Augustine and the

Disciplines
From Cassiciacum to Confessions

edited by
Karla Pollmann
and
Mark Vessey

1
 3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
© Oxford University Press 2005
The moral rights of the authors have been asserted
Database right Oxford University Press (maker)
First published 2005
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
ISBN 0–19–927485–1
EAN 9780199274857
1 3 5 7 9 10 8 6 4 2
Typeset by RefineCatch Limited, Bungay, Suffolk
Printed in Great Britain by
Biddles Ltd., King’s Lynn, Norfolk
 acknowledgements

This volume had its origins in a conference on ‘Augustine

and the Disciplines’ held at Villanova University, Pennsyl-
vania, 9–11 November 2000. Three other presentations
made during those days, by Peter Brown, Robert Markus,
and James O’Donnell, have already been published together
under the title of ‘The Study of Augustine, 1950–2000:
Evolving Disciplinary Contexts’, in Augustinian Studies 32
(2001), 177–206. For their generous sponsorship of the
original event, thanks are due to the Revd Edmund Dobbin,
OSA, President, and the Revd Kail Ellis, OSA, Dean of the
College of Liberal Arts and Sciences, Villanova University.
The editors are also grateful, respectively, to the British
Academy and the Social Sciences and Humanities Research
Council of Canada for grants in aid of research, and to the
Netherlands Institute for Advanced Study and the Institut
d’histoire de la Réformation (Université de Genève) for
material and technical support during the final stages of
preparation of this book. In Geneva, Marlène Jaouich made
all paths smooth: our special thanks to her. The Index
locorum is the work of Shelley Reid of the Department of
Classical, Near Eastern and Religious Studies at the
University of British Columbia.
K.P.
M.V.
 contents

Notes on contributors ix
Abbreviations x
1. Introduction 1
Ma r k Ve s s e y
PART I Honesta studia: Classrooms without Walls 23
2. Disciplines of Discipleship in Late Antique
Education: Augustine and Gregory Nazianzen 25
Ne i l McLy n n
3. The Duty of a Teacher: Liminality and disciplina
in Augustine’s De Ordine 49
Cat h e r i n e Co n y b e a r e
PART II Disciplinarum libri: The Canon in Question 67
4. Augustine’s Disciplines: Silent diutius Musae
Varronis? 69
Da nuta R. Sh a n z e r
5. Divination and the Disciplines of Knowledge
according to Augustine 113
Wi l l i a m E. Kl i n g s h i r n
6. The Vocabulary of the Liberal Arts in
Augustine’s Confessions 141
Ph i l i p Bu r to n
PART III Doctrina christiana: Beyond the
Disciplines 165
7. The Grammarian’s Spoils: De Doctrina Christiana
and the Contexts of Literary Education 167
Cat h e r i n e M. Ch i n
8. Augustine’s Critique of Dialectic: Between
Ambrose and the Arians 184
St e fa n He ß b r üg g e n-Wa lt e r
viii Contents
9. Augustine’s Hermeneutics as a Universal
Discipline!? 206
Ka r la Po l l m a n n
Bibliography 232
Index locorum 245
General index 255
 n ot e s o n c o n t r i b uto r s

Philip Bu r to n is Senior Research Fellow in Theology,

University of Birmingham, and Lecturer in Greek and
Latin, University of St Andrews.
Catherine M. Ch i n is Instructor in Church History, The
Catholic University of America.
Catherine Co n y b e a r e is Associate Professor of Greek,
Latin, and Classical Studies, Bryn Mawr College.
Stefan He ß b r üg g e n-Wa lt e r is wissenschaftlicher Mitar-
beiter at the Institut der Philosophie, Fernuniversität
Hagen.
William E. Kl i n g s h i r n is Professor of Greek and Latin,
The Catholic University of America.
Neil McLy n n is Professor in the Faculty of Law, Keio
University.
Karla Po l l m a n n is Professor of Classics at St Andrews
University.
Danuta R. Sh a n z e r is Professor of Classics and Medieval
Studies, University of Illinois at Urbana-Champaign.
Mark Ve s s e y is Professor of English, University of British
Columbia.
 a b b r e v i at i o n s

Titles of frequently cited works of Augustine

C. Acad. Contra Academicos
Beata Vita De Beata Vita
Civ. De Civitate Dei
Conf. Confessiones
C. Cresc. Contra Cresconium
C. Faust. Contra Faustum Manicheum
Disc. Chr. De Disciplina Christiana
Div. Qu. De Diversis Quaestionibus
LXXXIII
Doct. Chr. De Doctrina Christiana
En. Ps. Enarrationes in Psalmos
Ep. Epistulae
Gen. adv. Man. De Genesi adversus Manicheos
Imm. An. De Immortalitate Animae
Lib. Arb. De Libero Arbitrio
Ord. De Ordine
Quant. An. De Quantitate Animae
Retr. Retractationes
Serm. Sermones
Serm. Dom. Mont. De Sermone Domini in Monte
Sol. Soliloquia
Trin. De Trinitate
Util. Cred. De Utilitate Credendi
Ver. Rel. De Vera Religione
Titles of other ancient and medieval works are given in full
at their first appearance in the text or notes of a particular
chapter and are thereafter abbreviated in a way that should
be readily intelligible.

Titles of other frequently cited works

AL C. Mayer et al. (eds.), Augustinus-
Lexikon (1986– )
 Abbreviations xi
ANRW H. Temporini and W. Haase (eds.),
Aufstieg und Niedergang der römischen
Welt (1972– )
AugSt Augustinian Studies (1970– )
BA Bibliothèque Augustinienne (1936– )
CCSL Corpus Christianorum. Series Latina
(1953– )
CSEL Corpus Scriptorum Ecclesiasticorum
Latinorum (1866– )
GCS Die griechischen christlichen Schrift-
steller der ersten (drei) Jahrhunderte
(1897– )
GL Grammatici Latini, ed. H. Keil
(1857–80)
ILS Inscriptiones Latinae Selectae, ed.
H. Dessau (1892–1916)
JECS Journal of Early Christian Studies
(1993– )
JRS Journal of Roman Studies (1911– )
JTS Journal of Theological Studies, New
Series (1950– )
PL Patrologiae cursus completus. Series
Latina, ed. J. P. Migne (1844–64)
RE A. Pauly, Real-Encyclopädie der
classischen Altertumswissenschaft, ed.
G. Wissowa et al. (1893– )
REAug Revue des études augustiniennes
(1955– )
RechAug Recherches augustiniennes (1958– )
SC Sources chrétiennes (1942– )
StPatr Studia Patristica (1957– )
TLL Thesaurus Linguae Latinae (1900– )
 1
Introduction
Mark Vessey

As late as the 1950s, Peter Brown recently recalled, students

of the post-classical world ‘were still encouraged to sit in
on that most solemn and elevating of all track events: the
relay race of the formation of Western Christian civilization.
In this relay race, Augustine is seen to have picked up the
baton brought to him by Plotinus, all the way from Plato
and the ancient sages of Greece, and to pass it on tri-
umphantly to Boethius, and thence to Thomas Aquinas, to
Saint Bonaventure and now, who knows, to an Étienne
Gilson.’ Brown relates how, in the next decade, this edifying
spectacle was interrupted by new styles of scholarship,
more attentive to the social contexts of Augustine’s thought
as well as to his personal situation as a ‘man of Late
Antiquity’.1
It was not only Augustinian commentators who lost sight
of the baton around this time. Other stages of the Great
Western Relay were similarly affected in the 1960s, as long-
standing practices of intellectual history took a turn—often
but by no means always Marxian, Annaliste, or structuralist
in inspiration—towards subjects and methodologies associ-
ated with the social sciences. Spiralling academic production
and increasing specialization encouraged the trend towards
synchronic and loosely ethnographic modes of research.
Progressive politics led in the same direction. The study of
‘cultures’ in the plural replaced the doxologies of a singular

1
Introductory remarks at a conference on ‘Augustine and the Dis-
ciplines’ held at Villanova University, 9–11 November 2000, now in ‘The
Study of Augustine, 1950–2000: Evolving Disciplinary Contexts’, AugSt
32 (2001), 177–206, at 183.
2 Mark Vessey
(High) Culture. The epistemological challenge of so-called
postmodernism, when it came, may seem in retrospect little
more than the affirmation of a fait accompli. Already by
the early 1970s the grand narrative of ‘Western civilization’
was dissolving into a plethora of local descriptions and
micro-histories. While rarely a test bed for alternative modes
of cultural-historical analysis, the period and milieux of
Augustine’s life offered promising terrain for their further
development. For students venturing into it in the following
decades, the newly mapped and expanding ‘world of Late
Antiquity’ was open to every kind of interdisciplinary
inquiry, never having been claimed by any one discipline as
its own.
In fact, as Brown’s quietly polemical language indicates,
the study of Augustine’s role in the ‘transmission’ of intel-
lectual culture was neither concluded nor revolutionized in
the 1960s, but rather overtaken by competing presentations
of the man and his times. Among these, Brown’s own Augus-
tine of Hippo: A Biography has long occupied a place of
honour.2 Part of the wonder of that work, as its first readers
were quick to recognize, was that it could be so truly a
biography in the modern sense, so rich in contextual detail
and full of psychological insight. Both as biography and as a
contribution to the new historiography of the later Roman
Empire, Brown’s book stood in marked contrast to another
founding text of the twentieth-century science of Late
Antiquity, H.-I. Marrou’s Saint Augustin et la fin de la cul-
ture antique, to which it was none the less heavily indebted.3
There Augustine had been held up as an ideal representative
of the intellectual currents of his time and, in the best tradi-
tion of the relay-race approach to Western civilization, as
party to a historic succession. Marrou aimed to plot the
passage from classical intellectual culture, epitomized in
the ancient ideal of a unified scheme of the liberal arts (in
Greek, enkyklios paideia), to a ‘Christian culture of medieval

