he does not abandon it in Rereading, but he does make a concerted effort
to highlight the role of speech-in-character. Stowers identifies speech-in-
character in Rom 2:1–16, 17–29; 3:1–9; 3:27–4:2; and 7:7–8:2, amongst others.38
Stowers’s discussion of the rhetorical exercise provides a brief summary of
its conventions, such as composing speech that appropriately fits the charac-
ter and/or situation of the imaginary speaker. In fact, Stowers concludes that
“the reader and critic determine who is speaking [in a given text] by criteria of
appropriateness,” as I have also argued in Parts 1 and 2.39 What is more, Stowers
also argues that speech-in-character and the criterion of appropriateness
are particularly applicable to diatribe. Stowers comments: “Almost all of the
dialogical techniques characteristic of the so-called diatribe would be types
of speech-in-character,”40 and “one form of speech-in-character consists
of the speaker or writer simulating an imaginary dialogue with a fictitious

38  Ibid., 16–20. I completely agree that portions of 3:1–9; 3:27–4:2; and 7:7–8:2 contain
speech-in-character. Stowers’s identification of 2:1–16 and 17–29 as speech-in-character,
however, is highly problematic. Stowers argues that Rom 2:1–16 and 2:17–29 represent the
use of apostrophe, and that “Paul’s apostrophe so much resembles speech-in-character in
the diatribe that one must begin by assuming that it functions similarly.” Indeed, Stowers
equates apostrophe with a type or subset of προσωποποιία. Stowers, Rereading, 100–2, 144–
49. R. Dean Anderson Jr. has critiqued Stowers for these classifications of Rom 2:1–5/16
and 2:17–29 as προσωποποιία, but he is unable to pinpoint precisely Stowers’s motivation
for labeling these texts as such. In any case, Anderson correctly allows that 2:1–5 and 17–29
represent apostrophe, but he contends that they do not represent προσωποποιία, because
“no speech is put into the mouth of another party.” Anderson, Ancient Rhetorical Theory,
201–3, cf. 203n.34. In response, Stowers returns to his reading of Quintilian, Inst. 9.2.30–
33, which discusses the functions of speech-in-character, one of which is to “introduce
conversations.” For Stowers, “contra Anderson,” this means that “Quintilian tells us that
apostrophe, dialogue, and simulating the words of another person, including imaginary
objections, were seen as related phenomena and could all be included in the category
of προσωποποιία.” “Apostrophe,” 358. As argued in Part 1, Chapter 2, however, Stowers
has misidentified apostrophe as speech-in-character. As such, I side with Anderson that
Rom 2:1–5(/16) and 2:17–29 reperesent apostrophe but not speech-in-character. Naturally,
Rom 2 will be addressed in detail in Chapter 10 in preparation for the examination
of Rom 3:1–9.
39  Ibid., 19, emphasis mine.
40  Ibid., 20. Rather than suggesting “almost all” of the dialogical techniques of diatribe
(presumably including apostrophe), I limit this statement to actual attributions of speech
to another speaker.
Rescriptive Readings of the Dialogue in Romans 3:1–9 207

interlocutor. One finds this technique with great frequency in the diatribal
literature.”41
Without question, an application of speech-in-character to diatribal
dialogue as Stowers intimates is central to this project. Unfortunately, speech-
in-character makes little overt headway into Rereading’s analysis of Rom 3:1–9.
Stowers never demonstrates how the lines he assigns to the interlocutor ap-
propriately fit his understanding of the characterization of the interlocutor in
Rom 2:17–29, nor does he consider whether lines he attributes to Paul might
make better sense voiced by the interlocutor. For one example, if the interlocu-
tor is an opposing, pretentious, Jewish teacher who thinks “he can transform
the gentiles by getting them to do works from the law,”42 would it not be more
appropriate for this interlocutor to proclaim Rom 3:2’s affirmation of Jewish
advantage on account of their possession of the λόγια, rather than for Paul—
who preaches a gospel of impartiality and equality—to do so? As Stowers
describes the interlocutor, this rescription would at least seem to cohere better
with speech-in-character’s convention of appropriateness to characterization.
Stowers takes up the discussion of speech-in-character in more detail in yet
another offering, “Apostrophe, Προσωποποιία and Paul’s Rhetorical Education.”
This fourth and final piece from Stowers is his response to criticism from
R. Dean Anderson Jr. Anderson challenges Stowers’s identification of 3:1–9 as
an example of speech-in-character on the premise that speech-in-character
dialogues “are always of such a nature that the remarks of the two speakers
concerned are immediately identifiable without the aid of separate markers
in the text.”43 Because the supposed dialogical exchanges in Rom 3:1–9 are not
obvious from the text, Anderson argues that it is “much better characterized
as dialogue-like due to its use of αἰτιολογία.” Furthermore, “With respect to
dialogue, it should be noted that unless it is absolutely obvious from the text
which words belong to which person, the ancients indicated this in the text.”44
Anderson concludes, therefore, that 3:1–9 contains no dialogue and should be
understood as spoken by Paul from beginning to end.
But two rebuttals to Anderson’s argument are necessary. First, Anderson
notes Quintilian’s allowance that speech-in-character can occur without any
indication that a new speaker has began to speak (Inst. 9.2.37). Anderson, how-
ever, refuses to allow Quintilian to be applicable merely because “there is little
room for misunderstanding that προσωποποιΐα is being used” in the example

41  Ibid., 162.

42  Ibid., 151.
43  Anderson, Ancient Rhetorical Theory, 217n.59, emphasis mine.
44  Ibid., emphasis original.
208 chapter 9

Quintilian cites from Vergil.45 This refusal constitutes the sole opportunity for
Anderson to reject outright even the possibility of dialogue in Rom 3:1–9. In
fact, Quintilian’s use of Vergil supplies an excellent comparison to Rom 3:1–9;
what is missing in both instances is not an identification of the speaker but any
overt (such as a verb of speech) indication that another character is speaking
at all.46
Second, Anderson’s own premises are at least confusing if not self-
contradictory. Anderson writes, speech-in-character dialogues “are always of
such a nature that the remarks of the two speakers concerned are immedi-
ately identifiable without the aid of separate markers in the text.”47 Anderson
then claims, “With respect to dialogue, it should be noted that unless it is
absolutely obvious from the text which words belong to which person, the
ancients indicated this in the text.”48 Which is it? Are dialogical exchanges
always obvious from the text, or must an author or speaker specifically indicate
these transitions? Anderson is unclear. Consequently, Anderson’s contention
that Stowers’s division of 3:1–9 is not obvious from the text is to no avail as
evidence for dismissing the possibility of dialogue in Rom 3:1–9.
Speaking on his own accord, Stowers demonstrates that Anderson is dealing
with an overly narrow understanding of how a reader or auditor might rec-
ognize speech-in-character. Stowers argues that speeches-in-character can be
introduced in diverse ways. An overt verb of speech may occur but is not a
necessary feature of the exercise. For instance, an apostrophe might serve to
introduce an imaginary character to which speech is then applied.49 Stowers
argues that this is precisely the case in Rom 3:1–9. The passage contains no
verb of speech, but “the apostrophes of 2:1–16 and 17–29 are introductions for
the dialogue that follows and would have made the προσωποποιία clear for the
ancient reader.”50 In this way, Stowers dismantles Anderson’s primary ground
for contention, the notion that no introductory material paves the way for the
supposed dialogue in Rom 3:1–9.

45  Ibid., 203.

46  For my discussion of Quintilian’s use of Vergil, see Chapter 2.
47  Anderson, Ancient Rhetorical Theory, 217n.59, emphasis mine.
48  Ibid., emphasis mine.
49  Stowers correctly writes, speech-in-character “was sometimes identified in a passage by
the form of the characterization and not only with some language that said, in effect,
‘now I am going to shift into προσωποποιία.’” Stowers, “Apostrophe,” 353. For examples, see
Part 1.
50  Ibid., 362, 365.
Rescriptive Readings of the Dialogue in Romans 3:1–9 209

One final note about Stowers’s “Apostrophe” is in order. Richard Hays also
critiques Stowers’s Rereading, suggesting that Stowers’s argument for the dia-
logue of Rom 3:1–9 lacks “methodological controls.”51 Stowers avers,

I find this to be a somewhat silly criticism. The only criterion is sense.

When something that looks exactly like a dialogue follows the apostro-
phe of the Jewish teacher of gentiles in 2:17–29 and certain words make
excellent sense as words of that character, in light of the previous dis-
course, then it probably is a dialogue.52

Generally I agree with Stowers that one can identify 3:1–9 as dialogue based
on its similarity to other diatribal texts. Where Stowers is particularly want-
ing, however, is once again on the issue of neglecting to analyze the attrib-
uted speech in relation to the characterization of the imaginary speaker.
Stowers suggests the “words make excellent sense as words of that character,”
but he never demonstrates how this works textually, and in this sense I echo
Hays regarding a lack of methodological control. Might such examination
effectively compel Stowers to re-rescript the dialogue once more?
In conclusion, in his sequence of publications, Stowers makes genuine prog-
ress in the interpretation of Rom 3:1–9. Stowers demonstrates that 3:1–9 is not a
digression but is integrally tied to the rest of the letter. He reasonably accounts
for the interrogatives that expect a negative response. And he sufficiently
defends Rom 3:1–9 as dialogical and representative of speech-in-character.
Stowers’s shortcomings, however, limit his contributions. The lack of argument
concerning which lines in the discourse belong to which voice vis-à-vis the
convention of appropriateness in speech-in-character is particularly crippling.
As it stands, Stowers has convincingly identified 2:17–29 as apostrophe and
(parts of) 3:1–9 as speech-in-character, but such an observation in and of itself
is equally applicable to traditional and rescriptive readings alike, the only dif-
ference being which lines of the discourse count as speech-in-character. Thus,
Stowers’s overlaying of speech-in-character and diatribe promises much but
fails to produce a persuasive argument for his understanding of the script. The
following chapters aim to fill this argumentative gap in one direction or the
other. Before that, however, examinations of Neil Elliott and Douglas Campbell
are important.

51  Richard Hays, “‘The Gospel is the Power of God for the Gentiles Only’? A Critique of
Stanley Stowers’ A Rereading of Romans,” in CRBR 9 (1996): 27–44.
52  Stowers, “Apostrophe,” 365.
210 chapter 9

In The Rhetoric of Romans, Elliott revises the script of Rom 3:1–9 even more
than Stowers; Stowers inverts the middle verses in the dialogue leaving the
frame of the traditional reading intact, but Elliott goes the whole way and
rescripts the dialogue at every turn. Remarkably, Elliott’s argument for his
deviations from Stowers at 3:1–2 and 3:9 amounts to an exercise in using
Stowers’s own evidence against him. Elliott agrees completely with Stowers’s
script of 3:3–8c, though Elliott identifies 3:8d as the Jewish interlocutor’s
response to Paul’s leading questions in 3:7–8c.53 Concerning 3:1 and 9a, how-
ever, Elliott recognizes weaknesses in Stowers’s maintenance of these lines
for the interlocutor and echoes this project’s concerns, “Why … is Rom 3.1
necessarily an ‘objection’ raised by an interlocutor?”54 Elliott then appeals
to Stowers’s observation that interruptive objections from the interlocutor
are usually clearly marked. From this, Elliott concludes that an “interruptive
objection is to be distinguished, then, from the recapitulative ‘leading ques-
tion’ by which the teacher guides the student to the appropriate conclusion
by the Socratic procedure of question and answer.” For this type of leading
question, the regular marker is τί οὖν or an equivalent.55 Consequently, Elliott
argues that because nothing marks a change in speakers at 3:1, and because
3:1 begins with τί οὖν, “on purely formal grounds” the verse is best understood
as a recapitulative leading question in Paul’s own voice.56 Elliott continues,
“The same is true at 3.9,” which also begins with τί οὖν.57 The changes between
Stowers and Elliott in 3:1–2 and 9 are as follows:

Paul’s Leading Question: What, then, is the advantage of being a Jew?

Or what is the benefit of circumcision? (3:1)
Interlocutor’s Response: Much in every way! First, they were
entrusted with the oracles of God. (3:2)

Paul’s Leading Question: What then? Do we hold up anything as a

defense? (3:9a)
Interlocutor’s Response: Not at all! (3:9b)

53  Elliott, Rhetoric of Romans, 140.

54  Ibid., 138, italics mine.
55  Ibid.
56  Ibid., 139.
57  Ibid.
Rescriptive Readings of the Dialogue in Romans 3:1–9 211

So, in Elliott’s script, Paul raises leading questions in 3:1, 3, 5, 7–8c, and 9a, and
the interlocutor responds in 3:2, 4, 6, 8d, 9b.58
Like Stowers, Elliott argues that Paul leads into the dialogue in 3:1–9 through
the apostrophe in 2:17–29. Here, Paul addresses a Jew who “holds so high a view
of his or her possession of Torah.”59 In the manner of diatribal indictment, Paul
questions the Jew’s consistency, but not in such a way as to accuse. Rather,
Paul asks “penetrating questions, but not presumptive of the answer.”60 Paul’s
apostrophe to this Jew does not polemically set before his Roman audience a
depiction of “the paradigmatic braggart,” but it illustrates and proves for them
Paul’s teaching that no human being can escape accountability to God.61 The
point is, “If anyone enjoyed the privilege of exemption from God’s wrath, sure-
ly it must be the Jew” who has such great privileges, such as possession of the
Torah.62 Precisely because of these privileges, however, Paul argues, “The Jew,
above all others, cannot plead ignorance … is not exempt from God’s judgment
when he or she violates the very Torah that constitutes those privileges.”63
Following the apostrophe, the dialogue in 3:1–9 is Paul’s attempt to clarify
the Jew’s covenantal advantage. Elliott argues that the questions in 3:1–9 do
not stem from Paul’s denial that the “Jew has any real advantage … but because
Paul has shown that the Jew has no advantage that constitutes an exemption
from God’s righteous claim.”64 This observation makes sense of Paul’s com-
ments in 3:2 and 9; 3:2 asserts that Jews have a real advantage, but 3:9, recog-
nizing προεχόμεθα as a genuine middle, declares that these advantages do not
undermine God’s righteousness to judge Jewish disobedience.65 Jews cannot
cling to their advantages as a defense against God’s righteous judgment, and, if
even Jews cannot, neither can gentiles. Paul’s argument begins with the partic-
ular (the Jew) and moves to the universal (gentiles). In order to communicate
this point to his Roman audience, Paul draws upon the Jewish interlocutor in
a collaborative sense, so that the interlocutor works with Paul to make Paul’s
point.66

58  Ibid., 139–41. From Rhetoric, it is actually not entirely clear that 3:9b belongs to the inter-
locutor in Elliott’s assessment, as he describes 3:9b simply as “conclusion.” Elliott does,
however, make this clear in Arrogance, 205n.74.
59  Ibid., 128.
60  Ibid., 131.
61  Ibid., 130–31.
62  Ibid., 131.
63  Ibid.
64  Ibid., 132. See also 198–204.
65  Ibid., 132–33.
66  Ibid., 138–39, 202.
212 chapter 9

Elliott makes a number of excellent points. First, he demonstrates that

Stowers’s own terms can be employed in such a way as to reshape the script
of 3:1–9 even more than Stowers imagines. Obviously, Stowers and Elliott draw
from the same evidence but end with divergent results. As such, Elliott implic-
itly reveals that diatribal conventions alone might not be the strongest or most
consistent pool of evidence for analyzing dialogical exchanges. Second, Elliott
offers a reasonable argument for his rescriptions of 3:1 and 9 based on formal
categories. Elliott does not simply state his view, but he shows how his reading
adheres better to the more common criteria for objections and leading ques-
tions than Stowers’s own assessment does.
Problems, nevertheless, persist for Elliott as well. First, Elliott’s insistence
that Paul can affirm a real and actual advantage for Jews and at the same time
hold that this privilege does not confer any real salvific benefit in God’s divine
activity is representative of one line of thinking encountered in Chapter 8.
One wonders, however, what good is an advantage that ultimately (or escha-
tologically) conveys no benefit to Jews? Is such an advantage truly advanta-
geous, or is it a contradiction of terms? Is there a better way to formulate Paul’s
presentation of the relationship between God’s impartiality and faithfulness
that does not put so much tension on this dichotomy or require a breach be-
tween God and his called people (which Elliott emphatically and correctly
opposes)?67 Suffice it to say that this question will occupy one of the primary
threads in the discussions of Chapters 10 and 11. Second, though familiar with
Stowers’s Rereading, Elliott too seems unaware of the advantage speech-in-
character might play in analyzing a dialogue. As such, the foundation of Elliott’s
script rests on an argument as precarious as Stowers’s.68 Third, Elliott emends
the script of 3:7–8 so that 3:8d consists of the interlocutor’s response, but he
offers no reasoning for doing so. Fourth, Elliott argues that Paul indicts the Jew
in 2:17–29 and leads the interlocutor to make his case in 3:1–9. But, according
to Elliott’s interpretation, there is no development in the Jew’s responses; he
simply confesses what Paul wants him to say in each instance. Supposing that
Paul is guiding his interlocutor in the style of a Socratic questioner as Elliott
maintains,69 and momentarily granting Elliott’s script, might Paul collaborate
with his interlocutor in a different manner? Might Paul not simply use the
interlocutor as a witness for his argument but also simultaneously lead him

67  For instance, see Käsemann, Romans, 78; Rodríguez, If You Call, 68.
68  One’s argument can only be as good as the evidence, after all, and Elliott’s argument
is based on Stowers’s interpretation of a body of evidence that displays significant
variability.
69  Elliott, Rhetoric, 136–39.
Rescriptive Readings of the Dialogue in Romans 3:1–9 213

out of a place of folly and to a correct view in the manner of censure and pro-
treptic, which Stowers shows is so characteristic of diatribe?70
The last rescriptive reading under discussion is Campbell’s Deliverance of
God, which, for all of its novelty otherwise, is in one sense the least innovative
in terms of its view of the script of 3:1–9. This is because Campbell altogether
adopts Elliott’s script with no additional argumentation.71 Paul submits lead-
ing questions in 3:1, 3, 5, 7–8c, and 9a, to which the interlocutor responds in
3:2, 4, 6, 8d, and 9b. As a result, on the question of ‘who’s speaking when?,’
Campbell’s project is just as open to critique as Stowers’s and Elliott’s; none of
them (perhaps save Elliott on 3:1–2, 9) persuasively argue why various verses
are better heard in Paul’s or the interlocutor’s voice, respectively. This needs no
further discussion here, but it surely merits a remedy.
Campbell, however, raises questions of a different sort concerning Rom
1–3 and 3:1–9 in particular that do require consideration. First, Campbell does
not simply rescript 3:1–9; he also places 1:18–32 (and others) into the mouth of
the interlocutor, an opposing Teacher.72 This stems from Campbell’s view that
Rom 1–4 and 5–8 are fundamentally incompatible; Rom 5–8 represents Paul’s
apocalyptic, participatory, unconditional, and retrospective theology, but 1–4
proclaims a “Justification Theory” (i.e., justification by faith) that is antitheti-
cally opposed to Paul’s gospel.73 Campbell’s solution is to appeal to diatribe
and speech-in-character, to attribute all of the justification-by-faith language
to Paul’s interlocutor, and to argue that Paul engages in conversation with
the Teacher to prove to the Roman audience the weakness of the Teacher’s
justification-by-faith gospel.74

70  Stowers, Diatribe, 56–58. Elliott also addresses Rom 3:1–9 in The Arrogance of Nations.
This treatment, however, summarizes Elliott’s material in Rhetoric; he offers precisely the
same arguments about the apostrophes in Rom 2 and the dialogue in 3:1–9 as in Rhetoric.
One element deserves mention. In Arrogance Elliott highlights the fact that the dialogue
in Rom 3:1–9 does not target Paul’s Judean contemporaries but his largely non-Judean
audience in Rome. Elliott argues, “It is they who need to hear that Judeans do not, in fact,
presume on God’s grace and mercy to indulge their sins.” Consequently, Paul does not
indict but enlists his Judean colleague in order to make a point to the actual audience of
the letter, not to the interlocutor, and the actual audience is to interpret Paul’s point over
against Rome’s imperial claims. The interlocutor, therefore, agrees with Paul through and
through; “It would be hard to imagine a self-respecting Judean arguing against Paul” on
any of these points about God’s righteousness or faithfulness. Arrogance, 104–6.
71  Campbell, Deliverance, 572–74; 1088n.117.
72  Ibid., 542–44.
73  Ibid., 11–35, 65–66.
74  Ibid., 530–41.
214 chapter 9

Campbell’s problem with reading 1:18–32 in Paul’s voice is the passage’s

heavy language of retributive justice and judgment according to desert, which
Campbell sees as un-Pauline. Instead, the Teacher, who endorses Jewish priv-
ilege over gentiles because of the Jews’ possession of law and circumcision,
speaks these lines as condemnation of gentiles who abandon and are conse-
quently abandoned by God.75 Paul recognizes and responds to the Teacher’s
endorsement of God’s retributive justice in Rom 2 through a “universalization,”
which exploits the Teacher’s theology of Jewish privilege and soteriological
desert in order to use it against the Teacher himself.76 To this end, Paul offers
as examples bad Jews and good pagans; in the Teacher’s retributive theology,
bad Jews should be judged despite their privilege, and good pagans delivered
despite the fact that they do not possess the advantageous law. Consequently,
Paul argues that the Teacher’s whole system is not only unfair to gentiles who
do not possess the law, but it amounts to no solution at all since it has nothing
to offer them.77 Campbell argues, therefore, that by the end of Rom 2, Paul has
shown the Teacher’s mission to the gentiles to be completely bankrupt.78
In 3:1–9, Paul moves to show that the Teacher’s gospel is equally damning
for the Teacher and his followers, ultimately saving no one.79 Paul’s aim is to
enlist the Teacher to affirm in advance and “rather stupidly” the basic claims
that Paul makes in 3:9b–20, that “God will judge all for their sinfulness on the
basis of desert, and irrespective of any special privileges or pleading.”80 Thus,
Paul wants to employ the Teacher as chief witness in his own argument. So,
clinging to Jewish privilege, the Teacher answers Paul’s question about Jewish
advantage (3:1) in the affirmative on the basis of being entrusted with the law
(3:2), despite Paul’s earlier demonstration that the possession of the law and
circumcision “do not confer any decisive ethical” or “eschatological advantage”
(2:13, 25–29).81 Given the Teacher’s persistence, Paul makes it his task to elicit
from the Teacher that circumcision and law confer no advantage, so that “if a
Jew sins then, according to the Teacher’s gospel, he or she will be crushed by
a just God at the final assize.”82 Paul is successful; in the subsequent question-
ing, the Teacher clings to his central belief in judgment according to desert

75  Ibid., 543–44.

76  Ibid., 547–49.
77  Ibid., 552–57, 564, 568–69.
78  Ibid., 568–69, 571, 575.
79  Ibid., 572.
80  Ibid., emphasis original.
81  Ibid., 575.
82  Ibid.
Rescriptive Readings of the Dialogue in Romans 3:1–9 215

and, “perhaps a little shamefacedly,” relinquishes his hold on Jewish advantage

over the gentiles (3:4, 6, especially 9).83
The very next word in the letter, however, throws a significant obstacle
Campbell’s way, προῃτιασάμεθα (3:9). Other readings understand the προ-
prefix as referring to the material in 1:18–3:8, so that the preceding material is
summarized in 3:9 by Paul’s assertion that both Jews and Greeks are all under
sin. Such an interpretation would condemn Campbell’s reading in one fell
swoop, since in his view it is not Paul but the Teacher who condemns gentiles
in 1:18–32, and since Paul only charges “some” but not “all” Jews with being
under Sin in chapter 2. Campbell recognizes the problem fully and confesses
that if he is unable to “provide a plausible alternative construal of this verb,
my broader interpretive suggestion will founder significantly on this lexical
shoal.”84 Campbell’s proposed solution is to read the prefix not as temporal but
as spatial, to understand the aorist as a statement of verbal aspect rather than
temporality, and to explain the verb by what follows (3:10–18) rather than what
precedes (1:18–3:8).85 Therefore, προῃτιασάμεθα would not have the meaning of
‘charge or proclaim previously’ but ‘make a public accusation,’ which Campbell
argues occurs in 3:10–18.
Campbell has made significant contributions here. To begin, Campbell is
the first to suggest that Paul’s interaction with his interlocutor in 3:1–9 actually
follows the common diatribal pattern of censure and protreptic, whereby
the teacher (Paul) exposes inconsistencies in the student or interlocutor (the
Teacher) and seeks to cure them. For instance, Stowers’s interlocutor’s final
remark (in 3:1–9) is an additional objection (3:9a),86 and Elliott’s interlocutor
never disagrees with Paul in the first place.87 For Campbell, however, by the
end of the dialogue the interlocutor abandons her or his initial views (i.e.,
3:2) and adopts the supposedly correct view of the primary speaker (i.e., 3:9).
Second, Campbell’s emphasis on a liberative and loving God’s apocalyptic
initiation, enablement, and maintenance of the redemptive / salvific process
makes the most sense of Paul’s letters (especially Paul’s autobiographical

83  Ibid., 575–76.

84  Ibid., 579.
85  Ibid., 580.
86  Stowers, Rereading, 165–66. Jewett also suggests that the interlocutor’s viewpoint shifts
in the dialogue by taking up Paul’s argument about divine impartiality. Jewett’s interloc-
utor, however, only does so to “expose its inner contradictions” and therefore does not
represent a censure-protreptic model. Moreover, Jewett’s interlocutor “shifts” before the
dialogue actually begins, not as a result of the dialogue. Jewett, Romans, 240–41.
87  Elliott, Rhetoric, 136–41.
216 chapter 9

narratives, his comments about divine faithfulness, and his discussions of the
Spirit). So, with Campbell, it is appropriately placed in close proximity to (or
simply at) the center of Paul’s theology.
Concerns, however, persist at multiple points. The first problem is with
Campbell’s underlying methodological and motivational foundation. The root
cause of Campbell’s dissatisfaction with the juxtaposition of Rom 1–4 with
Rom 5–8 seems to be that he cannot accept a Paul whose theologizing displays
tension between certain points, much less a contradictory or incompatible
Paul and the ramifications that come with it. For Campbell, Paul must be
consistent. Essentially, Paul’s thought and his presentation of it, even in his
contextually contingent letters, must be without logical flaw. From a strictly
historical perspective, however, Campbell has no grounds on which to
prioritize or favor Paul’s abilities over those of the Teacher, or to expect Paul
to be completely free from inconsistency. Yet, Campbell is willing to accuse
the Teacher and his gospel with the crippling charge of inconsistency. At the
historical level, both Paul and the Teacher are equally capable of mistakes or
misrepresentations.88 As Campbell’s obligatory yet assumed premise stands,
however, Paul is on the verge of omniscience, and Campbell’s expectation of
Pauline perfection is historically indefensible. This is true whether one sees
Rom 1–4 as incompatible with 5–8 or not.
The second problem is Campbell’s introduction of the Teacher at 1:18–32.
Of course, Quintilian allows that speech-in-character can occur without
any indication of the speaker (Inst. 9.2.37). Recall, however, that Quintilian’s
example from Vergil does not actually demonstrate the rule. The identity of
the speaker is clear (“we Trojans”); what is missing is the indication that any-
one else begins to speak (see Chapter 2). Both features, however, are missing
from Rom 1:18–32. There is no identification or characterization of a new per-
sona on the scene. There is no verb of speech suggesting that Paul is no lon-
ger speaking in his own voice. So, though speech-in-character out of the blue
may be possible, it is far from probable in 1:18–32. Furthermore, the fact that
speeches-in-character almost always begin with characterizing and identifying
transitions further problematizes the lack of such elements in 1:18–32, espe-
cially when Rom 2 provides just such a passage in preparation for the dialogue
in 3:1–9. If Paul intends 1:18–32 as speech-in-character and knows he is going to
address the speaker in ch. 2, why not clarify the argument by moving the iden-
tifying and characterizing materials in Rom 2 to precede the imaginary speech
of Rom 1? Yet there is more. In light of the problems already on the table, it is
perhaps peculiar that the weight of Campbell’s argument teeters entirely on

88  For an example of a well-known innocent mistake in Paul, see 1 Cor 1:14–16.
Rescriptive Readings of the Dialogue in Romans 3:1–9 217

one of Quintilian’s uniquely attested elements, especially when the material

Campbell places in the mouth of the interlocutor is precisely the material he
finds so theologically objectionable if it belongs to Paul. Indeed, it is not sur-
prising that scholars like Moo suggest that Campbell’s argument “smacks of
special pleading.”89
Third, Campbell’s interpretation of προῃτιασάμεθα fails to convince; rather
than dismantling “Justification Theory,” Campbell in fact runs ashore over
the “lexical shoal” he saw looming before him. Campbell nowhere explains
why the script returns to Paul’s voice at the verb προῃτιασάμεθα. Does Paul
speak this line, or does the interlocutor continue to make Paul’s case? More
significantly, Campbell’s explanation that the first-person plural form is
“most likely an ‘apostolic’ we, as it is in 1:5,” fails to account for the use
of the first-person throughout the dialogue of 3:1–9.90 Why would the “we” of
προῃτιασάμεθα be differentiated from the others, especially the other instance
in 3:9a, προεχόμεθα? Whether προῃτιασάμεθα belongs on the interlocutor’s
lips or Paul’s, Stowers’s explanation of the “we” seems quite appropriate and
preferable; in its diatribal and dialogical context, the first-person plural most
likely refers in the first place to Paul and his dialogue partner.91 That is, Paul
and his interlocutor agree on this point, and Campbell argues precisely this
by illustrating how Paul moves the interlocutor to share his point of view
during the course of the dialogue. The collaboration between the interlocutor
and Paul in the first-person plural of προῃτιασάμεθα, therefore, throws
the door half-way open for a temporal reading of the verb that Campbell
attempts to slam shut. Merely granting for the sake of argument that the
interlocutor speaks the condemnation in 1:18–32, Paul’s collaboration with
the interlocutor in 3:9’s προῃτιασάμεθα positively aligns Paul with the
interlocutor’s previous accusation of idolatry and disobedience, so that
Paul indirectly participates in 1:18–32’s condemnation.
To clinch a temporal reading of προῃτιασάμεθα from Campbell’s grasp, how-
ever, one must also explain how Paul condemns all, rather than only some,
non-Jesus believing Jews. The key is noticing the criterion by which Paul
argues God’s judgment operates against humanity. God’s judgment is against
all human impiety κατὰ τὸ εὐαγγέλιόν μου διὰ Χριστοῦ Ίησοῦ, literally, “in
accord with my gospel through Christ Jesus” (2:16). It does not matter whether

89  Douglas J. Moo, “Review Article: The Deliverance of God: An Apocalyptic Rereading of
Justification in Paul by Douglas A. Campbell,” Journal of the Evangelical Theological Society
53.1 (2010): 148.
90  Campbell, 580.
91  Stowers, “Paul’s Dialogue,” 720. So also Fitzmyer, Romans, 325.
218 chapter 9

διὰ Χριστοῦ Ἰησοῦ attributively modifies Paul’s εὐαγγέλιον or adverbially

modifies the verb κρίνει; in either case, Christ is the contingent measure that
determines the direction of God’s judgment. By definition, any non-Jesus fol-
lower falls under God’s judgment as far as Paul is concerned, and, like Paul’s
gospel, this is inclusive of Jew and gentile. Regardless of 1:18–32, therefore, in
2:16 alone Paul has in fact “previously charged (προῃτιασάμεθα) both Jews and
Greeks all (πάντας) to be under Sin” and God’s judgment (3:9). Campbell’s quest
for an opposing interpretation of προῃτιασάμεθα, therefore, turns out to be all
for naught.

Preliminary Conclusions

Despite so many scholars’ efforts, honest engagements, and positive develop-

ments with the dialogue in Rom 3:1–9, the passage remains something of an
enigma. If the current analysis of the passage is to resolve any of these con-
cerns, it must be able to explain: (1) the use of μή in the interrogatives, (2) the
remarks about Jewish advantage, (3) the identity of the interlocutor, (4) how
the audience would understand the character of that interlocutor, and (5) a
number of exegetical concerns, such as the sense of προεχόμεθα in 3:9.
There is also the larger question of methodology, evidence, and argument.
It is true that appeals to diatribal tendencies rest on less-than-solid, though
not necessarily invalid, evidence. By nature, diatribe is a diverse and variable
phenomenon (see Part 2). It is hardly surprising to find that such diversity
and variability in diatribe lead to equal diversity and variability among
readings of Rom 3:1–9 that rely on such diatribal features alone. Traditional
and rescriptive scholars alike regularly appeal to the same pool of diatribal
evidence but consistently draw opposing conclusions. The discussion begs
for further methodological constraint, which is to be found in the rhetorical
conventions outlined for speech-in-character. Stowers recognizes this potential
methodological boon, but his analysis cuts short the advantages speech-in-
character might offer since he never actually shows how they apply to the script
of 3:1–9. This gives rise to yet another element a successful account of 3:1–9
must address—how is the script of 3:1–9 best arranged? Who speaks which
lines? More importantly, why does a particular line belong in a particular voice,
and why does it matter for understanding Romans? Answers that amount
to little more than assertions are unacceptable; argumentation is obligatory,
as is appeal to the best and most valid evidence. Perhaps at long last, this
project turns now to its own engagement with Romans and its attempt to meet
these needs.
 chapter 10

Romans 1–2: The Ethnically Inclusive and Impartial

Gospel and the Characterization of the Interlocutor

Romans 1:1–12

Romans 1:1 might seem an odd place to begin, as one might suppose it merely
contains introductory concerns common to epistles, such as identification of
the sender, the addressee, and maybe a stock thanksgiving.1 Such an assump-
tion would sell the introduction to Romans short, as Paul expands these in-
troductory matters in ways that impact readings of the whole letter. These
elaborations include detailed discussion of Paul’s gospel and quite surprising
descriptions of Paul’s audience that begin to prepare Paul’s readers and audi-
tors for the argument(s) to follow.
Paul begins his epistle to the Romans in customary fashion by introducing
himself, Παῦλος (1:1), but he does not stop there. Though it would hardly be
unfruitful to consider the anthropological and theological significance of
Paul identifying as a “slave of Christ Jesus” (δοῦλος Χριστοῦ Ἰησοῦ; 1:1), the
issue deserving immediate attention is rather that to which Paul also gives
immediate attention—the gospel. In fact, it is not an overstatement to suggest
that 1:1–5 is as much or more an introduction of the gospel than of Paul himself,
though the two are of course inseparable. When Paul concludes 1:1 by declaring
that he is set apart for the gospel of God (ἀφωρισμένος εἰς εὐαγγέλλιον θεοῦ), he
ceases to talk about himself and immediately begins defining for his Roman
audience characteristics about the gospel (1:2–5). Paul’s gospel is that which
God himself promised long ago through his prophets in the holy scriptures
(ὁ προεπηγγείλατο διὰ τῶν προφητῶν αὐτοῦ ἐν γραφαῖς ἁγίαις; 1:2; cf. 16:26)2 about
his Son (τοῦ υἱοῦ αὐτοῦ; 1:3, 9), who is physically of Davidic descent (1:3), but
who was declared the Son of God at his resurrection (τοῦ ὁρισθέντος υἱοῦ θεοῦ …
ἐξ ἀναστάσεως), namely, Jesus Christ, our Lord (Ἰησοῦ Χριστοῦ τοῦ κυρίου ἡμῶν;
1:5). Moreover, it is through Jesus that grace (χάριν) is mediated to humanity,
and it is through Jesus that Paul received his very apostleship and gospel to

1  For treatments of ancient letters, see Hans-Josef Klauck, Ancient Letters and the New
Testament: A Guide to Context and Exegesis (Waco: Baylor University Press, 2006); Stowers,
Letter Writing.
2  See also Gal 3:8 and the discussion of it in Chapter 5.

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_014

220 chapter 10

work toward πίστις among all the nations (ἐν πᾶσιν τοῖς ἔθνεσιν; 1:5).3 Paul, his
mission, and his gospel, therefore, are necessarily about and dependent on
God, Jesus, πίστις, and χάρις, and Paul primes his audience to think in this way
beginning with 1:1.
What, however, does Paul have to say about the identity of his audience?
The first thing to notice is that Paul views his epistolary audience as thoroughly
Christian.4 Paul writes to those in Rome who are “the called of Jesus Christ,”
“beloved of God,” “called saints,” and who share in the relationship of God as
Father and Jesus Christ as Lord (1:6–7). The members of Paul’s audience are
those who already have πίστις, who are ἀδελφοί (1:13), and with whom Paul can
be mutually encouraged (1:8, 12), even if it should come to light later that Paul
thinks they need correction on various points.
Second, Paul addresses his Christian audience as ethnically unspecified and
inclusive. Paul’s apostleship is directed toward all the nations (πᾶσιν τοῖς ἔθνεσιν;
1:5), and Romans is addressed to all (πᾶσιν) those beloved of God in Rome (1:7).
Paul does not say, “to all of you non-Jews” but “to all;” there is no reason to un-
derstand the phrase, πᾶσιν τοῖς ἔθνεσιν, in any sort of sense limited only to non-
Israelites.5 This does not suggest that the Roman congregations are comprised
of any particular ratio of Jews to non-Jews, though Stowers, Elliott, Rodríguez,
and others have persuasively argued that they most likely consist of a large
majority of non-Jews.6 The point is, Paul makes no actual distinction in the in-

3  I momentarily withhold discussion of πίστις until 1:16–17, where the πιστ- stem appears
four times in the context of God’s effecting salvation through the gospel. Instead of offer-
ing a specific translation, I often simply refer to πίστις terminology in the original language.
Especially since Hays’s The Faith of Jesus Christ, scholars have problematized and nuanced
the discussion of what Paul means by the πιστ- stem at a number of key points in his letters.
Objectively, does Paul have in mind human “faith” or belief in Jesus or God? Or, subjectively,
does Paul have in mind God’s or Jesus’ faithfulness? Often it is not overtly clear what nuance
Paul has in mind, so evidence and arguments must be based on a collaboration of texts and
one’s understanding of the bigger picture of Pauline theology. Waiting to discuss the term
until 1:16–17 allows for a slightly larger epistolary context to build, but it is also the point at
which one must begin to make decisions, since 1:14/16–17 represents the jumping off point for
the remainder of the letter. This will necessarily require a premature look ahead at the way
Paul uses the terms elsewhere.
4  I understand the anachronism in the use of “Christian,” but the term is nevertheless help-
ful (and less cumbersome) for identifying “Jesus believers,” “followers of Jesus,” and other
sobriquets.
5  Contra A. Andrew Das, Solving the Romans Debate (Minneapolis: Fortress Press, 2007), 54–60.
6  Peter Lampe, “The Roman Christians of Romans 16,” in The Romans Debate: Revised and
Expanded (Karl P. Donfried, ed.; Grand Rapids: Baker Academic, 1991), 224–25; idem., From
Paul to Valentinus: Christians at Rome in the First Two Centuries (Minneapolis: Fortress, 2003),
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 221

troduction between Jew and gentile concerning his gospel’s effectiveness,7 and
recognizing the ambiguity and inclusiveness Paul employs at multiple other
points in Romans will prove quite advantageous in determining more precisely
what Paul is arguing in those pericopae and Romans as a whole.
Third, Paul characterizes his ethnically unspecified Christian audience
in a very surprising fashion. Paul describes his audience as “called” (κλητοί,
κλητοῖς), “beloved” (ἀγαπητοῖς), and “holy” (ἁγίοις, 1:6–7). These are terms the
Septuagint applies to ethnic Israel as God’s chosen people. For instance, Isa
48:12 records, “Hear me, Jacob and Israel, whom I call” (καλῶ). Though LXX
Isa 48:12 records the active voice, the underlying Hebrew is a Pual participle
with first-person singular suffix, “who are called by me” or “whom I called”
(‫)מקראי‬. Similarly, Deut 7:6–8; 14:2; and Isa 41:8 utilize the synonymous terms
προαιρέω (I choose) and/or ἐκλέγομαι (I elect, select) to characterize Israel as
God’s chosen people. Then, concerning Paul’s Roman audience as “beloved,”
Isa 41:8 depicts Israel as the one whom God loved (ἠγάπησα), and Jer 38:3
envisions Israel as the one whom God “loved with an eternal love” (ἀγάπησιν
αἰωνίαν ἠγάπησά σε). Furthermore, Deut 7:6 and 14:2 portray Israel as a “holy”
(ἅγιος) people to the Lord God. Even more tellingly, Paul himself describes
Israel in these terms in Romans. Paul discusses Israel in terms of God’s “call”
in 9:12 (ἐκ τοῦ καλοῦντος) and 11:29 (ἡ κλῆσις τοῦ θεοῦ), in terms of “election” in
9:11 and 11:28 (ἐκλογήν), and as “beloved” by God in 9:13 (Ἰακὼβ ἠγάπησα) and
11:28 (ἀγαπητοὶ διὰ τοῦς πατέρας). Thus, Paul describes his intended audience in
1:6–7 not only as ethnically unspecified and inclusive Christians, but also in the
same terms scripture uses to illustrate God’s relationship with ethnic Israel.8
Paul confirms this characterization in 9:24–25 when he writes that God “called
(ἐκάλεσεν) us (ἡμᾶς) not only from the Jews but also from the nations, as it says

69–79. Stowers, Rereading, 21–33; Elliott, Rhetoric, 56; Das, Solving, 54–70; Thorsteinsson,
Paul’s Interlocutor, 87–122; Rodríguez, If You Call, 7–10; Jewett, Romans, 70–72.
7  Though Watson argues that Jewish Christians are the primary addressee of Romans, he never-
theless affirms that 1:6–7 does not exclude any on the basis of ethnicity. Francis Watson, “The
Two Roman Congregations: Romans 14:1–15:13,” in The Romans Debate (ed. Karl P. Donfried;
rev; Grand Rapids: Baker Academic, 1991), 214. Similarly, Matthew Black, Romans (2nd ed.;
NCBC; Grand Rapids: Eerdmans, 1989), 24, argues that Paul is “deliberately vague.”
8  Cf. Cranfield, Romans, 68–71; Fitzmyer, Romans, 238–39; Talbert, Romans, 34; Dieter Zeller,
Der Brief an die Römer (Regensburg: Friedrick Pustet, 1985), 37; Howard Rhys, The Epistle
to the Romans (New York: Macmillan, 1961), 16; Craig S. Keener, Romans: A New Covenant
Commentary (NCCS; Eugene: Cascade, 2009), 22; Jewett, Romans, 113–14; Dunn, Romans, 19–
20; Moo, Romans, 54–55; Hultgren, Romans, 51–52; Wright, Romans, 420–21; Witherington,
Romans, 36–37; Byrne, Romans, 46; Sarah Whittle, Covenant Renewal and the Consecration of
the Gentiles in Romans (SNTSMS 161; Cambridge: Cambridge University Press, 2015), 1–2.
222 chapter 10

in Hosea, ‘I will call (καλέσω) not my people my people, and not my beloved be-
loved (ἠγαπημένην).’” Though it would be premature to draw conclusions based
on the introduction alone, it will be important to consider how Paul develops
this portrayal as the letter progresses. Does Paul use these terms simply by hap-
penstance? Or, intentionally drawing on these terms’ scriptural and Pauline
application to ethnic Israel, is Paul saying something about all of humanity’s
relationship to God?
In any case, by the end of 1:1–7 Paul’s audience learns a great deal. Paul
introduces his inclusive gospel about Jesus for all nations, and he identifies
his audience in detail; they are Christian, ethnically inclusive, holy, called,
and loved by God. In light of reading Romans (especially 3:1–9) as diatribe,
these introductory points are quite important. Paul, as diatribal teacher,
wastes no time in beginning to set forth the content of his teaching, which he
nuances and from which he draws implications later in the argument. Also, the
dialogue in 3:1–9 is not a self-contained dialogue through which Paul isolates
and speaks only to his interlocutor. The dialogue is set within a much larger
letter composed to a specific audience for historically contingent reasons. The
dialogue is meant to communicate something not merely to the interlocutor
but, more importantly, to the actual audience of the letter. Knowing as much as
possible about the audience and how Paul creatively envisions them can only
help to clarify precisely what Paul is arguing. The discussions below continue
to unpack these points as Paul’s argument develops.

Romans 1:13–15

Scholars routinely offer Rom 1:16–17 as the “theme” of Romans.9 This may be
true, but it may be overstated.10 If 1:16–17 is not “the” theme, it nonetheless
resonates with the overarching goal of Romans. One problem with identifying
a particular verse or verses as the theme of Romans is the troubling decision
with where to identify the division in 1:8–17 between the epistolary proem and
the main body of the letter. For example, Nestle-Aland prints 1:8–15 as a textual
unit, and 1:16–17 as another, which benefits the standard identification of 1:16–
17 as the theme by linking all of 1:8–15 together under the epistolary thanksgiv-
ing. The primary advantage with locating the division between 1:15 and 16 is
that it keeps Paul’s discussion of his botched travel plans intact, where Paul

9  For instance, Stanley E. Porter writes, “The introduction to the body of the letter provides
the theme for the entire letter.” Porter, Romans, 57.
10  I am indebted to conversations with Beverly Roberts Gaventa for this hesitation.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 223

indicates that he planned to come strengthen the Romans, to be encouraged

with them, and to bear fruit among them, but he was hindered (1:8–15). The
primary disadvantage is that such an identification ignores other structural
markers that might suggest otherwise.
Transitions to the main body of a letter often begin with a disclosure for-
mula, “I want you to know,” or “I do not want you to be ignorant.” The tran-
sition in Philippians, for instance, reads, “I want you to know, brothers and
sisters …” (Γινώσκειν δὲ ὑμᾶς βούλομαι, ἀδελφοί; 1:12).11 Just such a formula oc-
curs in Rom 1:13, “I do not want you to be ignorant, brothers and sisters …” (οὐ
θέλω δὲ ὑμᾶς ἀγνοεῖν, ἀδελφοί). If the disclosure formula in 1:13 constitutes the
transition between epistolary sections, perhaps one might rather identify 1:14
as thematic, “To Greeks and barbarians, to the wise and unwise, I am one under
obligation” (Ἕλλησίν τε καὶ βαρβάροις, σοφοῖς τε καὶ ἀνοήτοις ὀφειλέτης εἰμί). This
arrangement favors the disclosure formula as a structural marker but divides
Paul’s discussion of his failed attempts to visit Rome. There are, it seems, six
eggs in one basket and half a dozen in the other, justifying a degree of hesi-
tancy when it comes to identifying with precision “Paul’s thematic statement
in Romans.”12
Again, this is not to say that 1:14 and 1:16–17 are opposed to one another. In
fact, quite the opposite is true; 1:16–17 explains and develops Paul’s preceding
comments in 1:14–15 through the particle γάρ. I withhold discussion of 1:16–17
until the following section. At this time, I only wish to illustrate how 1:13–15
continues Paul’s discussion of the scope of his ministry. As in 1:1–7, here too
Paul’s ministry is ethnically unspecified and altogether inclusive. Paul hoped
to have fruit among the Romans, as he did, no doubt hyperbolically, “among
the rest of the nations” (ἐν τοῖς λοιποῖς ἔθνεσιν; 1:13). Even more to the point is
1:14; Paul considers himself obliged to “Greeks and barbarians, to wise and un-
wise” (Ἕλλησίν τε καὶ βαρβάροις, σοφοῖς τε καὶ ἀνοήτοις). From the Greco-Roman
perspective in and to which Paul writes, “Greeks and barbarians” account for
the entirety of the world’s populations. If one supposes that “Greeks and bar-
barians” somehow exclude Jews from the argument, Paul’s equal obligation to
“wise and unwise” nevertheless accounts exhaustively for all peoples.13 Said

11  For example, Peter T. O’Brien, The Epistle to the Philippians (NIGTC; Grand Rapids:
Eerdmans, 1991), 85; Gordon D. Fee, Paul’s Letter to the Philippians (NICNT; Grand Rapids:
Eerdmans, 1995), 109.
12  On the issue of where the body of Romans begins, see the literature cited in Hultgren,
Romans, 85n.1, who favors 1:18 as the body opening.
13  Jewett, Romans, 130–33; Dunn, Romans, 32–33. Porter, Romans, 55, writes similarly, “It is
more likely that [Paul] uses the two radical oppositions as disjunctions to include the
224 chapter 10

otherwise, Paul considers himself, his apostleship, his gospel, and his God as
ethnically untethered and obligated to all people (1:14).14 How, then, does Paul

extremes and all in the middle.” Contra Wright, Romans, 422–23, and Hultgren, Romans,
66, who limit this to non-Jews. But see Hultgren’s comment that the “phrase typically
signifies all people in Hellenistic speech,” which is precisely my point.
14  Paul’s comments in 1:11–15 and 15:20–21 raise problems. What does Paul hope to accom-
plish with the Romans? In 1:11, Paul indicates that he aims to “impart some spiritual gift” to
strengthen them (ἵνα τι μεταδῶ χάρισμα ὑμῖν πνευματικόν), and 1:12 clarifies that he hopes
to experience mutual exhortation with them (συμπαρακληθῆναι ἐν ὑμῖν). Somewhat differ-
ently, Rom 1:13 seems to suggest that Paul wants to evangelize in Rome, as he writes, “So
that I might have some fruit among you,” and 1:15 speaks of Paul’s eagerness to preach the
gospel in Rome (ὑμῖν τοῖς ἐν Ῥώμῃ εὐαγγελίσασθαι). Thus, one possible reading casts Paul
as redundantly evangelizing Christians. Some manuscripts recognize this difficulty and
attempt to fix it. D* b and vgmss add ἐν before ὑμῖν, and G adds ἐπ’, both in order to suggest
that Paul aims to preach among or alongside (rather than to) the Roman Christians. The
established reading, however, is more than sufficiently supported by the strongest manu-
scripts, and the naked dative can be nuanced in numerous ways.
 A second problem surfaces at the end of the letter, as Paul argues that he “aspires to
preach the gospel where Christ has not been named,” so that he does not interfere with
someone else’s ministry (15:20). Does Paul hope to make Christians (of Christians!) in
Rome where Christ has been named as evidenced by the Roman Christians themselves,
or can the difficulties be explained some other way? The solution that solves both prob-
lems most simply is to allow the main verbs in the disclosure formula of 1:13 also to
govern 1:15. In 1:13, Paul informs the Romans that, in the past, he often planned (aorist,
προεθέμην) to come to them so that he might have some fruit among them. Again in the
past Paul was hindered (aorist passive; ἐκωλύθην) from carrying out his plans. Rom 1:15,
however, has no verb, so one must be supplied. On the surface, the most obvious answer
is to allow εἰμί (present) from 1:14 to govern 1:15, as it is the verb in closest proximity.
The content similarity between 1:13 and 15 concerning Paul’s desire to preach the gospel
and bear fruit in Rome, however, suggests that 1:15 should similarly be understood in
the past. The sense is then, “I, Paul, am obligated to preach the gospel to all people (1:14)
where Christ has not been named (15:20). Therefore (οὔτως), I planned to come to you
to bear fruit (1:13) as it was my eagerness to preach the gospel to you (1:15), but I was
hindered (1:13). Now, however, I hope to strengthen your Christian life and to be encour-
aged together with you (1:11–12).” Contra Günter Klein, “Paul’s Purpose in Writing the
Epistle to the Romans,” in The Romans Debate (ed. Karl P. Donfried; rev; Grand Rapids:
Baker Academic, 1991), 37–39. Peter Stuhlmacher, “The Purpose of Romans,” in The
Romans Debate (ed. Karl P. Donfried; rev; Grand Rapids: Baker Academic, 1991), 241n.25,
opposes Klein on the grounds of Rom 6:17 and 16:17. For solutions like mine, see Byrne,
Romans, 56; Stuhlmacher, “The Purpose of Romans,” 236–37; idem, Paul’s Letter to the
Romans: A Commentary (Scott J. Hafemann, trans.; Louisville: Westminster/John Knox,
1994), 26. Jewett makes the same argument connecting 1:13 and 15, but he applies it
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 225

advance the theme of ethnic exhaustiveness in his discussion of the gospel and
God’s δικαιοσύνη in 1:16–17?

Romans 1:16–17

Whether one ultimately views 1:14, 1:16–17, or some combination of them as

broadly thematic for Romans, the two pericopae overlap considerably, the only
difference being the perspective from which Paul states his case. In 1:14–15,
Paul speaks on his own behalf as an apostle obligated to share the gospel with
the world. In 1:16–17, Paul speaks about the gospel and its scope in terms of
God’s δικαιοσύνη at work in the world.
In Rom 1:16, Paul begins to explain his obligation to the world and his
consequent eagerness to preach in Rome (1:14–15) as it relates to the gospel. In
effect, Paul says that he is under obligation to the world, and thus he aimed to
preach the gospel in Rome, “for” or “because” (γάρ) he is not “ashamed of the
gospel” (ἐπαισχύνομαι τὸ εὐαγγέλιον; 1:16). Despite the gospel’s proclamation of
a crucified messiah and its counter-cultural implications,15 Paul argues that
the gospel actually “is the power of God for salvation to everyone characterized
by πίστις, to the Jew first and to the Greek” (δύναμις γὰρ θεοῦ ἐστιν εἰς σωτηρίαν
παντὶ τῷ πιστεύοντι, Ἰουδαίῳ τε πρῶτον καὶ Ἕλληνι).16 Paul does not yet disclose
exactly from what situation God powerfully effects salvation through the
gospel; he only begins to address that issue in 1:18. As such, Paul identifies
the gospel as the salvific solution to an as of yet unidentified problem, with
the only criterion being πίστις (cf. 1:5). What Paul does (re)indicate, however,
involves the scope of God’s salvific activity in the gospel. The gospel is not
applicable for Jews only or non-Jews only. God effects salvation through the

such that Paul never intended to preach to non-Christians in Rome but within Christian
congregations. Jewett, Romans, 134.
15  For instance, whether or not Paul intended them as such or whether they are implica-
tions generally extrapolated from Paul’s comments, it is easy to imagine a Roman hearing
Paul’s language of slavery to Christ, God’s gospel (1:1), Son of God (1:4), grace, peace, Lord,
Father, and God (1:7) as antithetical to Roman imperial claims. For example, see Elliott,
Arrogance, 59–85.
16  The phrase beginning with δύναμις either lacks a subject that takes δύναμις as a predi-
cate nominative, or it lacks a predicate nominative or adjective modifying δύναμις θεοῦ.
The explanatory function of γάρ suggests that the phrase beginning with δύναμις
is tied to the previous comment about the shameless gospel, and 1:17’s reference to
God’s righteousness being revealed “in it” (ἐν αὐτῷ) must refer back to τὸ εὐαγγέλιον.
Consequently, what is missing is the subject of the clause, “the gospel.”
226 chapter 10

gospel universally “for everyone characterized by πίστις” (παντὶ τῷ πιστεύοντι).

This necessarily includes both Jews and Greeks (Ἰουδαίῳ τε πρῶτον καὶ
Ἕλληνι). The scope of the gospel’s solution is, once again, ethnically inclusive,
just as Paul argues in 1:5–7 and 1:14.
In explaining why he is not ashamed of the gospel, Paul not only discusses
what the gospel is and for whom it is applicable; he also indicates why the
gospel is “the power of God for salvation,” though his explanation is terribly
truncated at this point in the letter. The gospel is God’s power for salvation,
Paul argues, because (γάρ) “in it, God’s δικαιοσύνη is apocalyptically revealed
on the basis of πίστις for the purpose of πίστις, as it is written, ‘The person
who is just on the basis of πίστις will live’” (δικαιοσύνη γὰρ θεοῦ ἐν αὐτῷ
ἀποκαλύπτεται ἐκ πίστεως είς πίστιν, καθὼς γέγραπται, ὁ δὲ δίκαιος ἐκ πίστεως
ζήσεται; 1:17).17 Paul’s explanation, however, raises as many or more questions
than it answers, partly because Paul only develops some of these key concerns
at later points in Romans, and partly because pinpointing the most accurate
connotation for some of these terms is, quite simply, terribly difficult. What
is “God’s δικαιοσύνη?” What does Paul mean by πίστις? What does the cryptic
phrase, ἐκ πίστεως εἰς πίστιν, mean? Whose πίστις is in view in the reference
from Hab 2:4? The main goal in the following (relatively brief) discussions is
not to review the prolific discussions on these questions but to look at the way
Paul develops these concepts, especially as it relates to the present task at hand
concerning Rom 3.
With regards to Paul’s use of πιστ- terminology, the first item to consider is
whether the concept applies to God or Jesus (subjectively, as in God’s or Jesus’
faithfulness) or to humans (objectively, as in humanity’s faith or trust in God
or Jesus).18 In different ways, the answer is both. Paul clearly has human πίστις
in mind when he speaks of the Romans’ “πίστις being proclaimed in all the
world” (1:8, see also 1:12), and, to some degree, when he claims that the gospel
is the powerful means by which God effects salvation “for everyone who is
characterized by πίστις” (παντὶ τῷ πιστεύοντι; 1:16). But Paul also speaks spe-
cifically of divine πίστις; Rom 3:3 asks, if some failed with regards to πίστις,
will that in any way nullify God’s πίστις (πίστιν τοῦ θεοῦ). Regardless of which
voice in the dialogue asks or answers the question, the answer is a resounding
“No!” God will be faithful and true. This is, therefore, a prime example of the
subjective genitive following πίστις, as confirmed by 3:4’s appeal for God always
“to be found true;” this is God’s own faithfulness under question.

17  An alternative translation of the Hab 2:4 citation is, “The just person will live on the basis
of πίστις.” I find the difference between the two translations somewhat marginal for my
purposes. In either instance, πίστις is the contingent factor that results in life. Cf. Rom 4:13.
18  Of course, see Hays, The Faith of Jesus Christ.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 227

Before returning to 1:16–17 and considering the πίστις language in the ab-
breviated phrase, ἐκ πίστεως εἰς πίστιν, two more questions are applicable.
First, to what is God invariably “faithful?” To stick with the discussion of God’s
faithfulness in 3:2–4, the context requires that, though some Jews have been
unfaithful (ἠπίστησαν) to the λόγια τοῦ θεοῦ with which “they were entrusted to
be faithful” (ἐπιστεύθησαν),19 God remains faithful specifically to those λόγια.20
This raises the question, what are the λόγια τοῦ θεοῦ? The λόγια τοῦ θεοῦ are the
Jewish scriptures, “the law and the prophets,” with their promises,21 such as
God’s promise to Abraham that he would be the heir of the world (τὸ κληρονό-
μον αὐτὸν εἶναι κόσμου) and the father of many nations (πατέρα πολλῶν ἐθνῶν;
4:1–17). These λόγια are the same Jewish scriptures in which God pre-promised
the gospel about his Son (1:2–3).22 Paul argues, therefore, that God will be
faithful to his promises to bless Abraham and his seed, to extend Abraham’s
family to the “world” at large (4:13) so as to include “many nations” (4:17), and
to his pre-promised Christological message (1:2–3) of salvation for all people
characterized by πίστις (1:16). That is, Paul’s God is invariably faithful to bring
his ethnically unspecified, nationally inclusive, and salvific promises to fruition
through the Christological gospel, and he will do so despite humanity’s faults
(Rom 3:3; Gal 3:8).
Second, granted that Paul can talk about πίστις as either human or divine,
is humanity the source of its own πίστις, or does human πίστις come from

19  On the sense of ἐπιστεύθησαν, see Paul’s usage in Gal 2:7; 1 Thess 2:4; 1 Cor 9:17, and
Cranfield, Romans, 178–9; Sam K. Williams, “The ‘Righteousness’ of God in Romans,” JBL
99.2 (1980): 267–68; Stowers, Rereading, 166–67; Jewett, Romans, 243; Wright, Romans, 453;
idem, “Romans 2:17–3:9,” 1–3; idem, Paul, 837–38. But see also my explanation of it in the
section on 3:3–4 in Chapter 11.
20  So also Byrne, Romans, 109.
21  On the meaning of λόγια, J. W. Doeve, “Some Notes with Reference to ΤΑ ΛΟΓΙΑ ΤΟΥ
ΘΕΟΥ in Romans III 2,” in Studia Paulina in honorem J. de Zwann (ed. J. N. Sevenster
and W. C. van Unnik; Haarlem: Bohn, 1953), 111–23; Dunn, Romans,130–31; Käsemann,
Romans, 78–79; Fitzmyer, Romans, 326; Black, Romans, 53; C. K. Barrett, A Commentary
on the Epistle to the Romans (BNTC; 2nd ed.; London: Black, 1957), 60; John Murray, The
Epistle to the Romans: the English Text with Introduction and Notes (NICNT; Grand Rapids:
Eerdmans, 1968), 92–93; Rhys, Romans, 35; Frank J. Matera, Romans (PCNT; Grand Rapids:
Baker, 2010), 81; Hultgren, Romans, 135; Longenecker, Eschatology, 195; Tobin, Rhetoric, 119.
22  As Richerd B. Hays writes, “[3:2, 21, 31] share in common a view of the Law as proleptic,
prefiguring the economy of salvation that is revealed in the gospel,” which he then ties
to Rom 1:2. Richard B. Hays, “Three Dramatic Roles: The Law in Romans 3–4,” in Paul and
the Mosaic Law (ed. James D. G. Dunn; WUNT 89; Tübingen: J. C. B. Mohr (Paul Siebeck),
1996), 158, 160, more generally 158–64. See also Stowers, Rereading, 171, 199. Contra
Cranfield (Romans, 55–57), Fitzmyer (Romans, 233), and Hultgren (Romans, 44) who take
1:2 and 3 as parallel but unconnected attributes of εὐαγγέλιον.
228 chapter 10

some other entity? To state the issue cumbersomely but more indicative of the
point—if, on the one hand, πίστις is the primary criterion by which the gospel
is applicable to all (Rom 1:16) and a person who is just on the basis of πίστις
will live (1:17), but, on the other hand, if Paul can ultimately claim that “no one
whatsoever is just” (οὐκ ἔστιν δίκαιος οὐδὲ εἷς; 3:10), the criterion for which is
πίστις according to Paul’s citation of Hab 2:4 (1:17), then how can the gospel
ever be effective? If no humans are δίκαιος ἐκ πίστεως in and of themselves (3:10;
cf. 1:17), then it stands to reason that no human actually has her or his own πίστις
by which the gospel is effective. So, what is the source of human πίστις that
allows Paul’s gospel to function? Paul preempts human πίστις by identifying its
origin as dependent on God’s initiative to act decisively on humanity’s behalf.
In Romans, Paul writes that “πίστις comes from hearing, and hearing comes
through the word of Christ” (ἡ πίστις ἐξ ἀκοῆς, ἡ δὲ ἀκοὴ διὰ ῥήματος Χριστοῦ;
10:17), so that πίστις arises out of an engagement with Christ. More to the point,
in 12:3, Paul plainly states that God “distributes to each a measure of πίστις”
(ἑκάστῳ ὡς ὁ θεὸς ἐμέρισεν μέτρον πίστεως; see also 12:6). Philippians perhaps
contains Paul’s clearest comment to this end, as he argues that πίστις is a gra-
cious gift from God, “It was graced to you … to have πίστις in him” (ὅτι ὑμῖν
ἐχαρίσθη … τὸ εἰς αὐτὸν πιστεύειν; Phil 1:29; cf. Rom 1:5, δι᾽ οὗ ἐλάβομεν χάριν …
εἰς ὑπακοὴν πίστεως). Plus, in Galatians, Paul identifies πίστις as one of the gifts
that accompany participation in the Spirit (5:22; cf. 3:5). For Paul, therefore, the
source of human πίστις is not humanity’s own volition, decision to believe, or
mental consent but God’s own gracious initiative to create and enable πίστις
within humanity.23
Though almost anyone (including myself) would admit that the shorthanded
phrase, ἐκ πίστεως εἰς πίστιν (1:17), is simultaneously highly enigmatic and
remarkably important for understanding Paul’s explanation of God’s δικαιοσύνη,
the two preceding conversations offer the best context from which to view this
cryptic little phrase. The phrase does not mean “with faith upon faith” or “with
exceeding amounts of faith” as some take the lemma as a statement of extent,
and as Paul, in a different grammatical construction, describes Abraham’s
hope in 4:18 (παρ᾽ἐλπίδα ἐπ᾽ἐλπίδι). The phrase explains how and why

23  This causes additional problems for Campbell’s argument. For Campbell, justification
by faith amounts to little more than an impossible legalistic theology. When one con-
siders God’s role in gracing humanity with πίστις, however, πίστις no longer approxi-
mates a human work, much less an impossible one. Quite the contrary, God’s gracing
humanity with πίστις fits snuggly in God’s apocalyptic deliverance of humanity through
Jesus’ faithfulness so characteristic of Rom 5–8. In this light, Rom 1–4 and 5–8 are not so
contradictory as Campbell imagines. Campbell, Deliverance, 55–61.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 229

God’s δικαιοσύνη is being apocalyptically revealed (ἀποκαλύπτεται). The phrase

works surprisingly well when read in view of God’s faithfulness and role as the
agent who effects human πίστις; God’s δικαιοσύνη is apocalyptically revealed
“on the basis of (ἐκ) [his] faithfulness (πίστεως), for the purpose of enabling
(εἰς) [human] πίστις (πίστιν).”24
To begin, understanding God’s δικαιοσύνη as stemming from his faithfulness
and enabling human πίστις functions logically in the soteriological discussion
of Rom 1:16–17. To paraphrase, Paul writes that ‘the gospel is God’s powerful
means of effecting salvation for everyone characterized by πίστις, since God’s
δικαιοσύνη is apocalyptically revealed in the gospel. God’s δικαιοσύνη extends
out of his faithfulness (ἐκ πίστεως) to his pre-promised gospel (1:1–3) and his
promises to Abraham and Abraham’s ethnically inclusive seed (3:2–4; 4:13–17),
which God’s δικαιοσύνη creates and maintains by gifting humanity with πίστις.’
Stated linearly, the gospel is God’s power for salvation to everyone characterized
by πίστις because God’s δικαιοσύνη actualizes human πίστις, which is the
gospel’s sole criterion. Paul’s citation of Hab 2:4 confirms this reading once
more; “the person who is just on the basis of πίστις will live” (1:17)—it is the
person who is just on the basis of πίστις for whom the gospel is God’s powerful
means of salvation.25
Furthermore, such a reading not only suggests a particular nuance for how
to understand God’s δικαιοσύνη in 1:17, but it suggests a nuance that is supported
elsewhere in Paul’s thought as well. Read in the context of 1:16–17, God’s δικαι-
οσύνη is at least as active as it is qualitative of God’s character.26 That is, God’s

24  Hays, Echoes, 39–41; Dunn, Romans, 44–45, 48; Wright, Romans, 425; Rhys, Romans,
21–22; Witherington, Romans, 48, 55–56. Cf. Douglas A. Campbell, “Romans 1:17—
A Crux Interpretum for the ΠΙΣΤΙΣ ΧΗΡΙΣΤΟΥ Debate,” JBL 113.2 (1994): 265–85, who
reads similarly, but with Christ’s faithfulness rather than God’s. For surveys of various ap-
proaches, see Colin G. Kruse, Paul’s Letter to the Romans (PNTC; Grand Rapids: Eerdmans,
2012), 75–78; and Charles L. Quarles, “From Faith to Faith: A Fresh Examination of the
Prepositional Series in Romans 1:17,” NovT 45 (2003): 1–21.
25  This understanding of “from faith for faith” in many ways renders the question of whose
faith is in view in the citation of Hab 2:4 a moot point; a person is just because God acts
faithfully to his promises and because God graciously creates πίστις for the person. This
explains the omission of any personal pronoun in the citation of Hab 2:4. Hays, Echoes,
39–41; Dunn, Romans, 44–46, 48–49.
26  For discussion of the nuances of God’s δικαιοσύνη, including its active sense, see Jewett,
Romans, 141–42; Wright, Romans, 397–406; idem, Paul, 841, 1055n.49; Hultgren, Romans,
75–76; 605–15; Moo, Romans, 70–75; A. Katherine Grieb, “The Righteousness of God in
Romans,” in Reading Paul’s Letter to the Romans (SBLRBS 73; ed. Jerry L. Sumney; Atlanta:
Society of Biblical Literature, 2012), 65–78; Campbell, Deliverance, 683–88; J. Louis
230 chapter 10

δικαιοσύνη does not simply describe God as “just” or “righteous,” though it does;
God’s δικαιοσύνη actively effects human πίστις and thereby enables the gospel.27
Paul does at times speak qualitatively of God’s character as “un/just” or “un/
righteous” using δικ- stem terminology. For example, following its juxtaposition
of God’s δικαιοσύνη with human ἀδικία, Rom 3:5 asks whether “God is not unjust”
(ἄδικος) when he brings wrath. Also, Rom 3:26 considers God’s “demonstration
of his δικαιοσύνη” in the present time as indicative that he is “just” (εἰς τὸ εἶναι
αὐτὸν δίκαιον). But Paul often uses this terminology in an active sense. In Rom
3:26, Paul also speaks about God as “the one who justifies” (εἰς τὸ εἶναι αὐτὸν…
δικαιοῦντα), as he similarly does in 3:30 (δικαιώσει), 4:5 (τὸν δικαιοῦντα τὸν ἀσεβῆ),
8:30 (ἐδικαίωσεν), and 8:33 (ὁ δικαιῶν). Concerning δικαιοσύνη, the form used
in 1:17, Paul employs it actively elsewhere as well. Romans 3:21–26 links God’s
δικαιοσύνη (twice in 3:21–22) to his salvific activity in Christ, so that those who
sin and lack God’s glory (3:23) “are justified freely by [God’s] grace” (δικαιούμενοι
δωρεὰν τῇ αὐτοῦ χάριτι; 3:24). This salvific activity amounts to a demonstration
of God’s δικαιοσύνη (twice in 3:25–26) in the present time, so that God’s δι-
καιοσύνη is largely synonymous (here) with his inclusive salvific initiative.
Additionally, in an equally salvific context, Rom 10:1–3 distinguishes between
God’s δικαιοσύνη and Israel’s endeavor to establish its own, the point being that
Israel did not recognize (ἀγνοοῦντες) God’s Christological plan for salvation.
Finally, numerous passive instances of δικαιόω imply God as an active agent,
such as 2:13; 3:20, 24. Consequently, one can say that God’s δικαιοσύνη is often
his “act of making just,” “act of right-wising,” or “act of rectifying” an otherwise
hopeless humanity (cf. 3:10–18), and 1:16–17 fits this category as well. Therefore,
reading ἐκ πίστεως εἰς πίστιν such that God’s “rectifying activity” (δικαιοσύνη)
is the actualization of his faithfulness to the λόγια because it effects and enables
the very human πίστις by which the gospel functions is in fact quite a strong
reading, despite the phrase’s abridged nature.28

Martyn, “God’s Way of Making Right What Is Wrong,” in Theological Issues in the Letters of
Paul (Nashville: Abingdon, 1997), 141–56; idem, Galatians, 263–75; Ernst Käsemann, “ ‘The
Righteousness of God’ in Paul,” in New Testament Questions of Today (W. J. Montague,
trans.; Philadelphia: Fortress, 1969), 168–82; Richard B. Hays, “Psalm 143 and the Logic of
Romans 3,” JBL 99.1 (1980): 107–15.
27  Thus, God’s faithfulness, righteousness, and truthfulness are not “virtual equivalents.”
Contra Williams, “Righteousness,” 268.
28  Recognizing the universal scope of God’s promises and the active nuance of God’s
δικαιοσύνη that is effective for all, manifested in the present, and Christological unsettles
readings that staunchly prioritize the applicability of God’s δικαιοσύνη to Israel by placing
it in a scriptural context. For instance, Hays insists that Paul’s use of scripture “[spirals] in
around a common focus: the problem of God’s saving righteousness in relation to Israel”
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 231

Before transitioning to Paul’s argument in 1:18 and following, one lingering

question remains. Given that in Paul’s argument God works through the gospel
in an equative and ethnically inclusive manner, what does Paul mean by the
phrase, “to the Jew first and to the Greek” (Ἰουδαίῳ τε πρῶτον καὶ Ἕλληνι)? It is
perhaps easier to say first what the phrase should not mean. If we follow Paul’s
argument linearly, the phrase should not mean that Paul attributes to Jews an
eschatological or salvific advantage over non-Jews, as Paul stresses multiple
times in 1:1–16 (and throughout Romans) that his gospel is equally applicable
to all (πᾶς). For Paul ultimately to allot the Jews salvific advantage in this way
would cause a breach of the anthropological equality he masterfully paints
elsewhere in Romans (and Gal 3:28; 1 Cor 12:13). Though a contradiction or ten-
sion on Paul’s part is possible, textual evidence suggests other interpretive op-
tions. For example, reading πρῶτον as a chronological rather than qualitative
marker not only avoids contradiction with the terms of Paul’s gospel, but Paul’s
earlier claims about the gospel also support such a view. In Rom 1:2, Paul argues
that God “pre-promised the gospel in advance through his prophets in the holy
scriptures” (προεπηγγείλατο διὰ τῶν προφητῶν αὐτοῦ ἐν γραφαῖς ἁγίαις), referring
of course to the Jewish scriptures. Paul suggests, therefore, that the message
about God’s salvific agenda for humanity was present in the Jewish scriptures
all along, though it only became recognizable in the Christ event (cf. Rom 10:1–
16). Thus, in 1:16 Paul simply states explicitly what he only implies in 1:2; Jews
had the first or earliest access to the gospel, but not in such a way as to privilege
them qualitatively on an eschatological or salvific basis.29

and that scripture “presses home a single theme relentlessly: the gospel is the fulfillment,
not the negation, of God’s word to Israel.” Hays, Echoes, 34. As John M. G. Barclay argues,
“Whatever the previous connotations of this motif, and however ‘the law and the prophets’
bear witness to it (3:21), its meaning emerges only in the good news itself.” Furthermore,
“This suggests that the meaning of this phrase for Paul can hardly be determined by its
contextual sense in biblical or Second Temple texts.” Barclay, Gift, 475–76, 476n.65, italics
original. Hays is of course correct that God’s δικαιοσύνη has implciations for Israel, but
Paul indicates that it has equal and synonymous implications for non-Israelites too (Rom
1–16, esp 4, 9–11; see below).
29  So, the burden is on readings that maintain a real salvific advantage for Jews over non-Jews
via the term “first” to show what salvific advantage Jews actually have before a just and im-
partial God. This will certainly involve discussion of Rom 9–11. Witherington, Romans, 51,
for instance, claims that the meaning of “[first] cannot be limited to the notion that Jews
were first offered salvation and then Gentiles were. Paul still sees that God has a plan for
the salvation of non-Christian Jews in the future.” In light of Rom 9–11, especially 11:25–32,
Witherington’s claim is true enough. What Witherington and other similar readings fail to
account for, however, is the degree to which God’s eschatological plan for Israel differs not
232 chapter 10

In fact, to foreshadow further conversation below, a chronological reading

of πρῶτον meshes quite well with Paul’s other comments about Israel, such
as his olive tree analogy (Rom 11). God calls Israel and commissions them to
exercise πίστις towards the λόγια, the Jewish scriptures (3:2), which contain
God’s pre-proclaimed gospel about Jesus (1:1–3). According to the metaphor,
because some of Israel lacks Christ-based πίστις (ἀπιστίᾳ),30 God breaks them
from the tree’s root, his gracious election (11:18; cf. Rom 4).31 In their place, God
grafts in non-Israelites who display πίστις (11:19–20). If the broken off Israelites

at all from his eschatological plan for non-Israel. Even in Rom 9–11, there is no difference;
what God does for Israel, God does for all. Jewish salvific advantage cannot be held in one
hand and divine equality and impartiality in the other. In light of Paul’s argument about
the universal Abrahamic promises, it is more true to say that all are advantaged. See the
supporting discussions of 9–11 in Chapter 11.
30  On the Christological center of Rom 11, see also Jonathan A. Linebaugh, God, Grace, and
Righteousness in Wisdom of Solomon and Paul’s Letter to the Romans (NovTSup 152; Leiden:
Brill, 2013), 207–26.
31  Identifying the root (ῥίζα) in the metaphor is tricky. In Rom 11:1 Paul begins to discuss a
remnant of faithful—Christian—Jews, including himself. This remnant exists because
of God’s gracious election (11:5), but God hardens the rest (11:7–10). Paul still hopes for
the salvation of those hardened, declaring that if the first portion is holy, so is the whole
lump, and if the root is holy, so are the branches (11:12–16). In 11:16, Paul “lumps” the rem-
nant and the hardened together; both constitute the whole lump and all the branches.
God, however, breaks the hardened branches from the tree due to a lack of πίστις (11:17,
20). Though unstated, the inference is that the remaining branches represent the faithful
remnant, and God grafts Christian-gentile branches in with them (11:17). The root, there-
fore, is something the Jewish remnant and Christian gentiles share through πίστις but
from which others are omitted, at least temporarily, due to a lack of πίστις (cf. 11:20, 23).
In Romans, Jews and non-Jews belong by πίστις to the people God graciously calls into
being through his promises to Abraham (Rom 4). Thus, it is God’s gracious call or election
as expressed in the Abrahamic promises that is identified as the ῥίζα that supports both
faithful Jews and non-Jews (cf. 9:24). See Barclay, Gift, 550–51. The implications of this
view are significant, as it suggests that Paul is trying to say less about the way gentiles are
incorporated into Israel’s history and more about the way gentiles were always within
God’s purview, even from the Abrahamic promises that began to call Israel into being.
Thus, both Israel and non-Israel were in view from the very beginning, though this is
only made recognizable in Paul’s post-Christ-apoaclypse present. Contra Wright, Paul,
1212n.589; Byrne, Romans, 341; Moo, Romans, 704; Keck, Romans, 274; Witherington,
Romans, 271, who suggest the “root” might generally refer to the patriarchs; and Hultgren,
Romans, 410, who argues the root relates more generally to “the Israelite/Jewish people’s
spiritual heritage.” Though Hultgren is generally right, more precision is possible. Jewett
is incorrect to identify the root holistically as Israel, since Israelites/Jews cannot be both
root and branches in the analogy. Jewett, Romans, 682–86.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 233

cease in their ἀπιστία and demonstrate πίστις, God will reattach them to the tree
(11:23). In Paul’s analogy, inclusion or exclusion from the tree is based solely on
the criterion of πίστις, just like Paul’s gospel. Israel possessed the earliest access
to God’s people and the pre-proclaimed gospel, but such chronological priority
does not afford them any eschatological or salvific advantage over against non-
Jews, as God willingly breaks πίστις-less Jews from the tree.32 What is more,
God will not graft faithless Israel back on to the tree unless they illustrate
πίστις. The criterion for Israel is no different than for non-Israel; the gospel is
God’s powerful means of salvation to the Jew first, but the stipulations and the
outcome remain the same for all.33 Appropriately, Paul concludes, “God locked
all (τοὺς πάντας) away in disobedience, so that he might show mercy to all (τοῦς
πάντας)” (11:32).34
So, to be sure, God’s call had implications for Israel “first,” but his promises
and call are not ethnically exclusive but inclusive, extending to all the “world”
and “many nations” (4:13, 17), and God deals equally with Israelite and non-
Israelite alike on the basis of πίστις. For these reasons, Paul’s description of his
ethnically unspecified audience in Rome as “called,” “beloved,” and “holy” (1:6–
7) seems more intentional than accidental, especially given Paul’s comments
in 9:24–25. Surely Paul is trying to say something about God, Israel, and non-
Israel, but what is it exactly? Paul continues to set forth his views about the
relationship between Jews and gentiles throughout much of Romans.
Romans 1:18 introduces into the letter a remarkable shift in emphasis, as
Paul turns from the apocalyptic revelation of God’s active δικαιοσύνη to the

32  Israel’s chronological priority also accounts for the distinction between “natural” and “un-
natural” branches. That is, Israelites are only the “natural” branches because they were the
earliest constituents to belong to the tree.
33  This project differentiates between what Paul thinks God is doing eschatalogically and
how God is achieving that goal in the present time. I emphasize the former. Concerning the
broader picture of salvation history (the “how”), or what I later refer to as “chronological
jostling,” see the nuances in Bruce W. Longenecker, “Different Answers to Different Issues:
Israel, the Gentiles and Salvation History in Romans 9–11,” JSNT 36 (1989): 95–123; idem.,
“Sharing in Their Spiritual Blessings? The Story of Israel in Galatians and Romans,” in
Narrative Dynamics in Paul: A Critical Assessment (Bruce W. Longenecker, ed.; Louisville:
Westminster John Knox, 2002), 58–84; idem., “Salvation History in Galatians and the
Making of a Pauline Discourse,” JSPL 2.2 (2012): 65–87. Note, however, how God not only
uses Israel on behalf of non-Israel but also non-Israel on behalf of Israel in Rom 9–11.
34  For yet another chronological progression, see also 11:30–31. “As you (non-Israelites) were
formerly disobedient to God but now have been shown mercy because of [Israel’s] dis-
obedience, so also [Israel] is now disobedient at the time when mercy is being shown to
you, so that they might be shown mercy (as well).”
234 chapter 10

equally apocalyptic revelation of his wrath (ὀργή). If up to this point Paul has
been describing how God’s δικαιοσύνη actualizes human πίστις and thereby
renders the gospel God’s powerful means for salvation, 1:18 begins to define
why and from what humanity needs such divine δικαιοσύνη.

Romans 1:18–32

Scholars struggle to explain the shift from God’s δικαιοσύνη in 1:17 to God’s ὀργή
in 1:18. Recall that Campbell finds such a stark contrast between the two that
he creatively imagines 1:18 transitioning into the voice of an opposing Teacher,
who proclaims his “gospel” of retributive justice and judgment according to
desert in 1:18–32.35 Weaknesses notwithstanding, Campbell’s argument finds a
way to explain the logic between 1:17 and 18, albeit unlikely on rhetorical and
exegetical levels (see Chapter 9).
Such radical solutions are unnecessary in this instance, as the logic
flows from 1:16–18 quite smoothly. The gospel is God’s powerful means of
salvation for everyone characterized by πίστις (1:16). The gospel is effective
for salvation because (γάρ) God’s δικαιοσύνη makes it so by enabling human
πίστις (εἰς πίστιν; 1:17), the gospel’s one criterion. The gospel is also God’s
salvific solution because (γάρ) God’s ὀργή is being apocalyptically revealed
(1:18), but God is invariably faithful to his promises (ἐκ πίστεως; 1:17). The
apocalyptic revelations of God’s δικαιοσύνη and ὀργή, therefore, both explain
Paul’s statement about the gospel as God’s powerful means for salvation. By
enabling human πίστις, God’s δικαιοσύνη (1:17) explains how the gospel is God’s
means for salvation, and God’s ὀργή (1:18) reveals why and from what the gospel
is God’s salvific solution.36 Paul confirms this latter point in 5:8–9; God’s love
is demonstrated in the fact that “while we were still sinners” and under wrath,
Christ made justification possible by dying on humanity’s behalf, and, as a
result, Christians will be saved “from ὀργή.”
More, of course, remains to be said about 1:18–32. First, who is the target
of God’s ὀργή? Scholars often correctly note various elements in 1:18–32 that
parallel accusations of idolatry that Wisdom of Solomon aims at non-Jews. For
example, Wis 13:1 proclaims,

35  Campbell, Deliverance, 542–47.

36  Cf. Jewett, Romans, 151–52; Moo, Romans, 99; Witherington, Romans, 63; Byrne, Romans,
65–66; Keck, Romans, 56–58.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 235

All people (πάντες ἄνθρωποι) in whom ignorance of God (ἀγνωσία θεοῦ)

exists are naturally foolish (μάταιοι), and they were unable to know (οὐκ
ἴσχυσαν εἰδέναι) the one who is (τὸν ὄντα) from the good things that are
visible (ἐκ τῶν ὁρωμένων), nor did they recognize (οὔτε … ἐπέγνωσαν) the
architect (τὸν τεχνίτην) when they paid attention to his works (τοῖς ἔργοις).

Paul writes quite similarly; the target of Rom 1:18–32 is able to know God from
creation but instead chooses to be ignorant and becomes foolish. Paul writes,

The knowledge of God (τὸ γνωστὸν θεοῦ) is evident among them, because
God appeared to them. For, his unseen attributes (τὰ … ἀόρατα αὐτοῦ)
are clearly visible (νοούμενα καθορᾶται) from the creation of the world
(ἀπὸ κτίσεως κόσμου) in the things that are made (τοῖς ποιήμασιν), both
his eternal power and deity, so that such people are without excuse
(ἀναπολογήτους), because, although they knew God (γνόντες τὸν θεόν),
they did not glorify or give thanks to him as God (οὐχ ὡς θεὸν ἐδόξασαν
ἢ ηὐχαρίστησαν), but they became futile in their thoughts (ἐματαιώθησαν
ἐν τοῖς διαλογισμοῖς) and their foolish heart was darkened (ἐσκοτίσθη ἡ
ἀσύνετος αὐτῶν καρδία). (1:19–21)

Additionally, both Wisdom and Paul indicate that the alternative to recognizing
God as Creator and worshipping him correctly is idolatry. Wisdom claims that
such people consider the natural elements to be “gods” (θεοὺς ἐνόμισαν; 13:2)
and that this extends to “likenesses of animals” (ἀπεικάσματα ζῴων; 13:10).
Paul also writes, “They exchanged the glory of the imperishable God for the
likeness of an image (ἐν ὁμοιώματι εἰκόνος) of perishable humanity (φθαρτοῦ
ἀνθρώπου) and birds and four-footed creatures and snakes,” (1:23) such that
“they worshipped and served (ἐσεβάσθησαν καὶ ἐλάτρευσαν) the creation (τῇ
κτίσει) rather than the Creator (τὸν κτίσαντα; 1:25).” Both texts even fire a litany
of accusations against their respective targets in extensive vice lists, includ-
ing charges of evilness, hypocrisy, murder, deception, lying, and the like (Wis
14:22–25; Rom 1:28–32). For many scholars, these parallels strongly indicate
that Paul, like Wisdom, specifically targets non-Jews in 1:18–32.37
It is peculiar, however, that Paul actually makes no ethnic distinctions
in 1:18–32. Paul never mentions gentiles, Jews, or Israel, much less does he

37  For instance, Fitzmyer, Romans, 269–70; Black, Romans, 39; Murray, Romans, 35; Keck,
Romans, 60–62; Byrne, Romans, 64–65; Witherington, Romans, 63–64.
236 chapter 10

distinguish between these groups. Rather, Paul talks about “them”;38 the
knowledge of God was evident “among them” (ἐν αὐτοῖς; 1:19); God appeared
“to them” (αὐτοῖς; 1:19); “they” (αὐτούς; 1:20) are without excuse; “they” became
futile (ἐματαιώθησαν; 1:21) in “their” (αὐτῶν; 1:21) thoughts and “their” (αὐτῶν;
1:21) foolish heart was darkened; God handed “them” (αὐτούς; 1:24) over; “they
who” (οἵτινες; 1:25) worshipped and served creation; and so forth. It is clear that
“they” are non-Christians, as “they” do not glorify or worship God, there is no
discussion of Christ or the Spirit, and God himself hands “them” over to dis-
obedience. But neither is there any talk of ethnicity. Why might Paul employ
conventional Jewish critique of gentiles yet neglect to identify his target as
such?
Though Paul withholds unveiling its full import until later in the letter (see
2:1–3:20), the logic of the argument up to this point prepares the reader for the
answer. Paul’s God, gospel, and apostleship are invested “in all the nations” (ἐν
πᾶσιν τοῖς ἔθνεσιν; 1:1–7) without any ethnic boundaries. As Paul claims, he is
exhaustively “under obligation to Greek and barbarian, to wise and unwise”
(1:14), because in the gospel God’s δικαιοσύνη effects salvation for all peoples
characterized by πίστις (παντὶ τῷ πιστεύοντι; 1:16–17). The gospel is the ethnical-
ly unencumbered salvific solution for all (παντί; 1:16–17), because (γάρ) God’s
ὀργή is also unhindered by ethnic boundaries (1:18). As Paul declares, “God’s
ὀργή is being apocalyptically revealed from heaven against all (πᾶσαν) ungodli-
ness and unrighteousness of humanity (ἀνθρώπων) who suppress the truth in
unrighteousness” (1:18). The solution and the problem are ethnically universal,
and thus 1:18–32’s discussion is ethnically vague and inclusive of all humanity.39
This is perhaps easier to see when one recognizes that 1:18–32 contains not
only conventional Jewish critique of non-Jews but also Jewish critique of Jews.
The discussion of “them” exchanging (ἤλλαξαν, μετήλλαξαν) God’s glory (δόξαν)
for idolatrous (ἐν ὁμοιῶματι εἰκόνος) images of people and animals (1:23) and
God’s truth for a lie (ἐν τῷ ψεύδει), so that they worshipped the creation rather
than the Creator (1:25), significantly overlaps Israel’s idolatry with the golden
calf as depicted in MT Ps 106.40 The text reads, “They exchanged (ἠλλάξαντο)

38  Cf. Rodríguez, If You Call, 28–29, who also notes Paul’s emphasis on “them,” but who ar-
gues that “them” does not apply to Paul’s previous self or his gentile audience.
39  See Kruse, Romans, 82n.1; Longenecker, Eschatology, 172–73; Ulrich Wilckens, Der Brief
an die Römer (EKK; 3 vols.; 3rd ed.; Neukirchen-Vluyn: Neukirchener Verlag, 1997), 121;
Jouette M. Bassler, Divine Impartiality: Paul and a Theological Axiom (SBLDS 59; Atlanta:
Scholars Press, 1982), 122, 135–36; Jonathan A. Linebaugh, “Announcing the Human:
Rethinking the Relationship between Wisdom of Solomon 13–15 and Romans 1.18–2.11,”
NTS 57.2 (2011): 214–37; idem., God, Grace, and Righteousness, 93–121.
40  Longenecker, Eschatology, 173–74; Jewett, Romans, 160–61; Wright, Romans, 433; Bassler,
Divine Impartiality, 122, 135, 195–97; Linebaugh, God, Grace, and Righteousness, 114.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 237

their glory (τὴν δόξαν αὐτῶν) for a likeness (ἐν ὁμοιὠματι) of a calf that eats
grass” (LXX Ps 105:20), which they “worshipped” (προσκύνησαν; LXX Ps 105:19).
It is an overstatement to classify Rom 1:18–32 as Paul’s condemnation specifi-
cally of gentiles; Paul himself says all … of humanity (πᾶσαν … ἀνθρώπων) and
draws on Jewish critique of both non-Jews and Jews to establish his point.41
Accordingly, Paul’s discourse in 1:18–32 is thoroughly ethnically exhaustive,
and he builds on this through 2:1–3:20.
Paul is not the only one to apply God’s ὀργή (and θυμός; 2:8) to Jews and non-
Jews alike, as his view finds ample precedence in Israel’s scriptures. On the one
hand, in the Septuagint, God’s ὀργή is poised against an idolatrous Israel in
texts like Deut 29:19–28; Isa 10:5–6; Jer 7:1–20; and Ezek 7:1–20. Synonymously,
God’s θυμός falls upon Israel for idolatry in Isa 1:19–29; Jer 7:1–20; Ezek 7:5/20;
8:18; and 9:8. On the other hand, God’s ὀργή targets non-Israelites in texts like
LXX Ps 58:6–14 (MT 59:5–13); and 78:1–7 (MT 79:1–7). It is, therefore, quite
unsurprising for Paul to poise God’s ὀργή against all humanity rather than only
against this or that nation.42
Second, why does God direct his ὀργή against such people? As in Wisdom
and MT Ps 106:19–20, the answer is idolatry, which Romans expands in terms
of hypocrisy. The targets of 1:18–32 have knowledge about God (1:19), but they
“suppress the truth” and neglect to glorify or give thanks to him as God (1:18,
21). That is, they refuse to give God due worship (1:25). Their knowledge and
behavior are inconsistent. Because of this hypocrisy, “they” are without excuse
(1:20) when they exchange God’s glorious presence for idols of people and
animals (1:23), the truth for a lie (1:25), and “natural use” for unnatural (1:26–
27). For these reasons, “God handed them over” (παρέδωκεν αὐτοὐς ὁ θεός) to
uncleanness (ἀκαθαρσίαν; 1:24), dishonorable passions (πάθη ἀτιμίας; 1:26), and
a useless mind (ἀδόκιμον νοῦν; 1:28).
In Paul’s argument, therefore, God is the orchestrator of humanity’s
stumbling (cf. 11:8) by handing them over to what Paul later discusses as
“slavery to Sin” (Rom 6), but he is also the faithful agent of their rescue. In
the gospel, God’s δικαιοσύνη saves humanity from slavery to Sin apocalyptically
conceived and, ultimately, from God’s own ὀργή (5:9). The argument of Rom
1:1–32 flows naturally along these lines; the distance between 1:16–17 and 1:18 is
not so broad as it is deep and mysterious (cf. 11:25–32). In Rom 2, however, Paul
begins to advance his argument through the use of other stylistic features. In

41  Contra Rodríguez, who denies the passage’s universal scope though he recognizes that it
is easy to find in the text “universalistic possibilities.” Rodríguez, If You Call, 28.
42  See Linebaugh for the intriguing argument that Paul’s use of Wisdom (or similar tradi-
tions) is in order to subvert it. Linebaugh, “Announcing the Human,” 214–37; idem., God,
Grace, and Righteousness, 93–121.
238 chapter 10

diatribal fashion, Paul directly addresses an imaginary persona and in this way
prepares for the dialogue beginning in 3:1. Understanding Paul’s argument in
Rom 2, therefore, is of critical importance for correctly analyzing the dialogue
in Rom 3:1–9.