2
Peter Brown, Augustine of Hippo: A Biography (London, 1967),
reissued with two new chapters in 2000.
3
H.-I. Marrou, Saint Augustin et la fin de la culture antique (Paris, 1938),
reissued with a ‘Retractatio’ in 1949.
 Introduction 3
4
type’ based on the Bible. For this purpose, the central
Augustinian text was the treatise De Doctrina Christiana
(‘On Christian Teaching’). Brown in his 1967 biography
(e.g., in the chapter ‘Doctrina Christiana’) was content to
adopt Marrou’s view of Augustine’s personal intellectual
culture and scientific programme, and to embed it more
securely in its socio-historical context.
Partly as a result of Brown’s deft assimilation of it,
Marrou’s account of Augustine’s place in the history of
Western theories of disciplinary knowledge has remained
definitive for most Anglophone scholarship down to the
present.5 Yet there are difficulties with the narrative of
Saint Augustin et la fin de la culture antique which later
adaptations, for all the refinements they have brought, have
not addressed. They concern the formation of a Christian
biblical culture in the late antique and early medieval West,
and the long history of the liberal arts or disciplines.
The basic problem is of Marrou’s own making. In the
1949 ‘Retractatio’ of his Saint Augustin he implicitly with-
drew his earlier claim that study of Augustine’s work
could help explain the emergence of a ‘Christian culture of
medieval type’. As he now saw things, the intellectual uni-
verse of Augustine and his contemporaries—their part in

4
He completed the project with his Histoire de l’éducation dans l’antiq-
uité (Paris, 1948); his later revisions, partly prompted by Brown’s work,
especially The World of Late Antiquity (London, 1971), can be found in
Décadence romaine ou antiquité tardive? IIIe–VIe siècle (Paris, 1977). By
then, the new focus on mentalités had made Marrou’s original approach to
the history of culture seem outdated in France, a turn of events confirmed
by the commissioning of a chapter on ‘L’antiquité tardive’ from Peter
Brown for the opening volume of an Histoire de la vie privée, i, ed. Paul
Veyne (Paris, 1985). For a recent mise à jour of Marrou’s project, see
H. Inglebert, Interpretatio Christiana: Les Mutations des savoirs (cos-
mographie, géographie, ethnographie, histoire) dans l’Antiquité chrétienne
30–630 après J.-C. (Paris, 2001).
5
Distinguished recent instances include R. A. Kaster, Guardians of
Language: The Grammarian and Society in Late Antiquity (Berkeley and
Los Angeles, 1988), 84–8; M. Irvine, The Making of Textual Culture:
‘Grammatica’ and Literary Theory 350–1100 (Cambridge, 1994), 178–89;
F. Young, Biblical Exegesis and the Formation of Christian Culture
(Cambridge, 1997), 270–7; C. Harrison, Augustine: Christian Truth and
Fractured Humanity (Oxford, 2000), ch. 2.
4 Mark Vessey
the ‘civilization of the Theopolis’ as he called it in 1949—
would not survive the collapse of Roman civic institutions in
the West. The Christian culture of the Latin Middle Ages
would therefore necessarily be of a different type. Although
it could still be traced back to the fourth century, its origins
lay ‘outside the cultural mainstream, in the sequestered
milieux of the Desert’.6
This revision of Marrou’s thesis has consequences for any
history that hews to its line. If Augustine’s work is not the
harbinger of a new order of learning, any more than it is the
undertaker of an old one, then we are obliged to consider
afresh what its historical significance might be. What in fact
was the relationship of Augustine’s to other initiatives in
late antique Christian education, including those that may
be supposed to have had a more immediate impact? To sus-
pend Augustine’s status as representative of his times and
ambassador to posterity is to pose the problem of establish-
ing his relative position.7 That challenge is especially sharp
with respect to a subject as open to multiple construction as
‘biblical culture’.8
Wherever they are finally located, the historical co-
ordinates of Augustine’s biblicism will be close to those
assigned to his theory (or theories) of disciplinary know-
ledge. A central piece of Marrou’s original argument was

6
Marrou, ‘Retractatio’, 692 n. 2, with a reference to the discussion of
early monastic schools in the just published Histoire de l’éducation dans
l’antiquité; M. Vessey, ‘The Demise of the Christian Writer and the
Remaking of “Late Antiquity”: From H.-I. Marrou’s Saint Augustine
(1938) to Peter Brown’s Holy Man (1983)’, JECS 6 (1998), 377–411, at
383–91.
7
For a response to the implicit challenge of Marrou’s ‘Retractatio’ on
this point, see R. Markus, The End of Ancient Christianity (Cambridge,
1991), the later chapters of which chronicle the rise in the fifth- and sixth-
century West of an ascetically conditioned ‘community of discourse . . .
shaped by the scriptures’ (p. 222), in which the teachings of Augustine at
first play only a modest role.
8
Augustine’s own ‘biblical culture’ has recently been the subject of sev-
eral collections: A.-M. la Bonnardière (ed.), Saint Augustin et la Bible
(Paris, 1986), and its transatlantic twin, P. Bright (ed.), Augustine and the
Bible (Notre Dame, Ind., 1999), volumes best read in combination with
others in their series; F. Van Fleteren and J. C. Schnaubelt (eds.), Augus-
tine: Biblical Exegete (New York, 2001).
 Introduction 5
the demonstration that Augustine’s intellectual repertoire,
despite innovations that he was led to make by his Christian
(and increasingly ‘scriptural’) faith, was bounded by the
ordinary cycle of studies known to educated Romans of his
and earlier times. This demonstration was the more con-
fident for being based on the conviction, further justified in
Marrou’s History of Education in Antiquity and elsewhere,
that the unitary scheme of the liberal arts that appears in
Augustine’s early philosophical dialogues and that can also,
with a little ingenuity, be extracted from the latter part of
Book 2 of De Doctrina Christiana, had by then already been a
standard feature of Graeco-Roman higher culture for several
centuries.9
Augustine’s own remarks on the subject could seem at
first glance to confirm as much. In his Retractationes, the
catalogue of his works that he drew up towards the end of his

9
Marrou, Saint Augustin et la fin de la culture antique, 211–27. Note esp.
pp. 212–13: ‘Trouvant le programme médiéval des arts libéraux déjà
formé chez saint Augustin, il nous faut rechercher dans quelle mesure ce
plan d’études est chez lui une innovation ou au contraire l’héritage d’une
tradition antérieure . . . Le rôle qu’il assigne aux arts libéraux est chez
lui un héritage de la tradition philosophique hellénistique.’ Cf. Marrou,
History of Education in Antiquity (London, 1956) on the enkyklios paideia:
‘[E]ssentially, in the restricted sense in which it was used by the philo-
sophers, it always meant the seven liberal arts which the Middle Ages were
to take over from the schools of Late Antiquity. These seven liberal arts,
which were finally and definit[iv]ely formulated in about the middle of
the first century B.C., between the times of Dionysius Thrax and of
Varro, were made up of the three literary arts, the Carolingian Trivium—
grammar, rhetoric and dialectic: and the four mathematical branches of
the Quadrivium—geometry, arithmetic, astronomy and theory of music—
the traditional division since the time of Archytas of Tarentum, if not
Pythagoras himself’ (p. 177). Marrou’s dating of the definitive form of the
sevenfold scheme of the arts derives from the prolegomena to F. Marx’s
edition of Celsus (Leipzig, 1915) and from F. Ritschl, ‘De M. Varronis
disciplinarum libris commentarius’, in his Opuscula Philologica, iii (Leip-
zig, 1877), 352–402. For the reduction of the scheme of the De Doctrina
Christiana to a version of ‘le cycle normal des études qui formaient la base de
la culture des lettrés de la fin de l’antiquité’ (his italics), see Saint Augustin et
la fin de la culture antique, 402–7, with the revealing disclaimer: ‘Dans tout
ce qui suit je résume l’enseignement du [traité] et groupe les références
se rapportant au même ordre de faits sans m’astreindre à une analyse
regulière du livre dont la composition est comme toujours assez complexe’
(404 n. 1).
6 Mark Vessey
life, he looks back on the period of his baptism at Milan in
387. At that time, he recalls, he had entertained the idea of
composing a series of ‘books of disciplines’ (‘disciplinarum
libri’) in dialogue form. The overall purpose of this work
would be to discover and describe a way for the mind to
ascend ‘by certain steps through things material to things
immaterial’.10 Of the whole imagined series, he completed
one book on the art of grammar and another six on music.
For the other five disciplines he meant to treat—namely (as
he now listed them) dialectic, rhetoric, geometry, arithmetic
and philosophy—he had got no further than first drafts.
Some idea of what he had in mind can be gleaned from
the dialogue De Ordine, composed at the country estate of
Cassiciacum outside Milan to which Augustine retired with
some friends in the autumn of 386. In Book 2 of that work
he presents a myth of the invention of the liberal arts by
Reason, bracketed with twin references to the Roman poly-
math and occasional dialogue partner of Cicero, Marcus
Terentius Varro.11 Since Varro is known to have written
Disciplinarum Libri (now lost), it would be possible (1) to
infer that his work was an important source for these ideas
and thus (2) confirm that the scheme presented in De Ordine,
which already corresponds closely to the medieval canon of
seven liberal arts, was established by the middle of the first
century b c.12
Marrou’s history of the liberal arts and the scholarly con-
sensus that formed around it was called into question by
Ilsetraut Hadot in an important 1984 study.13 She argued
that the ‘medieval’ ordering of the seven liberal arts was an
invention of late antique (or Middle) Platonism, which

10
Retr. 1. 6; full text and translation below 71, 141. See W. Hübner,
‘Disciplinarum libri’, AL ii. 485–7.
11
Ord. 2. 12. 35; 2. 20. 54.
12
A. Dyroff, ‘Über Form und Begriffsgehalt der augustinischen
Schrift De ordine’, in M. Grabmann and J. Mausbach (eds.), Aurelius
Augustinus: Festschrift der Görresgesellschaft zum 1500. Todestage des
Heiligen Augustinus (Cologne, 1930), 15–62. Marrou neither formally
adopts nor explicitly contests Dyroff’s hypothesis of Augustine’s reliance
on Varro.
13
I. Hadot, Arts libéraux et philosophie dans la pensée antique (Paris,
1984).
 Introduction 7
Augustine had probably found in a (lost) work of the
third-century philosopher Porphyry. Nothing so systematic
or conceptually coherent was likely to have been available
to him in Latin texts of earlier periods; in this respect, his
citations of Varro were mostly red herrings.14 Hadot also
drew attention to the fact, obvious enough to any reader of
De Doctrina Christiana, though somewhat obscured by the
manner of Marrou’s dealing with that text, that Augustine
had given up on the liberal arts as a Christian propaedeutic
by the mid-390s.15 A slightly(?) later African writer,
Martianus Capella, would show renewed enthusiasm for the
cycle of seven sciences in his Marriage of Philology and
Mercury.16 To see such a scheme entrenched in a specifically
Christian programme of learning, however, one would have
to wait until the mid-sixth century, when Cassiodorus,
ex-minister of the Ostrogothic regime in Italy, set about
reharmonizing ‘divine and human learning’ in the works of
his semi-monastic retirement at Scyllacium in Calabria.17