Romans 2

By the end of 1:32, Paul’s Christian audience in Rome likely feels none the worse
for wear; Paul seems aware that their πίστις is being proclaimed in all the world
(1:8), he wants to be mutually encouraged with them (1:12), and they have ex-
perienced the salvation which God’s δικαιοσύνη powerfully brings to fruition
in the gospel (cf. 1:16–17). Things are not, however, as they seem. Rom 1:1–32
certainly serves an important role in Paul’s argument, but in it he makes no
specific claims about the situation(s) he is addressing in Rome. Paul discusses
ethnic inclusivity and God’s salvific δικαιοσύνη for humanity, but he only plain-
ly sets forth the practical/ethical implications of his teaching as it relates to life
in Rome later in the letter. In Rom 2, Paul slowly begins to sharpen his focus,
pericope by pericope, as he turns in apostrophic fashion to address an imagi-
nary interlocutor in preparation for the dialogue that takes place in 3:1–9.43
As this project draws closer to Paul’s use of speech-in-character in 3:1–9, a
number of questions rise to the fore. Does Paul address one imaginary figure in
2:1–29, or does he engage two, one in 2:1–11/16, and one in 2:17–29? Who is (are),
and how does Paul characterize, his interlocutor(s)? What is the relationship
between Paul’s interlocutor(s) and his actual audience in Rome? Previous
issues return to discussion as well, such as Paul’s theme of equality and his
use of the phrase, “to the Jew first.” I address these issues (and others) as they
appear in the text.

Romans 2:1–11

After Paul’s discussion of God’s ὀργή against all ungodliness and unrighteousness
of humanity, Paul suddenly turns in 2:1 to address a figure in the second
singular. Paul writes, “Therefore, you are without excuse, O human, everyone
who judges” (διὸ ἀναπολόγητος εἶ, ὦ ἄνθρωπε πᾶς ὁ κρίνων). Though Paul uses
the second singular to address this figure, it is not as though he actually calls

43  For Rom 2 as apostrophe, see Stowers, Rereading, 100–4, 127–28; idem, “Apostrophe,”
357–65.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 239

out a specific person, true-to-life or imaginary. Once again, that adjective that
is already so significant in Paul’s argument returns—πᾶς.44 Paul employs the
second singular in typical apostrophic fashion, but it has an immediately
broader scope for any and everyone who judges (πᾶς ὁ κρίνων) in the way
he outlines momentarily. What is more, in 2:1–16 Paul never addresses this
interlocutor in ethnic terms.45 Paul talks about Jews and Greeks (2:9–10) and
“nations who do not have the law” (2:14), but he resists applying these terms to
his addressee. The scope of Paul’s argument, and the identity of his addressee,
is as broad and ethnically vague as ever.
So how does Paul characterize this addressee? The question cannot be
more important, since Paul’s characterization of his interlocutor must serve
as the measure by which to gauge each line in 3:1–9’s dialogue.46 Just as
those in 1:18–32 are “without excuse” (ἀναπολογήτους; 1:20), Paul lays the same
verdict against his interlocutor (ἀναπολόγητος; 2:1). Paul explains, “When47
you judge another, you condemn yourself, for you do the same things when you
judge”48 (ἐν ᾧ γὰρ κρίνεις τὸν ἕτερον, σεαυτὸν κατακρίνεις, τὰ γὰρ αὐτὰ πράσσεις
ὁ κρίνων; 2:1), and this is despite the fact that Paul and his interlocutor “know
(οἴδαμεν) that God’s judgment is true against those who do such things” (2:2).49
Furthermore, Paul rhetorically asks his hypocritical and judgmental addressee
whether he or she presumes to escape God’s judgment (σὺ ἐκφεύξῃ τὸ κρίμα
τοῦ θεοῦ) or ignorantly (ἀγνοῶν) thinks lightly of God’s kindness and patience
(2:3–4), because the interlocutor is storing up “wrath” (ὀργήν) “on the day of
wrath and the apocalyptic revelation of God’s just judgment” (ἐν ἡμέρᾳ ὀργῆς
καὶ ἀποκαλύψεως δικαιοκρισίας τοῦ θεοῦ; 2:5).
Paul, therefore, parallels the characterization of his interlocutor with five key
points in his charges against those in 1:18–32. First and second, Paul’s addressee

44  So Tobin, Paul’s Rhetoric, 111.

45  So also Tobin, Paul’s Rhetoric, 111.
46  Below I consider whether Paul addresses one or two interlocutors in Rom 2.
47  For reading the phrase temporally, see BDAG, ὅς, f. A few textual witnesses (C*vid 104
syh**) add κριματι after ἐν ᾧ, indicating that they take the phrase as something like “in the
way in which” or “with the same judgment by which.” Another possible translation is, “On
the issue which.”
48  Or, “For, you who judge do the same things.”
49  Barrett proposes that the interlocutor speaks 2:2. Barrett offers no argumentative support
for this unlikely view. Barrett, Romans, 42. First, Rom 2:2 makes little sense as an interlocu-
tor’s response to Paul’s accusation in 2:1 that he “does the same things.” One would expect
the interlocutor to deny the accusation, not confess that God’s judgment is true against
the behavior of which Paul accuses him. Second, if the interlocutor makes the confession
in 2:2, it renders Paul’s question in 2:3 redundant and senseless.
240 chapter 10

“does the same things” as those in 1:18–32. Paul critiques the interlocutor for
being judgmental, but judging does not appear as a concrete issue in 1:18–32.
The fundamental issues in 1:18–32 are the failure to worship God and idolatry
(1:21, 23). So too the interlocutor fails to worship God, specifically in God’s role
as judge. The interlocutor knows that God is judge and that God’s judgment
is true, but the interlocutor assumes God’s role and submits his or her own
judgments (2:1–3).50 Paul even wonders whether the interlocutor believes he
or she, as judge, will escape God’s judgment (2:3). Further support for under-
standing the interlocutor’s error in this way comes later in Romans. Exhorting
his Roman audience to stop passing judgment on one another (14:10–13), Paul
directly appeals to God’s office as judge. Paul writes,

But why do you judge (τὶ κρίνεις) your brother or sister?… For we will all
stand before the judgment seat of God (πάντες γὰρ παραστησόμεθα τῷ
βήματι τοῦ θεοῦ)… Each of us will give an account … Therefore, let us no
longer judge one another (μηκέτι οὖν ἀλλήλους κρίνωμεν).

God’s role as judge, therefore, constitutes a chief reason Paul’s Roman Christian
audience should not pass judgment on one another, and this is in clear contrast
to the behavior of Paul’s interlocutor in chapter 2. (For how Paul’s interlocutor
also succumbs to idolatry, see the section on 2:17–29.)
Third, Paul’s interlocutor is hypocritical. Those in 1:18–32 have access to the
knowledge of God (1:19–21, 32), but they act contradictory to this knowledge
(1:21–32). So also Paul’s interlocutor knows that “God’s judgment is true against
those who do such things” (2:2) but nevertheless “does the same things”
(2:1, 3).
Fourth, as noted above, Paul’s assessment in 1:18–32 parallels material in
Wisdom; Paul’s telling rhetorical questions in 2:3–5 parallel Wisdom as well.
Wisdom writes,

Though they have placed trust in lifeless idols and swear wickedly, they
do not expect to suffer any harm (ἀδικηθῆναι οὐ προσδέχονται). But, on
both accounts, just penalties will pursue them (μετελεύσεται τὰ δίκαια),
because they thought wickedly about God by paying attention to idols
(προσέχοντες εἰδώλοις), and in deceit they swore unjustly and thought
lightly of holiness (καταφρονήσαντες ὁσιότητος). (14:29–31)

50  Dunn writes similarly; by doing the same things, “Paul could mean that the very attitude
of passing judgment on others was in effect an attempt to usurp the role of the Creator
and so improper to the creature.” Dunn, Romans, 89.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 241

Just as Wisdom’s target expects not to suffer any harm, gives heed to idols, and
“thinks lightly” of holiness, Paul rhetorically asks his interlocutor whether he
or she presumes to escape God’s judgment (λογίζῃ … ὅτι σὺ ἐκφεύξῃ τὸ κρίμα τοῦ
θεοῦ; 2:3) and “thinks lightly” of God’s kindness (καταφρονεῖς; 2:4), all the while
storing up “God’s just judgment” (δικαιοκρισίας; 2:5). It becomes more clear why
the interlocutor thinks this way as Paul’s argument develops through Rom 2:29.
Fifth, God unleashes ὀργή against those in 1:18–32. No less seriously, God’s
ὀργή stands at the ready against Paul’s hypocritical and judgmental interlocu-
tor, who is storing up ὀργή “on the day of ὀργή (ἐν ἡμέρα ὀργῆς) and the apoca-
lyptic revelation of God’s just judgment (ἀποκαλύψεως δικαιοκρισίας τοῦ θεοῦ;
2:5).” Based on these five points, Paul holds his judgmental interlocutor to be
qualitatively no different from those accused in 1:18–32.
One final note before moving on—the thematic thread of divine impartial-
ity remains unbroken. As Paul portrays the gospel in 1:1–7, 14–17 and God’s ὀργή
in 1:18–32, Paul also couches his critique of the interlocutor on the founda-
tion of God’s ethnic impartiality.51 God does not only judge Jews, nor does he
only judge gentiles. God “will render to each according to his or her deeds”
(2:6). For “every human (ἐπὶ πᾶσαν ψυχήν ἀνθρώπου; 2:9) who does evil,” God
will mete out “wrath” (ὀργὴ; 2:8), “anger” (θυμός), “affliction, and distress” (2:9).
On the contrary, God will repay “everyone who does good” (παντὶ τῷ ἐργαζομένῳ
τὸ ἀγαθόν; 2:10) with “eternal life” (ζωὴν αἰώνιον; 2:7), “glory, honor, and peace”
(2:10). Such recompense operates, for the second and third times, “to the Jew
first, and to the Greek” (2:9–10), because God shows no favoritism (οὐ γάρ ἐστιν
προσωπολημψία παρὰ τῷ θεῳ; 2:11). God’s judgment, whether positive unto eter-
nal life, or negative unto wrath, functions with no ethnic bias whatsoever.52
Being Jewish affords no qualitative advantage at the judgment; equality, jus-
tice, (and, we will see, grace) are the currencies that rule that day. The chrono-
logical (rather than qualitative) argument posed for 1:16, therefore, offers the
best explanation here as well.53

51  So also, Longenecker, Eschatology, 182–84.

52  As Longenecker writes, Paul “finds it presumptuous to contain the principle of divine im-
partiality within an ethnocentric context.” Ibid., 184. So also Kruse, Romans, 128; Wright,
Romans, 438–40; Porter, Romans, 76.
53  If Paul’s own chronological comments in Rom 1 and 9–11 do not suffice (see above),
Israel’s own narrative also documents these points. God first judges according to their
deeds Adam, Eve, and Cain (Gen 3–4), and he first brings cataclysmic destruction on
Noah’s generation (Gen 6–8). But God also first rescues and makes a covenant with the
righteous Noah (Gen 6–8), and he first calls and blesses Abraham for the formation of a
people (albeit with universal implications; Gen 12). Interestingly, even the terms ὀργή and
θυμός as divine attributes first appear directed against Moses (Ex 4:14).
242 chapter 10

In sum, so far, Paul characterizes his interlocutor as ethnically unspecified,

judgmental, and hypocritical. The interlocutor neglects to worship God and
is on the fast track towards God’s ὀργή. Paul even appeals to the same Jewish
tradition critiquing those in 1:18–32 as idolatrous to characterize his interlocutor
as one who expects to escape God’s impartial judgment, though the reason for
this latter point remains momentarily unclear. Paul continues to narrow the
apostrophic characterization of his single (see below) interlocutor in 2:17–29,
but first he comments further on God’s impartiality and the standard by which
God judges.

Romans 2:12–16

In 2:12–16, continuing to address his interlocutor, Paul extends his discussion

of divine impartiality, yet the way the argument advances is more difficult to
track. Almost out of nowhere Paul begins to talk about law (νόμος), about those
who do νόμος being justified (2:13), and even about those who never had the law
somehow doing it successfully (2:14). All of this seems to oppose what Paul says
later about no one being justified by works of νόμος (3:20). How then is Paul’s
extension of divine impartiality to be understood in terms of law observance?
Following his declaration that God is impartial (2:11), Paul continues to ex-
plain why (γάρ) this is true. Paul argues that God knows no partiality because
God does not judge based merely on possession of νόμος but on successfully
doing νόμος; “It is not the hearers of νόμος who are just (δίκαιοι) before God, but
the doers (ποιηταί) of νόμος will be justified” (δικαιωθήσονται; 2:13). To illustrate
his case even further, Paul introduces “nations (ἔθνη) who do not possess νόμος”
who nevertheless “do the requirements of νόμος” (τὰ τοῦ νόμου ποιῶσιν; 2:14).
Paul argues that on the day of judgment, these νόμος-doers’ conscience will
bear witness and their thoughts will condemn or (in this case) defend them
before God, and they will be justified despite not actually possessing νόμος
(2:15–16). The standard for God’s judgment propels his impartiality even
further. God’s judgment for humanity (τῶν ἀνθρώπων) is based on the measure
of Paul’s gospel (κρίνει ὁ θεὸς … κατὰ τὸ εὐαγγέλιόν μου) through Christ Jesus (διὰ
Χριστοῦ Ἰησοῦ; 2:16).54 Though the phrase, κατὰ τὸ εὐαγγέλιόν μου, has “long been
a puzzle,” understanding it in this way takes it “in its most literal sense to mean

54  It does not matter for my argument whether διὰ Χριστοῦ Ἰησοῦ adjectivally modifies
τὸ εὐαγγέλιόν μου or adverbially modifies God’s act of judgment. The phrase κατὰ τὸ
εὐαγγέλιόν μου necessarily modifies κρίνει adverbially. So, whether διὰ Χριστοῦ Ἰησοῦ ad-
verbially modifies God’s judging or adjectivally modifies Paul’s gospel (which adverbially
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 243

that the gospel itself is the norm of judgment.”55 This fits within the context of
2:12–16 (see below). Paul argues, therefore, that God’s judgment for humanity,
regardless of ethnicity or possession of νόμος, is universally Christological and
gospel-of-justification-by-πίστις oriented.56
This, of course, raises a number of questions: what does Paul mean by “doing
νόμος,” and how does it result in justification? How is God’s Christological
and gospel-oriented judgment a fair measure of νόμος fulfillment? Does
2:12–16 contradict or square with Paul’s later comments about no one being
justified by works of νόμος (ἐξ ἔργων νόμου; 3:20)? These are difficult questions,
but a profitable way forward is possible by beginning with Paul’s categorical
statement about the measure of God’s judgment. God does not judge some
by one standard and others by another; God judges humanity (τῶν ἀνθρώπων)
on the basis of the Christological gospel (2:16). The points that require further
clarity, therefore, are how doing νόμος achieves justification and what it means
to be one who does νόμος. Evidence indicates that “doing νόμος” in 2:12–16 en-
tails being in a correct relationship with God as set out by the terms of the gos-
pel (cf. 1:1–5; 16–17), which would be possible whether one actually possessed
νόμος or not (2:14) and thereby maintains divine impartiality. This “Christian
gentile” reading not only allows one to read δίκαιοι and δικαιοθήσονται (2:13)
in their usual Pauline sense based on πίστις and God’s initiative in the gospel,
it also demonstrates how the Christological gospel is the just standard for
God’s judgment of humanity (2:16), both those ἀνόμως and those ἐν νόμῳ (2:12).
Reading νόμος in this way logically aligns all of the elements in 2:12–16 with
themselves and with the rest of Romans (cf. 9:30).57

modifies God’s judgment), the phrase διὰ Χριστοῦ Ἰησοῦ nevertheless sharpens the mea-
sure by which God judges.
55  Bassler, Divine Impartiality, 157, though she opts for a different nuance. See Porter, Romans,
79–80 for an endorsement of the gospel as the measure by which God judges.
56  Contra Black, Romans, 50; Matera, Romans, 65–67, who suggest two standards of judg-
ment; and contra Murray, Romans, 77–79, who argues this “cannot be taken as the uni-
versal rule of judgment,” since “this would contradict what the apostle said about those
who ‘sinned without the law.’” The error of each involves a misunderstanding of what Paul
means by “doing” (or failing to do) νόμος, as a Christological understanding of νόμος fulfill-
ment justifies 2:16 as a universal measure of judgment. Though we disagree regarding the
thrust of 2:14–15, Dunn writes similarly, “Paul’s addressees would probably understand the
phrase thus: it will be by measuring them against the gospel which will show the extent
to which these Gentiles ‘do the things of the law.’ ” Dunn, Romans, 103, 106. Dunn seems to
waver, however, in idem., Theology, 137.
57  For alternative views and the maintenance of the “Christian gentile” position, see the dis-
cussions and literature cited in Kruse, Romans, 136–40; Cranfield, Romans, 151–73; Wright,
244 chapter 10

First, so far, Paul only speaks of humans as just or justified in 1:16–17. There,
God’s δικαιοσύνη effects human πίστις and enables salvation in the gospel for
all, so that any person who is δίκαιος on the basis of πίστις shall live. Such an
understanding of justification melds perfectly with Paul’s other comments in
Romans (or Galatians), such as 3:19–31.58 It would surely seem strange for Paul
to suggest an alternative view in 2:12–16. Any responsible reading of 2:12–16
must explain what Paul means by δίκαιοι and δικαιοθήσονται in 2:13, especially if
it opposes Paul’s use at all other points in Romans. In this reading, 2:12–16 does
not diverge from Paul’s basic use, as the terms in 2:13 remain wholly contingent
on πίστις and God’s decisive act in the Christological gospel. Further support of
this reading requires a discussion of what Paul means by “doing νόμος.”
Second, recall that as early as 1:1–5 Paul begins to define the gospel about
God’s Son, the Lord Jesus Christ, through whom grace and apostleship result
(εἰς) in ὑπακοὴν πίστεως for all the nations (πᾶσιν τοῖς ἔθνεσιν). God pre-promises
this gospel “through his prophets in the holy scriptures” (διὰ τῶν προφητῶν
αὐτοῦ ἐν γραφαίς ἁγίαις), one of which is Moses (Deut 34:10).59 Paul’s only
other direct references to scripture in 1:6–2:11 are at 1:17, where Paul appeals
to Hab 2:4 in support of his gospel of justification on the basis of πίστις, and
2:6, where Paul draws on Prov 24:12 / Psalm 62:13 (LXX 61:13) to illustrate God’s
impartiality. Paul only begins to discuss νόμος at 2:12. Consequently, all of Paul’s
engagements with scripture to this point speak in one accord concerning the
scope and nature of God’s activity in the gospel. By the time Paul’s audience
arrives at 2:12–16, these references constitute the epistolary context by which
they might begin to understand his discussion of νόμος. Rom 2:15’s probable
reference to God’s activity in Jer 31:33 (LXX Jer 38:33)60—a text originally
addressing Jews but here applied to non-Jews—only provides additional
parallels to God’s impartiality, inclusivity, and sovereignty. Read in this light,

“The Law in Romans 2,” 131–50; idem, Romans, 441–42; Simon Gathercole, “A Law unto
Themselves: The Gentiles in Romans 2:14–15 Revisited,” JSNT 85 (2002): 27–49; idem,
Where is Boasting?: Early Jewish Soteriology and Paul’s Response in Romans 1–5 (Grand
Rapids: Eerdmans, 2003), 126–29; Jewett, Romans, 213–14; Watson, Paul, Judaism, and the
Gentiles: Beyond the New Perspective (rev.; Grand Rapids: Eerdmans, 2007), 211–13. See the
most recent support of this view in Barclay, Gift, 466–70. Cf. Charles H. Cosgrove, “What if
Some Have Not Believed?: The Occasion and Thrust of Romans 3:1–8,” ZNW 78 (1987): 102.
58  The difference between Rom 2:13–14 and Paul’s citation of Lev 18:5 in Gal 3:12 is that in
Rom 2:12–16 Paul expects the doers of νόμος to be justified. In Galatians, however, Paul’s
point is markedly the opposite. The difference must pertain to the portrayal of the Mosaic
law in each instance.
59  Cf. Cranfield, Romans, 55n.4.
60  For example, Jewett, Romans, 215.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 245

νόμος comprises one of the prophetic avenues by which God pre-promises

the Christological gospel (1:1–5) through which his salvific program operates
impartially on the basis of πίστις (1:16–17; 2:6), which Paul confirms in 3:21–24.
Furthermore, Paul’s discussion of Abraham’s πίστις-based δικαιοσύνη in Rom 4
(cf. Gal 3:8) provides the paradigmatic example of just how νόμος pre-promises
Paul’s πίστις-based gospel. Successfully “doing νόμος” in this context, therefore,
would entail being in a correct relationship with God as set out by Paul’s gospel,
i.e., being justified ἐκ πίστεως (1:16–17).
Third, this reading coheres with Romans’s larger argument about νόμος. By
doing the “things/requirements of νόμος” as it relates to the proclamation of
the Christological gospel, Paul writes that people demonstrate the singular
“work of νόμος written on their hearts” (τὸ ἔργον τοῦ νόμου γραπτὸν ἐν ταῖς
καρδίαις αὐτῶν; 2:15). Immediately in 2:17–29 Paul praises “circumcision of the
heart” (περιτομὴ καρδίας) over against literal and outward observance of, and
boasting in, νόμος. In 3:21–24, Paul argues that the νόμος and prophets bear
witness to God’s δικαιοσύνη, which functions through Christ’s faithfulness
(διὰ πίστεως Ἰησοῦ Χριστοῦ) for all people characterized by πίστις (εἰς πάντας
τοὺς πιστεύοντας) without distinction (οὐ γὰρ ἐστιν διαστολή; see also 3:27–31).
Similarly, Rom 8:3–4 connects God’s sending of the Son and defeat of Sin with
the fulfillment of the singular “righteous requirement of the law” (τὸ δικαίωμα
τοῦ νόμου). Rom 10:4 also makes the connection between the τέλος … νόμου and
Christ (cf. Gal 3:24). Paul’s claims connecting νόμος and Christology throughout
Romans, therefore, provide a basis for the view that being justified by “doing
νόμος” in 2:12–16 is ultimately Christological and based on God’s δικαιοσύνη in
the gospel.61 Questions about νόμος, answers about Χριστός indeed.62
Fourth, in Rom 13:8 (and Gal 5:14) Paul sums up the scope of νόμος fulfillment
simply as love for one another (ὁ γὰρ ἀγαπῶν τὸν ἕτερον νόμον πεπλήρωκεν).
Love for others, however, is contingent upon recognizing God as God of Jews
and non-Jews alike (3:29), so that God’s people constitute a unified and equal
whole. In fact, this is a main thrust of Paul’s argument in Rom 1–11, as it spe-
cifically relates to the concrete situation Paul is addressing among the Roman
Christians (see below). Additionally, νόμος fulfillment by loving others is only
possible through engagement with the gospel and the indwelling “Spirit of
Christ” (πνεῦμα Χριστοῦ), without which “it is impossible to please God” (θεῷ

61  Kruse, Romans, 138–40; Wright, Romans, 441.

62  Cf. Beverly Roberts Gaventa, “Questions about Nomos, Answers about Christos: Romans
10:4 in Context,” forthcoming. Gaventa should not be held accountable for my use, or
theft, of the title.
246 chapter 10

ἀρέσαι ού δύνανται; 8:8–9). Fulfillment of νόμος, therefore, is once again ethni-

cally inclusive, Christological, and gospel oriented.
Fifth, some have critiqued the “Christian gentile” reading based on
the grammatical structure of 2:14, in which the adverbial adjective φύσει
(“naturally”) stands between two phrases it could potentially modify. Critics
argue that φύσει should belong with what follows, (φύσει) τὰ τοῦ νόμου ποιῶσιν,
not with what precedes, τὰ μὴ νόμον ἔχοντα (φύσει), as the Christian gentile
reading requires.63 It is far from grammatically necessary, however, for φύσει
to modify what follows rather than what precedes,64 and arguments based on
the way Paul uses the term elsewhere (e.g., Gal 2:15; 4:8; Rom 2:27; 11:21, 24) to
discuss identity rather than behavior strongly tip the scale in favor of φύσει
modifying what precedes.65 Thus, 2:14 discusses not those who “naturally do”
νόμος but those who “do not naturally (i.e., by birth) have νόμος.”
Paul, therefore, does not imagine those who do not possess νόμος keeping
the whole of the Mosaic law (or some abridged version of it66) magically,
“naturally,” or otherwise by happenstance. Rather, Paul redefines what it
means to be a successful “doer of the law”—the very law through which God
pre-promises his Christological gospel of justification on the basis of πίστις
for all (1:2–4).67 In keeping with God’s active δικαιοσύνη, it is God’s ethnically
unbound provision of salvation in the gospel that renders one a “doer of the
law.”68 In this way, Paul extends his endorsement of divine impartiality in two
respects. On the one hand, these Christian gentiles testify to God’s inclusive
and impartial δικαιοσύνη in the gospel. On the other hand, the gospel is shown
to be the just measure of νόμος observance in 2:12–16, both for those “without

63  For example, Dunn, Romans, 98; Bassler, Divine Impartiality, 141–44.
64  Gathercole, “A Law,” 35–37; Wright, Romans, 441–42; Jewett, Romans, 213–14.
65  Gathercole, “A Law,” 35–37; Achtemeier, Romans, 45; Philip Maertens, “Une etude de Rm
2.12–16,” NTS 46.4 (2000), 510.
66  Gathercole, “A Law,” 35–37. Contra, for example, Hultgren, Romans, 117–18.
67  This also allows Paul to unsettle the distinctions of “Jew” and “(un)circumcision in 2:25–
29. Wright, Paul, 836–37.
68  As Barclay writes, “For both Gentiles and Jews it is the act of God that produces the nec-
essary human obedience, and in that act the God of Israel will display his righteousness
in spite of human sin.” Barclay, Gift, 467. It is at this point that Cosgrove’s argument wa-
vers a bit. Cosgrove argues that by the time the audience arrives at 3:1–8/9, “Paul has not
yet pointed to anything specific that would disqualify the inner and outer obedience
of the Torah-faithful across the board as authentic obedience.” In fact, the main thrust of
Cosgrove’s argument about Jewish ἀπιστία in 3:1–8 is presupposed in 2:12–16. Cosgrove,
“What if Some,” 90–105.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 247

the law” (ἀνόμως) and those “in the law” (ἐν νόμῳ; 2:12).69 In 2:17–29, however,
Paul returns to honing the characterization of his imaginary interlocutor, at
which time it becomes clear that the interlocutor holds to a different under-
standing of νόμος than Paul discusses in 2:12–16.

Romans 2:17–29

With regards to Paul’s interlocutor in Rom 2, Thorsteinsson argues that epis-

tolary interlocutors usually remain the same unless clearly indicated oth-
erwise, so Paul’s apostrophe should be understood as directed to a single
interlocutor throughout Rom 2.70 Though Rodríguez follows Thorsteinsson
on other major points, he disagrees on the number of interlocutor’s in Rom 2,
finding not one, but two. For Rodríguez, Paul addresses Rom 2:1–16 to one
interlocutor and 17–29 to another.71 Tobin even finds three interlocutors in
2:1–3:8.72 As noted earlier, evidence favors the view that Paul only engages one
interlocutor in Rom 2–3, so that “judging” in 2:1–16 is further defined in 17–29,
and this is for two reasons, each with multiple supporting points.
First, nothing in the text necessarily contrasts the identity of the addressee
in 2:1–16 with that of 17–29. To begin, in apostrophic style, Paul uses the sec-
ond singular in both pericopae.73 In 2:1, the second singular is broadened by
the accompanying πᾶς; similarly, Paul’s adaptation of LXX Isa 52:5 in Rom 2:24
broadens the scope of his address by using the plural (ὑμᾶς). Both passages,
therefore, address “all of you” or “anyone who …” Additionally, Paul includes
no strong contrastive related specifically to the addressee’s (or addressees’)
identity; there is no ἀλλά or μέν … δέ formula, such that Paul distinguishes
between “you who judge” (2:1) and “you who call yourself a Jew” (2:17). Rom
2:17 does have a δέ, but δέ takes a number of connotations. It might simply be a
connective particle on par with καί, such as “you who judge … and call y­ ourself

69  Contra Bassler, by appealing to Christian gentiles, Paul does not deviate from his view
of divine impartiality but in fact upholds it in light of Christ. Bassler, Divine Impartiality,
143–44.
70  Thorsteinsson, Paul’s Interlocutor, 144. I disagree, however, with the basis of Thorsteinsson’s
argument. Thorsteinsson bases his argument on what he finds to be a general rule for
epistolary interlocutors, namely, that they represent the letter’s recipients. Ibid., 136, 140–
41. As argued below, this is not the case in Rom 2–3.
71  Rodríguez, If You Call, 49–50. See also Keck, Romans, 73–78, 82–88; Witherington, Romans,
87; Porter, Romans, 80–81.
72  Tobin, Rhetoric, 119.
73  Cf. Byrne, Romans, 96.
248 chapter 10

a Jew.”74 Or, if contrastive, perhaps the contrast does not concern actual iden-
tity but the interlocutor’s views about νόμος, gospel, and divine impartiality,
which Paul discusses immediately prior in 2:12–16. Further investigation will
carry this forward. For the moment, let it simply suffice that δέ does not have
to contrast identities in 2:17.
Furthermore, in both pericopae, the σύ is ethnically vague (see above for
2:1). Rom 2:17–29 is, of course, more difficult to define in this way, since the in-
terlocutor calls him/herself a “Jew” (Ἰουδαῖος). Naturally, many scholars under-
stand the identity of the interlocutor as Jewish in a combined ethnic-religious
sense,75 but evidence indicates that actual ethnicity may not be an issue. If
nothing else, Paul does not address the interlocutor as someone who neces-
sarily and existentially “is” (e.g., εἶ, γίνου) a Jew, but as one who “calls [him/
herself] a Jew” (σὺ Ἰουδαῖος ἐπονομάζῃ).76 More to the point, Paul must define
what he means by “calling oneself a Jew.” Paul’s definition, however, has noth-
ing to do with ethnic but with what we might call religious emphases.77 Paul,
for example, defines himself in Rom 11:1 in terms of ethnicity; “I myself am an
Israelite (Ἰσραηλίτης), from the seed of Abraham, from the tribe of Benjamin.”
But none of these elements appear in Paul’s definition of “calling one’s self
a Jew (Ἰουδαῖος).” Instead, Paul highlights altogether different points, such as
one’s view of νόμος and God (2:17–18, 23–24, 25–27), one’s engagement in the
Jewish religious obligation to help and teach others (2:19–22), and the Jewish
religious practice of circumcision (2:25–29; cf. Phil 3:5–6, where Paul defines
himself in ethnic and religious terms). Paul’s definition of “calling oneself a
Jew,” therefore, contains religious or theological, rather than necessarily eth-
nic, qualifiers. In this way, anyone, Jew or non-Jew (i.e., a proselyte), can “call
oneself” or “dedicate oneself to being” religiously Jewish.78 So, not only does
Paul not overtly indicate a shift in addressee, but neither is ethnicity a distin-
guishing factor between 2:1–16 and 17–29.

74  BDAG, δέ, 1.

75  For instance, Jewett writes, “The Jew is ‘one who identifies with beliefs, rites, and cus-
toms of adherents of Israel’s Mosaic and prophetic tradition.’ The Jew was an ‘adherent
to the religion and member of the nation of the Jews.’ ” Jewett, Romans, 222. Cf. Byrne,
Romans, 99.
76  Porter notes that with this phrase Paul “implies that the label of being a Jew counts less in
his scheme than it might in the schemes of others.” Porter, Romans, 80.
77  Cf. Kruse, Romans, 146; Moo, Romans, 159.
78   For such treatments of ἐπονομάζῃ, see Thorsteinsson, Paul’s Interlocutor, 196–97;
Rodríguez, If You Call, 48–51; LSJ, ἐπονομάζω, especially entry 1; BDAG, ἐπονομάζω. Cf.
Stowers, Rereading, 148.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 249

Second, in 2:17–29, Paul accuses the addressee of the same problems as in

2:1–16, albeit more finely defined.79 In 2:1–16, Paul charges his interlocutor with
behavior that is incongruous to his knowledge, failure to worship God, idolatry,
and the presumption that one will escape God’s judgment though God’s ὀργή
mounts against him or her.80 No differently, Paul argues that the interlocutor in
2:17–29 finds rest in νόμος (ἐπαναπαύῃ νὀμῳ) and boasts in God (καυχᾶσαι ἐν θεῷ;
2:17) and the law (ἐν νόμῳ καυκᾶσαι; 2:23), he knows God’s will and approves the
superior things (2:18), he has devoted himself to help and teach others (2:19–
20), and he lives according to his teaching by not stealing, committing adultery,
or robbing temples (2:21–22).81 Nevertheless, despite knowledge of God and
the law, Paul accuses the interlocutor of inevitably transgressing νόμος. What is
more, the interlocutor’s incongruous views and actions have disastrous results.
The interlocutor fails to glorify and worship God as God, dishonors God (τὸν
θεὸν ἀτιμάζεις; 2:23), and causes God’s name to be blasphemed among the na-
tions (ἐν τοῖς ἔθνεσιν; 2:24).
The interlocutor’s error is rooted in his failure to understand νόμος cor-
rectly. Rather than finding in νόμος the proleptic promise of justification on
the basis of πίστις (1:1–5, 16–17; 2:12–16; 3:21–24), the interlocutor boastfully
holds up Jewishness, possession of νόμος, and circumcision as advantageous
over others.82 That the interlocutor thinks this way is strengthened by Paul’s
immediate discussion of whether any benefit accrues from circumcision
(περιτομὴ μὲν γὰρ ὠφελεῖ; 2:25–29) and by the probing questions about Jewish
advantage that open and close the dialogue in 3:1–9. “So, what advantage is
there to being Jewish (τὸ περισσὸν τοῦ Ἰουδαίου), or what benefit comes from
circumcision (τίς ἡ ὠφέλεια τῆς περιτομῆς; 3:1)?” “What then? Are we at any
advantage or disadvantage (προεχόμεθα; 3:9)?”83 The interlocutor, therefore,
clings to an exclusive possession of νόμος and circumcision in place of God’s

79  Cf. Keener, Romans, 42.

80  Cf. ibid., 46–48; Wilckens, Römer, 121.
81  For a positive treatment of 2:21–22, see Rodríguez, If You Call, 53–56, whose argument
is based on the οὐ in the interrogative, expecting a positive answer, “Yes, I teach myself.”
Contra, for only one example, Porter, Romans, 81–82. With others who argue similarly,
Porter fails to recognize that the interlocutor transgresses the law not by the actions men-
tioned in 2:21–22, but by treating status, νόμος, and God’s promises as exclusive rather
than universal.
82  See Fitzmyer, Romans, 297; Barrett, Romans, 43; Dunn, Romans, 108; Jewett, Romans, 221–
23; Longenecker, Eschatology, 174.
83  I address προεχόμεθα in my detailed discussion of 3:9. Dunn argues similarly regarding the
usefulness of 3:1 in this respect. Dunn, Theology, 117–18.
250 chapter 10

ethnically inclusive and divinely orchestrated plan of salvation through the

Christological gospel.84
Characterizing the interlocutor in this way sheds further light on Paul’s
charge that the interlocutor condemns himself when he judges because he
does the same things (2:1). Again, “doing the same things” implies failing to
worship God and engaging in idolatry, but how does the interlocutor engage in
idolatry? The interlocutor has “exchanged” (cf. 1:23–27) δικαιοσύνη ἐκ πίστεως
for a reliance on status, νόμος possession, and observance. Consequently, the
interlocutor “does the same thing” by treating status and νόμος in an idolatrous
manner, expecting that it will exempt him or her from undergoing God’s judg-
ment (2:1–4).85 To complete the parallel characterizations of the interlocutor,
though Paul does not use the language of “wrath” (cf. 2:5), he makes clear that
the interlocutor is subject to judgment (2:27) and that God does not praise the
interlocutor’s behavior (2:28–29).
When one recognizes that Paul’s characterization of the interlocutor in
2:1–11 aligns with his developing and more narrowly defined characterization
in 2:17–2986 on all five points—incongruous knowledge and actions, failure to
worship God, idolatry, privileged expectation to escape judgment, and God’s
displeasure—it becomes evident that the δέ of 2:17 does not contrast two dif-
ferent interlocutors. Instead, the δέ contrasts the interlocutor’s view of νόμος,
gospel, and divine impartiality with Paul’s presentation of those concepts.
Paul identifies his single interlocutor, therefore, as ethnically unspecified and

84  Wright, therefore, begins with the wrong foot forward in suggesting that Israel’s ἀπιστία
(cf. 3:3) has nothing to do with what one might call a Christian πίστις relating in some
way from and to God or Jesus. This, however, is the distinction drawn between Jews
and Christian gentiles in 2:12–16; non-Christian Jews fail to recognize in νόμος God’s
pre-promised gospel, but God makes these gentiles successful doers of νόμος through
the Christological gospel. Thus, a (if not the) major issue at stake is whether or not one
correctly understands God’s intent in νόμος for both Jews and non-Jews, and Wright
misunderstands this point just like the interlocutor. Wright, Paul, 838–39.
85  Bassler finds Jewish idolatry in 1:23’s allusion to Israel’s idolatry with the golden calf,
through which Paul “intended above all to superimpose references to Jews and Gentiles.”
Bassler, Divine Impartiality, 197. Dunn writes similarly, “For Paul, then, the power of sin
has manifested itself characteristically in misdirected religion. And that included not just
Gentile idolatry, but also the idolatrous misplaced confidence of his own people in their
own God-given religion and status before God.” Dunn, Theology, 118–19.
86  Byrne writes similarly, “In the first instance the address is couched in fairly general terms.
But as the sequence develops, the identity of the implied addressee emerges more and
more.” Byrne, Romans, 96. Cf. Dunn, Romans, 108; Hultgren, Romans, 112; Watson, Paul,
198–99; Gathercole, Where is Boasting, 197–200; Longenecker, Eschatology, 174.
Romans 1-2: Paul ’ s Gospel and Paul ’ s Interlocutor 251

religiously Jewish (defined in such a way that opposes Paul’s view),87 char-
acterized as follows: the interlocutor (1) is hypocritical; (2) does not worship
God according to the ethnically inclusive, impartial, and Christological gospel
of δικαιοσύνη ἐκ πίστεως; (3) believes that Jewish status and νόμος possession
and observance, rather than God’s decisive act of δικαιοσύνη, place an indi-
vidual in the right; (4) believes that status and νόμος preclude an individual
from undergoing God’s judgment and thus provide him or her an advantage
over others; and (5) is nevertheless confronted with God’s disapproval. This
characterization of Paul’s interlocutor is the measure by which to assess the
dialogical script of 3:1–9, since attributed speech must appropriately align with
the character of the imaginary speaker.
Before getting to 3:1–9 at long last, however, one final discussion is in order;
does the interlocutor directly represent Paul’s audience or are they different,
and what are the implications? The first question is easier. The interlocutor
cannot be a direct representative of the historical audience. Paul writes to
Christians in Rome. Paul’s interlocutor, however, clings to νόμος and circumci-
sion and not to the gospel through which God’s δικαιοσύνη operates; Paul’s eth-
nically unspecified interlocutor is not Christian but religiously Jewish. Paul’s
dialogue with the interlocutor, therefore, does not represent imaginary conver-
sation directly with the Roman Christians. Paul employs the interlocutor for a
different purpose, an educational purpose, in order to teach or communicate
a particular point (or points) to his Roman audience indirectly.88 What Paul
aims to teach his audience comes out more fully in the forthcoming dialogue
(3:1–9) and summaries (3:10–31). As we will see, the message Paul communi-
cates through the interlocutor directly pertains to the themes he develops so
far in the letter—divine impartiality, ethnic inclusiveness, and God’s δικαιοσύνη
actively taking effect through the Christological gospel of justification on the
basis of πίστις. This project is finally staged to hear the dialogue of 3:1–9 in light
of speech-in-character, diatribe, and the argument Paul is currently developing
in Romans.

87  So, I agree with Thorsteinsson and Rodríguez that the interlocutor identifies as religiously
Jewish, but I disagree that he or she is specifically gentile. Contra Rodríguez, If You Call,
32–50; Thorsteinsson, Paul’s Interlocutor, 188–204.
88  That Paul uses his interlocutor to educate his historical audience does not omit the ne-
cessity to consider the tones Paul takes with his interlocutor and audience. For instance,
in Disc. 2.20, Epictetus treats his interlocutor polemically but he nevertheless uses the
interlocutor to teach his students.
 chapter 11

Romans 3:1–9 and the Argument of Romans

Romans 3:1–9

(3:1) Therefore, what advantage comes from being Jewish, or what benefit
accrues from circumcision? (3:2) There is much [advantage] in every way!
To begin, they were entrusted with the oracles of God. (3:3) To what end?
If some lacked πίστις, their lack of πίστις will not nullify God’s πίστις, will
it? (3:4) Absolutely not! Instead, let God be true but every human a liar,
as it is written, “So that you might be justified in your words, and you
will overcome when you are judged.” (3:5a) But if our unrighteousness
proves God’s righteousness, what shall we say? God is not unjust when he
brings wrath, is he? (3:5b) I speak in a human way. (3:6) Absolutely not!
Otherwise, how will God judge the world? (3:7) But if God’s truthfulness
is increased for his glory by my lie, why am I still being judged as a sinner?
(3:8a) Why not, (3:8b) as we are slandered and as some claim that we say,
(3:8c) “Let us do evil so that good might come?” (3:8d) Their judgment is
justly deserved. (3:9a) What then? Are we advantaged or disadvantaged?
(3:9b) By no means! (3:9c) For we have charged both Jews and Greeks all
to be under Sin, (3:10) as it is written …

So who is speaking when and why does it matter? Given the terms of the meth-
od, must some lines be heard from one speaker, or do certain lines at least fit
better in a particular voice? How do different arrangements impact readings
of Romans? Recall that traditional readings generally envision the interlocutor
posing objections in the form of questions to Paul in 3:1, 3, 5a, 7–8a, 8c, and 9a
with only minor variations, so that Paul speaks 3:2, 4, (5b,) 6, 8b, 9b and follow-
ing. For some or all of the questions, however, rescriptive readers place Paul in
the role of Socratic questioner who guides the interlocutor to affirm a particu-
lar view. Few readings on either side of the spectrum offer any evidence in sup-
port of their dialogical arrangements, and when they do it is typically sparse
or otherwise problematic. But getting the script right matters, and in 3:1–9 this
has particular relevance to questions of Jewish advantage, divine impartial-
ity, and the overall message Paul communicates to his Roman audience. The
following analysis first engages each question-answer unit independently (as
much as possible) in order to define the shape of the script, then as a complete

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_015

Romans 3:1-9 and the Argument of Romans 253

diatribal dialogue. Afterwards, I consider the dialogue’s impact on the remain-

der of Romans.