14
For rebuttals of Marrou’s thesis on the early formation of the cycle of
seven liberal arts, see ibid., esp. pp. 52–7, and of Dyroff’s argument for
Varro as a source for the scheme in De Ordine, 132–6.
15
Ibid. 137: ‘Dans le domaine de l’éducation chrétienne qu’Augustin
décrit dans son traité De la doctrine chrétienne, le cycle des sept sciences
n’est plus retenu comme modèle . . . [S]i, auparavant, dans le traité
Sur l’ordre, l’étude des sept sciences avait été subordonnée à celle de la
philosophie, les études recommandées dans le traité De la doctrine chréti-
enne n’ont plus de valeur que dans la perspective de l’exégèse scripturaire,
c’est-à-dire qu’elles ne sont autorisées que dans la mesure où elles nous
aident à comprendre la Bible.’
16
Ibid. 137–55, again rejecting Varro as a likely source. See further
pp. 156–90, on ‘La Question Varronienne’, with critique of Ritschl,
pp. 162–76, concluding that ‘Quoi qu’on ait pu dire et écrire des Discipli-
narum libri de Varron, nous ne savons quasiment rien sur leur contenu et
sur leur ordre.’
17
Ibid. 191, and, for the hypothesis of a Greek source for Cassiodorus’
conception of the seven liberal arts, pp. 199–205, with the concession
that when Cassiodorus refers to the precepts of the ‘sancti patres’ at Inst.
2. 3. 22 (ed. Mynors) he could have been thinking of Augustine’s De
Ordine (p. 205). Hadot’s treatment of the Cassiodorian scheme of the arts
builds on and modifies that of P. Courcelle, Late Latin Writers and Their
Greek Sources, trans. H. E. Wedeck (Cambridge, Mass., 1969), 339–54.
Further discussion in J. W. Halporn and M. Vessey (eds.), Cassiodorus:
‘Institutions of Divine and Secular Learning’ and ‘On the Soul’ (Liverpool,
2004), 64–79.
8 Mark Vessey
Scholars of Augustine have been wary of accepting all
Hadot’s hypotheses, with good reason.18 As a corrective to
Marrou’s rather sweeping history of the liberal arts between
the Hellenistic and medieval periods, her critique none the
less retains its value. For one thing, it forces us to reckon
with the historical singularity of Augustine’s projected ‘dis-
ciplinarum libri’. By insisting on the representativeness of
Augustine as a lettré de la décadence, Marrou risked making
this enterprise appear merely a routine product of the
ambient culture. It was clearly anything but that. Whatever
sources may have been available to him that are lost to us,
Augustine is the first person known to have essayed a large-
scale theory of the liberal arts that would integrate them
into a Christian programme of education.19 Whether his in-
spiration was a fairly recent trend in Neoplatonic philosophy
(as Hadot maintains), a scholastic formula already en-
trenched in Latin texts of the classical period (as Marrou
held), or a measure of both, the result was at once something
uncommon and—as it turned out—full of consequence for
the later ‘tradition’.
The instruments of Augustine’s intellectual culture were
predominantly Latin, and it is worth underlining the fact
that the subsequent vogue for the liberal arts as a system
is likewise a Latin, Western phenomenon. Contrary to
Marrou’s revised thesis of the insuperable rupture caused by
a breakdown of Roman institutions in the West in the fifth
and sixth centuries, the kind of unified theory of the liberal
arts that we first see clearly outlined by Augustine was in fact
to prove remarkably durable and portable. After Martianus
Capella (Africa), Macrobius (an African(?) at Rome), and
Claudianus Mamertus (Gaul) in the fifth century, its chief
exponents are Boethius and Cassiodorus (Italy) in the sixth,
and Isidore (Spain) in the seventh. By then the ‘medieval’

18
See the excursus on ‘The Liberales Disciplinae’ in J. J. O’Donnell,
Augustine: Confessions, 3 vols. (Oxford, 1992), ii. 269–78, doubting in par-
ticular ‘[w]hether Hadot is correct in dissociating the disciplinae entirely
from the patronage of Varro’ (p. 270); M. Fussl, ‘Disciplinae liberales
(I–III)’, AL ii. 472–81 (a valuable synthesis) at 473.
19
On earlier, less ambitious formulations by Clement and Origen, see
Hadot, Arts libéraux et philosophie, 287–9 and 299–301 (Origenian scheme
incorporated by Cassiodorus).
 Introduction 9
canon, as we can retrospectively call it, was largely in place.
Even if modern-day defenders of the classical tradition may
still claim the Liberal Arts as a bequest from Hellenistic
Greece and Republican Rome, all the evidence points to later
Latin antiquity as the site of their definitive promulgation.
Augustine’s role in the process deserves further scrutiny,
despite his own mature judgement against the liberal arts as
a template for Christian education.
That seeming contradiction may indeed be the crux of the
matter. For if Augustine made a point of downgrading the
liberal arts in De Doctrina Christiana, with such complete
success that no one could now (re)constitute anything like
the plan of the medieval trivium and quadrivium from that
text of his alone, was it perhaps because he had made a point
of promoting the very same disciplines as the ideal trajectory
of a Christian education just a few years earlier—and had
since changed his mind? Public self-revision was a habit of
Augustine’s long before he set about his final Retractationes.
The more purposive and personal, the less straightforwardly
a reflex of the collective culture, we allow the project of the
Augustinian ‘disciplinarum libri’ to be, the more reason
there will be for Augustine himself deliberately to over-
write it. (We can assume that copies of the early dialogues
had circulated quite widely among his friends and clerical
colleagues in Italy and North Africa by the mid-390s,
making it impossible for him to retract them en bloc.) As a
head-on confrontation with the higher educational norms of
his time, De Doctrina Christiana would be an odd piece of
work. On the other hand, the brilliantly original taxonomic
critique of traditional ‘doctrinae’ in Book 2 makes excellent
sense as the second or latest thoughts of a writer who had
shown similar flair in an earlier reworking of received
categories.20 We know now that Augustine’s best efforts
could not prevent the assimilation of materials from or pre-
paratory to his ‘disciplinarum libri’ into medieval corpora of
the liberal arts. Such assimilation would thus occur both in

20
For an argument along these lines, see F. Van Fleteren, ‘Augustine,
Neoplatonism, and the Liberal Arts’, in D. W. H. Arnold and P. Bright
(eds.), De Doctrina Christiana: A Classic of Western Culture (Notre Dame,
Ind., 1995), 14–24.
10 Mark Vessey
spite and because of what he had written. This principle
of double agency may be taken to apply with special force to
De Doctrina Christiana, a work in which the critique of one
vision of disciplinary knowledge (‘liberal’) occurs at the
same time as the proclamation of another (‘biblical’).
Of the chapters in the section of Marrou’s thesis entitled
‘Doctrina christiana’, that on ‘La Bible et les lettrés de la
décadence’ later struck him as one of the least satisfactory in
the book.21 The collective plural of the chapter title points
once more to our problem. As long as Augustine is imagined
as speaking for an informal committee of his contem-
poraries, his works must either vaguely prefigure some future
ideal (the ‘biblical culture’ of the Latin Middle Ages, as
Marrou supposed in 1938) or else encapsulate a lost late
antique civilization (of the ‘Theopolis’, as he named it in
1949). For their historical import to be measured at all pre-
cisely, we must first stop to consider how representative in
fact any of them was.
Such consideration is impeded in this case by the extra-
ordinary prestige eventually accorded De Doctrina Chris-
tiana.22 It may—indeed should—come as something of a
shock, so late in scholarship, to realize how hard it still is to
account for this work in terms of its sources and analogues,
and hence how bizarre it is likely to have seemed to its
earliest readers, or such as there were of them.23 Again, we

21
‘Retractatio’, 646; the regrets have to do mainly with his earlier treat-
ment of Augustine’s view of the spiritual or allegorical interpretation of
Scripture.
22
See also C. Schäublin, ‘De Doctrina Christiana: A Classic of Western
Culture?’, in Arnold and Bright (eds.), De Doctrina Christiana, 47–67, a
notably clear-sighted essay.
23
On the early circulation of the text, consult M. Gorman, ‘The
Diffusion of the Manuscripts of Saint Augustine’s De doctrina christiana
in the Early Middle Ages’, Revue Bénédictine 95 (1985), 11–24, repr. in his
The Manuscript Traditions of the Works of St Augustine (Florence, 2001),
265–78; K. B. Steinhauser, ‘Codex Leningradensis Q.v.I.3: Some
Unresolved Problems’, in Arnold and Bright (eds.), De Doctrina Chris-
tiana, 33–43; J.-P. Bouhot, ‘Augustin prédicateur d’après le De Doctrina
Christiana’, in G. Madec (ed.), Augustin prédicateur (395–411): Actes
du colloque international de Chantilly (5–7 septembre 1996) (Paris, 1998),
49–61; I. Opelt, ‘Materialien zur Nachwirkung von Augustins Schrift
De doctrina christiana’, Jahrbuch für Antike und Christentum 17 (1974),
 Introduction 11
have no reason to think that Augustine created a new dis-
ciplinary programme (one we might now call ‘hermeneutics’)
out of thin air. Indeed, it would hardly have been wise for
him to do so. A newly appointed bishop could not afford
to be too innovative. Even if he was bluffing a bit when he
began his preface by stating matter-of-factly, ‘There are
certain precepts for the treatment of the Scriptures’, it was a
bluff that he would not have wished to hear called too loudly
at the time. The idea of a proper, more or less rigorous
method for interpreting and expounding the Bible was
gaining currency. Augustine would not be the first in the
field, even among Latin theologians.24 Yet, granting that, we
may also allow that he would now be as inventive a theorist
and advocate of biblical studies as he had once been of those
called liberal.25
Research into Augustine’s intellectual debts and affinities
in De Doctrina Christiana has advanced steadily in recent
years.26 As evidence accumulates of his theological and
other partis pris, the chances of his being taken or mistaken
for a ‘typical’ well-educated Christian of his time are
reduced. Tightly focused studies have made good the
inevitable defects of Marrou’s pioneering study. By their