Romans 3:1–2
Directly following Paul’s apostrophic characterization and critique of his
ethnically-unspecified but religiously-Jewish interlocutor who boasts in
his advantages rather than the gospel, the dialogue naturally opens as fol-
lows, “Therefore, what advantage comes from being Jewish, or what benefit
comes from circumcision?” (τί οὖν τὸ περισσὸν τοῦ Ἰουδαίου ἢ τίς ἡ ὠφέλεια τῆς
περιτομῆς; 3:1). The opening phrase, τί οὖν, connects 3:1 to the apostrophe, as it
regularly advances an argument by drawing on the content from the preceding
material.1 Here, the concepts of being Jewish (Ἰουδαῖος, Ἰουδαίου; 2:17, 28–29;
3:1), circumcision (περιτομή, περιτομῆς; 2:25–29; 3:1), and salvific advantage
(ὠφελεῖ, περισσόν, ὠφέλεια; 2:25; 3:1) comprise the content of the question.2
The question expects that being Jewish and undergoing circumcision confer
some positive result on the circumcised Jew, or it is addressed to someone who
holds such an expectation.3
But who asks the question? Balking at Paul’s critique in 2:23–29, does the
interlocutor fire this objection at Paul as traditionalists and some rescriptive
readers hold?4 Or does Paul, knowing his interlocutor’s character, submit a
leading question that compels the interlocutor to come face-to-face with his
own views, as other rescriptive readers contend?5 Based on Paul’s characteriza-
tion of the interlocutor, the answer is—both are possible, at least initially. First,
the interlocutor personally holds that an advantage accrues from being Jewish
and possessing νόμος. It could be perfectly appropriate for such an interlocutor
to interrupt Paul in order to ask about the implications of Paul’s comments for
God’s covenantal promises to Abraham and, thus, the interlocutor’s place with-
in those promises. Essentially the interlocutor would be asking, “Based on what
you are saying, Paul, does God offer no advantage for Jews like myself?” In this

1  Cf. Stowers, Rereading, 166; Jewett, Romans, 241.

2  I speak about 3:1 as a single question. It could with no ill effect be punctuated as two ques-
tions: “Therefore, what advantage comes from being Jewish? Or what benefit comes from
circumcision?”
3  Cf. Dunn, Romans, 130, “Both questions presuppose the standpoint of one who has hitherto
assumed that being a Jew is an advantage.”
4  For example, Tobin, Paul’s Rhetoric, 119; Song, Reading, 94; Rodríguez, If You Call, 64–65;
Jewett, Romans, 241–42; Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Porter,
Romans, 84.
5  Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73; Byrne, Romans, 106–14.
254 chapter 11

respect, traditional readings are off to a fair start. On the other hand, second,
Paul is thoroughly familiar with the character of the interlocutor (he crafted it,
after all6) and could just as surely pose the question in Socratic fashion to force
the interlocutor to explain his views in light of Paul’s argument. Furthermore,
as Elliott argues, there is no overt indication of a shift in speaker between 2:29
and 3:1 as might be expected,7 and τί οὖν often functions in diatribe as a marker
for leading questions in the primary speaker’s voice, as even Stowers notes but
fails to consider at this point.8 So, technically, both voices could reasonably
and appropriately speak 3:1, but there is more to consider.
The question in 3:1 is only reasonably posed by either speaker when consid-
ered in isolation from the rest of the dialogue, particularly from the response in
3:2. This is because, allowing the conventions for speech-in-character to guide
exegesis, only one voice most appropriately speaks 3:2—and it is not Paul’s.
To the question of Jewish advantage, Rom 3:2 replies, “Much in every way! To
begin,9 they were entrusted with the decrees of God” (πολὺ κατὰ πάντα τρόπον.
πρῶτον μὲν [γάρ] ὅτι ἐπιστεύθησαν τὰ λόγια τοῦ θεοῦ).10 The answer is positive;
Jews have an advantage as far as the speaker in 3:2 is concerned.11 Chapter 10,

6  Cf. Keck, Romans, 89–90, who correctly writes, “Since the interlocutor, in accord with the
diatribe style, is the speaker’s rhetorical device, the whole passage is Paul’s creation.” This,
however, hardly requires that Paul and the interlocutor agree at any given point, such
that it is unnecessary to determine who speaks which lines, as Keck argues. For instance,
Teles and Epictetus are responsible for crafting their respective interlocutors’ speeches,
but they regularly disagree with the views they place in their interlocutors’ mouths (see
Chapter 7).
7  Of course, such a shift is not absolutely necessary, according to Quintilian, Inst. 9.2.37.
And, no matter where one identifies speeches-in-character in 3:1–9, none of them are
introduced in any overt manner. Characterization and appropriateness must be the pri-
mary guides.
8  Elliott, Rhetoric, 138. Cf. Stowers, Diatribe, 125–29; Moo, Romans, 180.
9  πρῶτον μέν is to be taken as a phrase indicating the first example in a presumably longer
list. BDF 447.4. Otherwise, πρῶτον might be taken simply as an adverb modifying ἐπι-
στεύθησαν, so that the Jews were chronologically entrusted with the decrees “first.” A few
manuscripts advocate a chronological reading by attesting the adjectival textual variant
πρωτοι instead of πρῶτον (1739 .6; Eus). The weight of the witnesses for πρῶτον, however,
is overly substantial (‫א‬, A, B, D, etc.). Even if the phrase was intended chronologically, it
would only confirm my explanation for Paul’s use of the phrase “to the Jews πρῶτον” as
chronological rather than qualitative.
10  I address ἐπιστεύθησαν (entrusted) in the section on 3:3–4.
11  In Rhetoric, 133, 136–37, Elliott fails to consider the importance of the interlocutor speak-
ing 3:2. For Elliott, 3:1–2 claims that Jews have real privileges, but 3:3–9 reveals that these
privileges “do not undermine God’s righteousness when Jews are disobedient.” At the
Romans 3:1-9 and the Argument of Romans 255

however, demonstrated how Paul throughout Rom 1–2 repeatedly proclaims

the scope of God’s activity as ethnically inclusive and equally grounded on the
same Christological gospel for all humanity. Being Jewish, possessing νόμος, and
being circumcised give Jews no salvific advantage over non-Jews. Paul speaks
similarly in Gal 5:6, “In Christ Jesus, neither circumcision nor uncircumcision
amount to anything; what matters is πίστις taking effect through love,” and 6:15,
“circumcision and uncircumcision are nothing; what matters is new creation
(καινὴ κτίσις).” So, as Dodd noted so long ago though unable to recognize the
full import of his observation, if Paul responds to Rom 3:1’s question about
salvific advantage for Jews over non-Jews, “The logical answer on the basis of
[his] argument [in Rom 1–2] is, ‘None whatever!’ ”12 Eschatologically, Paul’s ar-
gument requires that Jews possess no special advantage over non-Jews before
an impartial God.13 For Paul to affirm Jewish advantage in 3:2 would be contra-
dictory to his argument and, essentially, out of character.14

same time, Elliott argues that Paul guides the discussion to a point of agreement, so that
3:9 represents the interlocutor’s capitulation to Paul’s point of view. If 3:9 constitutes the
point of agreement between the two dialogue partners, on what basis does 3:2, spoken
by the interlocutor, serve as Paul’s support of Jewish privilege? Though still quite brief,
Elliott is more accurate in Arrogance, 106, when he writes, “The Judean interlocutor is not
concerned to protect his privilege over against God’s claim; to the contrary, he enthusias-
tically agrees with Paul that Judeans enjoy no defense against God’s judgment.”
12  Dodd, Romans, 43.
13  On 3:1–2 as specifically salvific advantage, Fitzmyer, Romans, 326; Witherington, Romans,
93; Campbell, Deliverance, 573–78. Note the distinctions of Hultgren and Longenecker.
Hultgren argues, “The Jew has some advantages, even if no privileges.” Hultgren,
Romans, 135. Longenecker argues, “Because Paul considers the scriptures to contain the
promise that God’s righteousness will be revealed through faith … the advantage of ‘the
Jew’ is that he stands first in line to enter into right relationship with God when that
promise is fulfilled.” Nevertheless, “the advantage of ‘the Jew’ does nothing to remedy his
position in the state of sin.” Longenecker, Eschatology, 195–96.
14  Barrett argues that a negative answer to 3:1 would be “offensive” to Jews and theology.
Barrett, Romans, 59. But Barrett assumes that a denial of “Jewish privilege” also rejects
God’s faithfulness to the promises and diminishes or downgrades Jews rather than ele-
vates non-Jews. This is not the case; though it may seem bleak at the moment, things are
not always as they seem. I further discuss Paul’s views about God’s promises in the sec-
tions on 3:3–4 and 3:9, and more holistically in the final section in light of the letter as a
whole, especially Rom 4 and 9–11.
   See the section on chapters 9–11 for discussion of why 9:4–5 does not qualify as an af-
firmation of Jewish privilege over non-Jews either. As a foretaste, it seems quite difficult
to define 9:4–5 as comparative advantages for Jews over non-Jews when Paul indicates at
various points in Romans that Christian gentiles also partake in each of these qualities
described in 9:4–5 except for the biological connection to Christ.
256 chapter 11

The questioner in 3:1, therefore, is most reasonably Paul, because it is the

interlocutor who most appropriately speaks the response in 3:2.15 Clinging
to νόμος, circumcision, and Jewishness as advantageous over non-Jews, of
course the interlocutor would assert that much advantage and benefit accrue
to Jews, especially as it relates to the λόγια τοῦ θεοῦ, the decrees of God, i.e.,
scripture and God’s promises.16 As attributed speech, 3:2 fits Paul’s character-
ization of the interlocutor seamlessly. Paul’s leading question in 3:1, therefore,
essentially asks, “Based on what I have argued about God’s impartiality, eth-
nic inclusiveness, the gospel, and your transgression of νόμος, do you really
hold any advantage or benefit over non-Jews?” Contrary to the thrust of Paul’s
question, but appropriate for his character, the interlocutor affirms the exclu-
sive Jewish advantage in which he himself trusts.17 Thus, against traditional
readings, some version of a rescriptive reading takes hold in 3:1–2, and one that
is distinctly not Stowers’s.18

Romans 3:3–4
Based on the conclusion that Paul opens the dialogue in 3:1 with a leading
question for his “privileged” interlocutor to answer in 3:2, it would be easy
to assume that the rest of the dialogue progresses similarly, with Paul asking
questions for the interlocutor to answer. Though this very well might be the
case, such an assumption would fail to argue why the discourse is best heard
according to a particular script. Each line must receive due diligence.

15  Concerning 3:1–2, therefore, my treatment provides the argumentative basis for the script
of Elliott, Rhetoric, 139; idem., Arrogance, 105; Campbell, 572–73; Byrne, Romans, 106–9,
111–12.
16  For discussion of λόγια as scripture and God’s promises therein, see the section on 1:16–17.
17  For Stowers, Rereading, 166–67, Paul speaks 3:2, retorting that Jewish advantage is that
“God has entrusted Jews with the good news to the gentiles … His answer is a dismissive
rebuke, although he is quite serious about Jewish priority.” As Stowers is keen to note of
traditional readings elsewhere, his own reading does not work. According to Stowers’s
script, the interlocutor asks what salvific advantage belonging to the Jewish community
confers (3:1). Paul’s answer, however, has nothing to do with Jewish salvific advantage, but
with good news for the gentiles. Not only is Paul’s hypothetical answer a non sequitur, but
it is also nonsensical, as it is a Jewish advantage with no actual advantage for the Jew. Cf.
Keck, Romans, 90.
18  Though Stowers appeals to speech-in-character and understands the requirement of ap-
propriateness to characterization, by neglecting to consider the script in terms of these
conventions (at least in writing), he fails to adjust the dialogue appropriately. Stowers,
Diatribe, 129, 167; idem., Rereading, 165–66; idem., “Paul’s Dialogue,” 715. My reading simi-
larly disqualifies the scripts of those who simply follow Stowers, namely, Keck, Romans,
89; Witherington, Romans, 93.
Romans 3:1-9 and the Argument of Romans 257

Following the interlocutor’s affirmation of exclusive Jewish privilege, Rom

3:3 asks, “For what? If some lacked πίστις, surely their lack of πίστις will not
nullify God’s faithfulness, will it?” (τί γάρ; εἰ ἠπίστησάν τινες, μὴ ἡ ἀπιστία αὐτῶν
τὴν πίστιν τοῦ θεοῦ καταργήσει;). The opening phrase, τί γάρ, illustrates that 3:3
is an objection to the content of 3:2.19 The sense is, “For what purpose or goal
were they entrusted,” or more bluntly, “So what if they were entrusted with
the λόγια? Does it really matter? Surely their lack of πίστις concerning the
λόγια will not nullify God’s faithfulness to those λόγια, will it?” Furthermore,
as Godet and others observe, the μή in the interrogative clarifies that the ques-
tioner intends the respondent to answer negatively, “No, their lack of πίστις
will not nullify God’s πίστις.”20
Understanding the argument hinges on what 3:2 means by suggesting that
the Jews “were entrusted (ἐπιστεύθησαν) with the λόγια τοῦ θεοῦ.” As argued in
the section on 1:16–17, the λόγια τοῦ θεοῦ are the Jewish scriptures, in which
God promises to bless Abraham and his ethnically exhaustive seed, and in
which God pre-promises the Christological gospel through which his δικαιο-
σύνη effects salvation for all on the basis of πίστις (1:1–7; 16–17; 3:21–31; 4:1–25);
the two are of course one and the same in Paul’s theology (see Gal 3:6–18).
The term “entrusted” (ἐπιστεύθησαν) suggests that God intended the Jews to
exercise πίστις towards the λόγια. Paul uses the term elsewhere to speak of God
“entrusting” or “commissioning” him with the ethnically unhindered gospel,
for which the correct response is to proclaim that gospel (1 Cor 9:17; Gal 2:7; 1
Thess 2:4). The λόγια, therefore, and being entrusted with them, contain both a
message of redemption for all and the expectation to share that impartial and
inclusive message with all. The non-Christian Jews under discussion in 3:1–3,
however, fail in both respects regarding the λόγια. First, they neither recognize
nor respond in πίστις to God’s pre-proclamation of the ethnically unbound
gospel (cf. 2:12–16; 11:17–23),21 though this is largely God’s mysterious doing

19  Stowers, Diatribe, 119. Stowers, however, incorrectly suggests that the objection is in refer-
ence to 2:17–29 rather than 3:2. That 3:3 objects to the response in 3:2 is supported by the
use of πίστις terminology in both verses, as well as the general trend for the questions
in the discourse to build off the previous response. Plus, how can Paul object to his own
critique of the interlocutor in 2:17–29? Cf. Malherbe, “Μη Γενοιτο,” 236–37.
20  Godet, Romans, 134; BDF §427. “God’s faithfulness to the faithless” allows Barclay one of
many opportunities to speak of the “incongruity of the gift” that grounds Paul’s hope for
a world universally corrupted by Sin as well as Paul’s confidence that “God pays no regard
to ethnic background, moral upbringing, or access to the Law.” Barclay, Gift, 473.
21  Similarly, Moo, Romans, 184; Longenecker, Eschatology, 196; Jewett, Romans, 243–44;
Dunn, Romans, 131–32, 139–40; Cosgrove, “What if Some,” 90–105; Hall, “Romans 3.1–8,”
185–86. Readings that neglect or reject this aspect fail to recognize that the λόγια have
258 chapter 11

(11:25).22 Second, they do not advance God’s promises among the nations, at
least not correctly. As characterized, the Jewish interlocutor embodies his role
as a guide for the blind, a light for those in the darkness, an educator of the
simple-minded, and a teacher of infants (2:19–20).23 His behavior even coheres
with his teaching (2:21–22).24 Nevertheless, boasting in λόγια/νόμος possession
and observance over against non-Jews—and thereby failing to understand its
universal scope and Christological fulfillment—the interlocutor transgresses
the central Christological thrust of λόγια/νόμος and causes God’s name to be
blasphemed among the nations (2:23–25). The problem in 3:2–3 as it relates
to God’s promises for non-Jews is not so much whether or not the interlocutor
has proclaimed the message of λόγια/νόμος,25 but whether he has proclaimed
the correct message, which he largely has not. On the contrary, he considers
the λόγια/νόμος as exclusive advantages over non-Jews, thereby misrepresents
God’s intent, and ultimately places boundaries around who can or cannot be-
long to God’s people.26
Despite this lack of πίστις by some (i.e., non-Christian) Jews,27 the question
in 3:3 assumes that God nevertheless maintains his faithfulness to bring those
λόγια to completion,28 namely, to fulfill his promises to bless Abraham and
Abraham’s ethnically inclusive seed through the gospel. As characterized by
Paul, the interlocutor is wholly incapable of posing such a question; not only is

eschatalogical and salvific implications for all humanity (1:1–5, 16–17; 2:12–16; 4:16; 11:25–
32), not only for non-Jews. Contra Stowers, Rereading, 166–68, 170–71, but see 169; Wright,
Romans, 453; idem., “Romans 2:17–3:9,” 1–4; idem., Paul, 837–38, cf. 931; Keck, Romans, 91.
22  Susannah Ticciatti, “The Nondivisive Difference of Election: A Reading of Romans 9–11,”
JTI 6.2 (2012): 261–62.
23  Wright correctly notes Paul’s willingness to grant the interlocutor this praise, though he
fails to see it through to its logical end (see below). Wright, Paul, 837; idem., “Rom 2:17–
3:9,” 11–14.
24  Rodríguez, If You Call, 53–55.
25  It does not matter whether this is directly through “mission” work or indirectly through
the way Jews lived. See Wright, “Rom 2:17–3:9,” 14–15.
26  Consequently, readers like Wright who limit God’s purposes in the λόγια to non-Jews
achieve the same result as these privileged Jews, only in reverse order, ultimately boxing
out God’s purposes for Jews in the λόγια. For all parties involved, these privileged Jews and
readers like Wright, this stems from a misunderstanding of the scope of God’s activity in
the λόγια and its relation to the pre-promised gospel. Contra Wright, Romans, 453; idem.,
“Romans 2:17–3:9,” 1–3; idem., Paul, 837–38; Stowers, Rereading, 166–71; Keck, Romans, 91.
27  Stowers suggests this is the first instance where Paul accuses more than one Jew of being
unfaithful, but he overlooks the plural ὑμᾶς in 2:24’s scripture citation. Stowers, Diatribe,
168.
28  So also Byrne, Romans, 109.
Romans 3:1-9 and the Argument of Romans 259

3:3 an objection to the interlocutor’s response in 3:2, but the question depends
on an understanding of λόγια the interlocutor does not—yet—endorse. The
interlocutor is an unrepentant (2:4) transgressor of νόμος (2:23–29) and thus
one of those “unfaithful” (ἀπιστία; 3:3) Jews (religiously speaking) who fail to
exercise πίστις regarding the λόγια in Paul’s terms. At this point, the interlocu-
tor and Paul have different understandings of λόγια, νόμος, Jewish advantage,
and God’s impartiality.29 Such an understanding of God’s faithfulness, how-
ever, altogether agrees with Paul’s argument in Rom 1–2 (especially 1:16–17).
Rom 3:3, therefore, is most appropriately spoken in Paul’s voice as a second
leading question.30
Rom 3:4 follows with the expected negative response, “Absolutely not! Let
God be true and every human a liar, as it is written, ‘so that you might be justi-
fied in your words, and you will overcome when you are judged’” (μὴ γένοιτο·
γινέσθω δὲ ὁ θεὸς ἀληθής, πᾶς δὲ ἄνθρωπος ψεύστης, καθὼς γέγραπται· ὅπως ἂν
δικαιωθῇς ἐν τοῖς λόγοις σου καὶ νικήσεις ἐν τῷ κρίνεσθαί σε). Given that 3:3 is a
leading question from Paul, it would make sense for the response to come from
the interlocutor.31 Does 3:4 with its citation of LXX Psalm 50:6 fit his character?
Indeed it does. The interlocutor is religiously Jewish and boasts in God (2:17)
and νόμος (2:23); of course he would uphold God’s faithfulness and be familiar
with scripture. After all, it is advantageous for the interlocutor to exalt God’s
faithfulness, since his presumed privilege wholly relies upon it, especially
after being characterized by Paul as an unrepentant transgressor (2:4, 23–29).
Furthermore, the interlocutor judges the “unprivileged” (i.e., non-Jews; 2:1–4)

29  On the different levels of discussion between the interlocutor and Paul regarding how
they understand various terms differently, cf. Campbell, Deliverance, 577–78.
30  My argument, therefore, supports a rescriptive reading, all of which are identical at this
point in the dialogue. Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73;
Byrne, Romans, 106–14; Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck,
Romans, 89; Witherington, Romans, 93.
31  The interrogative construction with μή is in some ways a two-edged sword, especially as
a leading question intended to guide the interlocutor. On the one hand, it is helpful for
understanding the intent of the dialogue. On the other hand, it makes the thrust of the
question so similar to the answer that, technically, Paul himself could speak both ques-
tion and answer. This is not so much a problem with the method per se but with the gram-
matical construction. Sufficiently tipping the scale in favor of the interlocutor’s voice are:
the characterization of the interlocutor, the Socratic method of question and answer fre-
quent in diatribal dialogue to guide an interlocutor to a particular point of view, the com-
mon implementation of an interlocutor to provide the evidence in the primary speaker’s
argument, the logic of the argument in Rom 1–3, and the message Paul communicates to
his audience through the interlocutor (see below on several of these points).
260 chapter 11

and even seems to concede that he himself is a transgressor in keeping with

Paul’s characterization of him (at least, he does not object to it; 2:23–29). The
universalization from “some unfaithful non-Christian Jews” (3:3) to the false-
hood of “all humanity” (πᾶς δὲ ἄνθρωπος; 3:4), therefore, is possible within the
interlocutor’s characterization.32 Consequently, as guided, the interlocutor
correctly answers Paul’s question in 3:4.33
One further point bears mentioning. The first instance of μὴ γένοιτο belongs
not on Paul’s lips but the interlocutor’s. This significantly problematizes the
arguments of Malherbe, Song, and Matera (amongst others), who assume
that the phrase must be spoken by Paul (see my critique in Chapter 8).34
Furthermore, such arguments overlook the fact that Paul is responsible for
crafting and composing the whole dialogue; though many understand the
phrase to be particularly Pauline, nothing hinders him from applying it to any
voice he wishes. Rescriptive readers recognize this possibility, and Stowers
even shows how the same applies to the speech-in-character in Rom 7 as well.35
The instance in 3:5–6 must similarly be approached with openness of mind.

Romans 3:5–6
Rom 3:1–4 focuses on God’s faithfulness to the λόγια to bless Abraham and his
seed through the gospel. Following the interlocutor’s cue in 3:4,36 Rom 3:5–8
transitions to discuss God as judge. Rom 3:5 asks, “But if our unrighteousness
demonstrates God’s δικαιοσύνη, what will we say? Surely the God who brings
wrath is not unjust, is he?”37 (εἰ δὲ ἡ ἀδικία ἡμῶν θεοῦ δικαιοσύνην συνίστησιν,
τί ἐροῦμεν; μὴ ἄδικος ὁ θεὸς ὁ ἐπιφέρων τὴν ὀργήν;). Again, the μή indicates that
the questioner expects a negative answer, which 3:6 supplies, “Absolutely not!
Otherwise, how will God judge the world?” (μὴ γένοιτο· ἐπεὶ πῶς κρινεῖ ὁ θεὸς τὸν
κόσμον;).

32  Cf. Keck, Romans, 92, who argues that the interlocutor here condemns himself as a liar. Cf.
Dunn, Romans, 140.
33  Considering the conventions for speech-in-character again confirms rescriptive read-
ings. Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73; Byrne, Romans,
106–14; Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans, 89;
Witherington, Romans, 93.
34  Malherbe, “Μη Γενοιτο;” Song, Reading, 94–95, 95n.5; Matera, Romans, 78–80; Jewett,
Romans, 245n.61.
35  Stowers, Rereading, 264–69; idem., “Romans 7,” 191–202. See also Longenecker, Rhetoric,
88–93; King, “Rhetorical Chain-Link Construction.” Contra works like Timmins, “Romans
7,” 106, which simply assume the voice behind the μὴ γένοιτο declarations.
36  Malherbe, “Μη Γενοιτο,” 236–37.
37  The ἐπιφέρων could also be translated temporally, “when,” or conditionally, “if.”
Romans 3:1-9 and the Argument of Romans 261

Multiple points indicate that Paul once again plays the role of Socratic ques-
tioner. First, μή suggests that 3:5 is a leading question guiding one’s conversa-
tion partner in a particular direction.38 Second, Rom 3:5 expresses a correlative
relationship between God’s δικαιοσύνη and ὀργή by which God is both judge
and justifier; Paul draws this very connection between 1:16–17 and 1:18–2:16.39
Third, Paul’s previous discussion paves the way for him to pose such a lead-
ing question to the interlocutor, in which he presses his interlocutor to con-
sider Jewish privilege over non-Jews in light of Paul’s comments about God’s
ethnically inclusive δικαιοσύνη and ὀργή. Fourth, the previous discussion
about δικαιοσύνη and ὀργή also renders 3:5 redundant and unobjectionable if
voiced by the interlocutor, as Paul has already addressed it. What remains is
for the interlocutor to weigh in on the issue in response. Fifth, 3:5 is posed as a
conditional, like Paul’s objection in 3:3, thus creating an element of parallel-
ism both in style and, presumably, the speaker responsible. Sixth, hearing 3:5
in Paul’s voice omits the need to suppose that he awkardly or apologetically
interjects an authorial aside in 3:5b, “I speak humanly” (κατὰ ἄνθρωπον λέγω).40
Instead, the first-person speech in 3:5 hangs together as spoken by a single
voice, Paul’s (I consider below the use of first- and third-person speech in 3:1–9
as a whole).41
Whose voice does the response in 3:6 best fit? It certainly would fit Paul’s,
as expected with the use of μή. Plus, the language of God judging “the world”
(τὸν κόσμον) aligns with Paul’s discussion of God as judge of all of humani-
ty (1:18–2:16).42 But this too is expected in a diatribal discourse in which the

38  Stowers, Rereading, 170.

39  Contra Moo, Romans 189–90, who finds no reference to God’s active δικαιοσύνη here. See
Wright, “Rom 2:17–3:9,” 5n.5.
40  See the discussion of this in Chapter 8.
41  Dunn represents a case in point for Stowers’s critique that traditional readings fail in
respect to their treatment of the pronouns in the dialogue. Dunn argues that with the
first person plural, ἡμῶν, “the voice of the interlocutor is finally merged into Paul’s own
debate with himself … Paul identifies himself with the unfaithful Jews.” Dunn, Romans,
141. Dunn’s reading fails particularly at two points. It neglects the function of first-person
speech in diatribal discourse, which should push traditional readers to reconsider the ar-
rangement of the dialogue. More importantly, Dunn is unable to explain 3:1–8/9 in such a
way as to maintain a sustainable dialogue.
42  In attributing this verse to Paul, Dunn argues that “Paul has now boxed himself com-
pletely into a corner, from which he cannot escape. All he can do is hang on to these
two basic assertions of his faith: that God is eschatological judge … and that God has not
abandoned his purpose for Israel, his saving outreach through Israel to the world, and
his continuing faithfulness to Israel despite Israel’s unfaithfulness and unrighteousness.”
262 chapter 11

­ rimary speaker guides an interlocutor in a particular direction. Given that 3:5

p
is a leading question from Paul, does 3:6 fit the interlocutor as one might also
expect in a diatribal discourse? The interlocutor knows that God’s judgment is
true,43 and he even steps incongruously into God’s role and judges non-Jews,
for which he is without excuse (2:1–4). Furthermore, the interlocutor holds that
Jews have an advantage that will rescue them from God’s negative judgment
(2:3–4, 17–23; 3:2). Thus, the interlocutor can readily speak about God’s role as
judge, even if his own relationship to that judgment does not quite cohere with
Paul’s perception of it. Additionally, the μὴ γένοιτο rejection mirrors the phrase
in the interlocutor’s voice in 3:4. Based on appropriateness to characterization,
there is no reason to suppose that 3:6 does not represent the interlocutor’s re-
sponse to Paul’s leading question in 3:5.44 The interlocutor continues to fol-
low Paul’s lead and answers appropriately; despite human unrighteousness
(ἀδικία), God both exercises δικαιοσύνη and is not “unjust” (ἄδικος) when he
brings wrath.

Romans 3:7–8
In the final exchange before the discussion partners return to the question of
Jewish advantage or disadvantage that functions as bookends to 3:1–9, Rom
3:7–8c asks, “But if, because of my falsehood,45 the truth of God abounded for
his glory, why am I still being judged as a sinner? Should we not, as we are
slandered and as some claim that we say, ‘Let us do evil so that good might
come?’” (εἰ δὲ ἡ ἀλήθεια τοῦ θεοῦ ἐν τῷ ἐμῷ ψεύσματι ἐπερίσσευσεν εἰς τὴν δόξαν
αὐτοῦ, τί ἔτι κἀγὼ ὡς ἁμαρτωλὸς κρίνομαι; καὶ μὴ καθὼς βλασφημούμεθα καὶ καθώς
φασίν τινες ἡμᾶς λέγειν ὅτι ποιήσωμεν τὰ κακὰ, ἵνα ἔλθῃ τὰ ἀγαθά;). Once more,
the μή in the interrogative indicates that 3:8a is a leading question by which the

Dunn, Romans, 142. Dunn’s theology is more or less on point, as I argue below with nu-
ances. Dunn’s description of the nature of the dialogue, however, crumbles against the
diatribal norms for the primary speaker to be ever in control of the discourse and to guide
it towards a specific point. In Dunn’s analysis, Paul is no longer in control but must grasp
for whatever he can find to keep his head above water. Cf. Jewett, Romans, 247.
43  Stowers discusses this as a “shared axiomatic belief.” Stowers, Diatribe, 170. Cf. Byrne,
Romans, 110, 114; Dunn, Romans, 142.
44  Rom 3:5–6, therefore, also maintain a rescriptive reading. Elliott, Rhetoric, 139–41; idem.,
Arrogance, 105–6; Campbell, 572–73; Byrne, Romans, 106–14; Stowers, Rereading, 165–66;
idem., “Paul’s Dialogue,” 715; Keck, Romans, 89; Witherington, Romans, 93.
45  Wright’s contention that the first singular is a rhetorical way of talking indirectly
about Israel fails in light of the ethnic unspecificity so prevalent in Rom 1–3. Wright,
Romans, 454.
Romans 3:1-9 and the Argument of Romans 263

questioner expects the respondent to answer negatively.46 As in 3:5, attribut-

ing 3:7–8c to Paul in keeping with the question-response pattern so far in the
dialogue places all the first person speech in one mouth and omits any need to
presume that Paul awkwardly interjects his authorial commentary in the mid-
dle of the interlocutor’s objection.47 Furthermore, though 3:6 supplies the key
word (κρινεῖ; 3:4 also contains κρίνεσθαι) that advances 3:7–8 (κρίνομαι, κρίμα),48
both 3:5 and 7–8c draw potential but false and absurd conclusions from the
interlocutor’s response in 3:4. That is, Rom 3:7 begins with a consideration of
“God’s truth” (ἀλήθεια τοῦ θεοῦ) and human “falsehood” (ψεύσματι)—issues 3:4
raises. From this, 3:7–8c absurdly posits continuing in error in order to advance
God’s glory all the more. Rom 3:5 and 7–8c, therefore, stand together as leading
questions stemming from the interlocutor’s responses in 3:4, 6.
It would be altogether out of character for the interlocutor to consider the
viability of “doing evil so that good might come”;49 the interlocutor actively
judges those who do evil (1:18–2:4) and (albeit incorrectly) prides himself on
being in the right (2:17–23). Nor is there reason to suppose that the interlocutor
was slandered and accused of proclaiming such a view, as if he were respon-
sible for speaking the entirety of 3:7–8c. The slanderous charge makes little
sense if it targets those who cling to νόμος like the interlocutor. If, however, it
targets non-observers like Paul as the majority hold, it makes fine sense as a
law-contingent critique of Paul’s non-law-contingent gospel.50 Paul, therefore,
speaks 3:7–8c as a guiding question drawing on his own experience.51
This leaves 3:8d for consideration, and it is significant. The respondent
declares, “The judgment of them is just” (ὧν τὸ κρίμα ἔνδικόν ἐστιν). In con-
text, “them” does not refer to people who propose that the end justifies the
means52—in this script, there are no such people other than Paul, who uses
this as a false conclusion—“them” (ὧν) most reasonably refers to the “some”

46  So also Byrne, Romans, 114; Jewett, Romans, 251.

47  Contra traditional readings.
48  Malherbe, Μη Γενοιτο, 236–37.
49  Contra Stowers, who argues the saying “is the logical objection that a competitor like the
teacher might make to Paul’s explanation for the widespread Jewish failure to recognize
the gospel of Christ.” Stowers, Diatribe, 173. This view falls apart when one recognizes
that the interlocutor is rather one of those who accuse Paul of endorsing this position (see
below). Cf. Jewett, Romans, 251.
50  Jewett, Romans, 251.
51  Ibid.; Moo, Romans, 195; Hultgren, Romans, 139. For Paul’s considering himself a “sinner,”
see Rom 5:8; Gal 2:14–17. Cf. Hall, “Romans 3.1–8,” 193–94.
52  Contra Porter, Romans, 86.
264 chapter 11

(τινες) who slander Paul as a libertine.53 These are the same “some” who Paul
claims are “unfaithful” to the λόγια (3:3) though they cling to νόμος / λόγια as
advantageous (2:23–29; 3:2). As Jewett writes, “The key to this passage … is
that those who advocate such slanders are the very legalists represented by
the interlocutor.”54 This means that the respondent does not actually answer
Paul’s question about doing evil for the purpose of good. Instead, the respon-
dent condemns the actions of those who slander Paul as a libertine.55
Though Paul would agree with the response, the interlocutor must be the
respondent.56 Without 3:8d, the interlocutor never overtly makes the final step
that allows him to join with Paul in 3:9 (see below). He is religiously Jewish and
presumptuously privileged; of course he affirms God’s faithfulness, righteous-

53  Cf. Byrne, Romans, 110. It might be objected that ὧν is not a masculine reflexive pronoun
referring to τινες, but a neuter referring to the saying. Though this seems unlikely given
the presence of τινες as an extant antecedent and the interlocutor’s characterization as
one who judges others, the main thrust of my argument would still work with only minute
alterations, as the interlocutor would nevertheless side with Paul in 3:8d (see below).
   This is also the point where Paul J. Achtemeier’s structural analysis of 3:1–8 falters.
Achtemeier’s structural (not scriptive) analysis of 3:3 and 5 is accurate. Achtemeier
argues that 3:3, 5 begin with statements that are assumed true (εἰ), a false inference follows
(μή), and a strong denial rejects the false inference (μὴ γένοιτο) with some explanation.
In 3:7–8, this structure allows Achtemeier to suggest that 3:8d is not, contra Stowers, an
ad hominem retort, but an educational statement of fact explaining why one should not
play “fast and loose” with God’s eschatological justice. Achtemeier, “Romans 3:1–8,” 84–86.
Achtemeier’s analysis of 3:7–8, however, does not fit the mold. First, Achtemeier has to
supply the rejection that is so critical for establishing the structure of 3:3, 5 but otherwise
missing in 3:8. Second, for 3:8d to be an explanation for the rejection of the false infer-
ences Achtemeier proposes (he considers two), ὧν would have to refer to the statement
rather than to the τινες who falsely accuse Paul. Third, Achtemeier neglects to consider
how 3:8d might function in the interlocutor’s voice, which changes things significantly
and ironically resolves the problem Achtemeier finds with 3:8d in the first place.
54  Jewett, Romans, 251. Again, the convention of appropriateness confirms a rescriptive
reading for 3:7–8a. Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73;
Byrne, Romans, 106–14; Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck,
Romans, 89; Witherington, Romans, 93.
55  Campbell, therefore, is mostly right on this point. Campbell argues that the respondent
(the Teacher) condemns those who proclaim libertinism and those who slander Paul as
such, one of whom would be the Teacher himself. Campbell, Deliverance, 576–77.
56  Byrne, Stowers, Keck, and Witherington suggest that Paul speaks 3:7–8, which means the
interlocutor provides no response to 3:7–8c. Byrne, Romans, 106–14; Stowers, Rereading,
165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans, 89; Witherington, Romans, 93.
Arguing for a rescriptive reading with 3:8d in the interlocutor’s voice, however, validates
Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73.
Romans 3:1-9 and the Argument of Romans 265

ness, and judgment of the world—the latter of which sufficiently character-

izes him to speak of judgment in 3:8d, albeit not necessarily of “privileged”
Jews. Thus, Stowers’s script (with Keck and Witherington) fails at the begin-
ning (3:1–2) and end (3:7–9; see below on 9). Attributing the whole of 3:7–8
(especially 3:8d) to Paul’s voice creates a dialogue in which the interlocutor is
never compelled to rethink his view about Jewish salvific advantage over non-
Jews; the teacher, Paul, fails to guide the interlocutor to the correct view. This is
crystalized in such scripts by the interlocutor’s return to the question of Jewish
(dis)advantage in 3:9a, which Paul must answer for the interlocutor (according
to their script)—the interlocutor still does not understand the scope of God’s
impartiality and equality.57 That is, in Stowers’s arrangement, the interlocutor
does not actually join with Paul in 3:9, nor does he make the concluding point
in Paul’s argument. If, however, one recognizes that it is in 3:7–8 where the
interlocutor finally comes face-to-face with the contradictory nature of his
views and the implications of Paul’s leading questions throughout the dia-
logue, the teacher’s guidance is successful and the dialogical collaboration
near complete. Further explanation requires a discussion of the inappropriate
speech-in-character attributed to Paul in 3:8.
Paul claims that he is “blasphemed” or “slandered” (βλασφημούμεθα) by
some who attribute to him the proclamation, “Let us do evil, so that good
might come” (3:8). As far as speech-in-character is concerned, the hypothetical
speaker is identified (Paul et al.; ἡμᾶς), and a verb of speech (λέγειν) introduces
the attributed speech. Paul, however, avows that the attributed speech is out of
character, as it is falsely, slanderously, or otherwise inappropriately attributed
to him.58 But more is at stake. By slandering Paul in this way, the Jews speaking
confirm the interlocutor’s characterization (cf. Rom 2) by revealing that they
require νόμος possession and observance as a necessary prerequisite for be-
longing to God’s people. Thus, they hinder the advancement of God’s inclusive
promises and are “unfaithful” to the λόγια as defined by Paul (1:1–5; 16–17; 2:12–
16); they not only “blaspheme” Paul but cause God’s name to be “blasphemed”
among the nations as well (2:24). They are without excuse and deserving of
God’s judgment, which the interlocutor now recognizes. The discussion of 3:9
continues to fill out this argument.

57  Contra Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans, 89;
Witherington, Romans, 93. Cf. Byrne, Romans, 106–14; Jewett, Romans, 257–58.
58  This is further evidence that Paul understands the convention of characterization and
appropriateness when attributing speech-in-character to imaginary speakers.
266 chapter 11

Romans 3:9
In 3:9, Paul returns to the theme that begins the dialogue, τί οὖν; προεχόμεθα;59
As in 3:1, “What then?” (τί οὖν) prepares the interlocutor to consider a question
in light of the previous discourse. Essentially, the phrase compacts all of 3:1–8.
To paraphrase,

You (the interlocutor) believe that Jews have an exclusive advantage

over non-Jews because they were entrusted with the λόγια (3:1–2). But
is the possession of the λόγια really an advantage if you also affirm that
God’s faithfulness to the λόγια does not depend on Jewish πίστις towards
those λόγια (3:3–4)? You also maintain that God is the just judge of the
world’s unrighteousness, including Jewish ἀπιστία and ἀδικία (3:3–6) like
my (Paul’s) own previously (3:7). So what about those Jews, like you, who
hold up Jewishness and νόμος as a contingency for belonging to God’s
people? Your views seem to contradict. Recall that God’s actions in νόμος
/ λόγια are ethnically inclusive and according to πίστις, not possession
or observance (Rom 1–2; see also 3:10–29). Is it right for you to limit the
scope of God’s actions, slander me with such an absurdity, and cause
God’s name to be blasphemed among the nations (3:7–8c)? Of course
not; you answer correctly that you who limit the scope of God’s universal
activity are unfaithful to the λόγια and fall under God’s judgment (3:8d).
So, τί οὖν, what do you really think about Jews and non-Jews before an
inclusive and impartial God? Do either have an advantage (3:9a)?

Before considering the response, a quick note is due concerning Paul’s use of
προεχόμεθα. The verb is particularly difficult, but the difficulty, though real, is
only marginally significant. Most scholars view the verb as middle with active
force, meaning, “Do we have an advantage?” Some understand the verb as a
true middle, “Do we offer anything in defense.” Still others read προεχόμεθα as
a passive, “Are we at a disadvantage?”60 In light of what follows, however, how

59  Interpretive difficulties regarding προεχόμεθα, witnessed by ‫ א‬Β Κ and others, have
resulted in several variants attempting to improve the “difficult and probably original
wording προεχόμεθα οὐ πάντως.” P omits οὐ πάντως, A D2 L record προεχώμεθα, and D* G Ψ
and others record προκατέχομεν περισσόν. Jewett, Romans, 253.
60  For discussions of the verb, see the following and the literature cited there. Jewett,
Romans, 256–57; Stowers, Rereading, 173–74; Nils Alstrup Dahl, “Romans 3:9: Text and
Meaning,” in Paul and Paulinism: Essays in Honour of C. K. Barrett (Morna D. Hooker and
S. G. Wilson, eds.; London: SPCK, 1982), 184–204; Moo, Romans, 198–201.
Romans 3:1-9 and the Argument of Romans 267

one takes the verb matters little.61 The summary makes painfully clear that
neither Jews nor non-Jews are advantaged, disadvantaged, or have anything
to hold up in defense before God. All—Jews and Greeks—are under Sin and
on equal footing before a just, inclusive, and impartial God; all are justified by
God’s grace on the basis of πίστις (3:9–31; Gal 3:28). God is not the God of Jews
only, but equally of Jews and non-Jews (3:29–31). That said, I think the middle
with active force makes the most sense of the dialogue, as it creates an inclusio
between 3:1 and 9 and, in diatribal fashion, allows the interlocutor to recon-
sider his initial thoughts in view of Paul’s guidance.62
And the interlocutor does reevaluate his thoughts.63 Confronted again with
the question of advantage or disadvantage, the interlocutor affirms equality
and rejects the possibility outright, οὐ πάντως, “not at all” (3:9b).64 By holding
up νόμος as an exclusive prerequisite for belonging to God’s family rather than
recognizing in it God’s Christological plan for all based on πίστις, some Jews
transgress God’s intent in νόμος and fall under judgment no differently than
non-Jews. This confession, however, opposes the interlocutor’s characterization
as one who clings to Jewish advantage (2:17–29). What has happened? As illus-
trated in Chapter 7, the overarching category of diatribe, which often requires
an implied double-characterization, overcomes the conventions of speech-in-
character that work so well for 3:2, 4, and 6. In keeping with the Socratic ten-
dency to reveal a conversation partner’s errors (censure) and guide him/her to
the correct view (protreptic), the interlocutor recognizes his inconsistency and

61  Similarly, Robert H. Mounce, Romans (NAC 27; Nashville: Broadman & Holman, 1995), 107;
Kruse, Romans, 163; Murray, Romans, 101–2; Keener, Romans, 54n.11; Moo, Romans, 200–1.
62  Jewett argues that “the most serious weakness with this popular solution is that the dia-
tribal logic is thereby destroyed, because the Jewish interlocutor could hardly think of
himself sharing an advantage after the condemnation of … the result of the entire preced-
ing diatribe in vv. 1–8.” Romans, 257. Jewett’s judgment is based on his traditional script, in
which it would be ridiculous for the interlocutor to pose such a question. Jewett’s script,
however, handicaps his keen observation; Jewett does not realize the same argument sup-
ports a rescripted dialogue in which the interlocutor does not speak 3:9a but the rejection
in 3:9b.
63  Thus I affirm the scripts of Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell,
572–73; but not Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans,
89; Witherington, Romans, 93. Byrne ceases consideration of the dialogue at 3:8, so Paul
speaks all of 3:9 in his analysis. Byrne, Romans, 119. Furthermore, a rescriptive reading
resolves most of the problems raised by critics like Hall (see Chapter 9). Hall, “Romans
3.1–9,” 183–92.
64  In its diatribal context, οὐ πάντως carries the same force as μὴ γένοιτο. Jewett, Romans, 257;
Song, Reading, 96. Contra Porter, Romans, 88.
268 chapter 11

undergoes development. Though he originally endorses Jewish advantage over

non-Jews as characterized, by the end of the discourse Paul’s leading questions
guide him to forego such advantage, agree with Paul, and make Paul’s conclud-
ing point for him65—“Jews have no salvific advantage over non-Jews; all are
equal.”66 Speech-in-character’s single characterization, therefore, is effectively
supplemented by diatribe’s necessary double-characterization.