64–73. There is a helpful survey of the work’s later fortunes in the intro-
duction to R. P. H. Green’s edition of De Doctrina Christiana (Oxford,
1995), pp. ix–xxiii.
24
Apart from the ex-Donatist Tyconius, and in many ways more
imposing, there was Jerome. On the possible impact of the latter’s
scholarly propaganda on Augustine in the 390s, see M. Vessey, ‘Con-
ference and Confession: Literary Pragmatics in Augustine’s “Apologia
contra Hieronymum” ’, JECS 1 (1993), 175–213. G. Strauss, Schrift-
gebrauch, Schriftauslegung und Schriftbeweis bei Augustin (Tübingen, 1959)
remains fundamental.
25
K. Pollmann, Doctrina Christiana: Untersuchungen zu den Anfängen
der christlichen Hermeneutik unter besonderer Berücksichtigung von
Augustinus, De doctrina christiana (Fribourg, 1996), with comments by
C. Kannengiesser, B. Studer, and F. Van Fleteren, and a response by the
author, in AugSt 29 (1998), 99–137.
26
Review of scholarship by K. Pollmann, ‘Doctrina christiana (De-)’,
AL ii. 551–75. Note also B. Kursawe, Docere – Delectare – Movere: Die
officia oratoris bei Augustinus in Rhetorik und Gnadenlehre (Paderborn,
2000); G. Lettieri, L’altro Agostino: Ermeneutica e retorica della grazia
dalla crisi alla metamorfosi del De Doctrina Christiana (Brescia, 2001).
12 Mark Vessey
very specialization, however, they have put off confronting
the larger questions raised by the same scholar’s narrative
history of the disciplines, his own revision of it, and the
challenge offered by Hadot.
Those questions may be simply stated:

1. What is the historical significance of Augustine’s ‘con-

version’ from a theory of the liberal arts to a theory of
biblical interpretation?
2. What prompted and facilitated his change of perspective?
3. How did it fit or clash at the time with other fashions in
cultural theory and (Christian) pedagogy?
4. What were its immediate and longer-term effects?

These are issues for any history of Western intellectual cul-

ture that aims to be more than commentary on a relay race
of Great Thinkers. Even to begin to address them, we need
to keep both aspects of the problem in view—the history of
the ‘liberal arts’ and that of the ‘Bible’, or ‘biblical culture’.
The subject is by nature an interdisciplinary one. Its
inherent complexity invites scholarly collaboration. Such
considerations have shaped the present volume on ‘Augus-
tine and the Disciplines’. As a collection, it aims to shed new
light on the first three questions listed above and thereby
provide a more secure basis for future investigation of the
fourth. (Only incidentally is the reception of Augustine’s
disciplinary thinking treated here.)
As noted in the Preface, the essays presented below
(with the exception of Stefan Heßbrüggen-Walter’s) were
originally composed for a conference on ‘Augustine and the
Disciplines’ held at Villanova University, in November
2000. As soon as that meeting got under way, it became
clear that the most doubtful term in its title was the definite
article. To various degrees, almost all speakers pressed for a
loosening of our sense of ‘the disciplines’ spoken of in
conjunction with the name of Augustine. With respect to
scholarship since Marrou, two tendencies were confirmed:
(1) an interest in placing Augustine’s work as unceremoni-
ously as possible within a long history of ancient systems of
the arts and their medieval–modern successors, and (2) an
intuition that his texts both radically revise traditional dis-
 Introduction 13
ciplinary structures and participate in the formation of a
Christian ‘biblical culture’.
Like Marrou’s Saint Augustin et la fin de la culture
antique,27 though without the ambition of comprehensive-
ness, this volume is arranged in three parts which broadly
recapitulate Augustine’s personal experience from his
schooldays in the 360s and 370s to the simultaneously retro-
spective and prospective vision of his Confessions, begun
around 397 when he set aside the first draft of De Doctrina
Christiana.28
Part I (‘Honesta studia: Classrooms without Walls’) con-
tains two essays on the social, institutional, and ideological
parameters of late antique higher education, as first under-
gone and then reconstructed by Augustine. Reading the
Confessions against the equally unreliable memoirs of
Gregory Nazianzen and his student fraternity in Athens,
Neil McLynn makes a fresh estimate of the peculiarity
of Augustine’s case. As regards what would now count as
pedagogy or curriculum, the overriding impression obtained
from these student reminiscences is of disorder rather
than discipline. Augustine, it is argued, may have designed
his ‘disciplinarum libri’ to mark ‘a decisive break with the
day-to-day untidiness of the schoolroom’. A more lasting
analogy between East and West is suggested by McLynn’s
view of the Philocalia, the late antique collection of excerpts
from the works of Origen, as a kind of club-book for serious
young Christians like Gregory and Basil, containing ‘en-
couraging intimations of the possibility of an authentically
Christian paideia’. Denied such encouragement in his late
adolescence, Augustine looked elsewhere. Years would pass
before he encountered a biblical pedagogy comparable with
Origen’s, and many more before he himself would begin to
offer a fuller treatment of issues in biblical interpretation

27
The sections of Marrou’s book are entitled ‘Vir eloquentissimus ac
doctissimus’, ‘Studium sapientiae’, and ‘Doctrina christiana’.
28
The break occurs at Doct. Chr. 3. 25. 35/6; the treatise was finished
thirty years later (Retr. 2. 4). On the composition dates of the Confessions,
see now P.-M. Hombert, Nouvelles Recherches de chronologie augustinienne
(Paris, 2000), 9–23, who accepts the traditional date of 397 for the incep-
tion of the work and places its completion in 403.
14 Mark Vessey
than any since Book 4 of the latter’s Peri Archōn (‘On First
Principles’). Meanwhile, in the retirement of Cassiciacum, a
brilliant student again among students, he drew on resources
nearer at hand: those of the Ciceronian–Platonic philo-
sophical dialogue. In that most conventional and nostalgic
of didactic genres (in the early part of Book 2 of De Ordine)
Catherine Conybeare finds him devising a ‘theology’ of dis-
ciplinarity which, by undercutting the Neoplatonic distinc-
tion between the divine and the corporeal, already casts
doubt on the logic of his proposed Christianization of
the liberal disciplines. On this view, the abandonment of the
‘disciplinarum libri’ would reflect Augustine’s own recogni-
tion of the project’s conceptual inadequacy; his developing
thought now demanded a less constraining framework.
Bold ‘exercise in self-definition’ (McLynn), or obstacle to be
overcome (Conybeare)? Knowing that these ‘libri’ would
never exist as a finished series, we are free to see them as both.
Part II (‘Disciplinarum libri: The Canon in Question’)
transfers the title of Augustine’s first great ‘opus imperfec-
tum’ to the larger array of works in which he and other Latin
writers discussed or enumerated the liberal (and other) arts.
The opening essay by Danuta Shanzer is the hinge of the
present collection. Siding with Ritschl and Dyroff against
Hadot, she considers the hypothesis of a Varronian model
for the ‘medieval’ scheme of the liberal arts well founded,
and supplies new reasons for regarding Augustine as a
crucial witness to this line of descent. Her arguments have
important consequences both for our conception of Varro’s
original Disciplinarum Libri and for our history of the
liberal arts in the West from the fourth century onwards.
The restoration of Varro’s lost Musae—the personified arts
and the work that apparently also went by that name—makes
it easier to appreciate the difficulties Augustine faced in
attempting to adapt these traditional figures of disciplinary
thought to Christian purposes. At the same time, the newly
assembled evidence of the survival of Varronian themes
and motifs in Late Antiquity makes it harder to suppose
that any single author of that period, even one with the
powers of Martianus Capella, could ever have restructured
the tradition decisively. This is not to say that all problems
are solved. In modern scholarship, as in the imaginations of
 Introduction 15
former ages, it seems, an air of mystery will forever surround
the Muses, whether nine or seven.
Why did Augustine give up on the liberal arts? Among
possible ‘external’ factors, Shanzer mentions the influence
of Ambrose (on which, see now the essay by Heßbrüggen-
Walter). ‘But above all’, she writes, ‘there was the intimi-
dating realization that all of the Bible lay ahead of him for
exegesis, and that there would never be world enough and
time.’ The character here glimpsed is already that of the
impatient lover of Confessions 11.2.2, the one who has left
behind all arts, liberal (10. 8. 16), mechanical (10. 34. 53) and
superstitious (10. 35. 56), in an urgent desire to embrace the
only true Mediator between God and humankind (10. 42.
67–43. 70), Christ, the Truth who speaks in Scripture and of
whom the Scripture speaks (11. 2. 4). This is the Augustine
of c.396/7–400, who in the foreshortened, disciplinary-
historical perspective of this volume is also the Augustine of
the City of God (413–27), a work that disposes of the
accumulated disciplinary learning (‘litterae’) of the ancient
Graeco-Roman world in its first ten books without so much
as listing the liberal arts, then cleaves to the uniquely
authoritative text of Scripture (Civ. 11. 1).
At what stage of his intellectual and public career between
386 and 396/7—that is, between the serial dialogues of Cas-
siciacum and the solitary double labour of De Doctrina
Christiana and the Confessions—should we mark Augustine’s
turn from one kind of ‘litterae’ to another? Such datings are
crude approximations at best. Were we to pursue this one,
however, we might profitably pause over those linked works
De Vera Religione and De Utilitate Credendi, respectively
the last composed before his ordination in 391 and the first
completed after it (Retr. 1. 13. 14).29

29
Cf. O’Donnell, Augustine: Confessions, ii. 278; M. Cameron, ‘The
Christological Substructure of Augustine’s Figurative Exegesis’, in
Bright (ed.), Augustine and the Bible, 74–103. Augustine, Retr. 1. 11, on the
De Musica, last survivor of the projected ‘disciplinarum libri’, begins with
an explicit disavowal of the scheme of ascent via the corporeal to the
‘invisibilia dei’, to proclaim instead the necessity of a faith in Christ, ‘qui
unus mediator est dei et hominum’. On the importance of the De Vera
Religione for themes later developed in the City of God, as well as for the her-
meneutical concerns of De Doctrina Christiana, see G. O’Daly, Augustine’s
16 Mark Vessey
William Klingshirn takes his cue from a sermon preached
by Augustine in Hippo a few years later, shortly before he
became bishop, in which he provides ‘a comprehensive
survey of all the modes of divine communication with
humankind’. The one literally textual mode mentioned is
that constituted by the ‘instrument’ of the Divine Scrip-
tures. In his next breath, the preacher refers to several
procedures of divination. As this essay shows, Augustine’s
views on divination are a revealing index of his changing
sense of (the) discipline(s), including in due course the dis-
cipline of biblical interpretation. In effect, the pagan art of
divination becomes a kind of definitive ‘anti-discipline’ for
the purposes of Augustine’s theory. Thus he begins his
critical review of traditional forms of learning in Book 2 of
De Doctrina Christiana with an inventory of superstitious
practices, among which examples from technical divination
are to the fore. ‘This makes sense,’ observes Klingshirn,
‘since Book 2 is about knowledge and the interpretation of
unknown signs [i.e., in Scripture], and technical divination
dealt with both.’ We may wonder to what extent the semiotic
slant of Augustine’s guide to the new science of Scripture
was dictated by the prominence of sign theory in traditional
Graeco-Roman religion. Is it possible that the science that
we now call hermeneutics, and that Karla Pollmann (in her
essay here) finds first theorized in De Doctrina Christiana,
owes as much to the model of certain ancient mantic arts—
described by Martianus Capella, evoked by Macrobius and
other late antique commentators on Virgil—as it does to the
liberal arts of grammar, dialectic, and rhetoric?
The example of divination introduces a series of essays on
Augustine’s revision of received disciplinary structures. The
last in Part II completes the review of plenary lists of liberal
‘artes’ or ‘disciplinae’30 by considering his Latin vocabulary

‘City of God’: A Reader’s Guide (Oxford, 1999), 266–70. And note also
Augustine’s Ep. 21 to Bishop Valerius of Hippo, written soon after his
unexpected priestly ordination, in which he pleads for leisure for biblical
study, ‘ad cognoscendas divinas scripturas’.
30
For the sense of a distinction sometimes made in Latin between ‘ars’
and ‘disciplina’, the former translating the Greek technē, the latter
epistēmē, see Hadot, Arts libéraux et philosophie, 193–9, with special refer-
ence to dialectic.
 Introduction 17
for the canon, especially as presented by the Confessions.
Already in Book 2 of De Doctrina Christiana the sense of a
pre-ordained unity of learning (outside Scripture) is under-
mined by his deployment of a scheme of his own, in
which some disciplines (e.g., music, astrology) are sundered
from their usual congeners, while elements intrinsic to
another (grammar) are distributed across different classes.
Traditional structures are further weakened by Augustine’s
suppression of the usual names for these disciplines, a
device that must have been disconcerting for late fourth- or
early fifth-century Latin readers used to the ‘Varronian’
pattern and the appearance of solidity conferred on the
ensemble of the arts by their uniformly Greek titles. As if
to compound that effect, Augustine next proceeded to give
an account of his own education, in which the contours of
disciplinary knowledge were similarly blurred. As Philip
Burton demonstrates here, the idiom of intellectual activity
and accomplishment found in the Confessions strongly
favours Latin translations or near-equivalents of Greek
loan-words. Why is that? ‘In avoiding the traditional
[Greek] terminology of the arts’, Burton infers, ‘Augustine
was attempting to revalorize them’ for a Latin Christian
readership. This recoding, a sequel to the failed attempt to
recycle the enkyklios paideia in ‘libri disciplinarum’, aims at
more than just the effacement of antiquated institutions of
non-Christian learning. It also promotes the idea, explicitly
formulated in De Doctrina Christiana, that the useful and
salutary practice of arts otherwise known as liberal is
already exemplified in the sacred texts of Christianity.
The less narrowly ‘technical’, more broadly evocative (or
pre-disciplinary) vocabulary of the Confessions enables
Augustine to conduct his discussion of Christian intellectual
life in language concordant with Latin versions of Scripture.
By the time he finished his Confessions, Augustine’s
reflections on divine and human knowledge had reached
approximately the point from which he would begin to
elaborate his City of God a decade later.31 To follow his
31
Cf. M. Vessey, K. Pollmann, and A. D. Fitzgerald (eds.), History,
Apocalypse, and the Secular Imagination: New Essays on Augustine’s ‘City
of God’ (Bowling Green, Oh., 1999), 9 f. (also published as AugSt 30/2
(1999), same pagination).
18 Mark Vessey
course from there on, as Burton suggests, requires us to
engage with themes of ‘authority’ and ‘providence’ that are
vital to an understanding of his thought and its influence,
but which rapidly exceed the scope of ancient and modern
discourses of disciplinarity per se. Having come to a kind of
limit with the Confessions, we return instead to the conjugate
work De Doctrina Christiana. That is to say: we return and
advance at once, since De Doctrina Christiana was itself to be
completed, substantially ‘unretractated’, around the same
time as the City of God, almost at the end of Augustine’s life.
Part III (‘Doctrina christiana: Beyond the Disciplines’)
consists of three essays which try variously to capture this
sense of limits reached and passed. To the extent that they
do, they may be read as symmetrical to the opening two
essays with their emphasis on the shifting borderlines of
Augustine’s early dialogues and confessional self-definition.
Heßbrüggen-Walter’s contribution, moreover, takes us back
to questions about the ‘philosophical life’ previously raised
by Conybeare. Once again, strict chronology is not the goal.
Augustine was fond of images of disciplinary gradation.
Yet, as McLynn points out, we can rarely say exactly when
he ‘graduates’.
Despite differences of approach, Catherine Chin’s placing
of De Doctrina Christiana in the context of grammatical
education bears out McLynn’s earlier depiction of graduates
of the late Roman school system as carriers of heterogeneous
‘cargoes of culture’. Awkward and reductive as it would be
to construe Augustine’s treatise on biblical interpretation as
an ‘ars grammatica’, the difficulty we now experience in
assimilating this work to any single ancient model (see espe-
cially Pollmann’s essay) may be a sign of its special relation-
ship to grammar, the primary discipline for all practical
curricula of higher education in Augustine’s time. For it was
the grammarian’s function, Chin observes, ‘to redistribute,
piecemeal, the knowledge produced in other disciplines’. His
classroom was the first loading port for aspiring merchant-
men of Mediterranean paideia. In such important cognate
texts as Macrobius’ Saturnalia and Martianus Capella’s De
Nuptiis Philologiae et Mercurii, ‘[n]ot only is the grammarian
configured, ideally, as a polymath, but grammar itself is
understood to be the practice of breaking down larger fields
 Introduction 19
of knowledge into “mobile” decontextualized units’. In
pointing to the consonance of this habit of discursive
‘transferral’ with the famous image of the spoliation of the
Egyptians (Doct. Chr. 2. 40. 60), and interpreting the latter as
master trope of a discourse that produces the opposition of
‘Christian’ to ‘pagan’ culture through a process of de- and
re-contextualization, Chin seeks to explain the dynamics of
love (‘caritas’, but also ‘amor’) inscribed by the sign theory
of De Doctrina Christiana. Augustinian ‘grammar’, on her
analysis, becomes both an engine for subverting prior dis-
ciplinary formations and a self-constituting divine instru-
ment, mediating between human readers and the object
of their desire—thus homologous (if not identical?) with
Scripture, and with Christ. This is a surprising—indeed
barely tolerable—reading. By its own decontextualizing
strategy and selective use of modern theorists, it can perhaps
restore some of the shock value of Augustine’s own acts of
theory.
Ambrose of Milan is cited by Augustine as a great
‘despoiler’ of the Egyptians, and also as the teacher who
(according to the Confessions) decisively reorientated him to
the ‘auctoritas’ of Scripture.32 As Stefan Heßbrüggen-
Walter argues, he may also have been partly responsible for a
major revision in Augustine’s thinking about dialectic. Espe-
cially when rendered by more Latinate terms (see Burton’s
essay), dialectic continues to hold great sway in Augustine’s
mature thought. As the science of definition and division, it
may be thought to play at least as large a role as grammar in
the discursive transformations projected by De Doctrina
Christiana. But it is another aspect of dialectic, namely its
power to generate ontological truths, that is the main focus
of Heßbrüggen-Walter’s essay. Having granted dialectic
such power in early works like De Ordine, Augustine had
evidently changed his mind by the mid-390s. According to
De Doctrina Christiana, dialectic ‘is concerned with the for-
mal derivation of valid syllogisms, not the material truth of