Romans 3:1–9 Rescripted

Analyzing the lines in the dialogue in view of Paul’s characterization of the in-
terlocutor demonstrates that certain lines are appropriate for one speaker but
inappropriate for another (e.g., 3:2, 3, 5, 7–8c, 9a). It also shows that some lines
can be appropriate for either speaker (3:1, 4, 6, 8d, 9b). In these latter instances,
the opposite lines in the exchange, overlaid with diatribe’s Socratic method,
help indicate the proper arrangement of the script. For instance, 3:1 is best
heard in one voice because the other most appropriately speaks 3:2’s response;
or, that one speaker can appropriately speak both 3:3 and 4 is presupposed by
the grammatical formulation of 3:3 with μή. When the dust settles, the con-
ventions of speech-in-character and diatribe suggest that Paul plays the part
of Socratic questioner throughout the dialogue. Paul poses leading questions
in 3:1, 3, 5, 7–8c, and 9a for the interlocutor to answer in 3:2, 4, 6, 8d, and 9b,
respectively. As a historical-critical and methodological tool, the convention of
appropriateness in speech-in-character not only opposes traditional readings
but the rescriptive reading of Stowers, Keck, and Witherington too, since their
script is identical to traditional scripts at the beginning and end.67 Speech-in-
character also provides the argumentation to establish the script of Elliott and
Campbell (as well as the script of Byrne to 3:8 and the interior of the script of
Stowers, Keck, and Witherington).68 Charted, the various readings appear as
follows in Table 11.1:

65  On making Paul’s point for him, see Elliott, Rhetoric, 139; idem., Arrogance, 106; Campbell,
Deliverance, 574–76.
66  Though he distances himself from 3:9 as dialogue, Byrne agrees, “as Paul has maintained
all through, the law does not give the Jews any advantage” in terms of eschatological jus-
tification. “God is impartial.” Byrne, Romans, 116.
67  Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans, 89; Witherington,
Romans, 93.
68  Elliott, Rhetoric, 139–41; idem., Arrogance, 105–6; Campbell, 572–73; Byrne, Romans, 106–
14, excluding 3:9; Stowers, Rereading, 165–66; idem., “Paul’s Dialogue,” 715; Keck, Romans,
89; Witherington, Romans, 93.
Romans 3:1-9 and the Argument of Romans 269

table 11.1 Arrangements of the script of Rom 3:1–9 compared to my arrangement informed by
speech-in-character

Line in script Traditional Stowers Elliott King

3:1 Interlocutor Interlocutor Paul Paul

3:2 Paul Paul Interlocutor Interlocutor
3:3 Interlocutor Paul Paul Paul
3:4 Paul Interlocutor Interlocutor Interlocutor
3:5a Interlocutor Paul Paul Paul
(3:5b) Paul; authorial aside
3:6 Paul Interlocutor Interlocutor Interlocutor
3:7 Interlocutor Paul Paul Paul
3:8a
(3:8b) Paul; authorial aside
3:8c Interlocutor
3:8d Paul Interlocutor Interlocutor
3:9a Interlocutor Interlocutor Paul Paul
3:9b Paul Paul Interlocutor Interlocutor

This analysis, therefore, suggests the most appropriately scripted dialogue of

3:1–9 reads as follows:

Paul’s Guiding Question: Therefore, what advantage comes from being

Jewish, or what benefit accrues from circumcision? (3:1)
Interlocutor’s Response: There is much [advantage] in every way! To
begin, they were entrusted with the oracles of God. (3:2)

Paul’s Guiding Question: To what end? If some lacked πίστις, their lack of
πίστις will not nullify God’s πίστις, will it? (3:3)
Interlocutor’s Response: Absolutely not! Instead, let God be true but
every human a liar, as it is written, “So that you might be justified in
your words, and you will overcome when you are judged.” (3:4)

Paul’s Guiding Question: But if our unrighteousness proves God’s righ-

teousness, what shall we say? God is not unjust when he brings wrath, is
he? I speak in a human way. (3:5)
Interlocutor’s Response: Absolutely not! Otherwise, how will God
judge the world? (3:6)
270 chapter 11

Paul’s Guiding Question: But if God’s truthfulness is increased for his glory
by my lie, why am I still being judged as a sinner? Why not, as we are slan-
dered and as some claim that we say, “Let us do evil so that good might
come?” (3:7–8c)
Interlocutor’s Response: Their judgment is justly deserved. (3:8d)

Paul’s Guiding Question: What then? Are we advantaged or disadvan-

taged? (3:9a)
Interlocutor’s Response: By no means! (3:9b)

Paul’s Conclusion: For we have charged that both Jews and Greeks are all
under Sin, (3:9c) as it is written …

When viewed as a whole, additional evidence surfaces in support of this re-

scription. Paul is responsible for all the first-person speech. The texts tradition-
ally understood as authorial asides fit seamlessly in the flow of the dialogue
rather than being awkwardly interruptive. With the exception of the citation
of LXX Psalm 50:6, the interlocutor speaks entirely in the third person (until
3:27–31).69 Furthermore, Paul’s use of the first person in 3:1–9 mirrors his usage
of the first person elsewhere, especially with the phrase, τί οὖν ἐροῦμεν (4:1; 6:1;
7:6a, 13a).70
This brings us to several other questions about speech-in-character and
diatribe in 3:1–9. First, concerning speech-in-character, just how well does
Rom 3:1–9 fit with the bigger picture of the exercise in the primary literature
(see Part 1)? Paul’s use of speech-in-character in 3:2, 4, 6, 8d, and 9b71 mirrors
Quintilian’s use of Vergil’s Aeneid (Quintilian, Inst. 9.2.37; Vergil, Aen. 2.29).
Quintilian references Vergil as an example of speech-in-character without any
identification of the speaker. Recall, however, that Quintilian’s example actual-
ly demonstrates another rule; speech-in-character can occur without any overt
indication that another identified character actually begins to speak. Vergil

69  This script, therefore, makes even more sense of the pronouns than does Stowers’s, who
assigns first-person speech to both the interlocutor and Paul (cf. 3:9). Stowers, Rereading,
174–75.
70  This does not include the use of the first person in 7:7b–12, 13b–24, 25b, since these verses
are not spoken in Paul’s voice. Stowers, Rereading, 264–69; idem., “Romans 7,” 191–202;
Longenecker, Rhetoric, 88–93; King, “Rhetorical Chain-Link Construction.”
71  Again, not all of 3:1–9 is speech-in-character, contra Stowers. Only those verses where
Paul attributes speech to the interlocutor qualify. See the discussion of Stowers in
Chapter 9.
Romans 3:1-9 and the Argument of Romans 271

identifies and characterizes the Trojans on the scene, but no verb of speech (or
similar) marks the interjection of their wonder at the abandoned Greek camps.
Rom 2–3 is the same. Paul identifies his ethnically-unspecified, religiously-
Jewish interlocutor and characterizes him as judgmental, presumptuously ad-
vantaged, and dependent on but transgressor of νόμος (Rom 2). Then, without
any verbs of speech or overt textual markers, Paul simply attributes speech
to the interlocutor. The convention of appropriateness to characterization
is the only measure by which to determine that another speaker responds
and which lines he/she is responsible for speaking. Furthermore, Rom 2–3
follows the general trend for speech-in-character to progress in three steps:
(1) identification, (2) characterization, and (3) attributed speech-in-character.
Second, though it should be clear above, how does 3:1–9 cohere with diatribe’s
chief function of censure and protreptic? In diatribe, censure discloses a
student’s or interlocutor’s contradictions or errors, and protreptic guides her/
him to endorse the correct view. These features are readily identifiable in
Romans. Paul’s use of censure even begins in the characterization;72 Paul notes
that the interlocutor rests and boasts in νόμος but nevertheless transgresses it
(2:17–23). This continues in 3:1–9; the interlocutor endorses Jewish advantage,
but Paul demonstrates how affirming God’s faithfulness despite Jewish ἀπιστία
(3:3–4), and God’s judgment and δικαιοσύνη despite humanity’s ἀδικία (3:5–6),
ultimately contradicts Jewish advantage over others given the inclusive nature
of the divine promises contained in the λόγια. Thus, Paul censors the inter-
locutor’s contradictory views, but he also protreptically guides the interlocutor
in the right direction. Paul’s questions in 3:3, 5, 7–8c gradually lead the inter-
locutor closer to Paul’s view, so that the interlocutor is able to recognize his
error and join Paul’s side in 3:8d and 9b (see also 3:27–31). In 3:1–9, censure and
protreptic run their full course, dislodging the interlocutor from his previous
endorsement and winning him over for Paul’s view.
Considering censure and protreptic thusly gives rise to a third set of ques-
tions. What role does the imaginary interlocutor play for Paul’s historical
audience? How would the progymnasmatists and rhetoricians featured in
Parts 1 and 2 discuss Rom 3:1–9 in terms of speech-in-character’s functions?
These questions are naturally connected to even larger questions, such as
the purpose(s) of Romans, to which the final section of this chapter attends.
As argued above, Paul’s non-Christian, religiously-Jewish, presumptuously-
privileged interlocutor does not directly represent his Christian Roman

72  It is even arguable that censure begins in 1:18–32. Humanity is critiqued for behav-
ior being contradictory to knowledge (1:18–32), and the interlocutor “does the same
things” (2:1).
272 chapter 11

audience (though affinities appear in discussion of 9–11 and 12–15). Thus, 3:1–9
does not simply tell the Romans to stop holding νόμος or λόγια as advantageous
over non-Jews; that is not the point of the dialogue. Still, Paul communicates
something to the Romans through the interlocutor, and this directly relates to
the portryal of the gospel he develops beginning in 1:1. To repeat, the gospel is
God’s anthropologically universal and Christological means of demonstrating
his active δικαιοσύνη on the basis of πίστις. This gospel is the necessary solution
because God’s ὀργή is also being unleashed against all human unrighteousness.
The dialogue in 3:1–9 fits into Paul’s argument by showing that Jews too fall
under God’s impartial judgment—even if some suppose to escape it—and are
equally in need of God’s initiative in the gospel. Nobody has any salvific ad-
vantage over another before a just and impartial God, and Paul enlists the very
person who might (does) presume to have an advantage over others to make
this concluding point in the argument, someone religiously Jewish.73
Discussing the role of the interlocutor in diatribal terms, however, only ad-
dresses one side of a two-headed coin; the technical treatments of speech-
in-character also speak in this vein. Most notably, though he receives little
attention in Part 1 due to the brevity of his comments, Cicero writes about
speech-in-character in de Inventione, “At times you can sum up in your own
person … but at other times you can bring on the stage (inducere) some person
or thing (personaram aut rem aliquam) and let this actor sum up the whole
argument” (1.99 [Hubbell, LCL]). Quintilian would speak of 3:1–9 as a way to
introduce conversations between oneself and others in a credible manner,
and as a way “to provide appropriate characters for words of advice, reproach,
complaint, praise, or pity” (Inst. 9.2.30 [Russell, LCL]). Quintilian even
demonstrates how Cicero employs speech-in-character to set up characters
for certain rhetorical functions, specifically rebuke and encouragement, which
approximates diatribe’s use of censure and protreptic. Quintilian writes,
“Cicero in Pro Caelio makes Appius Caecus and Clodius, her brother, address
Clodia, the one to rebuke her vices, the other to encourage them” (Inst. 3.8.54
[Russell, LCL]). Finally, Theon would speak of Rom 3:1–9 as belonging to the
protreptic species of rhetoric (εἶδος… προτρεπτικῶν; Prog. 115.20–22; P 70), not-
ing how Paul exhorts the interlocutor towards a particular view. Diatribe and
speech-in-character, therefore, overlap considerably and can speak of the in-
terlocutor’s role in Rom 3:1–9 practically synonymously.

73  On this point, Elliott, Rhetoric, 141, writes, “The argument in 2.17–3.9 is not nakedly that
‘Not even the Jews have an excuse’, but that ‘Since not even the Jews have an excuse, no
one does.’”
Romans 3:1-9 and the Argument of Romans 273

Finally, fourth, what tone does Paul take with his audience and the inter-
locutor? Does Paul polemicize his religiously-Jewish interlocutor as most pre-
Stowers readings argue, or does he educate or collaborate with him as some
post-Stowers arguments adopt? Paul’s attitude towards his historical audi-
ence is easier to pinpoint. Because the interlocutor does not directly represent
the audience but communicates to them a more general or bigger premise in
the argument, Paul’s tone towards them is more educational than anything
else. In Rom 11 and 12–15, Paul shows his audience that there is more to the
interlocutor than meets the eye (still not polemically), but at the present he
only aims to illustrate the equality of all humanity and to validate the inclusive
gospel through which his impartial God procures salvation for all.
Paul’s tone with the interlocutor is a bit more difficult to deduce. According
to a traditional script, it is easier to hear polemic in Paul’s voice, as he curt-
ly rejects the interlocutor’s presumptuous objections time and time again.74
In a revised script, especially one that thoroughly situates Paul as Socratic
questioner, this is not the case. Instead, after pointing out and censoring the
interlocutor’s error (2:1–29), Paul simply asks questions for the interlocutor to
consider in light of some occasion (i.e., what if some Jews lacked πίστις). Paul
does not critique the interlocutor directly in 3:1–9 (though he does in 2:1–5,
23–29), nor does he call him “foolish,” an “idiot,” or the like. True, Paul guides
the interlocutor to answer in a certain way, but he refrains from attacking the
interlocutor for answering on the basis of erroneous views about λόγια / νόμος,
much less for being “Jewish.” After all, the thrust of 3:1–9 is not so much to
single out Jews as it is to make a point about all (3:4, 6, 9b–31).75 After the inter-
locutor responds, Paul just moves on to the next occasioned question until the
interlocutor connects the dots for himself between Paul’s previous discussion
(Rom 1–2), Paul’s leading questions (3:3, 5, 7–8c), and his own contradictory
views (3:2, 4, 6). Once connected, the interlocutor joins with Paul and issues
the concluding statement, “Nobody has any salvific advantage or disadvan-
tage.” In this sense, Paul’s tone with the interlocutor is primarily collaborative;
through Paul’s guidance, the two discussion partners ultimately work together
to make the intended point in Paul’s argument for his Roman audience.76

74  See, for instance, Käsemann, Romans, 68–85. One of the primary impasses between
scholars is the unfortunate fact that a consensus or standard definition of “polemic” does
not exist. What some might call polemic, others might simply call censure, and vice versa.
75  On the indirectness of what Luke Timothy Johnson calls “NT polemic against the messian-
ists’ fellow Jews,” see Luke Timothy Johnson, “The New Testament’s Anti-Jewish Slander
and the Conventions of Ancient Polemic,” JBL 108.3 (1989): 426.
76  Cf. Elliott, Rhetoric, 139; idem., Arrogance, 106; Tobin, Paul’s Rhetoric, 120.
274 chapter 11

As a result, it is not only the convention of appropriateness to characteriza-

tion that helps elucidate Rom 3:1–9. Diatribe and speech-in-character contain
reasonable explanations for the various additional features of the appropri-
ately scripted dialogue as well, such as the role of the interlocutor, Paul’s tone,
and the rhetorical function of the pericope. But 3:1–9 is only one piece of the
Romans puzzle; decisions made about the script of the dialogue and its mean-
ing carry significant weight for the argument of Romans as a whole. This is why
one must ask both “who is speaking when” and “why does it matter?” I turn
now to examine what impact this reading has on the bigger picture of Romans
and Paul’s thought.

Romans 3:1–9 and the Remainder of Romans

This final section considers how the proposed reading of 3:1–9 fits within and
effects the bigger picture of Romans on three issues in the study of the letter
and Paul’s thought: (1) the reoccurrence of questions and themes in Romans;
(2) the concepts of “advantage” and God’s promises as they relate to Jews and
non-Jews in Paul’s thought; and (3) the purpose(s) of Romans.

The Reoccurrence of Questions and Themes

Virtually all of the questions and themes raised in 3:1–9 reappear at some point
in Romans. Though this issue is not as massive as questions regarding God’s
faithfulness to his promises or the Romans debate, it plays an important role.
Many scholars argue that 3:1–9 raises issues briefly in order for Paul to address
them later in more detail, and such scholars are on the right track regarding
how the reoccurrences function.77 What remains to be demonstrated is to
show how these reoccurrences confirm this project’s reading of Rom 1–3, es-
pecially 3:1–9.78
Romans 3:10–31. Rom 3:10–31 is an obvious point of departure, as it summa-
rizes and confirms the ethnic impartiality and exhaustiveness of God’s ὀργή
and δικαιοσύνη in 1:1–3:9.79 Following 3:9’s determination that all—Jews and
Greeks—are under Sin, the catena and conclusion in 3:10–20 further confirm

77  Cf. William S. Campbell, “Romans III as a Key to the Structure and Thought of the Letter,”
NovT 23.1 (1983): 22–40.
78  These are not necessarily exhaustive treatments tracing the reoccurrences of themes
throughout Romans.
79  As Bassler argues, “‘No distinction’ applies both ways” regarding God’s judgment and
δικαιοσύνη. Bassler, Divine Impartiality, 156.
Romans 3:1-9 and the Argument of Romans 275

this point. “No one is just (οὐκ ἔστιν δίκαιος), not even one” (οὐδὲ εἷς; 3:10), “no
one” understands or seeks for God (3:11), “all turned away” (πάντες; 3:12), “no one
does kindness (οὐκ ἔστιν ὁ ποιῶν χρηστότητα), not even one” ([οὐκ ἔστιν] ἕως ἑνός;
3:12). What νόμος speaks to those under it, it speaks “so that every mouth might
shut and all the world might be accountable to God” (ἵνα πᾶν στόμα φραγῇ καὶ
ὑπόδικος γένηται πᾶς ὁ κόσμος τῷ θεῷ; 3:19). And, “From works of νόμος no person
at all will be justified before [God]” (ἐξ ἔργων νόμου οὐ δικαιωθήσεται πᾶσα σὰρξ
ἐνώπιον αὐτοῦ; 3:20). There is no talk of religion; there is no talk of ethnicity.
There is only talk of all.
The same is true concerning God’s δικαιοσύνη in 3:21–26. God’s active
δικαιοσύνη does not operate on the basis of Jewishness or νόμος possession
and observance, but “διὰ πίστεως Ἰησοῦ Χριστοῦ for all characterized by πίστις”
(εἰς πάντας τοὺς πιστεύοντας; 3:21–22). God’s salvific program operates in this
way for all because “there is no distinction” (οὐ γάρ ἐστιν διαστολή; 3:22); “all
sinned and lack God’s glory” (πάντες γὰρ ἥμαρτον καὶ ὑστεροῦνται τῆς δόξης τοῦ
θεοῦ; 3:23; cf. 5:12–14) and as a result “are being justified (δικαιούμενοι) freely
by [God’s] grace through the redemption available in Christ Jesus” (3:24; cf.
5:15–17).80 Furthermore, God’s act of sending Jesus on his salvific mission for all
is a manifestation (εἰς / πρός ἔνδειξιν) of God’s active δικαιοσύνη in the present
time (3:25–26). Such equality for all proves God to be just (δίκαιον) and justi-
fier (δικαιοῦντα) of those characterized by Jesus’ πίστις (3:26). Again, in keeping
with Paul’s gospel, there is only talk of all.81
Rom 3:27–31 resumes the dialogue and concludes chapters 1–3.82 Paul
asks the previously boastful interlocutor (2:17, 23), in light of our discussion
and your confession, “Where, then, is boasting?” The interlocutor must an-
swer, “It is excluded;” there is none. Paul continues, “Through what sort of
νόμος was boasting excluded? Through a νόμος characterized by works (τῶν
ἔργων)?” The interlocutor, “No, it is excluded through a νόμος characterized by
[Christological] πίστις (διὰ νόμου πίστεως),” like you have been saying (3:27; cf.
1:1–5; 2:12–16).83 Paul explains, “Good, for we consider [people] to be justified

80  Thus, despite his ethnic categories for 1:18–32 and 2:1–3:8, Fitzmyer is on the right track in
arguing that 1:18–3:20 depicts God’s “reaction to humanity without the gospel.” Fitzmyer,
Romans, 270.
81  See Barclay, Gift, 476, 476n.65.
82  Distributing the lines of 3:27–31’s dialogue is impossible in terms of speech-in-character’s
conventions, since Paul and his interlocutor seem to be in complete agreement here, be-
ginning with 3:8d and 9b. For this reason, I maintain the pattern of 3:1–9 with Paul as
Socratic questioner. Cf. Stowers, Rereading, 231–37.
83  Here, the νόμος under discussion is one and the same. The distinction Paul is making
involves how the Mosaic νόμος is fulfilled.
276 chapter 11

on the basis of πίστις and not works of νόμος” (λογιζόμεθα γὰρ δικαιοῦσθαι πίστει
ἄνθρωπον χωρὶς ἔργων νόμου; 3:28),84 and he asks, “Or (ἤ), is God only over Jews
(Ἰουδαίων ὁ θεὸς μόνον)? Is he not also over non-Jews (οὐχὶ καὶ ἐθνῶν; 3:29)?” As ex-
pected with οὐχί in an interrogative,85 the interlocutor responds, “Yes, he is also
over non-Jews, since God is One who will justify the circumcision on the basis
of πίστις and the uncircumcision through πίστις” (3:29). In the final exchange,
Paul asks, “Therefore, do we nullify νόμος through πίστις?” (3:31; cf. 3:8a–c). To
this, the previously νόμος-advantaged interlocutor concludes by proclaiming
the fullness of Paul’s argument; “Absolutely not (μὴ γένοιτο)! Justification on
the basis of πίστις is what really upholds νόμος,” as you have argued (3:31; cf.
1:1–5, 16–17; 2:12–16). Thus, 3:10–31 supplements the cosmic scope of God’s ὀργή
and δικαιοσύνη, upholds the primacy of πίστις, and once again allows the inter-
locutor to supply the concluding statement to Paul’s argument.
Romans 4: God’s faithfulness and the λόγια. Rom 3:3–4 argues that God will be
faithful to the λόγια despite Jewish ἀπιστία. God’s λόγια contain the ethnically
exhaustive promises to Abraham that God is bringing to fruition through the
πίστις-based gospel. Paul advances this theme in Rom 4.86 Paul argues that
Abraham was justified on the basis of πίστις (4:3–5, 22). David even speaks
about how blessed people are to whom God reckons justification apart from
works (λογίζεται δικαιοσύνην χωρὶς ἔργων; 4:6–8). Abraham received this bless-
ing while he was uncircumcised (4:9–10). As a result, Abraham became the fa-
ther of all people characterized by πίστις, both uncircumcised (πατέρα πάντων
τῶν πιστευόντων δι’ ἀκροβυστίας) and circumcised (πατέρα περιτομῆς; 4:11–12).
Given God’s promise (ἐπαγγελία) that Abraham and his seed would “be the heir
of the world” (τὸ κληρονόμον αὐτὸν εἶναι κόσμου), the promise cannot be contin-
gent on exclusive possession of νόμος but on δικαιοσύνη πίστεως (4:13), because
only a promise based on πίστις and χάρις can be effective (βεβαίαν) for all the
seed (παντὶ τῷ σπέρματι; 4:16).87 Said otherwise, only a promise based on divine
πίστις and χάρις addresses the universal scope of God’s redemtpive program.
Accordingly, God made Abraham “the father of us all” (πατὴρ πάντων ἡμῶν;

84  Rom 3:28 could also make sense as an explanation from the interlocutor, siding with
Paul’s view of justification. My decision to place it in Paul’s voice is based on the phrasing
of 3:29 with ἤ, which seems to presuppose a comment from the same speaker.
85  BDF §427.
86  Note that Rom 4 is also dialogical, though I pass over it now in the interest of brevity. I aim
to return to the dialogical aspect of Romans beyond Rom 3 in a future project.
87  Cf. Cranfield, Romans, 242; Kruse, Romans, 214–15; Barrett, Romans, 90; Jewett, Romans,
329–30.
Romans 3:1-9 and the Argument of Romans 277

4:16) and “a father of many nations” (πατέρα πολλῶν ἐθνῶν τέθεικά σε; 4:17).88
Furthermore, God continues to reckon this justification and blessing to those
characterized by πίστις (4:22–25). Paul’s discussion of the ἐπαγγελία (λόγια) in
Rom 4, then, is in perfect harmony with his definition of the gospel and this
investigation’s explanation of the dialogue in 3:1–9.
Romans 6: Shall we continue in Sin? In 3:8a–c, Paul asks whether it would be
better to do evil so that good might come, since God’s “truth” is increased by
human falsehood (cf. 3:4). The implied answer (μή) to the absurd false conclu-
sion is, “Absolutely not (μὴ γένοιτο)!” The interlocutor, however, does not an-
swer the question; instead, he comments about those who slander Paul, hold
out νόμος as an exclusive advantage, and hinder God’s inclusive λόγια (3:8d).
Paul, however, returns to the question in chapter 6. Like 3:4, 5a, and 7, Rom
5:20 declares, “Where Sin increased, grace super-abounded” (οὗ δὲ ἐπλεόνασεν
ἡ ἁμαρτία, ὑπερεπερίσσευσεν ἡ χάρις). Accordingly, like 3:5b and 8a–c, Rom 6:1
asks, “What, then, will we say?89 Shall we remain in Sin, so that grace might
increase?” (ἐπιμένωμεν τῇ ἁμαρτίᾳ, ἵνα ἡ χάρις πλεονάσῃ;). Though the inter-
locutor does not supply the obvious answer in 3:8d, Rom 6:2 does, μὴ γένοιτο!
Christians cannot continue in Sin because Sin is in league with Death (5:12–14).
Christians, however, are “baptized into [Jesus’] death,” and, just as “Jesus was
raised from the dead,” so also Christians “walk in new life” (6:3–4). Christians
“are not enslaved to Sin” (6:6) but in Christ are “dead to Sin” and “alive to God”
(6:11). Thus, “Sin will not rule as lord (ἁμαρτία … οὐ κυριεύσει)” over Christians
(6:14), for “[they] are not under Sin but grace” (6:14).
This latter affirmation gives rise to another formulation of the question.
“What then? Shall we sin, because we are not under νόμος but under grace?”
(6:15). The answer is another resounding μὴ γένοιτο (6:15). According to Paul,
the anthropological condition is such that humans are “slaves” (δοῦλοι) to
whomever they obey (6:16). In Christ, however, God frees Christians from slavery
to Sin and “enslaves” them to δικαιοσύνη and himself, so that they serve him
(6:17–23). On all accounts, the response to continuing in Sin is a reverberating
“absolutely not!” Humans can only have one L(l)ord; for a Christian to continue
serving Sin is a contradiction of terms in Paul’s thought (cf. 8:9–11).
Additional elements from 3:1–9 reappear later in Romans. For instance, 3:5’s
question of divine ἀδικία returns in 9:14, and 3:7’s imaginative consideration
of why God finds fault quickly follows in 9:19. Paul couches these questions,
however, within his narrower discussion of God, Israel, and the gentiles in
Rom 9–11. Because of the vast importance decisions about 3:1–9 have on these

88  Cf. Matera, Romans, 118–20.

89  Note the dialogical nature of Rom 6, too.
278 chapter 11

latter matters, and because of the vast importance these latter matters have on
Romans as a whole, they require their own section.

God’s Inclusive and Universal Promises: Rom 3:1–9 and 9–11

The biggest “advantage” of traditional readings of 3:1–9 in which Paul overtly
supports Jewish advantage of some sort is that explaining Paul’s discussion of
Israel in 9–11 becomes easier—not easy, easier—but not necessarily better.
Nevertheless, the best traditional readings still have to wrestle with how to
explain Paul’s ceaseless appeal to divine impartiality with his endorsement of
Jewish advantage over non-Jews. For example, Longenecker argues that Paul
himself wrestles with this tension throughout Romans 1–11. Beginning in 1:16,
Paul qualifies “the universalism of his gospel” by noting that it is “to the Jew
first,” whereby “we can already notice a peculiar tension in his perspective, a
tension which includes an ethnic advantage or priority within his universalis-
tic outlook.”90 Paul resolves this issue in 9–11, where he draws a distinction be-
tween the ages.91 “In the present age, ethnic lineage has no part in determining
the membership of the community of grace; with the culmination of this age,
however, all Israel will be turned to faith in Christ.”92 By itself, Longenecker’s
conclusion is fine and plays a significant part in the following discussion. What
such readings miss, however, is the degree to which God does nothing for Israel
that he does not equally do for non-Israel. God does not show mercy only to
ethnic Israel, but to all (πάντας; 11:32) and to all fully conceived (πλήρωμα; 11:12,
25). Thus, if one must use the language of “advantage,” rather than limiting
eschatological redemptive benefits to Israel over non-Israel, it would be more
accurate to say that all are “advantaged” or “privileged” in light of God’s πίστις
and his promises’ universal scope.
Allowing speech-in-character’s convention of appropriateness to guide
interpretation of Rom 3:1–9 helps make this point more clear, since it requires
the νόμος-boastful interlocutor to proclaim the affirmation of Jewish advantage
over non-Jews in 3:2, not Paul. Such an affirmation would conflict with the eth-
nic vagueness and incessant inclusivity to which Paul appeals in Rom 1–3, but it
constitutes the very identity of the interlocutor. Paul, therefore, never suggests
that Jews have any salvific edge over non-Jews. But what does this say about
Paul’s conception of God’s relationship to the promises? Λέγω οὖν, μὴ ἀπώσατο
Παῦλος τὴν διαθήκην τοῦ θεοῦ; Μὴ γένοιτο! Rather, Paul appeals to the impartial
and universal core of God’s promises and presses them to their extremes.

90  Longenecker, Eschatology, 168, italics original.

91  Ibid., 168, 195–99, 256–57.
92  Ibid., 257.
Romans 3:1-9 and the Argument of Romans 279

This conversation of course comes to a head in Rom 9–11. In Rom 3, Paul

can only guide his interlocutor to dismiss Jewish salvific advantage in the in-
terest of equality; in 9:1–5, however, Paul speaks quite positively about what
we might call “ethnic” Israel. Paul lists a number of would-be “advantages” for
Israel, which include: possessing adoption (ὧν υἱοθεσία), the glory/presence (ἡ
δόξα), the covenants (αἱ διαθῆκαι), the giving of the law (ἡ νομοθεσία), the min-
istry (ἡ λατρεία), the promises (αἱ ἐπαγγελίαι), and the ancestors (οἱ πατέρες;
9:4). Furthermore, the Messiah genealogically descends from Israel (ἐξ ὧν ὁ
Χριστός; 9:5).93
Alas, Israel’s possession of these would-be “advantages” does not make one
iota of difference for them over against non-Israel. At the time of writing, Paul
believes unbelieving Israel is in dire straits, so much so that he expresses his
great grief for them (9:1–3) and feels compelled to address the question of
whether God’s word has failed (9:6). Something is definitely amiss. Yet, in his
discussion of the problem, Paul never says, “Don’t worry, you possess νόμος,” or
“everything will be okay, you serve in the temple,” much less “God deals with
Israel differently than the rest of the nations.”94 Nor is it important that Israel
was chronologically privy to the λόγια / νόμος with its διαθῆκαι and ἐπαγγελίαι.
What matters is what God does to Israel,95 but what God does to Israel, he does
equally to all.96 Consequently, in Rom 9–11’s discussion of Israel, Paul demon-
strates over and over how God deals with Israel and non-Israel on the same
basis, as the following discussions illustrate.
Romans 9. Given Israel’s state of affairs, Paul posits whether God’s word
has failed, or whether God has been unfaithful, contra 3:3–4. The answer is of
course, “no,” and Paul draws out the scope of God’s “word” (ὁ λόγος; 9:6) in two

93  Given the history of interpretation, it is striking that Paul does not actually speak about
9:4–5 as περισσόν or ὠφέλεια, nor does he include the λόγια in this list.
94  I.e., there are no “two ways.” Longenecker, Eschatology, 257n.1.
95  Fitzmyer, therefore, advances the right theology in the wrong place when he argues that
“Jewish advantage” rests on what God does for Jews but applies this view to 3:2, since
presumed advantage in 2:1–3:2 involves exclusivist possession rather than the universal
applicability of the λόγια as Paul insists. Fitzmyer, Romans, 325.
96  This does not suggest that “Israel” in Romans equals “all.” Instead, Israel is one constitu-
ent within the “all.” Contra Wright, Romans, 690–93; idem, The Climax of the Covenant:
Christ and the Law in Pauline Theology (Minneapolis: Fortress Press, 1993), 250. So also
Rodríguez, If You Call, 226; Dunn, Theology, 527, whose opposing argument is based on
11:28–29; Beker, “The Faithfulness of God and the Priority of Israel in Paul’s Letter to the
Romans,” HTR 79.1–3 (1986): 10–16; Ticciatti, “Nondivisive,” 266, who argues, “At the escha-
tological climax … difference remains without division.”
280 chapter 11

ways.97 First, Paul argues that God’s word is only applicable to the “children of
the promise” (τὰ τέκνα τῆς ἐπαγγελίας; 9:8). It is not mere genealogical descent
from Abraham that constitutes his “seed,” since God promises “a seed (σπέρμα)
will be called [for him] in Isaac” (9:7; Gen 21:12). Paul’s discussion of Abraham
helps explain what he means in the infamously cryptic comment about “not
all these from Israel are Israel” (Rom 9:6b); just as Abraham’s seed is limited,
neither do all of Israel’s descendants constitute God’s Israel (9:6).98 Second,
Paul illustrates that God elects impartially. Before Isaac and Rebekah’s chil-
dren do anything good, evil, or otherwise, God chooses to love Jacob, not Esau
(9:11–13). God’s choice of Jacob maintains the “election” (ἐκλογήν; 9:11) aspect of
his plan, so that what matters for belonging to God’s people is not one’s “works”
(ἐξ ἔργων) but God’s call (ἐκ τοῦ καλοῦντος; 9:12).99 Thus, God’s word has not
failed; it has divinely selective applicability in accord with God’s promise, elec-
tion, and call.
God’s impartiality, however, sparks the reoccurrence of 3:5’s question about
divine ἀδικία. Given that human ἀδικία makes room for God’s active δικαιοσύνη
(3:5), Paul’s leading question in 3:5 asks whether God is “unjust” (ἄδικος) for
exerting wrath. Similarly, 9:14 asks, “What, then, shall we say? There is no injus-
tice with God, is there?” (τί οὖν ἐροῦμεν; μὴ ἀδικία παρὰ τῷ θεῷ;).100 The question
asks whether God is unjust for impartially choosing some but not others. As
in 3:6, 9:14’s answer is, “Absolutely not” (μὴ γένοιτο)! God’s impartial election
is part and parcel in keeping with his character; he even informs Israel that
he will show mercy and compassion to whomever he wishes (9:15; Ex 33:19).
Consequently, belonging to God’s people does not depend on the human who
wills or runs, but on God’s preemptive mercy (τοῦ ἐλεῶντος θεοῦ; Rom 9:16), for
he hardens or shows mercy to whomever he wishes (9:17–18).
This depiction of God as orchestrator of both election and non-election
prompts another question from Rom 3:1–9. Rom 3:7 imaginatively considers
why Paul is considered a sinner during his period of “falsehood” (i.e., πίστις-
lessness), since that falsehood magnifies God’s “truth” (i.e., faithfulness). In an
interlocutor’s voice (ἐρεῖς μοι), Rom 9:19 asks, if God is responsible for harden-
ing humanity, “Why does he still find fault? Who has opposed his will?” Paul
is more blunt here than in 3:1–9; Paul essentially answers, “God is God and

97  Similarly, Longenecker, Eschatology, 253.

98  On this latter point, see Beverly Roberts Gaventa, “On the Calling into Being of Israel:
Romans 9:6–29,” in Between Gospel and Election (WUNT 257; Florian Wilk and J. Ross
Wagner, eds.; Tübingen: Mohr Siebeck, 2010), 255–69.
99  Thus, “election” and “call” are roughly synonymous.
100  Note the dialogical style.
Romans 3:1-9 and the Argument of Romans 281

who are you to question his decisions?” (9:20–21). God can demonstrate ὀργή
on some and ἔλεος on others prepared for this purpose, namely, “us whom
he called not only from Jews but also from non-Jews” (οὓς καὶ ἐκάλεσεν ἡμᾶς
οὐ μόνον ἐξ Ἰουδαίων ἀλλὰ καὶ ἐξ ἐθνῶν; 9:24). Again, what matters—for Jews
and non-Jews—is nothing short of God’s call, and Paul documents this from
scripture. Concerning God’s call of non-Jews, Paul appeals to Hosea, “I will call
(καλέσω) not my people (τὸν οὐ λαόν μου) my people, and not my beloved (τὴν
οὐκ ἠγαπημένην) beloved … they will be called (ἐκληθήσονται) children of the
living God (υἱοὶ θεοῦ ζῶντος; 9:25–26; LXX Hos 2).”101 At the same time, Isaiah
proclaims concerning Israel that though most perish, God will preserve a
remnant (Rom 9:27–29; Isa 1:9; 10:22; cf. Hos 2:1 LXX). Once more, divine initia-
tive trumps ethnicity and any other presumed advantage.
In drawing Rom 9 to a close, 9:30–33 allows for additional discussion of how
God interacts identically with Israel and non-Israel. Rom 9:30 asks, if God “calls
not my people my people” (9:25) and only saves a remnant of Israel (9:27–29),
“What, then, will we say?” The answer is quite reminiscent of 2:12–16;102 the na-
tions (ἔθνη) not pursuing δικαιοσύνη obtain it on the basis of πίστις, but Israel’s
pursuit of “law-righteousness” (νόμον δικαιοσύνης) fails because their pursuit
is not based on πίστις but ἔργα (9:30–31). In this respect, Israel stumbles be-
cause they do not understand God’s δικαιοσύνη. This, however, is not primarily
Israel’s fault. God places the stumbling and scandalizing stone in Israel’s path,
over which safe passage requires divinely graced πίστις. God, however, does
not grace to (most of) Israel a vehicle with appropriate ground clearance—
πίστις—at least not yet. Thus, God orchestrates Israel’s crash (9:32–33; Isa 8:14;
28:16).103
Rom 9, therefore, argues that inclusion in God’s people depends on nothing
short of God’s impartial promise, election, call, and gift of πίστις, not νόμος,
ἔργα, or ethnicity.104 For any and all, divinely orchestrated lack of πίστις only
results in stumbling over the stumbling stone (cf. 1:24, 26, 28). In this way,
Rom 4 and 9 speak in one accord that the criterion is one and the same for all
humanity.

101  As Gaventa writes, “The capstone comes in v. 25 with the citation of Hosea’s words to
Israel now applied to both Israel and the Gentiles. Just as God called Israel into being, God
has called into being this people from those who were ‘not my people.’” Gaventa, “On the
Calling,” 267.
102  Thus, 9:30–33 provide additional support for the “Christian gentile” reading of 2:12–16.
103  Gaventa, “Questions about Nomos.”
104  Though not identical, I see God’s promise, election / call, and graced-πίστις all as inter-
related parts of a single Pauline creational and salvific matrix.
282 chapter 11

Romans 10. Rom 9 identifies Israel’s problem as a lack of πίστις. When Paul
grieves for Israel, therefore, he grieves for “unbelieving,” “non-Christian,” or
“non-πίστις” Israel (see 10:16; 11:1–2). Paul further explains Israel’s dilemma in
Rom 10. In Rom 1:18–32, humanity has knowledge of God but does not wor-
ship him because they chase after idols; Paul argues practically the same for
Israel in Rom 10. Israel has “zeal for God but not according to knowledge”
(κατ’ ἐπίγνωσιν; 10:2). Instead, “Being ignorant (ἀγνοοῦντες) of God’s δικαιοσύνη
and seeking to establish their own, they [are] not subjected (ὑπετάγησαν) to
God’s δικαιοσύνη” (10:3). This is because God’s δικαιοσύνη is Christological and
effective for all based on πίστις (παντὶ πιστεύοντι), not on “doing” or “possessing”
νόμος, as the speeches-in-character in 10:6–8 denote.105 Such a relationship
with νόμος essentially exalts νόμος as an idol rather than recognizing in it God’s
salvific plan for all humanity (cf. Rom 1:18–2:29).106 As scripture says, πίστις is
the basis for “all” (πᾶς; 10:11; Isa 28:16), because there is “no distinction between
Jew and Greek (οὐ γάρ ἐστιν διαστολὴ Ἰουδαίου τε καὶ Ἕλλενος). The same Lord is
over all” (ὁ γὰρ αὐτὸς κύριος πάντων; Rom 10:12), and he offers salvation equally
to “all” (πᾶς; 10:13; LXX Joel 3:5).
So far, however, Rom 10 seems to suggest that Israel is simply ignorant. Rom
10:14–21 corrects this potential misconception. Rom 10:16 indicates that Israel’s
problem is related to the gospel (τῷ εὐαγγελίῳ), but 10:18–19 imaginatively
attempts to alleviate this problem by considering whether Israel has not
had an appropriate encounter with the gospel, μὴ οὐκ ἤκουσαν; (10:18) and
μὴ Ἰσραὴλ οὐκ ἔγνω; (10:19). As indicated by μή, the answer is, “No, Israel heard
and knew.” The proclamation went out “to all the earth (εἰς πᾶσαν τὴν γῆν)”
and “to the corners of the world” (10:18; LXX Ps 18:5). God even tells Israel
he will make them jealous by what is “not a nation” and that he was “found
by those not seeking him … all the while he held out his hand to a disobedi-
ent and obstinate people” (Rom 10:20–21; Deut 32:21; LXX Isa 65:1–2). Indeed,
Israel knew and heard; their “ignorance” is no less culpable than humanity’s in
Rom 1, but neither is it any less divinely orchestrated, as Rom 11’s conclusion
puts so succinctly.
Romans 11. Israel’s precarious situation provokes yet another question,
“Surely God did not reject his people, did he?” (μὴ ἀπώσατο ὁ θεὸς τὸν λαὸν

105  For a discussion of 10:6–8, see Part 1, Chapter 5.

106  See Dunn, Theology, 114–19. Though not connecting this to idolatry, Longenecker
writes, “Israel has stumbled by their lack of faith (as Isaiah prophesied), for they con-
sidered the covenant to be restricted to the community of the Jews, whereas Scripture
foretold that no such restrictions would apply in the eschatological age.” Longenecker,
Eschatology, 254.
Romans 3:1-9 and the Argument of Romans 283

αὐτοῦ;). “Absolutely not (μὴ γένοιτο; 11:1)!” Paul offers himself as proof that God
has not rejected his people, “I myself am an Israelite (Ἰσραηλίτης), from the
seed of Abraham (ἐκ σπέρματος Ἀβραάμ), from the tribe of Benjamin” (11:1).
Lest anyone suppose that Paul’s autobiographical comments support some
sort of ethnic advantage, consider what Paul says about these qualifications
elsewhere. In Phil 3, Paul declares that he could have more confidence in the
flesh than others; he was

circumcised on the eighth day, from the race of Israel, from the tribe
of Benjamin, a Hebrew from Hebrews, a Pharisee according to the
law, according to zeal a persecutor of the church, and according to
righteousness in the law blameless. (3:4–6)

Because of Christ, however, Paul rejects these previous “advantages” (κέρδη;

3:7) and considers them “dung” (σκύβαλα). Rather than rely on these qualifi-
cations, Paul claims that justification comes not from νόμος but διὰ πίστεως
Χριστοῦ and from God, based entirely on πίστις (3:9). No differently, Gal 1 re-
veals how God’s apocalyptic revelation of the Son altogether halts Paul’s “pre-
vious conduct in Judaism” (1:13–17) and points him toward justification διὰ
πίστεως Ἰησοῦ Χριστοῦ on the basis of πίστις for all people (2:15–16). Only in
this way can Paul equally and impartially proclaim, “There is neither Jew nor
Greek, slave nor free, male nor female; for all (πάντες) are one in Christ” (3:28).
Rom 11 further clarifies that Paul’s ethnic qualifications provide him no ad-
vantage over non-Jews before a just and impartial God, as God operates on a
different standard. As in Elijah’s day (Rom 11:2–4), so also in Paul’s, God pre-
serves a remnant based not on ἔργα but on his gracious election (λεῖμμα κατ’
ἐκλογὴν χάριτος; 11:5–6; so also 9:6–13), not only from Jews but also from non-
Jews (9:24). Israel failed to obtain what it was seeking, but the “elect” (ἐκλογή)
obtained it (11:7). The rest—the “non-elect,” “not-yet-elect,” or perhaps the
“momentarily-hardened-elect”—God hardened (11:7); “God gave them (ἔδωκεν
αὐτοῖς ὁ θεός) a spirit of stupor, blind eyes, and deaf ears” (11:8–10; cf. Deut 29:3;
Isa 6:9; 29:10; LXX Psa 68:23). Even more clearly than tripping Israel over the
stumbling stone (9:32–33), God orchestrates non-Christian Israel’s failure.
God’s hardening of many Jews prompts Paul finally to consider the ultimate
fate of these Israelites; “Surely they did not (μή) stumble permanently, did they”
(11:11)? The answer provides a beginning glimpse of hope for Israel; absolutely
not (μὴ γένοιτο), “their fullness” (τὸ πλήρωμα αὐτῶν) is still possible (11:12, cf.
11:15). Non-Christian Israelites were “broken off” because of ἀπιστία, and non-
Israelites were included because of πίστις (11:20). If, however, Israel “[does] not
persist in ἀπιστία,” they will be reattached (11:23). Israelite or non-Israelite, Jew
284 chapter 11

or Greek, slave or free, male or female, divinely-graced Christological πίστις

or lack thereof is the sole criterion for whether or not one belongs to God’s
people.107
In 11:25–36, Paul expands on 11:11–24’s glimmer of hope and makes a final
plea for the equality of Jew and non-Jew in God’s salvific program. As in 11:7–
10, Paul argues that Israel has been hardened (πώρωσις ἀπὸ μέρους τῷ Ἰσραὴλ
γέγονεν). Now, however, Paul temporally qualifies this hardening; it only lasts
“until the fullness (πλήρωμα) of the nations enter” God’s people (11:25).108 In
this progression, “all Israel will be saved” (πᾶς Ἰσραὴλ σωθήσεται) when the res-
cuer comes in accord with the covenant and forgives Israel’s sins (11:26–27; cf.
Isa 27:9; 59:20; Jer 31:33).109 So, though Israel is currently an enemy with respect
to the gospel, they remain “beloved” (ἀγαπητοί) with respect to “election” (κατὰ
δὲ τὴν ἐλκογήν; Rom 11:28) because God’s “gracious gifts and calling are irrevo-
cable” (ἀμεταμέλητα γὰρ τὰ χαρίσματα καὶ ἡ κλῆσις τοῦ θεοῦ; 11:29).
From Paul’s perspective this is truly hopeful news for Israel, but this does
not change the fact that God engages Israel and non-Israel equally.110 God’s
calling (κλῆσις) is irrevocable (11:29), but God does not only call Jews; God
calls (ἐκάλεσεν) a “fullness” (πλήρωμα) from both Israel and non-Israel (9:24–
25; 11:12, 25). And, as 11:32 concludes, “God confined all (πάντας) to disobe-
dience, so that he might show mercy to all (πάντας)” (cf. 3:9). Furthermore,
though Israel’s “fullness” (πλήρωμα; 11:12) remains a possibility, their “full-
ness” depends on non-Israel’s “fullness” (πλήρωμα; 11:25). Whatever Israel’s
“fullness” means (11:12), it means the same for non-Israel’s “fullness” (11:25).
If “fullness” (πλήρωμα) means “all” (πᾶς) for Israel (11:26), it means “all” for non-
Israel.111 Paul makes no distinction. Instead, as Paul argues from Rom 1:1–11:36,
God’s graced-πίστις, promise, and election / call remain the valid currency, and
these are impartially and exhaustively operable for all. Indeed, what God does
for Israel, God does equally for all.

107  Longenecker, Eschatology, 258.

108  J. Ross Wagner, Heralds of the Good News: Isaiah and Paul “In Concert” in the Letter to the
Romans (NovTSup 101; Leiden: Brill, 2002), 277.
109  Käsemann correctly observes that Paul’s hope for Israel does not devolve into “speculative
fantasy,” but that one “must leave it in the [eschatological] horizon which determines it.”
Käsemann, Romans, 314.
110  Dunn observes, “[Paul] does not weaken his commitment to the Gentiles in any degree.
But he holds forth a hope for the final salvation of Israel …” Dunn, Theology, 528.
111  For the parallel treatment of πλήρωμα in 11:12, 25 and interpretive options, see Cranfield,
Romans, 558, 575–76; Fitzmyer, Romans, 611, 621–23; Kruse, Romans, 428–29, 443, 448–51;
Barrett, Romans, 206; Rhys, Romans, 146; Jewett, Romans, 678; Dunn, Theology, 527.
Romans 3:1-9 and the Argument of Romans 285

The import of Rom 3:1–9 and 9–11. Though not exclusively, Rom 9–11 focuses
on God’s dealings with Israel, at least on the surface. God’s dealings with Israel,
however, are no different than his engagement with non-Israel. To be sure, there
is some chronological jostling in both directions as God uses Israel on behalf
of non-Israel and non-Israel on behalf of Israel. God, however, ultimately—
eschatologically—deals the same with all humanity.112 In conjunction with
3:1–9, this has a number of implications for understanding Romans. First, it is
quite significant that Paul does not speak the affirmation of Jewish advantage
in 3:2.113 Based on Rom 1–2, it must be concluded that Paul would reject any
view of Jewish privilege that affords Jews a salvific advantage over non-Jews;
God’s promises to Abraham are universal, and according to Paul’s gospel God’s
active δικαιοσύνη faithfully saves all people characterized by πίστις (1:16–17).
The fact that Jews were “entrusted” with the λόγια confers to them no salvific
edge over non-Jews, as Paul leads even the “privileged” interlocutor to conclude
in 3:9. The same is true of 9–11. The list of attributes Paul attributes to Israel
that some might construe as “advantages” (9:4–5) also belong to Christian
gentiles;114 neither has an inherent step ahead of the other. What truly matters
is God’s choice, and God chooses from Jews and non-Jews alike (9:24, 30–33).
He condemns all, and he shows mercy to all (11:32). Paul’s thought on the ques-
tion of Jewish “advantage” neither contradicts nor displays a paradox between
Rom 1–2 and 3:1–9, between 3:1–2 and 3:9, or between 3:1–9 and 9–11. Paul cat-
egorically affirms that God deals equally with all. Consequently, what Paul says
about his own “gains” in Phil 3:4–11 could also apply to the interlocutor’s pre-
sumptuous “advantages” in 3:1–2 and Israel’s (exclusively conceived) in 9:4–5.

112  See Ticciatti, “Nondivisive,” esp. 261. Ticciatti is right to caution that universalism does
not reduce distinction to sameness, and that distinction and uniqueness do not equate
to inequality. By my reading, however, Ticciatti’s limited scope (Rom 9–11) casts her con-
clusions slightly off center. For instance, Ticciatti does not discuss the universal scope
of God’s promises to Abraham (Rom 4), and her identification of Christ as the τέλος
specifically of Israel’s law (10:4) fails to account for Paul’s discussion of gentiles who
Christologically do it (2:12–16). As a result, Ticciatti’s argument reads as though in Romans
Paul is more interested in saying something about Israel than he is in saying something
about God’s election and call of non-Israel or, more correctly, all.
113  Again, contra Keck, Romans, 90, who suggests understanding the arrangement of the
script is of little import.
114  These traits include adoption (8:15, 23); glory (8:18–21); covenants and promises (4:13–
16); νόμος (8:2–4); worship (12:1); the ancestors (4:16–18); and of course the Messiah, the
Lord over all (10:12), though not on a bio-geneological basis. See Keener, Romans, 116–17;
Wright, Climax, 237; Dunn, Theology, 503–4; Whittle, Covenant Renewal, 36–37.
286 chapter 11

Second, τί οὖν; μὴ ἀπώσατο Παῦλος τὴν διαθήκην τοῦ θεοῦ; Μὴ γένοιτο! Paul’s
silence regarding salvific advantage for Jews over non-Jews does not mean that
Paul denies God’s faithfulness to his promises to Abraham as well. The exact
opposite is true. Paul staunchly maintains that God will be faithful to the λόγια
(3:3–4) and that he continues to operate based on the promises (ἐπαγγελίαι; 4;
9:6–13; Gal 3:15–18). Paul’s post-Christ-apocalypse understanding of the prom-
ises is simply infinitely broader than the interlocutor’s. Given the amount of
time and space Paul spends developing this theme, it seems clear that it repre-
sents one of his main aims in Romans.
Third, Paul’s novelty in this respect, and the way he maintains God’s faith-
fulness to the promises despite avoiding Jewish superiority, is that he does not
view the promises as applicable only to Jews. Unlike the interlocutor, Paul be-
lieves the promises have a universal outlook. It seems that when Paul says “all,”
he really means “all.” 115 This is what Paul argues through Rom 4. Justification
and belonging to Abraham’s promised “seed” are based on πίστις and God’s
grace for all, uncircumcised and circumcised (4:11–12), such that “the prom-
ise (τὴν ἐπαγγελίαν) is effective (βεβαίαν) for all the seed (παντὶ τῷ σπέρματι),
not only those under νόμος but also those with Abrahamic πίστις” (4:16). In the
truest sense of the word, Paul demonstrates that the promises are applicable
on a cosmic scale, as Abraham and his “seed” will be the “heir of the κόσμος”
(4:13). Thus, according to the promise, God made Abraham the “father of many
nations” by calling (κληθήσεται, ἐκάλεσεν) for him a “seed” not only from Jews
but also from non-Jews (4:17–18; 9:6–9, 24). Paul’s silence concerning exclusive
Jewish “advantage” does not abandon God’s concern for Israel—it upholds it
by proclaiming the ethnically unhindered and cosmic intent of God’s prom-
ises, Israel included.116 As Paul writes in 15:8–9, “Christ became a servant of the
circumcision for the sake of God’s truthfulness (i.e., faithfulness), in order to
actualize the promises made to the fathers (εἰς τὸ βεβαιῶσαι τὰς ἐπαγγελίας τῶν
πατέρων), so that the nations (τὰ ἔθνη) might glorify God.”117

115  I am indebted to Beverly Roberts Gaventa for this way of phrasing Paul’s conception of
“all” in Romans, though she should not be held accountable for my application of it.
116  Contra Barrett, who finds such a view “offensive.” Barrett, Romans, 59.
117  The δέ requires explanation. Many take the conjunction as a connective; Jesus became a
servant for this and (δέ) for this. The thrust of Romans, however, especially 4:16, suggests
the conjunction should be read as explanatory, shedding further light on Jesus’ effectuat-
ing the promises. So, “Jesus became a διάκονον of the circumcision … in order to make
effective the promises made to the fathers, that is, for the nations to glorify God.” See
Wallace, Greek Grammar, 673. For various approaches, see Moo, Romans, 875–78; Jewett,
Romans, 890–93. See especially Wright, “Rom 2:17–3:9,” 21–22.
Romans 3:1-9 and the Argument of Romans 287

Fourth, recognizing that a significant thread in Rom 1–11 aims to stitch to-
gether the ethnically inclusive “seed” God promised to Abraham requires a
return to Paul’s interesting characterization of his Roman Christian audi-
ence in the prescript (see Chapter 10). Paul identifies his audience as “called
(κλητοί, κλητοῖς),” “beloved (ἀγαπητοῖς), and “holy (ἁγίοις)” (1:6–7)—terms the
Septuagint and Paul use for Israel as God’s chosen people. Given the way Paul
unveils the universal intent of God’s promises and thereby dismantles the
boundaries imposed on them by the interlocutor, it seems all the more certain
that Paul’s characterization of his audience in this way is intentional. By de-
scribing his ethnically unspecified but gentile-majority audience no differently
than as Israel, Paul demonstrates that the breadth of God’s promises does not
stop with Israel but extends to all God’s people. Jews and non-Jews, Israel and
non-Israel, equally reside within the reach of the promises and God’s “called,”
“beloved,” and “holy” people.118 What is more, positioned at the beginning of
the letter, this ethnic equalizing sets the course for the remainder of the letter
and primes its audience to read and hear as such.119
One final note on this point is necessary. Paul’s universalizing of God’s prom-
ises does not shame, downgrade, or reject Israel’s place in those promises. It el-
evates non-Israelites to their equal and rightful place (as Paul sees it). God still
acts on behalf of Israel, but he acts all the same on behalf of non-Israel as well.
In view of God’s promises, the relationship between Jew and non-Jew is either
one of no advantage or of universal “advantage.” This is why Paul demonstrates

118  Whittle’s basic conclusion that Jews and gentiles are included in God’s people on the
same Christological basis, therefore, is basically accurate. Her argument, however, re-
mains open to misunderstanding, if not misrepresentation. Whittle argues that Paul’s use
of select “covenant renewal” texts in 9–11 explains how Paul can identify his Roman audi-
ence as “holy,” despite that there is “little or no explicit evidence [elsewhere] in the letter
to connect Israel’s consecration at the mountain to the consecration of the Gentiles, or
[to] explain how Paul might arrive at such a conclusion.” Beginning for all intents and
purposes in Rom 9, Whittle far too quickly passes over Rom 4, where Paul’s overt discus-
sion of Abraham solidly establishes non-Jews within the scope of God’s promises—the
same promises that begin God’s relationship with Israel. This is why there is no overt or
extended discussion in Romans of Sinai or covenant renewal; Paul understands gentile
inclusion in God’s people to be even more fundamental than Whittle’s presentation sug-
gests, grounded not on the events at Sinai but on the promises to Abraham “that came
four hundred thirty years earlier” (cf. Gal 3:17) and are pre-promised in scripture (Rom
1:2). Whittle, Covenant Renewal, 31–75, cf. 146–51.
119  On the function of the prescript, see Philip L. Tite, “How to Begin, and Why? Diverse
Functions of the Pauline Prescript within a Greco-Romans Context,” in Stanley E. Porter
and Sean A. Adams, eds., Paul and the Ancient Letter Form (Pauline Studies (Past) 6;
Leiden: Brill, 2010), 59.
288 chapter 11

throughout Romans that non-Israelites share the same would-be “privileges”

as Israelites in 3:2 and 9:4–5: adoption (8:15, 23); glory (8:18–21); covenants and
promises (4:13–16); νόμος (8:2–4); worship (12:1); the ancestors (4:16–18); and of
course the Messiah, the Lord over all (10:12), though not on a bio-genealogical
basis but on a spiritual basis.120 The interlocutor initially denies non-Jews a
place in the promises. Paul’s response is not to expel Israel, but to demonstrate
how non-Jews are no less a part of God’s created people than the Jews121 and
how all belong through the sole criterion of God’s call and divinely-graced
πίστις. In this light, Paul is able to proclaim, “Rejoice, (you) nations (ἔθνη), with
his people” (15:10; LXX Deut 32:43), and “Praise the Lord, all the nations (πάντα
τὰ ἔθνη), praise him, all peoples (πάντες οἱ λαοί)” (Rom 15:11; Ps 117:1).122

A Purpose of Romans
A final, but relatively brief, discussion of the “Romans debate” begins to
bring this project to a close. This project highlights the dialogue in 3:1–9, but
conversation necessarily and inevitably extends into much of Rom 1–11. The
results of these investigations, with consideration of 12–15, have particular
bearing on why Paul wrote the letter. To be sure, the “purpose” of Romans is
pluriform, with explanations generally focusing on Paul’s trip to Jerusalem,
Spain, or Rome (15:22–29), the latter of which is often more narrowly tied
to some disunity and/or potential misunderstanding of the gospel in Rome
(cf. 1:15).123 It is on the latter of these points that this study is particularly
relevant.124

120  Keener, Romans, 116–17; Wright, Climax, 237; Dunn, Theology, 503–4; Whittle, Covenant
Renewal, 36–37.
121  On Paul’s intentionality, see Dunn, Theology, 504.
122  One potential reading of the Deuteronomy citation allows a relatively strong distinction
between “nations” and “his people.” The citation from Ps 117, however, indicates that such
a reading is off course. Rom 15:11 / Ps 117:1 harmoniously joins Jews and non-Jews in prais-
ing God, as the parallelism between “all nations” and “all peoples” suggests.
123  As entry points to the consideration of the purpose of Romans, see Donfried, The
Romans Debate; A. Andrew Das, Solving; Williams, “Righteousness,” 245–55; or most any
commentary.
124  Despite Karl Paul Donfried’s need to distance Romans from diatribe (at least in generic
terms) in view of the Romans Debate, his insistence that the use of rhetorical patterns in
no way precludes the likelihood that Romans is addressed to particular and contingent
situations in Rome is correct. Karl Paul Donfried, “False Presuppositions in the Study of
Romans,” in The Romans Debate (Karl P. Donfried, ed.; rev.; Grand Rapids: Baker Academic,
2011), 118–19. Stowers’s work shows, and my examples above add to, the probability that
diatribe and diatribal dialogue operate on multiple levels, one with the interlocutor, and
Romans 3:1-9 and the Argument of Romans 289

As with 3:1, discussion in this respect begins at 1:1–7. Chapter 10 argued that
much of Romans evinces an ethnically inclusive nature, such that it is often in-
determinable whether Paul is addressing Jews, non-Jews, or both. For instance,
Paul’s gospel and apostleship are for all nations (1:5). Paul writes Romans to
all God’s beloved, called, and holy people in Rome (1:6–7). The gospel is God’s
powerful means of effecting salvation for all (1:16–17), just as his wrath is un-
veiled against all humanity’s wickedness (1:18–32). All are confined under Sin
(3:9). God calls a seed for Abraham from Jews and non-Jews (4:11–12, 13, 16;
9:24). Additionally, Paul never identifies the interlocutor in any ethnic terms;
the interlocutor is anyone who religiously considers oneself a Jew (ch. 2).
Such ethnic vagueness is present in most of Romans. The few instances
where Paul approaches ethnicity become occasions not to stress difference
or distinction but equality. God procures salvation through the gospel for Jew
and non-Jew alike (1:16–17). God’s positive and negative judgment falls impar-
tially on Jew and non-Jew (2:9–11). When Paul specifically addresses non-Jews
in 9–11’s discussion of Israel, it is to stifle haughtiness and arrogance over non-
Christian Israel (11:13–24). But at this point we perhaps get slightly ahead of
ourselves; there is more to say about Rom 1.
The beginning of ancient letters, especially epistolary thanksgivings, often
reveal information about the contents of the letter and approximate some-
thing of a table of contents.125 For example, the thanksgiving in Philippians
mentions their repeated “partnership in the gospel” (ἐπὶ τῇ κοινωνίᾳ ὑμῶν εἰς
τὸ εὐαγγέλιον; 1:5), which one later learns is a technical term to refer to their
financial support of Paul’s mission (4:10–20).126 Given the ethnic ambiguity
and equality of Romans, one comment in the thanksgiving calls for additional
consideration. Paul writes that he wants to strengthen the Romans by impart-
ing to them some spiritual gift (1:11). But Paul quickly qualifies this claim. Paul
writes, “I mean, I want to be mutually encouraged (συμπαρακληθῆναι) with you
(ἐν ὑμῖν) through each other’s πίστις (διὰ τῆς ἐν ἀλλήλοις πίστεως), both yours
and mine” (ὑμῶν τε καὶ ἐμοῦ; 1:12). Though easily missed on a first, second, or

one with the actual audience. Thus, the “typical” of diatribal dialogue cannot help but ad-
dress something of the “particular pedagogical needs of the audience,” even in Romans.
Stowers, Diatribe, 180–81.
125  Cf. Matera, Romans, 26.
126  Gerald Peterman, Paul’s Gift from Philippi: Conventions of Gift Exchange and Christian
Giving (SNTS 92; Cambridge: Cambridge University Press, 1997), 90–103.
290 chapter 11

perhaps tenth reading/hearing, this qualification takes on new meaning when

one considers it in light of Rom 11 and 12–15.127
In Rom 1, Paul, a Jewish Christian—perhaps the Jewish Christian par
excellence—expresses his desire to engage mutually with a gentile Christian
majority. In short, Paul models that Jewish and non-Jewish Christians can and
should profitably co-exist—and “co-exist” puts the matter too lightly.128 At
least some of the Roman Christians, however, seem to be acting otherwise.
Again, the interlocutor in Rom 2–3 does not immediately represent Paul’s
actual audience, but similarities nevertheless exist between 2–3 and 11–15.
For instance, chapters 2–3 engage a judgmental, privileged, and religiously-
Jewish interlocutor who boasts over non-Jews, Rom 11 addresses Christian
pride over non-Christian Israel, and 12–15 exhorts against Christians judging
other Christians. That is, Rom 2–3:2 portrays non-Christian Jews who “judge”
(κρίνεις, ὁ κρίνων, cf. κατακρίνεις, τὸ κρίμα; 2:1–3) non-Jews because of a sense
of privileged and boastful superiority. Quite similarly, in 11:13–24, Paul advises
his gentile-majority, Christian audience not to boast over non-Christian Israel
(μὴ κατακαυχῶ τῶν κλάδων; 11:18) or to think too highly (μὴ ὑψηλὰ φρόνει; 11:20)
given their status within God’s people.129 Rom 12–15 translates this judgmen-
talism into the Christian communities. In Rom 12, Paul exhorts (παρακαλῶ;
12:1) all Christians in Rome (παντὶ τῷ ὄντι ἐν ὑμῖν; 1:7) not to think arrogant-
ly (μὴ ὑπερφρονεῖν) but moderately (ἀλλὰ φρονεῖν εἰς τὸ σωφρονεῖν), because,
though there are many members of Christ’s body, all are equally important
members of the unified body (12:1–8; cf. 1 Cor 12; Gal 3:28). Rom 14 makes this
rather general exhortation concrete; some Roman Christians—also ethnically
unspecified130—are quarrelling over trivial disputes (as Paul sees it), such as
what to eat. Like the non-Christian interlocutor who once clung to a restricted
understanding of νόμος and hindered God’s promises from taking full effect,
some Roman Christians cling to the freedom that comes from their “measure
of faith” and potentially cause other Christians to stumble (Rom 14). In order

127  Similarly, Marty L. Reid, “A Consideration of the Function of Rom 1:8–15 in Light of Greco-
Roman Rhetoric,” JETS 38.2 (1995): 181–91, esp. 189–91.
128  Rom 1:12, therefore, is more than a “humble apology” or “diplomatic correction” for 1:11.
Contra, for examples, Barrett, Romans, 26; Kruse, Romans, 63; Mounce, Romans, 67–68;
Bruce, Romans, 72; Fitzmyer, Romans, 246.
129  Paul addresses “you nations” (ὑμῖν δὲ λέγω τοῖς ἔθνεσιν; 11:13). Given Paul’s frequent refer-
ence to all nations in Romans, and given the focus in 9–11 on non-πίστις Israel, might
Paul’s use of “nations” in 11:13 include Jewish Christians like Paul within its scope?
130  Jewett, Romans, 70ff.
Romans 3:1-9 and the Argument of Romans 291

to resolve these divisive conflicts, Paul directs both sides of the dispute not to
judge one another (σὺ δὲ τί κρίνεις; μηκέτι οὖν ἀλλήλους κρίνωμεν), because
God is the judge of all (πάντες γὰρ παραστησόμεθα τῷ βήματι τοῦ θεοῦ; 14:10–13;
cf. 12:19–21).131 Instead, the Roman Christians should “love one another” (13:8–
10), “welcome” others (14:1; 15:7), not cause others to stumble (14:13), and the
truly strong should sacrifice their rights for the benefit of others (14:15–15:1).
Romans, therefore, is at minimum a letter of unification, and Paul prepares
his audience for the exhortation in 12–15 from the very beginning. Paul’s display
of mutual encouragement, the religio-ethnic ambiguity, the inclusive scope of
the gospel, and God’s exhaustive and impartial ὀργή and δικαιοσύνη seek to
remove any sense of exclusivity within the Christian communities in Rome.132
One might even say that Paul’s argument in 1–11 sets a trap for his actual audi-
ence, which he finally springs when he begins to address the divisiveness of
the Roman congregations in 12–15. In this way, Paul seeks the united, singular,
corporate (ὁμοθυμαδόν) glorification of God from all God’s people (15:6). All are
to rejoice together (15:10). All nations (πᾶντα τὰ ἔθνη) and all peoples (πάντες οἱ
λαοί) are to praise God (15:11).
Correcting the inner-Christian divisiveness among the Roman congrega-
tions is, therefore, almost certainly one facet of the purpose of Romans. Rom
3:1–9 plays only a small role in the large-scale argument of Rom 1–15, but it
plays a significant role nonetheless by arguing for ethnic equality before a just
and impartial God, even for Jews. Correctly understanding the script, there-
fore, is actually quite significant. How could Paul argue for equality and unity
yet affirm that Jews have a salvific advantage over non-Jews, as traditional
readings must paradoxically affirm or hold in tension? Μὴ γένοιτο! Rescripting
the dialogue in 3:1–9 altogether alleviates this problem. Rom 3:1–9 and 9:4–5
no longer appear as paradoxical or in tension with Paul’s sweeping concern for
equality and impartiality. Instead, Paul’s Socratic guidance of the interlocutor
towards equality firmly aligns 3:1–9 with the rest of Romans and puts forth a
united front concerning the relationship between Jews and non-Jews. Said in

131  See also Bassler, Divine Impartiality, 162–63, who argues that a “real sociological spin-off”
of Paul’s concept of divine impartiality in eschatological judgment is visible. “As a re-
sult of God’s impartial treatment of the two groups the immediate social consequence
is proclaimed that no distinction is to be made within the community itself.” Instead,
“Christians are now to refrain from judging each other.” Bassler, Divine Impartiality, 163,
italics original.
132  Bassler is correct, therefore, to surmise that Paul’s ethnic evasiveness is not accidental.
Bassler, Divine Impartiality, 136.
292 chapter 11

Pauline terms, a correctly rescripted dialogue of 3:1–9 frees the passage from
slavery to confused contradiction and enslaves it in service of the pervasive
message of God’s impartial δικαιοσύνη and the unity of the body.

Conclusion

Before the reader ever arrives at Rom 3, Paul makes substantial progress
defining his gospel and characterizing the interlocutor (Rom 1–2). Paul ar-
gues that when God’s δικαιοσύνη brings about human πίστις, this actualizes
the saving power of the gospel, which ultimately rescues its objects from
God’s wrath. This divine initiative knows no bounds, ethnic or otherwise; God’s
δικαιοσύνη is impartial and thoroughly so. But so is God’s wrath, and in Rom 2
Paul begins to address an imaginary persona who thinks quite differently
than Paul. Paul argues that failure to worship God and idolatry fall squarely
under God’s judgment, and he accuses the interlocutor precisely of these
faults. But the interlocutor presumes to have a trump card over God’s judg-
ment, namely, his exclusive privilege of Jewish identity and possession of
νόμος, and this characterization sets the stage for Rom 3:1–9’s ensuing dialogue.
Despite the majority view in the passage’s history of interpretation, the tran-
sitions in the conversation between Paul’s voice and the interlocutor’s are a
far cry from obvious or easy to distinguish. Adhering to the proposed method
in which characterization and attributed speech should appropriately cor-
relate, however, presents a new way forward. Since attributed speech should
be appropriate to an imaginary speaker’s characterization, characterization
can serve as the interpretive key to determine which lines in a discourse be-
long to which speaker, especially when the transitions are not overtly marked.
Applied to the dialogue in Rom 3:1–9, this methodological approach exposes
problems within the traditional reading and provides the argumentative basis
for the rescripted arrangement of the discourse by Neil Elliott, which is oth-
erwise presented quite shallowly. As a result, the conventions for speech-in-
character and diatribe suggest that it is not the imaginary interlocutor who
fires objections at Paul in 3:1, 3, 5, 7–8c, 9a, but Paul who, in the role of Socratic
questioner, presents leading questions to censure his interlocutor and guide
him from his original, faulty, point of view towards Paul’s more correct un-
derstanding. Through the course of the dialogue, the interlocutor amends his
answer and comes to Paul’s aid. Though he initially affirms exclusive advantag-
es for Jews (3:2), he eventually relinquishes that hold (3:8–9) in view of God’s
faithfulness, justice, and impartiality (3:3–6). Jews possess no salvific edge over
Romans 3:1-9 and the Argument of Romans 293

non-Jews; all equally fall within the scope of God’s judgment and within the
scope of his merciful and graciously incongruous δικαιοσύνη.133
A correct arrangement of the script, especially attributing 3:2 and 9 to the in-
terlocutor, has a remarkable impact on holistic readings of Romans. Traditional
readings must often argue that Paul holds in tension his view of divine impar-
tiality for Jew and gentile with his belief that Jews possess a unique eschato-
logical advantage with God. Such tension, if not contradiction, is resolved if
Paul is not responsible for speaking the affirmation of Jewish advantage in 3:2.
Rom 3:1–9 then falls seamlessly into Paul’s central emphases on divine impar-
tiality and anthropological equality. This is buttressed by Paul’s demonstration
that non-Jews are privy to the same would-be advantages as Jews (Rom 1, 4,
6, 8, 9), and even where Paul addresses God’s dealings with Israel at length
(9–11), he trusts that God’s eschatological plan for Israel is “fully” synonymous
with his eschatological plan for non-Israel. Indeed, God hardens all, so that he
might show mercy to all. This in no way diminishes Israel’s relationship with
God; quite the contrary, it maintains Israel’s position as beloved of God and el-
evates gentiles to their rightful place as equal objects of the divine promises to
Abraham that call into being the people of God from both Jews and non-Jews.
Additional implications from rescripting 3:1–9 are discussed in the following
conclusion to this project as a whole.

133  On the “incongruous” nature of God’s grace, see Barclay, Gift.
 chapter 12

Conclusion

This project focuses on the questions of who is speaking which lines in the dia-
logue of Rom 3:1–9—Paul or his imaginary, diatribal interlocutor—and why
it matters for understanding Romans. Though evincing diverse conclusions,
previous traditional and rescriptive readings of Rom 3:1–9 display a remark-
able but unacceptable consistency, namely, the preponderance of assump-
tions rather than arguments based on valid and/or persuasive evidence. None
are exempt or without excuse. Even at a recent annual meeting of the Society
of Biblical Literature,1 which included several papers featuring discussions of
Rom 3:1–9, this reality remained the norm (at least for those papers I was able
to attend, each of which understood 3:1–9 along traditional lines without pro-
viding any supporting arguments).2
This absence of methodological constraint and persuasive argumenta-
tion, however, proves surprisingly refreshing, since it allows both traditional
and rescriptive readings to receive an equal hearing. From the outset, there
is no reason to show partiality to either view as far as sustainable and per-
suasive arguments are concerned. Interestingly, this project ends on a similar
note, with one outstanding difference. As previous traditional and rescriptive
readings succumb to a common error, so also in Romans Paul holds that all
of humanity—Jews and non-Jews—equally, exhaustively, and impartially fall
under the common judgment of God and the power of Sin. Whereas a certain
rescriptive reading ultimately overcomes traditional (and other rescriptive)
readings when viewed in light of the conventions for speech-in-character, Paul
consistently maintains that neither Jews nor non-Jews ever salvifically or es-
chatologically surpass the other. God deals equally with all humanity, just as he
promised to Abraham (Gen 12, 15; Rom 4). This means that whatever God plans
to do for the fullness of all Israel, God plans to do equally for the fullness of all
non-Israel (Rom 11). But getting to this point took some doing, and Rom 3:1–9
played a critical role. In this final chapter, I briefly summarize my conclusions.

1  Atlanta, 2015.
2  For two examples, Joshua Garroway, “Under Sin: Finding the Antecedent for Paul’s Charge in
Rom 3:9b” (paper presented at the annual meeting of the SBL, Atlanta, November 21, 2015);
Michael T. Graham Jr., “An Examination of Paul’s Use of Ps 51:4 in Rom 3:4 and Its Implications
on Rom 3:1–8” (paper presented at the annual meeting of the SBL, Atlanta, November 22,
2015).

© koninklijke brill nv, leiden, 2018 | doi 10.1163/9789004373297_016

Conclusion 295

This project began in Part 1 by examining the primary literature for the
rhetorical practice of speech-in-character and Paul’s use of the skill in his letters.
The primary sources agree that speech-in-character must appropriately align
with the characterization of the imagined speaker. This essential convention of
appropriateness to character undergirds the whole project, as it constitutes the
plumb line by which one can often determine whether an imagined speaker
could or could not speak given lines.
Largely following Stowers’s reassessment of diatribe as an educational
technique by which a teacher guides students to adopt better views through
censure and protreptic, Part 2 demonstrated the remarkable similarity between
speech-in-character and diatribe’s dialogical passages. In fact, every instance
of diatribal dialogue utilizes speech-in-character in the creation of staged
discourses. Consequently, when the transitions between speakers in diatribal
dialogues are unclear or otherwise difficult to distinguish, readers may refer
to the conventions for speech-in-character to help confirm or define which
speaker is responsible for which lines in the imaginary conversation. Thus, in
addition to Pauline studies, this project also makes contributions to rhetorical,
Progymnasmata, and diatribal studies.
Finally, Part 3 applies the method developed in Parts 1 and 2 to Paul’s dia-
logue with an interlocutor in Rom 3:1–9. Traditional readings of the passage
generally imagine the interlocutor posing objections to Paul in 3:1, 3, 5, 7–8c,
and 9a, which Paul responds to in 3:2, 4, 6, 8d, and 9b. Rescriptive readings,
however, rearrange the dialogical exchanges primarily in one of two ways. One
model begins with the interlocutor posing objections in 3:1 and 9a but inverts
the middle of the dialogue so that Paul poses the questions in 3:3, 5, 7–8. The
other arrangement completely reverses the traditional script, so that Paul rais-
es questions in 3:1, 3, 5, 7–8c, and 9a for the interlocutor to answer in 3:2, 4,
6, 8d, and 9b.
Evidence demonstrates that all previous traditional and rescriptive read-
ings fail to persuade. Most readings neglect to provide any argument at all,
opting instead simply to assume a given script. The readings that try to sup-
port their dialogical arrangements do so equally unsatisfactorily. Such “argu-
ments” either reduce to assumptions in their own right (such as the required
speaker of the μὴ γένοιτο rejections), or they rely on diatribal evidence that
is routinely recognized as diverse and inherently inconsistent. Given the va-
riety of views available, not to mention the significance of their unique im-
plications for understanding Romans, assumptions and invalid evidence
cannot be allowed to pass as acceptable. As such, all previous readings fall
within the category of weighed and wanting as far as arguments offered (or
not offered) are concerned. To remedy this problem, a more consistent body
296 chapter 12

of evidence must be incorporated into discussion of the diatribal dialogue in

Rom 3:1–9. This body of evidence involves the rhetorical figure of speech-in-
character. Because diatribe must employ speech-in-character in its composi-
tion of imaginary dialogue, and because Romans is diatribal, the introduction
of speech-in-character is hardly onerous or problematic from a methodologi-
cal point of view.
In my assessment of Romans, beginning with the very epistolary prescript,
Paul gradually builds toward the dialogue in Rom 3:1–9. In chapter 1, Paul
emphasizes the universal scope of his gospel; it is applicable to all without
exclusion. Accordingly, Paul addresses the letter to all the Christians in Rome.
Paul even applies to this ethnically unspecified audience terms the LXX and
Paul himself later use to refer to Israel as God’s chosen people. All God’s people
are called, beloved, and holy; this includes Jew as well as non-Jew, Israel as
well as non-Israel. Paul’s disclosure in the proem that he hopes to experience
mutual encouragement with the Roman Christians further buttresses this
claim, as it models for his audience the equal status and unified relationship
that Paul argues should exist between Jewish and non-Jewish Christians.
Paul is able to maintain this equality between Jew and non-Jew because, due
to God’s faithfulness to his universal promises to Abraham, God’s δικαιοσύνη
activates πίστις within humanity and thereby causes the gospel to effect
salvation for all people characterized by such πίστις. This salvation rescues
humanity from God’s wrath, to which all of humanity are also subjected due to
their disposition towards idolatry. Left to their own devices and despite their
knowledge of God, humanity does not worship him as God, and for this reason
God hands them over.
Romans 2 represents Paul’s characterization of the interlocutor with whom
he engages in conversation in 3:1–9. Contrary to Paul’s description of God’s
impartial wrath, Paul’s religiously Jewish interlocutor supposes that, though
others fall under God’s judgment, he will not. This is because the interlocutor
believes that his status as a Jew and possession and observance of νόμος provide
him with a salvific, eschatological advantage over non-Jews. Paul, however, ac-
cuses the interlocutor of “doing the same things” and consequently falling no
less completely within the breadth of God’s wrath. A stark difference, there-
fore, exists between Paul’s presentation of God’s impartial gospel and wrath
in Rom 1 and the characterization of the interlocutor’s views in Rom 2. Paul’s
message is one of equality, impartiality, and inclusivity; the Jewish interlocu-
tor’s view is one of privilege, partiality, and exclusivity, and this characteriza-
tion sets the stage for the ensuing dialogue.
Arriving at Rom 3:1–9, using Paul’s characterization of the interlocutor as
the measure against which to determine the appropriateness of lines in the
Conclusion 297

imaginary dialogue concludes that a rescriptive reading of the dialogue is in

order. Traditional readings assume that the interlocutor poses the questions
in 3:1 to Paul, and that Paul answers in 3:2. Thus, in traditional readings, when
the interlocutor asks about Jewish advantage, Paul affirms that Jews possess
an advantage over non-Jews. When one considers the conventions for speech-
in-character, however, a different picture develops. In Rom 1–2, Paul holistical-
ly argues for anthropological equality and divine impartiality. The interlocutor,
however, maintains that Jews possess a salvific advantage over non-Jews based
on Jewishness and the possession and observance of νόμος. Given that at-
tributed speech must be appropriate to the characterization of an imaginary
speaker, it must be concluded on methodological grounds that the interlocutor
most appropriately speaks the affirmation in 3:2. Similar arguments take root
in 3:3–9. In the end, the conventions for speech-in-character mount to suggest
that Paul plays the role of Socratic questioner (or diatribal teacher) throughout
3:1–9’s dialogue, posing leading questions that compel the interlocutor to re-
think his previously held privileged and exclusive views. This analysis provides
actual argumentation for the rescriptive model that casts Paul speaking the
questions in 3:1, 3, 5, 7–8c, and 9a and the interlocutor responding in 3:2, 4, 6,
8d, and 9b.
This rescriptive reading offers numerous advantages over other readings.
First, it stems from and accounts for a body of evidence that is much more con-
sistent and widely documented than the diatribal evidence referenced by pre-
vious scholars. Because the prosopopoetic material is intrinsically tied to the
creation of diatribal dialogue, incorporating this material is the most natural
solution to the search for a stronger base of evidence relevant to the diatribal
script of Rom 3:1–9. Second, my rescriptive reading makes better sense of the
dialogue. Paul and the interlocutor maintain consistent roles throughout, and
all the first-person speech is attributed to Paul. Moreover, what the traditional
reading identifies as authorial asides no longer appear as awkward interjec-
tions into the interlocutor’s speech since they align with Paul’s guiding ques-
tions. Third, this rescriptive reading removes unnecessary contradictions or
tensions from Paul’s argument regarding divine impartiality and anthropologi-
cal equality. Other readings must explain why Paul affirms Jewish advantage
in 3:2 but rejects it in 3:9b. My reading avoids this issue because these lines are
attributed to the interlocutor. In fact, for the interlocutor to respond differently
in 3:9 than in 3:2 is altogether expected in a diatribal discourse interested in
transformation through the use of censure and protreptic. The interlocutor be-
gins with an erroneous view of Jewish eschatological advantage over non-Jews,
but Paul exposes the inconsistencies in the interlocutor’s view and leads him
to endorse Paul’s view of universal, eschatological equality. The interlocutor
298 chapter 12

even makes for Paul the concluding point in his argument. Thus, this work not
only answers the question of who is speaking when, but it also addresses the
issue of why it matters.
This approach is also able to explain better the dialogue’s function in the
larger argument. Rom 3:1–9 is not a digression, aside, or passage otherwise in
need of excision. Rom 3:1–9 fills a particular role in arguing for exhaustive and
universal divine impartiality and anthropological equality. If not even Jews are
excused from undergoing God’s judgment, no one is. All are equal. This point
is further developed as the questions Paul raises in 3:1–9 reappear in nuanced
forms later in the letter, but it is most prominent in Paul’s discussion of Israel
in Rom 9–11. Because Paul largely focuses on Israel in these chapters, scholars
often allow Paul’s discussion of the gentiles to fall out of sight. Paying atten-
tion to what Paul says about God, Israel, and non-Israel in Rom 9–11, however,
demonstrates Paul’s consistency regarding divine impartiality and anthropo-
logical equality. Interestingly, what Paul believes God will do for Israel, Paul
declares that God will do the same for non-Israel. Is the fullness of Jews going
to enter into God’s people? Not until the fullness of non-Jews enter in as well.
Did God cause most Jews to stumble over the stumbling stone? Yes, but God
equally confined all to disobedience. Is God going to show mercy to Israel? Yes,
but he will show mercy equally to all. If (or since) God’s promises to Abraham
have salvific and eschatological implications for Israel, Paul insists the prom-
ises’ universal scope demands that they have the same salvific and eschatologi-
cal implications for non-Israel (cf. Rom 4). There is no qualitative hierarchy;
none is salvifically advantaged over another. The rescripted dialogue of 3:1–9 is,
therefore, in accord with Paul’s argument about Israel and all in Rom 9–11 and
throughout Romans. Whatever God does to Israel, Paul argues, God will do for
non-Israel. In this light, Paul’s declaration in Gal 3:28 rings true for Romans as
well and brings this study to an—appropriate—conclusion:

Οὐκ ἔνι Ἰουδαῖος οὐδὲ Ἕλλην,

οὐκ ἔνι δοῦλος οὐδὲ ἐλεύθερος,
οὐκ ἔνι ἄρσεν καὶ θῆλυ·
πάντες γὰρ ὑμεῖς εἷς ἐστε ἐν Χριστῴ Ἰησοῦ.
Bibliography

Primary Texts and Translations

Aelius Aristides. In Defence of Oratory. Translated by C. A. Behr. Loeb Classical Library

458. Cambridge: Harvard University Press, 1973.
Aristotle. The Art of Rhetoric. Translated by J. H. Freese. Loeb Classical Library.
Cambridge: Harvard University Press, 1926.
Baldwin, Charles Sears. “The Elementary Exercises (ΠΡΟΓΥΜΝΑΣΜΑΤΑ) of
Hermogenes.” Pages 23–38 in Medieval Rhetoric and Poetic (to 1400): Interpreted from
Representative Works. New York: Macmillan, 1928.
Butts, James R. “The Progymnasmata of Theon: A New Text with Translation and
Commentary.” Ph.D. diss., The Claremont Graduate School, 1987.
Cicero. De Inventione; De Optimo Genere Oratorum; Topics. Translated by H. M. Hubbell.
Loeb Classical Library 386. Cambridge: Harvard University Press, 1949.
Cicero. De Oratore. Translated by H. Rackham. Loeb Classical Library 349. Cambridge:
Harvard University Press, 1992.
Cicero. De Partitione Oratoria. Translated by H. Rackham. Loeb Classical Library 349.
Cambridge: Harvard University Press, 1992.
Cicero. In Catilinam. Translated by C. MacDonald. Loeb Classical Library 324.
Cambridge: Harvard University Press, 1977.
Cicero. Pro Caelio. Translated by R. Gardner. Loeb Classical Library 447. Cambridge:
Harvard University Press, 1958.
Cornelius Nepos. On Great Generals: Militiades. Translated by J. C. Rolfe. Loeb Classical
Library 467. Cambridge: Harvard University Press, 1929.
Ctesias. History of Persia. Translated by Lloyd Llewellyn-Jones and James Robson, in
Ctesias’ History of Persia: Tales of the Orient. London: Routledge, 2010.
Ctesias. La Perse. Translated by Dominique Lenfant, in Ctésias de Cnide: La Perse: L’Inde:
Autres Fragments. Paris: Les Belles Lettres, 2004.
Demetrius. On Style. Translated by Doreen C. Innes. Loeb Classical Library 199.
Cambridge: Harvard University Press, 1995.
Demosthenes. Private Orations: Against Neaera. Loeb Classical Library 351. Translated
by A. T. Murray. Cambridge: Harvard University Press, 1939.
Dio Chrysostom. Translated by J. W. Cohoon and L. Crosby. 5 vols. Loeb Classical
Library. Cambridge: Harvard University Press, 1932–51.
Diodorus Siculus. Library of History: Book IX–XII.40. Loeb Classical Library 375.
Translated by C. H. Oldfather. Cambridge: Harvard University Press, 1946.
Epictetus. Discourses. Translated by W. A. Oldfather. 2 vols. Loeb Classical Library.
Reprint. Cambridge: Harvard University Press, 2000.
300 Bibliography

Gibson, Craig A., trans and ed. Libanius’ Progymnasmata: Model Exercises in Greek
Prose Composition and Rhetoric. Society of Biblical Literature Writings from the
Greco-Roman World 27. Atlanta: Society of Biblical Literature, 2008.
Hense, Otto, ed. C. Musonii Rufi reliquiae. Leipzig: Teubner, 1910.
Hense, Otto. Teletis reliquiae. Tübingen: Teubner, 1889. 2nd ed., 1909.
Herodotus. The Persian Wars. 4 vols. Translated by A. D. Godley. Loeb Classical Library
117–120. Cambridge: Harvard University Press, 1998–2001.
Homer. The Iliad. Translated by A. T. Murray. Revised by William F. Wyatt. 2 vols. Loeb
Classical Library. Cambridge: Harvard University Press, 1999, reprinted 2001.
John Chrysostom. Hom. I–XLIV in 1 Cor. Migne, PG 61.9–382.
Kennedy, George A, trans and ed. Progymnasmata: Greek Textbooks of Prose Composition
and Rhetoric. Society of Biblical Literature Writings from the Greco-Roman World
10. Leiden: Brill, 2003.
Kirkstrand, Jan Frederick. Bion of Borysthenes: A Collection of the Fragments with
Introduction and Commentary. Studia Graeca Upsaliensia 11. Stockholm: Almquist
& Wiksell, 1976.
Livy, History of Rome: Books 1–2. Translated by B. O. Foster. Loeb Classical Library 114.
Cambridge: Harvard University Press, 1919.
Lutz, Cora. Musonius Rufus, “The Roman Socrates.” Yale Classical Studies 10. New
Haven: Yale University Press, 1947.
Maximus of Tyre. Philosophumena. Translated by Hermann Hobein. Leipzig: Teubner,
1910.
Menander. The Principal Fragments. Translated by Francis G. Allison. Loeb Classical
Library. Revised. Cambridge: Harvard University Press, 1930.
Musonius Rufus. Reliquiae. Translated by Otto Hense. Leipzig: Teubner, 1910.
Nadeau, Ray. “The Progymnasmata of Aphthonius in Translation.” Speech Monographs
19 (1952): 264–85.
O’Neil, Edward, trans and ed. Teles (the Cynic Teacher). Missoula: Scholars Press, 1977.
Patillon, Michel and Giancarlo Bolognesi, eds. Aelius Theon: Progymnasmata. Paris: Les
Belles Lettres, 2002.
Philo. Translated by F. H. Colson, G. H. Whitaker, and R. Marcus. 12 vols. Loeb Classical
Library. Cambridge: Harvard University Press, 1949–61.
Plato. Laws: I–VI. Loeb Classical Library 187. Translated by R. G. Bury. Cambridge:
Harvard University Press, 1961.
Plutarch. Lives: Aristides. Loeb Classical Library 47. Translated by Bernadotte Perrin.
Cambridge: Harvard University Press, 2006.
Plutarch. Moralia. Translated by F. C. Babbit, W. Helmbold, et al. 14 vols. Loeb Classical
Library. Cambridge: Harvard University Press, 1926–69.
Quintilian. Institutio Oratoria. Translated by Donald A. Russell. 5 vols. Loeb Classical
Library. Cambridge: Harvard University Press, 2001.
Bibliography 301

Rabe, Hugo, ed. Hermogenis Opera. Rhetores Graeci VI. Lepizig: Teubner, 1913.
Rhetorica ad Herennium. Translated by Harry Caplan. Loeb Classical Library.
Cambridge: Harvard University Press, 1954.
Seneca. Epistulae Morales. Translated by Richard M. Gummenre. 3 vols. Cambridge:
Harvard University Press, 1917–25.
Seneca. Letter to Lucilius. Translated by E. Phillips Barker. 2 vols. Loeb Classical Library.
Cambridge: Harvard University Press, 1932.
Seneca. Tragedies. Translated by Frank Justus Miller. 2 vols. Loeb Classical Library.
Cambridge: Harvard University Press, 1960–61.
Spengel, Leonhard von, ed. Rhetores Graeci. 3 vols. Reprint. Frankfurt: Minerva, 1966.
Suetonius. De Grammaticus et Rhetoribus. Edited and translated by Robert A. Kaster.
Oxford: Clarendon Press, 1995.
Vergil. Aeneid I–VI. Translated by H. Rushton Fairclough. Loeb Classical Library.
London: William Heinemann, 1916, reprinted 1930.