32
For strong arguments against undervaluing Ambrose’s early influence
on Augustine’s approach to Scripture, see now M. Dulaey, ‘L’Apprentis-
sage de l’exégèse biblique par Augustin’, REAug 48 (2002), 267–95 and
49 (2003), 43–84.
20 Mark Vessey
propositions, which in matters of the faith . . . can only be
ascertained by appeal to the revealed truth of Scripture’. It
seems that Augustine may have been helped to this position
by arguments or gestures used by Ambrose in polemic
against some of the more skilful Arian dialecticians, who
allegedly sought to derive theological truths independently
of divine revelation. As later for Kant, Heßbrüggen-Walter
concludes, dialectic for Augustine (c.396/7) is ‘not an orga-
non, but a canon for cognition’. The Christian’s only true
organon now consisted of the texts collectively known, in
Latin since Tertullian, as ‘instrumenta’ (sc., ‘vetus et
novum’). Of a canon in the looser and more ambitious sense
implied by our subtitle for Part II, no clear trace remains in
Augustine’s later work.
Dialectic, Heßbrüggen-Walter reminds us (after Augus-
tine), was sometimes defined as a discipline ‘instrumental’
to other branches of rational inquiry, a ‘disciplina disciplin-
arum’ in the sense that none of them could proceed without
it. It will be clear by now that the instrumentality imputed
by Augustine to the activity he calls ‘tractatio scripturarum’
is of another, more transcendent order. Although itself dis-
pensable in certain cases, as he points out in the prologue
to De Doctrina Christiana and elsewhere, the ‘science’ of
biblical interpretation (2. 7. 9) subserves no other intel-
lectual discipline. In fact, no other intellectual discipline
receives formal or institutional notice within the Christian,
ecclesial, theological world-view of De Doctrina Christiana,
whatever allowances are made for the continuing vitality of
traditional, non-Christian forms of learning in a separate
sphere.
Would it then be wrong to speak of Augustine’s settled
conviction of the late 390s as a ‘reductio omnium artium
ad scientiam solam scripturarum’? What qualifications are
required for such a formula to fit the case? The final essay,
by Karla Pollmann, addresses these issues under the title:
‘Augustine’s Hermeneutics as a Universal Discipline!?’ As
an account of De Doctrina Christiana, set against the back-
drop of ancient theorizing on language, the interpretation of
texts, and the relation of human to divine knowledge, with
an eye to Augustine’s earlier initiatives from Cassiciacum
onwards and to the related work of his Confessions, this piece
 Introduction 21
draws together the main threads of our collective inquiry.
It could serve equally well as an alternative introduction.
The combination of an exclamation mark and a question
mark in Pollmann’s title should, in any case, be understood
as applying to the whole of the present subject-matter. More
than for any exemplarity that can be credited to it, despite
whatever partial models and analogues can be suggested for
it, the hermeneutic ‘discipline’ announced by Augustine in
De Doctrina Christiana is perhaps most impressive in the end
for its singularity—its strange, historical eventfulness. This
volume will have achieved its goal if it succeeds in communi-
cating a sense of shared surprise at so belated a (re)discovery.
 pa r t i
Honesta studia: Classrooms without Walls
 2
Disciplines of Discipleship in Late Antique
Education: Augustine and Gregory Nazianzen
Neil McLynn

1. c h r i st i a n s i n t h e c la s s r o o m
Cassiciacum was not Augustine’s first experiment in recon-
ciling traditional patterns of education with Christian
faith. Seventeen years earlier, at Carthage, he had tried
to match his schoolbooks with the Bible, and had sought
love both among his fellow students and in the city’s
churches. Augustine’s student experiences have an obvious
relevance to the programme of studies that he would later
devise in the laboratory conditions of Cassiciacum—the
two represent, indeed, the formal termini of a long, continu-
ous engagement in (so-called) ‘honesta studia’. Moreover,
Augustine’s narrative of his period at Carthage is the longest
surviving account of the experience, common to well-born
young men across the whole empire, of being herded
through the teeming rhetorical schools.1 Yet it has remained
comparatively neglected by historians.2 This is partly due,
no doubt, to its deeply idiosyncratic character, as more a
‘tissue of meditative abstractions’ than a straightforward

1
‘Per totum orbem rhetorum scholae adolescentium gregibus pers-
trepant’: Augustine Util. Cred. 7. 16, remarking upon the meagre results
of such application.
2
B. Stock, Augustine the Reader: Meditation, Self-knowledge, and the
Ethics of Interpretation (Cambridge, Mass., 1996), 3, identifies Augus-
tine’s ‘comments on ancient methods of instruction in literate disciplines’
as a possible approach to his theme, but declines to pursue them in any detail;
I. Hadot, Arts libéraux et philosophie dans la penseé antique (Paris, 1984),
contains many valuable reflections upon the schools (pp. 215–61), but
operates at a higher level of abstraction than is attempted here.
26 Neil McLynn
3
narrative; since Henri Marrou, moreover, it has become so
instinctive to measure Augustine’s educational attainments
in relation to the cultural inheritance of the classical past that
we easily lose sight of those less formally academic criteria
that may have made Augustine the student a child of his
own distinctive time and place.4 The central purpose of this
essay is therefore to explore, at least in outline, Augustine’s
experiences at the receiving end of late antique education,
in order to obtain a clearer impression of which features
might have been standard for students across the Empire,
and which peculiar to Carthage or personally to Augustine;
on the basis of this it will be possible to suggest some fresh
perspectives upon the disciplinary programme that he con-
ceived at Cassiciacum. The scope of the essay allows no
more than a sketch of a vast subject;5 to maintain as close a
focus as possible upon Augustine’s account, the approach
adopted here involves a sustained comparison between his
reminiscences and those of his older contemporary Gregory
of Nazianzus, another Christian bishop who would review
his experiences in the rhetorical schools.6
The rhetor’s classroom was anything but isolated from the
city outside it.7 And Augustine himself invites us to consider
his education in its broader social context, for although a
whole book of his Confessions is devoted to his experiences in

3
J. J. O’Donnell, Augustine: Confessions, 3 vols. (Oxford, 1992), ii. 145.
The point is made in reference to Conf. 3. 1. 1 but applies more widely.
4
H.-I. Marrou, Saint Augustin et la fin de la culture antique, 4th edn.
(Paris, 1958).
5
Our understanding of Roman rhetorical education has recently been
much enhanced by papyrological studies. T. Morgan, Literate Education in
the Hellenistic and Roman Worlds (Cambridge, 1998), helpfully presents
much material on pedagogical practice; R. Cribiore, Gymnastics of the
Mind: Greek Education in Hellenistic and Roman Egypt (Princeton, 2001),
combines the papyri with evidence from Libanius to great effect in a
discussion of the rhetor’s classroom: ‘Learning to Fly: Rhetoric and
Imitation’, 220–44.
6
The interpretation of Gregory Nazianzen presented here is more
fully developed in work currently in progress. There is a fine account of
Gregory’s Athenian experience in J. M. McGuckin, Saint Gregory of
Nazianzus: An Intellectual Biography (Crestwood, NY, 2001), 47–83.
7
P. Brown, Power and Persuasion in Late Antiquity: Towards a Christian
Empire (Madison, 1992), 43–4.
 Augustine and Gregory Nazianzen 27
Carthage, very little of it concerns his formal schooling.
We meet him successively sizzling in the frying-pan of
love (Conf. 3. 1. 1), bewitched by the illusions of the stage
(3. 2. 2–4), and doing dark business in church (3. 3. 5), before
we are introduced to him in the schoolroom, strutting arro-
gantly before his peers, the friend (at a safe distance) of the
diabolical ‘eversores’ (3. 3. 6). When he reads a textbook,
moreover, he does so alone (3. 4. 7); then he embarks upon a
self-chosen reading programme which leads him into the
hands of the Manichees (3. 5. 9–6. 10). In all this, there is
no reference whatever to his teachers, who are mentioned
only in the following book, in a context where Augustine
has left them far behind and—aged 26—is himself a teacher
of rhetoric and an aspiring author. Recalling his own literary
début, he reflects on his aptitude for learning, and remem-
bers his encounter with Aristotle’s Categories—how the
Carthaginian rhetor, his master, and other learned men
would puff their cheeks proudly at the name of the book, as
they recalled their own struggles to understand it, even with
the help of ‘most erudite masters’ and complex diagrams;
yet he had mastered it by himself (4. 16. 28).
The chief historical difficulty posed by this material is to
establish a relationship between the different activities that
Augustine describes: how the star pupil related to the eager
lover, or the theatre-goer to the young father. All such
reconstructions are necessarily provisional—as has been
brought home vividly by the impact of a single piece of new
evidence, the passage in a recently discovered sermon in
which Augustine recalls how as a young student at Carthage
he would attend vigils where ‘women were mixed up with the
licentiousness of men’.8 Behaviour that had once, on the
basis of the Confessions alone, seemed like shy gallantry in
a medieval cathedral is suddenly, almost shockingly con-
temporary: the calculated opportunism of a sexual predator.9

8
Sermo de Oboedientia 5: F. Dolbeau, ‘Nouveaux Sermons d’Augustin
pour la conversion des païens et des Donatistes (III)’, REAug 38 (1992),
50–79, at 59.
9
P. Brown, ‘A New Augustine’, New York Review of Books, 24 June
1999, modifying the presentation in Brown, Augustine of Hippo: A
Biography (London, 1967), 41.
28 Neil McLynn
If, moreover, Augustine’s reference to how young men—‘the
licentious and impudent slaves’—would gather at the church
entrances to ‘begin what they would later try to complete’
is (as the parallelism with his language in the Confessions
might suggest10) autobiographical, the plural raises a further
disquieting possibility. This mauling of matrons seems to
have been a team sport: did Augustine, one wonders, attend
church with the ‘eversores’? That such shifts of perspective
are even conceivable suggests the importance of relating
our Augustinian evidence to other sources on late antique
paideia.
Hence the relevance for an understanding of the young
Augustine of the account—indeed, the two overlapping
accounts—provided by Gregory Nazianzus of his own
encounter with secular education. Gregory arrived in
Athens, by far the most celebrated educational centre of the
late antique empire, a quarter-century before Augustine
came to Carthage. His fullest description of his adventures
there (which lasted more than a decade) is a famous section
of his memorial speech for Basil, where he explains how he
first met his great friend (Or. 43. 14–24).11 Young men at
Athens, he explains, were as mad for sophists as in other
cities they were for the circus, seizing new arrivals and
initiating them into their own group (l. 15). These initiation
rituals included a parade to the public bath; Gregory secured
Basil an exemption from this (l. 16), and came to his rescue

10
Compare Sermo de Oboed. 5, ‘[ne] inciperent quod postea perficere
molirentur’ with Conf. 3. 3. 5, ‘ausus sum . . . concupiscere et agere
negotium procurandi fructus mortis’.
11
The speech is edited by J. Bernardi, Grégoire de Nazianze: Discours
42–43, SC 384 (Paris, 1992). For the context see N. B. McLynn, ‘Gregory
Nazianzen’s Basil: The Literary Construction of a Christian Friendship’,
StPatr 37 (2001), 173–93 at 179–83. For a useful survey of the passage see
J. Bernardi, ‘Un Regard sur la vie étudiante à Athènes au milieu du IVe
siècle après Jésus-Christ’, Revue des Études Grecques 103 (1990), 79–94.
Valuable points are made in two contributions to T. Hägg and P. Rousseau
(eds.), Greek Biography and Panegyric in Late Antiquity (Berkeley and Los
Angeles, 2000): F. W. Norris, ‘Your Honor, My Reputation: St Gregory
of Nazianzus’ Funeral Oration for Basil the Great’, 140–59; D. Konstan,
‘How to Praise a Friend: St Gregory of Nazianzus’ Funeral Oration for
Basil the Great’, 160–79.
 Augustine and Gregory Nazianzen 29
when some Armenians, members of the same diatribe,
engaged the newcomer in a rhetorical duel designed to
humiliate him (l. 17). The focus then softens: a mutual con-
fession of ‘yearning’ for the philosophical life led to their
becoming ‘all in all’ to one another. They were one soul in
two bodies: rooming together, dining together, and spending
their days together—sharing a love that transcended the
transience of erotic attraction, being solidly based upon
mutual aspiration for higher things (l. 19).12 The cultivation
of ‘excellence’ that was their sole business is defined in
religious (but not exclusively Christian) terms as the prepara-
tion for escaping from the world: which meant in practice
preferring peaceable to quarrelsome companions, and the
finest courses of study to the more pleasant (§20). Thus far,
teachers (of whatever sort) are absent from Gregory’s
account: in introducing these efforts to live out a godly
life, it does not occur to him to invoke any institutional or
even informal guidance. The only compasses available to
himself and Basil were ‘the commandment’ (that is, the
prescriptions of Scripture)13 and the standard that each set
for the other, as measuring-rod and rule. Gregory thus
presents us with an experiment in Christian life that is
essentially home-made. This offers a point of contact not
only with Augustine’s account of his own experiments at
Carthage but with the formative religious experiences of
other young men.14
Like Augustine in Carthage, Gregory attended church.
Two roads alone were known to Basil and himself, the nobler
of them leading to ‘the sacred house of ours and the teachers

12
For the erotic charge that runs through Gregory’s account, see
J. Børtnes, ‘Eros Transformed: Same-Sex Love and Divine Desire’, in
Hägg and Rousseau (eds.), Greek Biography and Panegyric, 180–93.
13
The same expression recurs in the speech (Or. 43. 50, 70) and else-
where in Gregory’s writings with a large variety of biblical connotations.
The generalized appeal to scriptural authority might be compared to
Augustine’s requirement as a student that his preferred doctrine include
the name of Christ (Conf. 3. 4. 8).
14
Our richest source on the diverse currents of late antique student
religiosity comes from a later period: Zacharias Scholasticus, Vita Severi,
ed. and trans. M. A. Kugener, Patrologia Orientalis 2 (Paris, 1907)).
30 Neil McLynn
there’ and the other to ‘the educators outside’ (Or. 43. 21).15
Augustine too would instinctively juxtapose church and
schoolroom (Conf. 3. 3. 5–6). Like Augustine, too, Gregory
recognizes other possible destinations (including the theatre)
but shows himself and Basil dutifully shunning all roads
which did not lead to excellence. Aretē, however, was enacted
in public. Gregory remarks upon the different labels that are
derived from a man’s paternity or origin, or his preferred
avocations: he and Basil took pride in being—and in being
called—‘Christians’. The wording makes it clear that the
label was pinned upon them not by their fellow Christians
in the congregation of Athens but by their (largely pagan)
fellow students. They revelled in being Christians in a
dangerously unchristian world. And this self-conscious
flaunting of a label might remind us of the Carthaginian
‘eversores’.
Gregory continues to think in terms of gangs in the fol-
lowing passage, where he declares that ‘the finest thing’
about their experience at Athens was the ‘brotherhood’ of
young men who gathered around Basil (Or. 43. 22). Basil
and Gregory became famous not only among these fellow
students but also among their teachers. As in Augustine’s
account, teachers are introduced here almost as an after-
thought, and serve a purely ancillary function. They provide
an index to Gregory’s and Basil’s fame: these teachers were
as widely known as Athens itself, and the two students
were as widely known as their teachers, as ‘a pair not without
a name’.
This latter phrase recurs also in Gregory’s other account
of his schooldays, in his autobiographical poem De Vita Sua
(ll. 211–64). Here too, the emphasis is upon gangs. Whereas
the rest of the young bloods ‘whirled around in different
brotherhoods’ (ll. 214–15), Gregory resisted their pull and
instead attracted others, drawing his friends to ‘higher
things’ (ll. 219–20). Here again, friendship is presented as
being central: he shared with Basil ‘literary studies, lodgings,

15
McGuckin, Saint Gregory of Nazianzus, 62, interprets this as a
reference to Gregory’s teacher Prohairesius and his ‘place in the official
life of the Athenian church’: but the expression seems to distinguish
between two separate bodies of teachers, sacred and secular.
 Augustine and Gregory Nazianzen 31
thoughts’ (ll. 226–7), but sees no need to mention his teachers
until the moment of his planned departure (l. 250). The core
of his Athenian experience, as reported in the poem, is the
comradeship that provided the space for the cultivation of
religious instincts.16

2. g e n i u s lo c i
Gregory’s Athens, unlike Augustine’s Carthage, is illumi-
nated for us from several distinct angles. An epigram
reveals that his teacher was the famous Prohairesius.17
And Eunapius, another pupil of Prohairesius, has left an
admiring portrait of the master; while the surviving orations
of Prohairesius’ great rival Himerius (who was also claimed
by a fifth-century source as Gregory’s teacher) provide a
complementary perspective.18 So, for example, Eunapius’
description of the small, spartan house that Prohairesius had
inherited from his teacher Iulianus, where statues of past
pupils and a private auditorium of polished marble fostered
a sense of almost religious intimacy and intensity, provides
essential background for what otherwise seems the over-
blown claustrophobia of Himerius’ orations, which lavish
the same sort of attentions upon newly arrived students as

16
McGuckin’s ingenious interpretation (Saint Gregory of Nazianzus,
63) of Gregory’s command at De Vita Sua 212–13 that ‘others should
speak of matters there’ as a veiled allusion to baptism at Athens (and
hence implying a deeper involvement in the local church than is argued
here) underplays the grammatical connection with the activities described
in the following lines.
17
Gregory of Nazianzus, Epit. 5.
18
Eunapius’ account of Prohairesius is well discussed by R. J. Penella,
Greek Philosophers and Sophists in the Fourth Century A.D. (Leeds, 1990),
79–84. Himerius’ career has been reconstructed by T. D. Barnes,
‘Himerius and the Fourth Century’, Classical Philology 82 (1987), 206–25;
we still await a fuller appreciation of his oratory. The report that Gregory
and Basil were taught by Himerius as well as Prohairesius (Socrates,
Hist. Eccl. 4. 26. 6) must be assessed in the light of the historian’s
accompanying claim that they both subsequently also attended Libanius’
school at Antioch. See also R. Goulet, ‘Prohérésius le païen et quelques
remarques sur la chronologie d’Eunape de Sardes’, Antiquité Tardive 8
(2000), 209–22.
32 Neil McLynn
upon visiting proconsuls.19 When Himerius speaks as a
shepherd to his flock, reminding his pupils of their responsi-
bilities as ‘initiates’ of the Muses, he provides a glimpse of
the relevance of traditional education for future bishops.20
Similarly, as a teacher, Augustine would cure Alypius of his
obsession with the circus by a single well-chosen simile in
class (Conf. 6. 7. 12). The vehemence with which he insists,
in retrospect, that his admonition had not been a sermon
suggests how unclear, in practice, the distinction may have
been. The degree of pastoral involvement suggested by some
of the surviving texts is striking; Himerius celebrates the
birthdays of his students and hymns their weddings.21 One
wonders whether Athenian teachers were party to arrange-
ments among their students equivalent to Augustine’s semi-
formal relationship with his nameless concubine, or the birth
of his son.
Eunapius and Himerius both convey the ties between
masters and pupils, where Gregory concentrates on relations
between pupils; but these are both part of the same nexus.
For a striking feature of Eunapius’ portrait is that he never
once shows Prohairesius teaching; likewise, Himerius’
orations do not so much teach rhetoric as exhibit it. And
perhaps the rhetor’s principal job was to facilitate the sort of
bond that united Gregory and Basil, and to create the sort of
atmosphere where (as in an Islamic madrasah22) young men
felt encouraged to educate one another. In becoming a
Pylades to Basil’s Orestes (Or. 43. 22), Gregory was follow-
ing an established pattern. His teacher Prohairesius had

19
Eunapius, Vitae Sophistarum (hereafter VS) 483, ed. W. C. Wright,
Philostratus and Eunapius: Lives of the Sophists (Cambridge, Mass., 1921),
466; Himerius, Or. 14 (Egyptian students), Or. 17 (Cypriots), Or. 18
(Cappadocians), Or. 26 (Ephesians and Mysians), Or. 27 (a student
from his own city, Prusias), Or. 59–60 (Ionians), Or. 69. 8 (a Bithynian, a
Galatian, and some Egyptians).
20
Himerius, Or. 69. 7–9, for the trope of initiation; Or. 54. 2, for the
student body as a ‘flock’.
21
Himerius, Or. 44 (birthday); Or. 9 (an epithalamium for a recent
student, Severus).
22
On peer learning in the madrasah, see D. F. Eickelman, ‘The Art of
Memory: Islamic Education and its Social Reproduction’, Comparative
Studies in Society and History 20 (1978), 485–516. I am grateful to Peter
Brown for the reference.
 Augustine and Gregory Nazianzen 33
pooled his possessions with his friend Hephaestion, so that
like an attenuated Geryon they seemed ‘to be two men in
one’; similarly, during his time at Athens Libanius had
shared with Chromatius a single table, under the same roof,
with the same pleasures and thoughts, and they had sharp-
ened each other by acting as judges.23 Two features of these
relationships require special emphasis. The first is that they
were enacted before an audience. Gregory’s claim that he
and Basil were ‘notable to our teachers and fellows, notable
to all Hellas’ (Or. 43. 22) thus had a specific connotation.
More particularly, his account of his intervention in the dis-
putation between the malicious Armenians and the newly
arrived Basil—participating initially to save the house’s
‘honour’, but ending up defending Basil’s—presupposes an
audience of Prohairesius’ other students (43. 17).
The same incident also introduces our second factor: these
houses were competitive crucibles. Friendship between
aspiring rhetoricians was necessarily agonistic, since their
education would have required them to engage in regular
duels. And in an important passage Gregory acknowledges
that he and Basil were competitors: a world of tensions
is implicit in his claim that they overcame the pressure of
phthonos, and each took delight in the other’s victory rather
than his own (Or. 43. 20).24 More typical, one suspects, was
the situation described by Eunapius: the young Prohairesius
‘competed with’ his friend Hephaestion both in poverty and
in logoi, but the latter eventually retired from the field,
to clear the path for Prohairesius’ advancement; similarly,
Chromatius faded from being Libanius’ critic to being his
cheerleader.25