Romans Commentaries

Achtemeier, Paul J. Romans. Interpretation. Atlanta: John Knox Press, 1985.

Barrett, C. K. A Commentary on the Epistle to the Romans. 2nd ed. Black’s New Testament
Commentaries. London: Black, 1957.
Barth, Karl. The Epistle to the Romans. Translated by D. H. van Daalen. Richmond: John
Knox, 1933.
Barth, Karl. A Shorter Commentary on Romans. 2nd ed. Richmond: John Knox Press,
1960.
Black, Matthew. Romans. 2nd ed. The New Century Bible Commentary. Grand Rapids:
Eerdmans, 1989.
Bray, Gerald, ed. Romans. Ancient Christian Commentary on Scripture: New Testament
6. Downers Grove: InterVaristy, 1998.
Bruce, F. F. The Letter of Paul to the Romans: An Introduction and Commentary. 2nd
ed. The Tyndale New Testament Commentaries. Leicester: Inter-Varsity Press, 1985.
Byrne, Brendan. Romans. Sacra Pagina 6. Collegeville: Liturgical Press, 1996.
Cranfield, C. E. B. A Critical and Exegetical Commentary on the Epistle to the Romans. 2
Vols. International Critical Commentary. London: T & T Clark, 1975–79.
Dodd, C. H. The Epistle of Paul to the Romans. Moffatt New Testament Commentary.
London: Hodder and Stoughton Limited, 1932.
Dunn, James D. G. Romans. 2 vols. Word Biblical Commentary 38a–38b. Dallas: Word
Books, 1988.
Edwards, James R. Romans. Peabody: Hendrickson, 1992.
302 Bibliography

Fitzmyer, Joseph A. Romans: A New Translation with Introduction and Commentary.

Anchor Bible 33. New Haven: Yale University Press, 1993, reprinted 2008.
Godet, Fredric Louis. Commentary on St. Paul’s Epistle to the Romans. Translated by A.
Cusin. Edinburgh: T & T Clark, 1886–87.
Haacker, Klaus. Der Brief des Paulus an die Römer. Theologischer Handkommentar
zum Neuen Testament 6. Leipzig: Evangelische Verlagsanstalt, 2006.
Heil, John P. Paul’s Letter to the Romans: A Reader-Response Commentary. New York:
Paulist Press, 1987.
Hultgren, Arland J. Paul’s Letter to the Romans: A Commentary. Grand Rapids: Eerdmans,
2011.
Jewett, Robert. Romans: A Commentary. Hermeneia. Minneapolis: Fortress Press, 2007.
Johnson, Luke Timothy. Reading Romans: A Literary and Theological Commentary.
Reading the New Testament. New York: Crossroad, 1997.
Käsemann, Ernst. Commentary on Romans. Translated by Geoffrey W. Bromiley. Grand
Rapids: Eerdmans, 1980.
Keck, Leander. Romans. Abingdon New Testament Commentaries. Nashville:
Abingdon, 2005.
Keener, Craig S. Romans: A New Covenant Commentary. New Covenant Commentary
Series. Eugene: Cascade, 2009.
Kruse, Colin G. Paul’s Letter to the Romans. The Pillar New Testament Commentary.
Grand Rapids: Eerdmans, 2012.
Leenhardt, Franz J. L’épitre de Saint Paul aux Romains. Commentaire du Nouveau
Testament 2. Geneva: Labor et Fides, 1995.
Lohse, Eduard. Der Brief an die Römer. Kritisch-exegetischer Kommentar über das
Neue Testament 4. Göttingen: Vandenhoeck & Ruprecht, 2003.
Longenecker, Richard N. The Epistle to the Romans: A Commentary on the Greek Text.
The New International Greek Testament Commentary. Grand Rapids: Eerdmans,
2016.
Matera, Frank J. Romans. Paideia Commentaries on the New Testament. Grand Rapids:
Baker, 2010.
Moo, Douglas J. The Epistle to the Romans. The New International Commentary on the
New Testament. Grand Rapids: Eerdmans, 1996.
Morgan, Robert H. Romans. New Testament Guides. Sheffield: Sheffield Academic
Press, 1995.
Morris, Leon. The Epistle to the Romans. Pillar New Testament Commentary. Grand
Rapids: Eerdmans, 1988.
Mounce, Robert H. Romans. The New American Commentary 27. Nashville: Broadman
and Holman, 1995.
Murray, John. The Epistle to the Romans: the English Text with Introduction, Exposition
and Notes. The New International Commentary on the New Testament. Grand
Rapids: Eerdmans, 1968.
Bibliography 303

Nygren, Anders. Commentary on Romans. Translated by Carl C. Rasmussen.

Philadelphia: Fortress Press, 1949.
Porter, Stanley E. The Letter to the Romans: A Linguistic and Literary Commentary. New
Testament Monographs 37. Sheffield: Sheffield Phoenix, 2015.
Rhys, Howard. The Epistle to the Romans. New York: Macmillan, 1961.
Sanday, William, and Arthur Headlam. A Critical and Exegetical Commentary on the
Epistle to the Romans. 5th ed. International Critical Commentary. Edinburgh: T & T
Clark, 1902.
Schreiner, Thomas R. Romans. Baker Exegetical Commentary on the New Testament 6.
Grand Rapids: Baker Academic, 1998.
Stuhlmacher, Peter. Paul’s Letter to the Romans: A Commentary. Translated by Scott J.
Hafemann. Louisville: Westminster/John Knox Press, 1994.
Talbert, Charles H. Romans. Smyth & Helwys Bible Commentary. Macon: Smyth and
Helwys, 2002.
Wilckens, Ulrich. Der Brief an die Römer. 3rd ed. 3 Vols. Evangelisch-Katholischer
Kommentar zum Neuen Testament. Neukirchen-Vluyn: Neukirchener Verlag, 1997.
Witherington III, Ben, with Darlene Hyatt. Paul’s Letter to the Romans: A Socio-
Rhetorical Commentary. Grand Rapids: Eerdmans, 2004.
Wright, N. T. “The Letter to the Romans.” Pages 393–770 in The New Interpreter’s Bible.
Vol. 10. Edited by Leander E. Keck. Nashville: Abingdon, 2002.
Zeller, Dieter. Der Brief an die Römer. Regensburg: Friedrick Pustet, 1985.
Ziesler, J. A. Paul’s Letter to the Romans. London: SCM Press, 1989.

Other Works

Achtemeier, Paul J. “Romans 3:1–8: Structure and Argument.” Anglican Theological

Review 11 (1990): 77–87.
Amato, Eugenio and Jacques Schamp, eds. ἨΘΟΠΟΙΙΑ: La representation de carac-
teres entre fiction scolaire et réalité vivante a l’époque imperial et tardive. Cardo 3.
Salerno: Helios, 2005.
Anderson Jr., R. Dean. Ancient Rhetorical Theory and Paul. Revised. Contributions to
Biblical Exegesis & Theology 18.Leuven: Peeters, 1999.
Anderson Jr., R. Dean. Glossary of Greek Rhetorical Terms Connected to Methods of
Argumentation, Figures and Tropes, From Anaxamenes to Quintilian. Contributions
to Biblical Exegesis and Theology 24. Leuven: Peeters, 2000.
Aune, David E. The New Testament in its Literary Environment. Library of Early
Christianity. Philadelphia: Westminster Press, 1987.
Aune, David E. “Romans as a Logos Protreptikos.” Pages 278–96 in The Romans Debate.
Edited by Karl P. Donfried. Revised. Grand Rapids: Baker Academic, 1991.
304 Bibliography

Aune, David E. The Westminster Dictionary of New Testament and Early Christian
Literature and Rhetoric. Lousiville: Westminster John Knox Press, 2003.
Baldwin, Charles Sears. Ancient Rhetoric and Poetic: Interpreted from Representative
Works. Gloucester: Peter Smith, 1924, reprint 1959.
Barclay, John M. G. Paul and the Gift. Grand Rapids: Eerdmans, 2015.
Barrett, C. K. The First Epistle to the Corinthians. Harper’s New Testament Commentaries.
New York: Harper & Row, 1968.
Barrett, C. K. The Second Epistle to the Corinthians. Harper’s New Testament
Commentaries. New York: Harper & Row, 1973.
Bassler, Jouette. Divine Impartiality: Paul and a Theological Axiom. Society of Biblical
Literature Dissertation Series 59. Chico: Scholars Press, 1982.
Bates, Matthew W. The Hermeneutics of the Apostolic Proclamation: The Center of Paul’s
Method of Scriptural Interpretation. Waco: Baylor University Press, 2012.
Bauer, Walter, Frederick William Danker, W. F. Arndt, and F. W. Gingrich, eds. A Greek-
English Lexicon of the New Testament and other Early Christian Literature. 3rd edi-
tion. Chicago: University of Chicago Press, 2000.
Beker, J. Christiaan. “The Faithfulness of God and Priority of Israel in Paul’s Letter to
the Romans.” Harvard Theological Review 79 (1986): 10–16.
Betz, Hans Dieter. 2 Corinthians 8 and 9. Hermeneia. Philadelphia: Fortress Press, 1985.
Betz, Hans Dieter. Galatians: A Commentary. Hermeneia. Philadelphia: Fortress Press,
1979.
Blass, F. and A. Debrunner. A Greek Grammar of the New Testament and Other Early
Christian Literature. Translated and Revised by Robert W. Funk. Chicago: University
of Chicago Press, 1961.
Bonner, Stanley F. Education in Ancient Rome: From the elder Cato to the younger Pliny.
Berkeley: University of California Press, 1977.
Bornkamm, Günther. Die Vorgeschichte des sogennanten Zweiten Korintherbriefes.
Heidelberg: Winter, 1961.
Bruce, F. F. Commentary on Galatians. New International Greek Testament
Commentary. Grand Rapids: Eerdmans, 1982.
Bultmann, Rudolf. Der Stil der paulinischen Predigt und die kynisch-stoische Diatribe.
Forschungen zur Religion und Literatur des Alten und Neuen Testaments 13.
Göttingen: Vandenhoeck and Ruprecht, 1910. Repr., Göttingen: Vandenhoeck and
Ruprecht, 1984.
Campbell, Douglas A. The Deliverance of God: An Apocalyptic Rereading of Justification
in Paul. Grand Rapids: Eerdmans, 2009.
Campbell, Douglas A. “Romans 1:17—A Crux Interpretum for the ΠΙΣΤΙΣ ΧΡΙΣΤΟΥ.
Debate.” Journal of Biblical Literature 113.2 (1994): 265–85.
Campbell, W. S. “Romans III as a Key to the Structure and Thought of the Letter.”
Novum Testamentum 23.1 (1983): 22–40.
Bibliography 305

Cancik, Hildegard. Untersuchungen zu Senecas epistulae morales. Spudasmata 18.

Hildesheim: Georg Olms, 1976.
Capelle, Wilhelm, and Henri-Irénée Marrou. “Diatribe.” Pages 990–1009 in Reallexicon
für Antike und Christentum: Sachworterbuch zur Auseinandersetzung des
Christentums mit der antiken Welt. Theodor Klauser, ed. Vol. 3. Stuttgart: Anton
Hiersemann, 1957.
Clark, Donald Lemen. Rhetoric in Greco-Roman Education. Westport: Greenwood Press,
1957.
Conzelmann, Hans. 1 Corinthians. Hermeneia. Philadelphia: Fortress Press, 1975.
Cosgrove, Charles H. “What if Some have not believed: the Occasion and Thrust of
Romans 3:1–8.” Zeitschrift für die neutestamentliche Wissenschaft und die Kunde der
älteren Kirche 78 (1987): 90–105.
Dahl, Nils A. “Romans 3:9: Text and Meaning.” Pages 184–204 in Paul and Paulinism:
Essays in Honour of C. K. Barrett. Edited by M. D. Hooker and S. G. Wilson. London:
SPCK, 1982.
Das, A. Andrew. “The Gentile-Encoded Audience of Romans: The Church Outside the
Synagogue.” Pages 29–46 in Reading Paul’s Letter to the Romans. Edited by Jerry L.
Sumney. Resources for Biblical Study 73. Atlanta: Society of Biblical Literature, 2012.
Das, A. Andrew. Solving the Romans Debate. Minneapolis: Fortress Press, 2007.
De Boer, Martinus C. Galatians: A Commentary. New Testament Library. Louisville:
Westminster John Knox, 2011.
Doeve, J. W. “Some Notes with Reference to ΤΑ ΛΟΓΙΑ ΤΟΥ ΘΕΟΥ in Romans III 2.”
Pages 111–23 in Studia Paulina in honorem J. de Zwann. Edited by J. N. Sevenster and
W. C. van Unnik. Haarlem: Bohn, 1953.
Donfried, Karl P. “False Presuppositions in the Study of Romans.” Pages 102–25 in The
Romans Debate: Revised and Expanded Edition. Edited by Karl P. Donfried. Peabody:
Hendrickson, 1991.
Donfried, Karl P., ed. The Romans Debate: Revised and Expanded Edition. Peabody:
Hendrickson, 1991.
Dunn, James D. G. The Epistle to the Galatians. Black’s New Testament Commentary.
Peabody: Hendrickson, 1993.
Dunn, James D. G. The Theology of Paul the Apostle. Grand Rapids: Eerdmans, 1998.
Dunn, James D. G., ed. Paul and the Mosaic Law. Wissenschaftliche Untersuchungen
zum Neuen Testament 89. Tübingen: J. C. B. Mohr (Paul Siebeck), 1996.
Elliott, Neil. The Arrogance of Nations: Reading Romans in the Shadow of Empire. Paul in
Critical Contexts. Minneapolis: Fortress Press, 2008.
Elliott, Neil. The Rhetoric of Romans: Argumentative Constraint and Strategy and Paul’s
Dialogue with Judaism. Journal for the Study of the New Testament Supplement 45.
Sheffield: Sheffield Academic Press, 1990. Repr., Minneapolis: Fortress Press, 2007.
306 Bibliography

Fee, Gordon D. The First and Second Letters to the Thessalonians. The New International
Commentary on the New Testament. Grand Rapids: Eerdmans, 2009.
Fee, Gordon D. The First Epistle to the Corinthians. The New International Commentary
on the New Testament. Grand Rapids: Eerdmans, 1987.
Fee, Gordon D. Paul’s Letter to the Philippians. The New International Commentary on
the New Testament. Grand Rapids: Eerdmans, 1995.
Fitzgerald, John T., Thomas H. Olbricht, and L. Michael White, eds. Early Christianity
and Classical Culture: Comparative Studies in Honor of Abraham J. Malherbe.
Supplements to Novum Testamentum 105. Leiden: Brill, 2003.
Garland, David. 1 Corinthians. Baker Exegetical Commentary on the New Testament.
Grand Rapids: Baker Academic, 2003.
Garroway, Joshua. “Under Sin: Finding the Antecedent for Paul’s Charge in Rom 3:9b.”
Paper presented at the annual meeting of the Society of Biblical Literature. Atlanta,
November 21, 2015.
Gathercole, Simon J. “A Law unto Themselves: The Gentiles in Romans 2:14–15
Revisited.” Journal for the Study of the New Testament 85 (2002): 27–49.
Gathercole, Simon J. Where is Boasting? Early Jewish Soteriology and Paul’s Response
in Romans 1–5. Grand Rapids: Eerdmans, 2002.
Gaventa, Beverly Roberts. First and Second Thessalonians. Interpretation. Louisville:
John Knox, 1989.
Gaventa, Beverly Roberts. “God Handed Them Over.” Pages 113–23 in Our Mother
St. Paul. Louisville: Westminster John Knox, 2007.
Gaventa, Beverly Roberts. “Is Pauline Theology Just a ‘Guy Thing’?” Pages 63–75 in Our
Mother St. Paul. Louisville: Westminster John Knox, 2007.
Gaventa, Beverly Roberts. “On the Calling-into-Being of Israel: Romans 9:6–29.” Pages
255–69 in Between Gospel and Election. Edited by Florian Wilk and J. Ross Wagner.
Wissenschaftliche Untersuchungen zum Neuen Testament 257. Tübingen: Mohr
Siebeck, 2010.
Gaventa, Beverly Roberts. “Questions about Nomos, Answers about Christos: Romans
10:4 in Context.” Forthcoming.
Gaventa, Beverly Roberts. “‘Where Then is Boasting?’: Romans 3:27 and its Contexts.”
Proceedings (1985): 57–66.
Gibson, Craig A. “Prosopopoeia in the New Testament: Where should we look, and what
should we expect to find?” Paper presented at the annual meeting of the Society of
Biblical Literature, Philadelphia, PA, November, 2005.
Gignac, Alain. “Proces de l’human ou proces de Dieu? Le jeu intertextuel entre Rm 3,
1–9 et Ps 50 (LXX).” Revue Biblique 112 (2005): 46–62.
Graham Jr., Michael T. “An Examination of Paul’s Use of Ps 51:4 in Rom 3:4 and Its
Implications on Rom 3:1–8.” Paper presented at the annual meeting of the Society
of Biblical Literature. Atlanta, November 22, 2015.
Bibliography 307

Grieb, A. Katherine. “The Righteousness of God in Romans.” Pages 65–78 in Reading

Paul’s Letter to the Romans. Society of Biblical Literature Resources for Biblical
Study 73. Edited by Jerry L. Sumney. Atlanta: Society of Biblical Literature, 2012.
Halbauer, Otto. De Diatribis Epicteti. Leipzig: Robert Noske Bornen, 1911.
Hall, David R. “Romans 3:1–8 Reconsidered.” New Testament Studies 29.2 (1983): 183–97.
Hays, Richard B. The Conversion of the Imagination: Paul as Interpreter of Israel’s
Scripture. Grand Rapids: Eerdmans, 2005.
Hays, Richard B. “ ‘The Gospel is the Power of God for the Gentiles Only’? A Critique
of Stanley Stowers’ A Rereading of Romans.” Critical Review of Books in Religion 9
(1996): 27–44.
Hays, Richard B. Echoes of Scripture in the Letters of Paul. New Haven: Yale University
Press, 1989.
Hays, Richard B. The Faith of Jesus Christ: the Narrative Substructure of Galatians 3:1–4:11.
2nd ed. The Biblical Resource Series. Grand Rapids: Eerdmans, 2002.
Hays, Richard B. The Letter to the Galatians: Introduction, Commentary, and Reflections.
New Interpreters Bible 11. Nashville: Abingdon, 2000.
Hays, Richard B. “Psalm 143 and the Logic of Rom. 3.” Journal of Biblical Literature 99.1
(1980): 107–15.
Hays, Richard B. “Three Dramatic Roles: The Law in Romans 3–4.” Pages 151–64 in Paul
and the Mosaic Law. Edited by James D. G. Dunn. Wissenschaftliche Untersuchungen
zum Neuen Testament 89. Tübingen: J. C. B. Mohr (Paul Siebeck), 1996. Reprinted
in The Conversion of the Imagination: Paul as Interpreter of Israel’s Scripture. Pages
85–100. Grand Rapids: Eerdmans, 2005.
Heath, Malcom. “Theon and the History of the Progymnasmata.” Greek, Roman, and
Byzantine Studies 43 (2002/3): 129–60.
Heinrici, Carl Friedrich Georg. Das erste Sendschreiben des Apostel Paul an die Korinther.
Berlin: Wilhelm Hertz, 1880.
Heinrici, Carl Friedrich Georg. Das zweite Sendschreiben des Apostel Paulus an die
Korinther. Berlin: Wilhelm Hertz, 1887.
Heinrici, Carl Friedrich Georg. Der litterarische Charakter der neutestamentlichen
Schriften. Leipzig: Durr, 1908.
Heusch, Christine. “Die Ethopoiie in der griechischen und lateinischen Antike: von
der rhetorischen Progymnasma-Theorie zur literarischen Form.” Pages 11–33 in
‘ΗΘΟΠΟΙΙΑ: La représentation de caracteres entre fiction scolaire et réalité vivante a
l’époque imperial et tardive (edited by Eugenio Amato and Jacques Schamp; Cardo
3; Salerno: Helios, 2005).
Hirzel, Rudolf. Der Dialog: Ein literarhistorischer Versuch. 2 Volumes. Leipzig: S. Hirzel,
1895.
Hock, Ronald F. “The Rhetoric of Romance,” pages 445–65 in Handbook of Classical
Rhetoric in the Hellenistic Period: 330 B.C.–A. D. 400. Edited by Stanley E. Porter.
Leiden: Brill, 1997.
308 Bibliography

Hock, Ronald F., and Edward N. O’Neil. The Chreia in Ancient Rhetoric: Volume I: The
Progymnasmata. Society of Biblical Literature Texts and Translations Series 27.
Graeco-Roman Religion Series 9. Atlanta: Scholars Press, 1986.
Hooker, Morna D. and S. G. Wilson, eds. Paul and Paulinism: Essays in Honour of C. K.
Barrett. London: SPCK, 1982.
Johnson, Luke Timothy. “The New Testament’s Anti-Jewish Slander and the Conventions
of Ancient Polemic.” Journal of Biblical Literature 108.3 (1989): 419–41.
Käsemann, Ernst. “‘The Righteousness of God’ in Paul.” Pages 168–82 in New Testament
Questions of Today. Translated by W. J. Montague. Philadelphia: Fortress Press, 1969.
Keener, Craig S. 1–2 Corinthians. The New Cambridge Bible Commentary. Cambridge:
Cambridge University Press, 2005.
Kennedy, George A. The Art of Rhetoric in the Roman World. Princeton: Princeton
University Press, 1972.
Kennedy, George A. Greek Rhetoric Under Christian Emperors. A History of Rhetoric 3.
Princeton: Princeton University Press, 1983.
Kennedy, George A. A New History of Classical Rhetoric. Princeton: Princeton University
Press, 1994.
King, Justin. “Rhetorical Chain-Link Construction and the Relationship between
Romans 7:1–6 and 7:7–8:39: Additional Evidence for Assessing the Argument of
Romans 7–8 and the Identity of the Infamous ‘I.’” Journal for the Study of the New
Testament 39.3 (2017): 258–78.
King, Justin. “Speech-in-Character, Diatribe, and Romans 3:1–9: Who’s Speaking When
and Why It Matters.” Ph.D. diss., Baylor University, 2016.
Klauck, Hans-Josef. Ancient Letters and the New Testament: A Guide to Context and
Exegesis. Waco: Baylor University Press, 2006.
Klein, Günter. “Paul’s Purpose in Writing the Epistle to the Romans.” Pages 29–43 in The
Romans Debate. Edited by Karl P. Donfried. Revised. Grand Rapids: Baker Academic,
1991.
Krauss, Fritz. Die rhetorischen Schriften Plutarchs und ihre Stellung im Plutarchischen
Schriftkorpus. Nürnberg: J. L. Stitch, 1912.
Kustas, George L. Diatribe in Ancient Rhetorical Theory. Protocol of the colloquy of the
Center for Hermeneutical Studies in Hellenistic and Modern Culture 22. Berkeley:
Center for Hermeneutical Studies, 1976.
Lampe, Peter. From Paul to Valentinus: Christians at Rome in the First Two Centuries.
Minneapolis: Fortress Press, 2003.
Lampe, Peter. “The Roman Christians of Romans 16.” Pages 216–30 in The Romans
Debate. Edited by Karl P. Donfried. Revised. Grand Rapids: Baker Academic, 1991.
Lausberg, Heinrich. Handbook of Literary Rhetoric: A Foundation for Literary Study.
Translated by Matthew T. Bliss, Annemiek Jansen, and David E. Orton. Edited by
David E. Orton and R. Dean Anderson. Leiden: Brill, 1998.
Bibliography 309

Linebaugh, Jonathan A. “Announcing the Human: Rethinking the Relationship be-

tween Wisdom of Solomon 13–15 and Romans 1:18–2:11.” New Testament Studies 57.2
(2011): 214–37.
Linebaugh, Jonathan A. God, Grace, and Righteousness in Wisdom of Solomon and Paul’s
Letter to the Romans: Texts in Conversation. Supplements to Novum Testamentum
152. Leiden: Brill, 2013.
Longenecker, Bruce W. “Different Answers to Different Issues: Israel, the Gentiles and
Salvation History in Romans 9–11. Journal for the Study of the New Testament 36
(1989): 95–123.
Longenecker, Bruce W. Eschatology and the Covenant: A Comparison of 4 Ezra and
Romans 1–11. Journal for the Study of the New Testament, Supplement 57. Sheffield:
JSOT Press, 1991.
Longenecker, Bruce W. “Pistis in Romans 3.25: Neglected Evidence for the ‘Faithfulness
of Christ’?” New Testament Studies 39 (1993): 478–80.
Longenecker, Bruce W. Rhetoric at the Boundaries: The Art and Theology of New
Testament Chain-link Transitions. Waco: Baylor University Press, 2005.
Longenecker, Bruce W. “Salvation History in Galatians and the Making of a Pauline
Discourse.” Journal for the Study of Paul and His Letters 2.2 (2012): 65–87.
Longenecker, Bruce W. “Sharing in Their Spiritual Blessings? The Story of Israel in
Galatians and Romans.” Pages 58–84 in Narrative Dynamics in Paul: A Critical
Assessment. Edited by Bruce W. Longenecker. Louisville: Westminster John Knox,
2002.
Longenecker, Bruce W. The Triumph of Abraham’s God: the Transformation of Identity in
Galatians. Edinburgh: T & T Clark, 1998.
Longenecker, Richard N. Galatians. Word Biblical Commentary 41. Nashville: Thomas
Nelson, 1990.
Maertens, Philip. “Une etude de Rm 2.12–16.” New Testament Studies 46.4 (2000): 504–19.
Maier, Harry O. Picturing Paul in Empire: Imperial Image, Text and Persuasion in
Colossians, Ephesians and the Pastoral Epistles. London: Bloomsbury, 2013.
Malherbe, Abraham. “Hellenistic Moralists and the New Testament.” Principat 26.1
(1992): 313–20.
Malherbe, Abraham. The Letters to the Thessalonians: A New Translation with
Introduction and Commentary. Anchor Bible 32B. New York: Double Day, 2000.
Malherbe, Abraham. “Μη Γενοιτο in the Diatribe and in Paul.” Harvard Theological
Review 73 (1980): 231–40.
Malherbe, Abraham. Social Aspects of Early Christianity. Baton Rouge: Louisiana State
University Press, 1977.
Marshall, Peter. Enmity in Corinth: Social Conventions in Paul’s Relations with the
Corinthians. Wissenschaftliche Untersuchungen zum Neuen Testament 2.23.
Tübingen: J. C. B. Mohr, 1983.
310 Bibliography

Martin, Ralph P. 2 Corinthians. Word Biblical Commentary 40. Waco: Word Books, 1986.
Martyn, J. Louis. Galatians. Anchor Bible 33A. New York: Doubleday, 1997.
Martyn, J. Louis. Theological Issues in the Letters of Paul. Nashville: Abingdon Press,
1997.
Matera, Frank J. II Corinthians: A Commentary. The New Testament Library. Louisville:
Westminster John Knox, 2003.
Mitchell, Margaret M. Paul and the Rhetoric of Reconciliation: An Exegetical Investigation
of the Language and Composition of 1 Corinthians. Louisville: Westminster / John
Knox Press, 1992.
Mitchell, Margaret M. “Paul’s Letters to Corinth: The Interpretive Intertwining of
Literary and Historical Reconstruction.” Pages 307–38 in Urban Religion in Roman
Corinth, Interdisciplinary Approaches. Edited by Daniel Showalter and Steven J.
Friesen. Cambridge: Harvard University Press, 2005.
Morris, Leon. The First Epistle of Paul to the Corinthians. Tyndale New Testament
Commentaries 7. Grand Rapids: Eerdmans, 1980.
Myers, Alicia D. Characterizing Jesus: A Rhetorical Analysis on the Fourth Gospel’s Use of
Scripture in its Presentation of Jesus. Library of New Testament Studies 458. London:
T & T Clark, 2012.
Norden, Eduard. Die Antike Kunstprosa vom VI. Jahrhundert v. Chr. Bis in die Zeit der
Renaissance. Leipzig: Teubner, 1898.
O’Brien, Peter T. The Epistle to the Philippians. New International Greek Testament
Commentary. Grand Rapids: Eerdmans, 1991.
Oakes, Peter. “Re-mapping the Universe: Paul and the Emperor in 1 Thessalonians and
Philippians.” Journal for the Study of the New Testament 27.3 (2005): 301–22.
Palmer, Michael. “Ti oun: the Inferential Question in Paul’s Letter to the Romans with
a Proposed Reading of Romans 4:1.” Pages 200–218 in Discourse Analysis and Other
Topics in Biblical Greek. Edited by Stanley E. Porter and D. A. Carson. Sheffield:
Sheffield Academic Press, 1995.
Perkins, Pheme. First Corinthians. Paideia. Grand Rapids: Baker Academic, 2012.
Pernot, Laurent. Rhetoric in Antiquity. Translated by W. E. Higgins. Washington D. C.:
Catholic University of America Press, 2005.
Peterman, Gerald. Paul’s Gift from Philippi: Conventions of Gift Exchange and Christian
Giving. Society for New Testament Studies Monograph Series 92. Cambridge:
Cambridge University Press, 1997.
Piwonka, Mario Puelma. Lucilius und Kallimachos. Frankfurt: V. Klostermann, 1949.
Pogoloff, Stephen M. Logos and Sophia: The Rhetorical Situation of 1 Corinthians. Society
of Biblical Literature Dissertation Series 134. Atlanta: Scholars Press, 1992.
Polster, Carol. “A Conversation Halved: Epistolary Theory in Greco-Roman Antiquity.”
Pages 21–51 in Letter Writing Manuals and Instruction from Antiquity to the Present.
Edited by Carol Polster and Linda C. Mitchell. Historical and Bibliographic Studies.
Columbia: University of South Carolina Press, 2007.
Bibliography 311

Porter, Stanley E. “The Argument of Romans 5: Can a Rhetorical Question Make a

Difference.” Journal of Biblical Literature 110.4 (1991): 655–77.
Porter, Stanley E., ed. Handbook of Classical Rhetoric in the Hellenistic Period: 330 B.C.–
A.D. 400. Leiden: Brill, 1997.
Quarles, Charles L. “From Faith to Faith: A Fresh Examination of the Prepositional
Series in Romans 1:17.” Novum Testamentum 45 (2003): 1–21.
Rahn, Helmut. Morphologie der antike Literatur: Eine Einführung. Darmstadt:
Wissenschaftliche Buch-Gesellschaft, 1969.
Räisänen, Heikki. Paul and the Law. 2nd ed. Wissenschaftliche Untersuchungen zum
Neuen Testament 29. Tübingen: J. C. B. Mohr (Paul Siebeck), 1987.
Räisänen, Heikki. “Zum Verständnis von Röm 3, 1–8.” Pages 185–203 in The Torah and
Christ: Essays in German and English on the Problem of the Law in Early Christianity.
Publications of the Finnish Exegetical Society 45. Helsinki: Finnish Exegetical
Society, 1986.
Reichel, Georg. “Questiones Progymnasmaticae.” Ph.D. diss., Leipzig, 1909.
Reid, Marty L. “A Consideration of the Function of Rom 1:8–15 in Light of Greco-Roman
Rhetoric.” Journal of the Evangelical Theology Society 38.2 (1995): 181–91.
Rodríguez, Rafael. If You Call Yourself a Jew: Reappraising Paul’s Letter to the Romans.
Eugene: Cascade, 2014.
Roetzel, Calvin J. 2 Corinthians. Abingdon New Testament Commentaries. Nashville:
Abingdon, 2007.
Rosenmeyer, Patricia A. Ancient Epistolary Fictions: The Letter in Greek Literature.
Cambridge: Cambridge University Press, 2001.
Sampley, J. Paul. The First Letter to the Corinthians: Introduction, Commentary, and
Reflections. The New Interpreter’s Bible 10. Nashville: Abingdon, 2002.
Sampley, J. Paul. The Second Letter to the Corinthians: Introduction, Commentary, and
Reflections. The New Interpreter’s Bible 11. Nashville: Abingdon, 2000.
Schellenberg, Ryan S. Rethinking Paul’s Rhetorical Education: Comparative Rhetoric and
2 Corinthians 10–13. Early Christianity and its Literature. Atlanta: Society of Biblical
Literature, 2013.
Schmeller, Thomas. Paulus und die “Diatribe:” Eine vergleichende Stilinterpretation.
Neutestamentliche Abhandlungen 19. Münster: Aschendorff, 1987.
Schmidt, E. G. “Diatribe and Satire.” Wissenschaftliche Zeitschrift der Universität Rostock
15 (1966): 507–15.
Schouler, B. “L’éthopée chez Libanios ou l’evasion esthétique.” Pages 79–92 in
‘ΗΘΟΠΟΙΙΑ: La représentation de caracteres entre fiction scolaire et réalité vivante a
l’époque imperial et tardive (edited by Eugenio Amato and Jacques Schamp; Cardo
3; Salerno: Helios, 2005).
Sinko, Tadeusz. “On the So-called Cynic-Stoic Diatribe” (in Polish). Eos 21 (1916): 21–63.
Smith, Abraham. The First Letter to the Thessalonians: Introduction, Commentary, and
Reflections. New Interpreters Bible 11. Nashville: Abingdon, 2000.
312 Bibliography

Smyth, Herbert Weir. Greek Grammar. Revised by Gordon M. Messing. Cambridge:

Harvard University Press, 1984.
Song, Changwon. Reading Romans as a Diatribe. Studies in Biblical Literature 59. New
York: Peter Lang, 2004.
Stanley, Christopher D. Arguing with Scripture: The Rhetoric of Quotations in the Letters
of Paul. New York: T & T Clark International. 2004.
Stegemann, Willy. “Theon (5),” PW 5A (1934): 2036–54.
Stendahl, Krister. “Paul and the Introspective Conscience of the West.” Pages 78–96 in
Paul Among Jews and Gentiles and Other Essays. Philadelphia: Fortress Press, 1976.
Stowers, Stanley K. “Apostrophe, Προσωποποιια, and Paul’s Rhetorical Education.” Pages
351–69 in Early Christianity and Classical Culture: Comparative Studies in Honor of
Abraham J. Malherbe (eds. John T. Fitzgerald, Thomas H. Olbricht, and L. Michael
White; NovTSup 105; Leiden: Brill, 2003).
Stowers, Stanley K. “Diatribe.” Pages 190–93 in The Anchor Bible Dictionary. Vol. 2.
Edited by David Noel Freedman. New York: Doubleday, 1992.
Stowers, Stanley K. “The Diatribe.” Pages 71–83 in Greco-Roman Literature and the New
Testament: Selected Forms and Genres. Edited by David E. Aune. Atlanta: Scholars
Press, 1988.
Stowers, Stanley K. The Diatribe and Paul’s Letter to the Romans. Society of Biblical
Literature Dissertation Series 57. Chico: Society of Biblical Literature, 1981.
Stowers, Stanley K. Letter Writing in Greco-Roman Antiquity. Library of Early Christianity
5. Philadelphia: Westminster Press, 1989.
Stowers, Stanley K. “Paul’s Dialogue with a Fellow Jew in Romans 3:1–9.” Catholic
Biblical Quarterly 46 (1984): 707–22.
Stowers, Stanley K. A Rereading of Romans: Justice, Jews, and Gentiles. New Haven: Yale
University Press, 1994.
Stowers, Stanley K. Review of T. Schmeller, Paulus und die ‘Diatribe’: Eine vergleichende
Stilinterpretation. Journal of Biblical Literature 108.3 (1989): 538–40.
Stowers, Stanley K. “Romans 7:7–25 as a Speech-in-Character Prosopopoiia.” Pages
180–202 in Paul in his Hellenistic Context. Edited by Troels Engberg-Pederson.
Minneapolis: Fortress Press, 1995.
Stuhlmacher, Peter. “The Purpose of Romans.” Pages 231–44 in The Romans Debate.
Edited by Karl P. Donfried. Revised. Grand Rapids: Baker Academic, 1991.
Talbert, Charles H. Reading Corinthians: A Literary and Theological Commentary.
Reading the New Testament. Revised edition. Macon: Smyth & Helwys, 2002.
Thiselton, Anthony C. The First Epistle to the Corinthians: A Commentary on the
Greek Text. The New International Greek Testament Commentary. Grand Rapids:
Eerdmans, 2000.
Thorsteinsson, Runar M. Paul’s Interlocutor in Romans 2: Function and Identity in the
Context of Ancient Epistolography. Coniectanea Biblica New Testament Series 45.
Stockholm: Almqvist and Wiksell International, 2003.
Bibliography 313

Thrall, Margaret. A Critical and Exegetical Commentary on the Second Epistle to the
Corinthians. 2 vols. The International Critical Commentary. London: T & T Clark,
2000.
Thromm, Hermann. Die Thesis. Rhetorical Studies 17. Paderborn: F. Schöningh, 1932.
Ticciatti, Susannah. “The Nondivisive Difference of Election: A Reading of Romans
9–11.” Journal of Theological Interpretation 6.2 (2012): 257–78.
Timmons, Will. “Romans 7 and Speech-in-Character: A Critical Evaluation of Stowers’
Hypothesis.” Zeitschrift für die Neutestamentliche Wissenschaft 107.1 (2016): 94–115.
Tobin, Thomas H. “Controversy and Continuity in Romans 1:18–3:20.” Catholic Biblical
Quarterly 55 (1993): 298–318.
Tobin, Thomas H. Paul’s Rhetoric in its Contexts: The Argument of Romans. Peabody:
Hendrickson, 2004.
Tonstad, Sigve. “The Revisionary Potential of ‘Abba! Father!’ in the Letters of Paul.”
Andrews University Seminary Studies 45.1 (2007): 5–18.
Usener, Herman. Epicurea. Berlin: Teubner, 1887.
Vouga, François. “Romains 1.18–3.20 comme narration.” Pages 145–61 in La Narration:
Quand le récit deviant communication. Edited by P. Bühler and J.-F. Habermacher.
Lieux Théologiques 12. Geneva: Labor et Fides, 1988.
Wagner, J. Ross. Heralds of the Good News: Isaiah and Paul “In Concert” in the Letter to
the Romans. Novum Testamentum Supplements 101. Leiden: Brill, 2002.
Wallace, Daniel B. Greek Grammar Beyond the Basics: An Exegetical Syntax of the New
Testament. Grand Rapids: Eerdmans, 1996.
Wallach, Barbara P. “A History of the Diatribe from its Origin up to the First Century B.C.
and a Study of the Influence of the Genre upon Lucretius.” Ph.D. diss., University of
Illinois, 1974.
Watson, Francis. Paul and the Hermeneutics of Faith, London: T & T Clark, 2004.
Watson, Francis. Paul, Judaism, and the Gentiles: Beyond the New Perspective. Revised
and expanded. Grand Rapids: Eerdmans, 2007.
Watson, Francis. “The Two Roman Congregations: Romans 14:1–15:13. Pages 203–15
in The Romans Debate. Edited by Karl P. Donfried. Revised. Grand Rapids: Baker
Academic, 1991.
Weber, Heinrich. De Senecae philosophi dicendi genere Bioneo. Marburg: Friederici
Soemmering, 1895.
Weima, Jeffrey A. D. “‘Peace and Security’ (1 Thess 5:3): Prophetic Warning or Political
Propaganda?” New Testament Studies 58.3 (2012): 331–59.
Weiss, Johannes. “Beiträge zur paulinischen Rhetorik,” in Theologische Studien:
Bernhard Weiss ze seinem 70. Geburtstag dargebracht. Göttingen: Vandenhoeck &
Ruprecht, 1897.
Weiss, Johannes. Der erste Korintherbrief. Göttingen: Vandenhoeck & Ruprecht, 1910.
Weiss, Johannes. Die Aufgaben der neutestamentlichen Wissenschaft. Göttingen:
Vandenhoeck & Ruprecht, 1908.
314 Bibliography

Welborn, L. L. “On the Discord in Corinth: 1 Corinthians 1–4 and Ancient Politics.”
Journal of Biblical Literature 106 (1987): 83–113.
Wendland, Paul. Die hellenistische-römische Kultur in ihren Beziehungen zu Judentum
und Christentum. Handbuch zum Neuen Testament 1. Tübingen: J. C. B. Mohr (Paul
Siebeck), 1907.
Wendland, Paul and Otto Kern. Beiträge zur Geschichte der griechischen Philosophie
und Religion. Berlin: Georg Reimer, 1895.
White, Joel R. “‘Peace and Security’ (1 Thessalonians 5:3): Is it Really a Roman Slogan?”
New Testament Studies 59.3 (2013): 382–95.
White, John L. The Apostle of God: Paul and the Promise of Abraham. Peabody:
Hendrickson, 1999.
Whittle, Sarah. Covenant Renewal and the Consecration of the Gentiles in Romans.
Society for New Testament Studies Monograph Series 161. Cambridge: Cambridge
University Press, 2015.
Wilamowitz-Moellendorff, Ulrich von. “Excurs 3: Der kynische Predifer Teles.” Pages
292–319 in Antigonos von Karystos. Philologische Untersuchungen IV. Berlin:
Weidmannsche Buchhandlung, 1881.
Williams, Sam K. “The ‘Righteousness’ of God in Romans.” Journal of Biblical Literature
99.2 (1980): 241–90.
Winter, Bruce W. After Paul Left Corinth: The Influence of Secular Ethics and Social
Change. Grand Rapids: Eerdmans, 2001.
Witherington III, Ben. 1 and 2 Thessalonians: A Socio-Rhetorical Commentary. Grand
Rapids: Eerdmans, 2006.
Witherington III, Ben. Grace in Galatia: A Commentary on Paul’s Letter to the Galatians.
Grand Rapids: Eermans, 1998.
Wright, N. T. The Climax of the Covenant: Christ and the Law in Pauline Theology.
Minneapolis: Fortress, 1993.
Wright, N. T. Paul and the Faithfulness of God: Book II. Christian Origins and the
Question of God. Vol. 4. Minneapolis: Fortress Press, 2013.
Wright, N. T. “Romans 2:17–3:9: A Hidden Clue to the Meaning of Romans?” Journal for
the Study of Paul and His Letters 2.1 (2012): 1–25.
Wright, N. T. “The Law in Romans 2.” Pages 131–50 in Paul and the Mosaic Law. Edited
by James D. G. Dunn. Wissenschaftliche Untersuchungen zum Neuen Testament 89.
Tübingen: J. C. B. Mohr (Paul Siebeck), 1996.
Modern Authors Index

Achtemeier, Paul J. 3n5, 170n33, 246n65, Dahl, Nils Alstrup 266n60

264n53 Das, Andrew 178, 220n5, 288n123
Anderson Jr., R. Dean 16n2, 29n19, 33n28, De Boer, Martinus C. 71n36, 72n43
90n88, 97n102, 206n38, 207–8 Dodd, C. H. 9, 165, 195, 255
Aune, David E. 16n2, 122n138, 124n149 Doeve, J. W. 227n21
Donfried, Karl Paul 288n123, 288n124
Baldwin, Charles 15n1, 16n2 Dunn, James D. G. 6, 71n36, 73n45, 89n85,
Barclay, John M. G. 7, 230n28, 232n31, 92n92, 169, 183–85, 190, 192, 194–95,
243n57, 246n68, 257n20, 275n81 240n50, 243n56, 250n85, 253n3, 257n21,
Barrett, C. K. 75n48, 85n75, 239n49, 255n14, 261n41, 261n42, 279n96, 282n106,
286n116 284n110, 285n114
Bassler, Jouette M. 236n39, 236n40, 243n55,
247n69, 250n85, 274n79, 291n131, Elliott, Neil 3n6, 4–6, 120n123, 124n149,
291n132 191–92, 191n184, 199, 201n19, 210–13,
Bates, Matthew W. 16n2, 69n31, 70n33, 213n70, 215, 220, 225n15, 253n5, 254,
90n88, 90n89, 93n97, 94n98 254n11, 256n15, 264n56, 268, 268n65,
Beker, J. Christiaan 279n96 272n73, 291
Betz, Hans Dieter 71n36, 72n41, 85n75
Black, Matthew 221n7, 243n56 Fee, Gordon D. 64n14, 66n24, 75n52, 81n69
Bolognesi, Gioncarlo 16n2, 38 Fitzmyer, Joseph A. 89n85, 124n149, 217n91,
Bonner, Stanley 15n1 255n13, 275n80, 279n95
Bornkamm, Günther 85n75
Bruce, F. F. 72n42 Garland, David E. 74n46, 75n49, 78n61,
Bultmann, Rudolf 2–3, 81n67, 90n88, 101, 79n63, 83n73, 84n74
103–6, 104n14, 104n15, 116–18, 124, 165, Gathercole, Simon 243n57, 246n64, 246n65
169–71, 170n37, 189, 191–93 Gaventa, Beverly Roberts 64n15, 71n35,
Butts, James R. 15n1, 40n5 89n83, 89n85, 90n86, 90n88, 92n92,
Byrne, Brendan 3n3, 169n28, 224n14, 195n204, 222n10, 245n62, 280, 281n101,
250n86, 253n5, 256n15, 264n56, 268, 286n115
268n66 Gibson, Craig A. 16n2, 41n11, 53n30, 59–62
Godet, Frederic Louis 9, 165–68, 166n7,
Campbell, Douglas A. 3n6, 5, 16n2, 94n98, 167n14, 188, 189n169, 191, 204, 257
120n123, 191, 191n184, 199, 213–18,
228n23, 234, 253n5, 255n13, 256n15, Halbauer, Otto 106–7, 117n108, 118
259n29, 264n55, 264n56, 268, 268n65 Hall, David R. 3n3, 197–99, 204, 257n21
Campbell, William S. 274n77 Hays, Richard B. 71n36, 209, 220n3, 226n18,
Cancik, Hildegard 107, 114 227n22, 230n28
Capelle, Wilhelm 107 Headlam, Arthur C. 165, 168, 191
Caplan, Harry 15n1 Heath, Malcom 38n3
Clark, Donald Lemon 16n2 Heinrici, Carl Friedrich Georg 169n23
Conzelmann, Hans 75n48, 76n53, 78n60 Hense, Otto 104n13, 109n47, 111n63
Cosgrove, Charles H. 243n57, 246n68, Heusch, Christine 16n2
257n21 Hirzel, Rudolf  104n16, 108
Cranfield, C. E. B. 89n85, 90n88, 243n57 Hock, Ronald F. 16n2
316 Modern Authors Index

Hultgren, Arland 90n88, 195n206, 223n12, Matera, Frank J. 88n82, 90n88, 192–93,

232n31, 255n13 243n56, 260
Hyatt, Darlene 3n3, 16n2, 89n85, 123n148, Mitchell, Margaret M. 74–77, 78n60, 81n67,
169n28, 231n29, 256n18, 264n56 85n75
Moo, Douglas J. 7, 89n85, 184n137, 186n161,
Jewett, Robert 7, 89n85, 90n88, 90n90, 189n169, 217, 257n21, 261n39, 266n60,
94n98, 169, 186n160, 215n86, 224n14, 267n61
232n31, 236n40, 243n57, 246n64, 248, Morris, Leon 75n48
257n21, 264, 266n60, 267n62, 267n64 Mounce, Robert H. 267n61
Johnson, Luke Timothy 90n88, 273n75 Murray, John 243n56
Myers, Alicia D. 16n2
Käsemann, Ernst 90n87, 90n88, 212n67,
273n74, 284n109 Nadeau, Ray 15n1
Keck, Leander 3n3, 89n85, 90n88, 169n28, Norden, Eduard 169n23
254n6, 256n18, 257n21, 258n26, 260n32,
264n56, 265, 268 Oakes, Peter 65–66, 67n28
Keener, Craig S. 74n46, 75n49, 81n69, 88n82, O’Neil, Edward 133n8
267n61, 285n114
Kern, Otto 114n85 Patillon, Michel 16n2, 38n2
Kennedy, George A. 8n26, 15n1, 16n2, 39n4, Perkins, Pheme 76n54, 78n61
59n4 Pernot, Laurent 16n2
Kindstrand, Jan Frederik 109n47, 110 Peterman, Gerald 289n126
King, Justin 193n194, 260n35 Piwonka, Mario Puelma 107n39
Klauck, Hans Josef 219n1 Pogoloff, Stephen M. 76n53
Klein, Günter 224n14 Porter, Stanley 5n16, 7, 127n152, 222n9,
Krauss, Fritz 112n74 243n55, 248, 263, 267n64
Kruse, Colin G. 229n24, 236n39, 243n57,
245n61, 267n61 Quarles, Chrales L. 229n24
Kustas, George L. 107. 129
Rahn, Helmut 107
Lampe, Peter 220n6 Räisäinen, Heikki 184n136
Lausberg, Heinrich 16n2 Reid, Marty L. 290n127
Linebaugh, Jonathan A. 232n30, 236n39, Rodríguez, Rafael 6, 122n138, 124n149, 169,
236n40, 237n42 178–82, 179n97, 180n111, 191n181, 191n182,
Longenecker, Bruce 33n27, 72, 126, 183n131, 192, 194, 196n209, 212n67, 220, 236n38,
193n194, 195n207, 233n33, 236n39, 237n41, 247, 248n78, 249n81, 251n87,
236n40, 241n51, 241n52, 255n13, 257n21, 253n4, 258n24, 279n96
260n35, 278, 280, 282n106 Roetzel, Calvin 85n75
Longenecker, Richard 70n34, 198n4 Rosenmeyer, Patricia A. 114n89
Lutz, Cora E. 111n63 Russell, Donald A. 15n1

Maertens, Philip 246n65 Sampley, J. Paul 75n49, 75n51, 78n61, 88n82

Malherbe, Abraham 6, 63n12, 64–65, 66n26, Sanday, William 165, 168, 191
103, 107, 123n146, 124n149, 169, 171–72, Schellenberg, Ryan S. 97n102
171n40, 189–92, 191n182, 260 Schmeller, Thomas 103, 117–20, 118n112, 124
Marrou, Henri-Irénée 107 Schmidt, E. G. 107n39
Martin, Ralph 85n75 Sinko, Tadeusz 106–7
Martyn, J. Louis 71n36, 72n37 Smith, Abraham 64n15
Modern Authors Index 317

Smyth, Herbert Weir 76n56, 83n72, 133n8 Usener, Herman 104n13, 117

Song, Changwon 5, 103, 120n123, 122–24,
123n142, 169, 171–73, 191n184, 192, 201n19, Wagner, J. Ross 284n108
253n4, 260, 267n64 Wallace, Daniel B. 83n72
Stanley, Christopher D. 70n33 Watson, Francis 94n98, 221n7, 243n57
Stowers, Stanley K. 2–6, 8, 10, 15n1, 16n2, Weima, Jeffrey A. D. 64n13, 65–66, 66n21,
29n19, 32n25, 90n88, 101, 103, 106–117, 66n26, 67n27
117n105, 118n112, 124, 129, 133n8, 136, 141, Weiss, Johannes 169n23
149n25, 153n28, 154n31, 163, 169, 171n40, Welborn, Larry L. 76
178, 191–92, 191n181, 193n194, 199–210, Wendland, Paul 113
212, 215, 217, 220, 248n78, 254, 256, White, Joel R. 65–66, 66n21, 124n149
256n17, 256n18, 257n19, 257n21, 258n26, Whittle, Sarah 285n114, 287m118
260, 261n41, 262n43, 263n49, 264n53, Wilamowitz-Moellendorf, Ulrich von 
264n56, 265, 266n60, 268, 270n70, 104n10, 108–9
288n124 Wilckens, Ulrich 236n39
Stuhlmacher, Peter 224n14 Williams, Sam K. 227n19, 230n27, 288n123
Winter, Bruce 78n62
Talbert, Charles 16n2, 75n48, 85n75, 88n82 Witherington III, Ben 3n3, 16n2, 64n15,
Thiselton, Anthony C. 75n48, 76n56, 83n72 89n85, 123n148, 169n28, 231n29, 256n18,
Thorsteinsson, Runar M. 103, 120–21, 264n56, 265, 268
120n123, 122n138, 124, 180n111, 247, Wright, N. T. 7, 9, 89n85, 165, 169, 183,
247n70, 248n78, 251n87 185–87, 185n151, 186n157, 190, 195,
Thrall, Margaret 85n75 199n12, 227n19, 232n31, 236n40, 243n57,
Thromm, Hermann 107 245n61, 246n64, 250n84, 257n21,
Ticciati, Susannah 258n22, 279n96, 285n112 258n23, 258n26, 262n45, 279n96,
Timmins, Will  33n27, 193n194, 260n35 285n114
Tite, Philip L. 287n119
Tobin, Thomas H. 5, 16n2, 90n87, 90n88, 103,
119n121, 120n123, 121–22, 122n138, 124,
169, 173–78, 189–91, 189n169, 191n184,
194–95, 205n36, 239n44, 239n45, 247,
253, n4
Tonstad, Sigve 72n43
Subject and Name Index

2 Corinthians’ composition history 85–87, mood see also tone

85n75 primary sources 104, 108, 118n112
relation to speech-in-character 97, 101,
Abraham 69–71, 71–73, 196, 227–29, 232n31, 123n148, 129–31, 136–38, 141–45, 149,
245, 253, 257–58, 260, 276–77, 280, 156–59, 206–7, 212, 267–68, 272–74, 295,
286–87, 298 297
Aelian 42n14 Stowers’s model 108–17
Aelius Aristides 46 tone 105–6, 111, 113, 115–16, 125–27,
Alciphron 42n14 135–36, 141, 145, 147–49, 154, 158, 191, 201,
Aphthonius 16, 38n1, 41n10, 46, 52, 52n29, 61 211, 251n88, 265
apostrophe 11, 29n19, 172, 202, 204, 208–9, Dio Chrysostom 112, 116
238–39, 247
attributed speech see speech-in-character eidolopoiia / εἰδολοποιία 8n26, 46–48, 51
see also speech-in-character
Bion 104, 109, 109n47, 109n48 Epictetus 104, 110–13, 116, 122, 138–59,
149n25, 171, 192–93, 254n6
censure / refutation / ἐλεγκτικός 111–14, 116, ethopoiia / ἠθοποιία 8n26, 15, 28n17,
121–22, 125–27, 129, 135, 141, 144–45, 46–48, 51
147–49, 154, 158–59, 215, 267–68, 272, see also speech-in-character
297
characterization 8–9, 11, 20, 22–26, 31, ficta oratio 33
34–35, 54–55, 64, 66n26, 67–68, 70, see also speech-in-character
72–73, 75, 77, 79–80, 82, 87–88, 130, 137, fictiones personarum 27
141–42, 144, 155–56, 262 see also speech-in-character
chreia 39, 40n5, 62
Cicero 19, 19n1, 30n20, 31, 34 [Hermogenes] 11, 18, 38n1, 38n3, 45–51, 61,
De inventione 19n3  96
Pro Caelio 30n20
[Cicero] 19 interlocutor 2–5, 8, 11, 105, 110–15, 120–23,
conformatio 19, 22–27, 37 125, 129, 131–34, 140–41, 145, 147–48,
see also speech-in-character 150–58, 163, 171

detractionem 32, 32n26 Libanius 61

diatribe 2–5, 8n27, 7–11, 97, 101–59, 163, Livy 34
169–218, 267, 295
Bultmann’s model 103–6 Maximus of Tyre 112, 116
dialogical 1–7, 104–6, 108, 109n48, methodology 16–18, 33, 46, 53n30, 55–56,
110–125, 129–59, 163, 168–218, 252–74 56n3, 58–63, 74–75, 74n47, 85–86,
diversity / variability 107, 110–11, 113–16, 86n76, 101, cf. 120, 128, 130, 132, 143–45,
121, 125, 127, 131, 148, 157–58, 173, 192, 157–59, 170, 206, 212, 212n68, 218, 254n7,
199–200, 205, 212n68, 218 268, 271, 292
genre 103–4, 106–9, 111n57, 116–19, Musonius Rufus 111–13, 111n63, 116
117n108, 122, 122n138, 127, 127n152,
173–74, 190 Nicolaus 16, 38n1, 46, 52, 52n27, 52n28, 61
Subject and Name Index 319

Paul rectification
affirmation of Jewish advantage 160, 162 see also dikaiosune / δικαιοσύνη
“all” 69–71, 78n60, 78n62, 81, 89n85, 175, rhetorical acumen 10–11, 18, 54, 60–63,
190, 195–96, 217–18, 220, 222, 224–26, 67–68, 70–71, 73, 77, 80, 82–84, 86n76,
230n28, 231, 233, 236–37, 239–41, 95–97, 189, 219–98, 265n58
244–45, 247, 255, 257, 257n21, 260–61, righteousness
267, 272–92, 294–98 see also dikaiosune / δικαιοσύνη
anthropological equality 195, 205, right-wising
219–98 see also dikaiosune / δικαιοσύνη
apocalyptic 226, 228n23, 229, 233–34, Spirit / πνεύμα 69, 71n36, 72–73, 78–81,
237, 239, 241 236, 245–46
Christ / Messiah 69–70, 71–73, 74–77, universal outlook 226, 230, 233, 236, 239,
76n53, 90, 90n89, 219, 225, 234, 236, 244, 242–47, 258, 274–88, 285n112, 296–98
255, 257, 272, 275, 282–84 Philo of Alexandria, 113–14, 116
dikaiosune / δικαιοσύνη 69–71, 89–95, Philostratus 42n14
89n85, 90n89, 93n97, 167, 225–26, Plutarch 112–13, 116, 125, 125n150
228–30, 230n28, 234, 236–38, 244–47, Pompeii 126–27
250–51, 257, 260–62, 271–72, 274–76, progymnasmata 11, 38–53, 59
282–88, 291–93, 296 prosopopoiia / προσωποποιία 8n26, 15,
divine impartiality 10, 190, 195–96, 205, 26–28, 28n17, 37, 38n3, 40–48, 51, 55
212, 219–98 prosopopoeia 27n16
divine sovereignty 88–95, 89n85, 232n31, see also speech-in-character
237–38, 243–47, 249–50, 257–58, 277, protreptic / exhortation / προτρεπτικός 
280–88 111–14, 116, 121–22, 125–27, 129–30, 135,
ethnic inclusivity 70n34, 219–98 141, 154, 158–59, 204, 215, 267–68, 272,
ethnic unspecificity / vagueness 70n34, 297
190, 219–98 pseudepigraphy 42n14
faith(fullness) see also epistolary speech-in-character
see also pistis / πίστις Pseudo-Demetrius 19n1
gospel 219, 221, 225–26, 233, 242–47, 251, pseudonymity
255–57, 276, 291 see also pseudepigraphy
grace 219–20, 241, 244, 267, 275–98
justification Quintilian 11, 15, 18, 19n3, 26–37, 38n3, 71, 81,
see also dikaiosune / δικαιοσύνη 96, 270–71
law / νόμος 69, 89–95, 89n84, 90n89, Institutio oratoria 59
93n97, 242–47, 249–50, 275–76, 275n83,
281–82 Rhetorica ad Herennium 11, 18–27, 36–37,
Christological fulfillment 242–47, 38n3, 59, 129
249, 258, 267–68, 275 rhetorical theory 15–57
me genoito / μὴ γένοιτο 171–73, 189–95, in practice 58–60, 62
259–62, 276–77 Roman empire 63–71, 225n14
participationism 69–70, 71, 86 Pax Romana 65
pistis / πίστις 69–71, 70n34, 71–73, 89–95, peace and security / pax et
90n89, 93n97, 166, 212, 220, 220n3, securitas 63–71
225–30, 232–34, 232n31, 236, 238, Romans, epistle 
242–47, 249–51, 257–60, 266–67, 271, cohesion with conventions for speech-in-
274–98 character 238–98, esp. 270–74
source of 227–29, 284, 288 ambiguity 219–98
320 Subject and Name Index

Romans, epistle (cont.) appropriateness 20–26, 29–31, 34–37, 40,

audience 178–79, 220–22, 233, 238, 271–72, 41n11, 42–44, 48–49, 51, 54–55, 72–73,
287 80, 130, 137, 142, 144, 154, 156, 256, 262,
dialogue 1–11, 252–74, 259n31 278
diatribal tone 273–74 core conventions 10, 15, 54–57, 58, 193,
hypocrisy 234–51, 256 268, 292
idolatry 234–51, 256 epistolary 42n14
Jewish advantage? / privilege? 166, in the New Testament 1, 18, 53n30, 58–97
195–96, 219–98, 258n26 length 61–62
Paul’s interlocutor / relation to diatribe 97, 101, 123n148,
characterization 180, 238–74, 259n31 129–31, 136–38, 141–45, 148–49, 156–59,
eventual agreement with Paul  206–7, 212, 267–68, 272–74, 295, 297
263–74, 276 summary of conventions in
appropriateness to characterization, ad Herennium 26
219–98 summary of conventions in
differences with Paul 250–51, 258–59, [Hermogenes] 51
258n26, 262, 271, 296 summary of conventions in
pre-Bultmannian readings of Quintilian 36
3:1–9 166–68 summary of conventions in Theon 45
rescriptive readings of dialogue 1–7, 163, synthesis of core conventions in primary
199–218, 252, 256–74, 292, 295, 297 sources  36–37, 51–52, 54–55
Romans Debate 274, 288–92, 288n123, synthesis of unique conventions in
288n124 primary sources 36–37, 52–53, 55–57
traditional readings of dialogue 1–7, 163, unique elements 10, 15, 55–57, 58–59
169–96, 218, 252, 278, 292, 295 Stilpon 131
successful arguments for the traditional
Seneca  114–16 reading of Rom 3:1–9 160, 162
sermocinatio 19–27, 37 sweeping assumption, just one 192–94
sermocinatio consequentes 21–22, 25, 129
see also speech-in-character Teles 104, 109–10, 109n47, 109n48, 132–38,
Socrates / Socratic questioning 113, 121, 125, 156, 254n6
149n25, 153, 212, 252–74, 259n31, 275n82, Theon 15, 18, 38–45, 38n1, 96, 138n16
291–92, 297 Progymnasmata 59
speech-in-character 8–11, 8n26–27, 15–97,
129, 136–38, 141–45, 154–59, 163, 206, Vergil 31, 207–8, 216, 270–71
206n38, 252–74, 270n70, 295
Ancient Sources Index

Aphthonius, Progymnasmata 3.14 25, 56

34.2–18 52n29 4.53 74
34.15 23n7 4.54 20
35.11–13 49n23, 62 4.55 8n26, 20–21
35.13–14 53n30 4.65 8n26, 20–21, 123,
35.15–36.20 41n10, 61 123n148
4.66 74
Aristides, In Defense of Oratory
2.319 50 Cicero, Topica 
2.319–20 50 45 55n2
2.319–343 48
2.321ff 50 Cornelius Nepos, Militiades
5.4 42n13
Aristotle, Ars Rhetorica 9
3.7.1–7 43n15, 55n1 Ctesias, History of Persia 
13.22 42n12, 42n13
Bion, Fragment 16A 119 13.25 42n12, 42n13

Bion, Fragment 17 119 Demosthenes, Against Neaera

94 42n13
Bion, Fragment 21 119
Deuteronomy 
Bion, Fragment 68 119 7:6 221
7:6–8 221
Chrysostom, Homily in 1 Cor 3.2 8:17 90n89, 92
61.24 75 8:17a 92
9:4 90n89, 92
Cicero, De Inventione 9:4a 92
1.99 28n18, 272 14:2 221
1.99–100 55n2 29:2–30:20 92n94
29:3 283
Cicero, De Oratore 9 29:19–28 237
3.205 28n18 30 92
30:11–13 93n97
Cicero, De Partitione Oratoria 30:11–14 92, 93n97
55 55n2 30:12 92–93
30:12–13 90n89
Cicero, In Catilinam 30:13 92–93
1 34 30:12–14 90n89
1.17–18 31, 34 30:14 90n89, 92–93,
1.27 31, 34 93n97
32:21 187, 282
[Cicero], Rhetorica ad Herennium 32:43 (LXX) 288
1.1 59n2 34:10 244
322 Ancient Sources Index

Dio Chrysostom, περὶ λύπης, 119 2.23.2 149

2.23.2–4 152
Diodorus Siculus, Library of History 2.23.2–5 156n32
10.27.1–3 42n13 2.23.3–4 149
11.2.2 42n13 2.23.5 156n32
2.23.5–7 156n32
Epictetus, Discourses 5 2.23.6 149
1.2.29 151 2.23.7 149
1.12.8–35 119 2.23.9 149
1.23 111, 116, 125 2.23.9–15 149
2.12 149n25, 153 2.23.16 150–51, 154–56
2.20 111, 116, 125, 136, 2.23.17–19 150
138–49, 158, 174n62, 2.23.18–19 149
201, 251n88 2.23.20 150, 154
2.20.1 138, 148 2.23.20–22 155
2.20.1–5 139–42 2.23.21 153
2.20.2 139 2.23.21–22 150, 154
2.20.3 139 2.23.21–23 155
2.20.4 139, 141 2.23.22 151, 154, 156
2.20.4–5 140 2.23.23 152–53, 156n32
2.20.5 140, 141 2.23.23–24 153, 156
2.20.6 142–43 2.23.23–25 154
2.20.6–9 145 2.23.23–26 152–53, 155–56
2.20.6–20 142–45 2.23.24 156n32
2.20.7 142 2.23.24–25 156n32
2.20.7–14 144 2.23.24–26 153
2.20.8–14 142 2.23.25 156n32
2.20.10 142 2.23.26–27 155
2.20.15 142, 144 2.23.26–29 153
2.20.15–16 144–45 2.26.4 111n59
2.20.16 143–45 3.23.28–29 112
2.20.17 145
2.20.17–20 143 Exodus
2.20.21 141, 148 4:14 241n53
2.20.21–37 145–48 33:19 280
2.20.22 145–48
2.20.23 146, 148 Ezekiel
2.20.24 147 7:1–20 237
2.20.25 147 7:5 237
2.20.26–27 147 7:20 237
2.20.28 147 8:18 237
2.20.29–31 147 9:8 237
2.20.35 147 13:10 65–66
2.20.37 147 13:15 65–66
2.21 125n150
2.21.15–22 125n150 Genesis
2.23 111, 116, 125, 149–58, 3–4 241n53
192, 201 6–8 241n53
Ancient Sources Index 323

12 241n53, 294 1.201–16 44, 44n16

12:3 69–70, 69n32, 1.205 44–45
70n34 1.206 44–45
15 294 1.208 44
15:6 69 1.209–210 44
17 69 6.111–120 42n13, 44n16
18:18 69–70, 69n32,
70n34 Homer, Iliad
21:12 280 18.2 50
18.22–35 50
Habakkuk 18.78–93 50
2:4 174, 226, 228–29, 18.97–126 50
229n25, 244 22.437 50
[Hermogenes], Progymnasmata 22.437–476 50
20.7–8 46 22.477–514 50
20.8 48
20.8–9 47, 50 Hosea
20.9–10 47 2:1 (LXX) 281
20.10 48 2:23 222, 281
20.10–11 48
20.10–12 47 Isaiah
20.11–12 48 1:9 281
20.11–18 48 1:19–29 237
20.12–14 47 6:9 283
20.14 47 8:14 281
20.14–16 47 10:5–6 237
20.14–18 48 10:22 281
20.16–18 48 27:9 284
20.19 48 28:16 281–82
20.19–21 48 29:10 283
20.21–23 48 41:8 221
20.24 48 48:12 221
20.24–21.2 48 52:5 (LXX) 247
21.1 48 59:20 284
21.1–4 48 65:1–2 (LXX) 282
21.2–4 48
21.4–5 48 Jeremiah 
21.6–7 48–49 6 66
21.7–9 49 6:14 65–66
21.10–18 57 6:24 65–67
21.14 50 7:1–20 237
21.15–18 50 31:33 244, 284
21.19–22.3 49, 57 38:3 221
22.4–5 49 38:33 (LXX) 244

Herodotus, The Persian Wars Joel

1.201 44 3:5 (LXX) 282
1.201–205 45
324 Ancient Sources Index

Leviticus 3:22 75

18:5 90, 90n89, 3:22–23 75
244n58 3:23 75
4:14–21 78n62
Livy, History of Rome 5:1–2 78n62
1.9.1–3 35 5:6–11 78n62
1.9.3 34 6:1–11 78n62
6:11 79
Menander 6:19 79
Fragment 545 47 6:19–20 77
7:23 77
Musonius Rufus, Discourse IX 12 9:11 78n60
9:17 186, 227n19, 257
Musonius Rufus, Discourse XV 12 10:28 63n11, 95–96
11:23–26 63n11
Musonius Rufus, Fragment IV 111 12 290
12:1 78, 78n62
Musonius Rufus, Fragment XVII 119 12:1–3 78n62
12:2 78n62
Nicolaus, Progymnasmata 12:2–3 78n60
64.1–3 52n29 12:3 78–81, 78n60,
64.5–13 52n28 78n62, 95–96
64.12 50 12:3a 78n62, 80, 82
64.20–65.10 52n29 12:3b 96
65.11–66.9 53n30 12:3c 95
66.9–15 49n23, 62 12:4 78n60, 79
66.16–67.9 52n27 12:4–11 78n60, 81
12:4–31 78n62
Paul, 1 Corinthians 12:4–14:40 78n60
1:1–3 74 12:7–13 79
1:4–9 74 12:12 81–82
1:10 74 12:13 81, 231
1:10–11 75 12:13–14 75
1:10–12 77 12:14 81–83
1:11 74 12:15 82
1:12 74–77, 12:15–16 81–85, 95–96
74n47, 12:15a 83n72
76n53, 12:15b 83n72
95–96 12:15c 83n72
1:12b 74 12:16 82
1:13 75, 76n53 12:16a 83n72
1:14–16 216n88 12:16b 83n72
3:1 77 12:16c 83n72
3:1–3 77 12:17 82
3:3 77 12:17–20 81, 83
3:1–4 77 12:18 82
3:4 74–77, 95–96 12:19 82
3:16 79 12:21 81–85, 95–96
Ancient Sources Index 325

12:21a 82 10:18 86

12:21b 82 11–12 61
12:22–26 83 11:30 85
12:27–31 81 11:31 86
14:1 78n60 12:1 85
15:28 89n85 12:2–4 85
15:35 63n11, 95–96 12:5 85
15:46 78n60 12:6 85
16:22 78n62 12:7 85, 88n81
12:7–8 88
Paul, 1 Thessalonians 12:8 85
2:4 186, 227n19, 257 12:8–9 85
5:1–2 63–64 12:9 85–88, 86n76, 88n81,
5:3 63–71, 87, 95–96 95–96
5:3a 64, 67 12:9a 86n77
5:3a-b 67 12:9b 86n77
5:3b 64, 67 13:3–4 86–87
5:3c 64 13:4 86
5:4 63–64
5:4–5 63–64 Paul, Galatians
5:4–8 67 1:6–7 192
5:4–10 67 1:16 70n34
5:5 64 1:8–9 78n62
5:5–6 64 1:13–17 283
5:6–7 64 2 70n34
5:6–8 63 2:2 70n34
5:8 63–64 2:7 186, 227n19, 257
5:8–10 64 2:8 70n34
5:9–10 63 2:9 70n34
5:11 63–64 2:12 70n34
2:14 70n34
Paul, 2 Corinthians 2:14–17 263n51
1–9 86, 87n78 2:15 70n34, 246
1:12 85 2:15–16 283
1:14 85 3:1–5 71
5:12 85 3:2 72–73
7:4 85 3:2–5 69, 79
7:14 85 3:5 68, 72–73
8:9 87 3:6 69
8:24 85 3:6–9 71
9:2–3 85 3:6–18 257
10–12 85 3:7 69–70
10–13 86, 87n78 3:8 68–71, 70n34, 95,
10:1 86 219n2, 227
10:7 86 3:8a 70n34
10:8 86 3:8b 70n34
10:10 63n11, 95–96 3:9 69
10:14 86 3:10–14 69
326 Ancient Sources Index

Paul, Galatians (cont.) 1–15 291

3:10–22 69 1–16 230n28
3:12 244n58 1:1 219–20, 225n15, 272
3:14 70n34, 72–73 1:1–3 229, 232
3:14–16 69–70 1:1–5 219, 243–45, 249,
3:15–18 286 257n21, 265, 275–76
3:16 71 1:1–7 222–23, 236, 241, 257,
3:17 287n118 289
3:24 245 1:1–12 219–22
3:26 72 1:1–16 231
3:26–27 71 1:1–32 237–38
3:26–29 71, 75 1:1–3:9 274
3:28 71, 231, 245, 267, 283, 1:1–11:36 284
290, 298 1:2 219, 227n22, 231,
3:29 71 287n119
4:3 71 1:2–3 227
4:4 71, 73 1:2–4 246
4:5 71 1:2–5 219
4:6 71–73, 71n36, 79, 95–96 1:3 219, 227n22
4:7 72 1:4 225n15
4:8 246 1:5 217, 219–20, 225, 228,
5:6 255 289
5:14 245 1:5–6 179n97
5:22 228 1:5–7 226
6:15 255 1:6–7 220–21, 221n7, 233, 287,
289
Paul, Philippians 1:6–2:11 244
1:5 289 1:7 220, 225n15, 290
1:12 223 1:8 220, 225–26, 238
1:29 228 1:8–15 222–23
2:6–11 87, 93 1:8–17 222
3:4–6 283 1:8–2:4 263
3:4–11 285 1:9 219
3:5–6 248 1:11 224n14, 289, 290n128
3:7 283 1:11–12 224n14
3:9 283 1:11–15 224n14
4:10–20 289 1:12 220, 224n14, 226, 238,
289, 290n128
Paul, Romans 1:13 179n97, 220, 223,
1 89n85, 167, 241n53, 224n14
282, 289–90, 293, 296 1:13–15 222–25
1–2 195, 219–51, 255, 259, 1:14 220n3, 223–25, 224n14,
266, 273, 285, 292, 297 226, 236
1–3 176, 179, 190, 194, 213, 1:14–15 223, 225
259n31, 262n45, 1:14–17 241
274–75, 278 1:15 222, 224n14, 288
1–4 213, 216, 228n23 1:16 222, 225–28, 231, 234,
1–11 245, 278, 287–88, 291 241, 278
Ancient Sources Index 327

1:16–17 174, 179, 220n3, 222–23, 2:1–16 180, 206, 206n38, 208,
225–34, 236–38, 238–39, 247–49
243–45, 249, 257, 2:1–29 238, 273
257n21, 259, 261, 265, 2:1–3:2 279n95, 290
276, 285, 289 2:1–3:8 275n80
1:16–18 234 2:1–3:20 236–37
1:16–11:36 173 2:2 239–40, 239n49
1:17 225n16, 226, 228–30, 2:3 175, 239n49, 240–41
234, 244 2:3–4 239, 262
1:18 223n12, 231, 233–34, 2:3–5 240
236–37 2:4 241, 259
1:18–32 174–76, 179–80, 183, 2:5 239, 241, 250
213–18, 234–42, 271n72, 2:6 241, 244–45
282, 289 2:6–11 175
1:18–2:16 261 2:7 241
1:18–2:29 178, 282 2:8 237, 241
1:18–3:8 215 2:9 241
1:18–3:20 174, 275n80 2:9–10 175, 239, 241
1:19 236–37 2:9–11 195, 289
1:19–21 235, 240  2:10 241
1:19–32 183 2:11 175, 180, 241–42
1:20 236–37, 239 2:12 243–44, 246–47
1:21 236–37, 240 2:12–16 92, 175, 180, 183, 186,
1:21–32 240 242–47, 244n58,
1:23 179, 235–37, 240, 248–49, 250n84, 257,
250n85 257n21, 265, 275–76,
1:23–27 250 281, 285n112
1:24 89n85, 236–37, 281 2:12–29 175–76
1:25 235–37 2:13 214, 230, 242–44
1:26 89n85, 237, 281 2:13–14 2434n58
1:26–27 237 2:14 239, 242–43, 246
1:28 89n85, 237, 281 2:15 244–45
1:28–32 235, 275n80 2:15–16 242
1:32 238, 240 2:16 217–18, 242–43, 243n56
2 11, 30, 166–67, 173, 176, 2:17 163, 179n97, 185,
179, 183–84, 185n151, 247–50, 253, 259, 275
205, 214–16, 237–51, 2:17–18 248
265, 271, 289, 292, 296 2:17–22 181
2–3 205, 247, 271, 290 2:17–23 262–63, 271
2:1 172, 175, 180, 238–40, 2:17–29 175–78, 180–81, 187,
239n49, 247–48, 250, 202–4, 206–9, 206n38,
271n72 211–12, 238, 242, 245,
2:1–3 240, 290 247–51, 257n19, 267
2:1–4 250, 259, 262 2:17–3:9 165, 205, 272n73
2:1–5 206n38, 273 2:18 249
2:1–11 176, 176n75, 180, 185, 2:19–20 249,258
238–42, 250 2:19–22 248
328 Ancient Sources Index

Paul, Romans (cont.) 278, 279n95, 285, 288,

2:21–22 181, 249, 249n81, 258 292–93, 295, 297
2:23 181, 249, 259, 275 3:2–3 185n151, 186, 186n161,
2:23–24 248 258
2:23–25 258 3:2–4 227, 229
2:23–29 181, 253, 259–60, 264, 3:3 3–6, 166, 168, 170–73,
273 176–77, 181–82, 186–89,
2:24 247, 249, 265 191, 191n181, 195, 200,
2:25 253 202–5, 211, 213, 226–27,
2:25–27 248 250n84, 252, 256–61,
2:25–29 92, 181, 186, 214, 257n19, 264, 264n53,
246n67, 248–49, 253 268–74, 292, 295, 297
2:27 246, 250 3:3–4 194, 256–60, 266, 271,
2:28–29 250, 253 276, 279, 286
2:29 241, 254 3:3–6 176, 266, 292
3 10, 33n28, 56n3, 173, 3:3–8 203
181, 186n157, 205, 226, 3:3–8c 210
279, 292 3:3–9 254n11, 297
3–4 184 3:4 3–6, 166–68, 170–73,
3:1 3–6, 164, 166, 168, 176–77, 181–82, 184,
170–73, 176–77, 181–83, 186–88, 192–94, 197,
186–88, 200, 202–3, 200–3, 205, 211, 213,
210–11, 213–14, 238, 215, 226, 252, 256–60,
249, 252–56, 255n14, 262–63, 267–74, 277,
256n17, 266–74, 289, 295, 297
292, 295, 297 3:5 3–5, 166–68, 171–73,
3:1–2 4, 172, 203, 205, 210, 176–77, 182, 189, 191,
213, 253–56, 254n11, 191n181, 194–95,
255n13, 256n15, 197–98, 200, 203–5, 211,
265–66, 285 213, 230, 260–63,
3:1–3 257 264n53, 268–74, 277,
3:1–4 260 280, 292, 295, 297
3:1–8 163–218, 185n151, 3:5a 6, 181, 187–88, 252, 269,
254–98, 261n41, 277
264n53, 267n62 3:5a-b 176
3:1–9 1–11, 56, 60, 97, 101, 3:5b 6, 181, 187–88, 195, 252,
104n16, 128, 163, 261, 269, 277
165–218, 254–98, 3:5c 176, 181
261n41, 270n70 3:5–6 176, 187, 194, 260–62,
3:1–20 176–77 262n44, 271
3:2 2–6, 168, 171–73, 3:5–8 260
176–77, 181–84, 186–88, 3:6 3–6, 167–68, 170–73,
195, 200, 202–3, 205, 176–77, 181–82, 184,
207, 211, 213–15, 187–88, 192–95, 200,
227n22, 232, 252–57, 203, 205, 211, 213, 215,
254n11, 256n17, 257n19, 252, 260–63, 267–74,
259, 262, 264, 267–74, 280, 295, 297
Ancient Sources Index 329

3:6–7 172n49 3:9–31 267

3:7 6, 172, 181–82, 187, 3:10 228, 252
198–99, 252, 262–66, 3:10–18 182, 215, 230
268–74, 277, 280 3:10–20 274
3:7–8a 187, 198, 252 3:10–29 266
3:7–8c 3, 5, 168, 171, 173, 176, 3:10–31 251, 274–76
187–88, 200, 203, 3:11 275
210–11, 213, 262–63, 3:12 275
264n56, 266, 268–74, 3:19 275
292, 295, 297 3:19–31 244
3:7–8 4, 167, 172, 177, 203–5, 3:20 230, 242–43, 275
212, 262–65, 264n53, 3:21 227n22, 230n28
295 3:21–22 92, 230, 275
3:7–9 4, 265 3:21–24 245, 249
3:8 167, 172, 181–82, 189, 3:21–26 187, 204, 230, 275
191, 194, 199, 262–65, 3:21–31 257
268–74 3:23 230, 275
3:8–9 292 3:24 230, 275
3:8a 6, 252, 262, 269 3:25–26 230, 275
3:8a-c 276–78 3:26 230, 275
3:8b 6, 187, 252, 269 3:27 275
3:8c 6, 252, 269 3:27–31 245, 270–1, 275, 275n82
3:8d 3, 5–6, 168, 171, 173, 176, 3:27–4:2 206
178, 187–88, 200, 203, 3:28 276, 276n84
210–13, 252, 263–66, 3:29 245, 276, 276n84
264n53, 264n56, 3:29–31 267
268–74, 275n82, 277, 3:30 230
295, 297 3:31 193, 227n22, 276
3:9 2, 178, 184, 195–96, 205, 4 196, 230n28, 232,
210–11, 213, 215, 218, 232n31, 245, 276–77,
249, 254n11, 264–68, 276n86, 281, 285n112,
274, 284–85, 289, 293 286, 287n118, 293, 294,
3:9a 4–6, 173, 176, 178, 298
181–82, 187, 200, 203–5, 4:1 170
210–11, 213, 215, 252, 4:1–17 227
265–74, 267n62, 292, 4:1–25 257
295, 297 4:2 170
3:9a-b 203 4:3–5 276
3:9b 4–6, 176, 181, 187, 200, 4:3–6 92
203–5, 210–11, 213, 252, 4:5 230
267–74, 267n62, 4:6–8 276
275n82, 295, 297 4:9–10 92, 276
3:9b–10 188 4:11 92
3:9b–20 173, 200, 205, 214 4:11–12 276, 286, 289
3:9b–31 273 4:13 92, 226n17, 227, 233,
3:9c 252 276, 286, 289
3:9–20 163, 173 4:13–16 285n114, 288
330 Ancient Sources Index

Paul, Romans (cont.) 8:4–6 79

4:13–17 229 8:8–9 245–46
4:16 257n21, 276–77, 286, 8:9–11 277
289 8:9–16 79
4:16–18 285n114, 288 8:15 285n114, 288
4:17 227, 233, 277 8:15–16 73
4:17–18 286 8:18–21 285n114, 288
4:18 228 8:20 89n85
4:22 276 8:23 285n114, 288
4:22–25 277 8:30 230
5 89n85 8:33 230
5–8 213, 216, 228n23 9 88, 279–82, 287n118,
5:5 79 293
5:8 263n51 9–11 2, 182, 184, 187, 190–91,
5:8–9 234 196, 199, 205, 230n28,
5:9 237 231n29, 233n33,
5:12–14 275, 277 241n53, 277–89,
5:15–17 275 285n112, 287n118,
5:17 92 290n129, 293, 297
5:20 277 9:1–3 279
6 199, 237, 277–78, 9:1–5 279
277n89, 293 9:4–5 255n14, 279n93, 285,
6:1 270, 277 288, 291
6:2 193, 277 9:6 279–80
6:3–4 277 9:6b 280
6:6 277 9:6–9 286
6:11 277 9:6–13 283, 286
6:14 277 9:6–10:3 89n85
6:15 193, 277 9:7 280
6:16 277 9:7–13 184
6:17 224n14 9:8 280
6:17–23 277 9:11 221, 280
7 33n27, 56n3, 61, 92n93, 9:11–13 280
193, 260 9:12 221, 280
7:6 79 9:13 221
7:6a 270 9:14 193, 277, 280
7:7 193 9:15 280
7:7–25 1 9:16 280
7:7–8:2 206 9:17–18 280
7:7b–12 1, 57, 92, 270n70 9:18 184
7:12 92n93 9:19 277, 280
7:13 193 9:20 95–96
7:13a 270 9:20–21 281
7:13b–24 1, 57, 92, 270n70 9:24 232n31, 281, 283–86,
7:15–21 92 289
7:25b 1, 57, 92, 270n70 9:24–25 221, 233
8 89n85, 293 9:25 281, 281n101, 284
8:2–4 285n114, 288 9:25–26 281
8:3–4 245 9:27–29 281
Ancient Sources Index 331

9:30 92, 243, 281 11:1 193, 232n31, 248,

9:30–31 281 282–83
9:30–33 89, 281, 281n102, 285 11:1–2 282
9:32–33 198, 281, 283 11:2–4 283
10 89, 282 11:5 232n31
10:1 89 11:5–6 283
10:1–3 230 11:7 283
10:1–16 231 11:7–10 232n31, 284
10:2 89, 282 11:8 195, 237
10:3 89, 89n84, 89n85, 282 11:8–10 283
10:4 89–90, 90n89, 245, 11:11 193, 283
285n112 11:11–24 284
10:4–5 89n84 11:12 278, 283–84, 284n111
10:4–8 90n89 11:12–16 232n31
10:4–21 282 11:13 179n97, 290n129
10:5 90, 90n89, 92–93, 11:13–24 289–90
93n97 11:14 187
10:5–8 90 11:15 283
10:6 90n89, 90n90, 92–93 11:16 232n31
10:6a 90n89 11:17 186n161, 232n31
10:6b 90n89 11:17–23 257
10:6b–8 94n98 11:18 232, 290
10:6c 90n89, 93n97, 94n98 11:19–20 232
10:6d 90n89 11:20 232n31, 283, 290
10:6–7 93n97 11:21 246
10:6–8 1, 88–95, 90n88, 11:23 232n31, 233, 283
90n89, 93n97, 95–96, 11:24 246
282 11:25 257–58, 278, 284,
10:7 90n89, 93 284n111
10:7a 90n89 11:25–32 89n85, 182, 231n29,
10:7b 90n89, 93n97, 94n98 237, 257n21
10:7c 90n89 11:25–36 284
10:8 93 11:26–27 284
10:8a 90n89, 94n98 11:28 221, 284
10:8b 90n89, 93n97 11:29 221, 284
10:8c 90n89, 94n98 11:30–31 233n34
10:9 80 11:32 187, 195, 233, 278,
10:11 282 284–85
10:12 282, 285n114, 288 12 290
10:13 282 12–15 273, 288, 290–91
10:16 282 12:1 285n114, 288, 290
10:17 228 12:1–8 290
10:18 282 12:3 228
10:18–19 282 12:6 228
10:19 282 12:19–21 291
10:20–21 282 13:8 245
11 167, 187, 232, 232n30, 13:8–10 291
273, 282–84, 290, 294 14 290
11–15 290 14:1 291
332 Ancient Sources Index

Paul, Romans (cont.) Pseudo-Hermogenes, 11

14:10–13 240, 291 Progymnasmata
14:13 291 20.7–18 25n10
14:15–15:1 291 20.10 23n7
15:6 291
15:7 291 Quintilian, Institutio 26–27
15:8–9 286 oratoria
15:10 288, 291 1.1.1 59n3
15:11 288, 288n122, 291 1.6 59n3
15:15–16 179n97 1.8.2 35
15:20 224n14 1.8.3 15n1, 27n16, 35, 56
15:20–21 224n14 3.8.49 31
15:22–29 288 3.8.50–51 31, 38n3
16:17 224n14 3.8.54 30n20, 272
16:26 219 4.1.28 31n23, 34n29
6.1–5 28n18
Plato, Laws 6.1.25–27 31n21
3.698.c 42n13 9.2 8n26, 27
9.2.1 27
Plutarch, Aristides 9.2.29 26n11, 27–28
5.1.1 42n13 9.2.29–37 35
9.2.30 28, 56, 68, 73–74,
Plutarch, De recta ratione audiendi 77, 81, 84, 88, 95,
43E 113 130, 138, 142, 156,
47A 113 272
9.2.30–33 206n38
Plutarch, περὶ ἀρετῆς καὶ κακίας 119 9.2.30–37 28n17, 30
9.2.31 25n10, 26, 26n11,
Proverbs  27–28, 27n16, 31,
24:12 244 56, 73
9.2.32 25n10, 28, 31, 38n3,
Psalms 55, 84
18:5 (LXX) 282 9.2.34 34–35, 56
50:6 (LXX) 259, 270 9.2.34–35 34
51:4 198 9.2.36 23n7, 31–32, 34
58:6–14 (LXX) 237 9.2.37 27, 27n16, 32, 34, 56,
62:13 244 62, 74, 207, 216,
68:23 (LXX) 283 254n7, 270
78:1–7 (LXX) 237 9.2.58 28n17, 37n32
105:19 (LXX) 237 9.3.58–64 32n26
105:20 (LXX) 236–37 11.1.39–41 31n21
106 236 11.1.41 31n22
106:19–20 237
106:26 93 Seneca, epistulae morales 114
117 288n122
117:1 288, 288n122 Sibylline Oracles
3:8–15 174n65
Pseudo-Demetrius, On Style
265–266 28n18, 74
Ancient Sources Index 333

Suda Theon, Progymnasmata 11

Θ 206 38, 40n7 65.10–11 24n9
96.18–22 62n10
Suetonius, On Grammarians and Rhetors  115.12 38n3
25.4 40 115.12–14 40
115.14–15 41
Teles, Fragment 1, περὶ τοῦ 87, 130, 132–38, 115.15–16 41
δοκεῖν καὶ τοῦ εἶναι 158, 174n62 115.16–17 41
1–3 132 115.17–18 41
2–3 133 115.18–19 41
4 133 115:20–21 88
5–6 133 115.20–22 42, 57, 68, 84,
7 133 130, 157, 272
8–9 133 115.22–24 43
10 133 115.22–27 43, 145
11–12 133 115.22–116.22 38n3, 57
13 133 115.27–28 43
14–17 133 115.28–116.22 43
17–25 134 116.22–117.6 135, 138n16
19 133 116.22–117.30 57
21 133 116.27 88
29 133–34 116.27–117.4 57, 68, 84
29–30 137 116.28–32 44
31–32 134, 137 116.32–117.4 44
32 133 117.4 68
33 133–34 117.4–5 43
34–35 134–35 117.5–6 44
37–38 137 117.6–13 88
45–47 134, 136 117.6–24 44
47 136 117.15–16 88
117.24–28 44
Teles, Fragment II, 109n47, 119 117.28–30 44
περὶ αὐταρκείας 117.30–32 44, 57
19–44 109n48
Vergil, Aeneid 
Teles, Fragment III, 109n47 2.1–29 34
περὶ φυγῆς 2.25 32
9–31 131 2.25–28 33
13 131n7 2.29 32, 32n25, 62,
16 131n7 270
21 131n7
60 109 Wisdom of Solomon 183, 237, 240–41
13–15 174n65
Teles, Fragment IVA, ἐπιτομή 109n47 13:1 234–35
13:2 235
Teles, Fragment V, περὶ τοῦ 110n49, 126 13:10 235
μὴ εἶναι τέλος ἡδονήν 14:22–25 235
14:29–31 240
Teles, Fragment VI, 110n49
περὶ περιστάσεων