23
Eunapius, VS 487 (ed. Wright, 484–6); Libanius, Ep. 390. 5.
24
Competitive instincts clearly died hard: note how in Or. 43. 22
Gregory begins by emphasizing Basil’s leadership of their group,
including himself among those who ran on foot behind his ‘Lydian car’,
then in the next sentence identifies the ‘we’ who became famous among
teachers and fellow students not as the group as a whole but as Basil and
himself alone, until by the end of the paragraph he has climbed aboard
Basil’s chariot, as his identical twin and partner in suffering.
25
Eunapius, VS 487 (ed. Wright, 488); Libanius, Ep. 390. 6–7.
34 Neil McLynn
Each rhetorical house in Athens thus offered in micro-
cosm the competitively collaborative world of élite civic
politics. What held these fragile communities together, in
good classical manner, was shared hostility to other similar
houses. That is to say, the student brawls for which Athens
was notorious were an integral part of the training.26 For the
élite being moulded at Athens needed to develop the physical
self-assertion that was central to public life and a pre-
requisite for the effective use of fine words: one of the ways
they learned to measure their worth was in confrontation
with students of rival houses. Learning to be a man thus
meant learning to stare down rivals in the street even when
blows were threatened or given; the rhetor’s job was to
impart the eloquence that would match such poise. Even
those who shunned the fights and the feasts had to take sides.
The professors were fully implicated. Libanius sourly noted
that the Athenian schools turned out soldiers, not speakers;
and Himerius is a witness for the prosecution, for he makes
Homeric heroes of his battered street-fighters, rousing them
with a hymn to victory.27 Iulianus’ house, where Prohairesius
gave his lectures, had the feel of a fortress. We see here the
consequences of an arrangement that required professors to
remain rivals: they therefore needed to keep their students in
a state of permanent mobilization.28
Even those who shunned the fights had to take sides. The
initiation parade to the baths which Gregory describes, for
example, should be understood as a means to advertise
publicly each new recruit and to proclaim his allegiance.
When Gregory and Basil publicly paraded to church (pre-
sumably wearing the distinctive red cloaks of students29)

26
Brawls: Eunapius, VS 483–5 (ed. Wright, 468–76); Libanius
Or. 1. 19.
27
Libanius, Ep. 715; Himerius, Or. 65.
28
Gregory Nazianzen would sarcastically evoke the ‘war’ fought by
a sophist of Caesarea against his rival, ‘and this in the midst of your
partisans, who cheer you on’ (Ep. 192. 3).
29
Red cloaks are not specifically mentioned before the fifth century
(Olympiodorus, Fr. 28: R. C. Blockley, The Fragmentary Classicising
Historians of the Later Roman Empire (Leeds, 1983), ii. 192); but
Gregory’s reference to students having ‘taken the gown’ (Or. 43. 17)
implies that the uniform was already established.
 Augustine and Gregory Nazianzen 35
they were creating much the same effect—and consciously
so. It is significant that Gregory recalls the road leading them
to church, but not the services themselves. There was an
intrinsically muscular quality to late antique paideia that we
neglect at our peril.
All this raises questions about Augustine’s experience at
Carthage. For if the structure of education was the same, the
conditions seem very different. His complaint about his own
teaching situation there concerns essentially the impossi-
bility of replicating the same mystique as was possible at
Athens. ‘Scholastici’ lurch into other teachers’ classes
‘proterve’, with furious faces (Conf. 5. 8. 14): in Augustine’s
experience, at least, the boundaries that at Athens made each
school a quasi-religious community were not maintained.
Part of the fault was perhaps Augustine’s own: in a famous
anecdote, already mentioned in another context, he shows
himself paying special attention to a late-comer who was not
even on his roster of pupils, but who happened to be an
acquaintance from his home town (6. 7. 11–12). The schools
of Carthage, one suspects, were slightly too close to the stu-
dents’ homes. When Alypius was in trouble, an acquaintance
of a senatorial family friend, not his teacher, would vouch
for him (6. 9. 15); for the latter part of his student career,
Augustine had his mother living with him.30 The sheer size
of the city also diminished the profile of the student popula-
tion.31 When a student declaimed a piece by Libanius in the
Athenian Lyceum, he was set upon by exponents of rival
texts;32 at Carthage, Alypius was able to prepare his exercises
in the forum, blithely oblivious to the presence of another
student, who in turn was too absorbed in petty larceny to
notice Alypius (6. 9. 14).
We are dealing here with what might be called micro-
climates of paideia. The difference between Athens and

30
Eunapius, VS 483 (ed. Wright, 470) shows a sophist in a courtroom
alleging concern for his ‘children’; Prohairesius, more tellingly, ordered
that Eunapius be treated ‘as if he were my own son’: VS 486, cf. 493 (ed.
Wright, 482, 512). P. Petit, Les Étudiants de Libanius (Paris, 1955), 138–44,
shows Libanius standing in loco parentis to young students from distant
provinces.
31
As Gregory remarked of Alexandria: Or. 7. 6.
32
Libanius, Or. 14. 35.
36 Neil McLynn
Carthage was not merely one of size. Whereas Alypius had
to share the forum of Carthage with lawyers and goldsmiths,
the students of Athens had the ancient Agora and Lyceum to
themselves, as their own reserved playground.33 They could
therefore re-create the classical works they were studying
in situ: the procession to the baths (for example) may well
have led down the Panathenaic Way, past the Painted Stoa
which illustrated so many of the themes of their own com-
positions.34 No wonder that the students of Athens thought
of themselves as gods, considering those of Alexandria (as
Synesius complained) mere ‘donkeys’ by comparison.35 And
as Alexandria was to Athens, so perhaps was Augustine’s
Carthage to Rome. In the ancient capital the Forum housed
a still growing statue gallery of famous orators, and students
gathered not to rehearse their own arguments but to applaud
the current masters of the art.36 This makes one wonder to
what extent the shadow of Rome, just out of reach, loomed

33
A. Frantz, The Athenian Agora 24: Late Antiquity A.D. 267–700
(Princeton, 1988), 12–48, discusses the physical fabric of the fourth-
century Agora, noting the failure to restore administrative buildings
destroyed in the Herulian sack. Construction in the second half of the
fourth century of sophists’ homes (and therefore schools) on the Areopagus
(ibid. 44–8) would suggest consolidation of academic dominance of the
centro storico; but see G. Fowden, Journal of Roman Archaeology 3 (1990),
495–6, for reservations about this identification.
34
Frantz, Athenian Agora, 26–8, dates the reconstruction of the Dipy-
lon Gate and Panathenaic Way to ‘the reign of Constantine and a little
later’, just before Gregory’s arrival; the most natural route ‘through the
agora’ (Or. 43. 20) would have followed this route, terminating at the
Southwest Bath (lavishly refurbished in the mid-fourth century, at exactly
this period: Frantz, Athenian Agora, 32–3). In later life Gregory refers
instinctively (Ep. 233, 235) to Callimachus and Cynagirus, whose exploits
were portrayed in the Painted Stoa; Himerius, Or. 2, describes the
painting. For a fictive debate between the two heroes’ claims, see Polemon,
Declam. in Callimachum, Declam. in Cynagirum. The fullest evocation of
the inspiring effect of the Athenian monuments upon students is Cicero,
De Finibus 5. 1–5.
35
Synesius, Ep. 54; at Ep. 136 he insists that Athens was honoured only
by its bee-keepers.
36
Jerome, Comm. ad Galat. 2. 11. Augustine, Conf. 8. 2. 3 reports
the statue of Marius Victorinus (cf. Jerome, Chron. s.a. 353); Eunapius,
VS 492 (ed. Wright, 506–8), that of Prohairesius, who visited c.343; cf.
Libanius, Ep. 278.
 Augustine and Gregory Nazianzen 37
over the schools of Carthage; whether the latter were self-
consciously second-rank. Augustine’s delusive dreams of
better teaching conditions at Rome were fuelled, he says,
by his friends (Conf. 5. 8. 14): a suggestive comment on the
topics of conversation among ambitious young men in the
African metropolis. It is perhaps significant, in this context,
that his cue for mentioning his own teachers is his sending a
copy of his own first book to Rome (4. 13. 21)—to the orator
Hierius, the teacher that he had never had.

3. r e l i g i o n s o f t h e b o o k ?
Mastering a syllabus seems to have been incidental to the
student experiences of Gregory and Augustine, both of
whom describe instead an open-ended quest for wisdom.
Nor is there any reason to think that they were unusual,
since Himerius and Eunapius also suggest the primacy to
education of what might be called the spiritual dimension.
Where Gregory emphasizes friendship, Augustine identifies
books as his source of inspiration: an encounter with a
school text, Cicero’s Hortensius, led to an extracurricular
reading programme—first the Bible, then the books of the
Manichees.
Here too, however, Gregory might yet provide a point of
comparison. In his account of his studies with Basil he
does not, indeed, mention any shared reading. Yet one other-
wise mysterious text might be associated with Gregory’s
Athenian period—and if so, would provide an important
parallel with Augustine’s experience. Gregory first mentions
the Philocalia, an anthology of extracts from various works
of Origen, a quarter-century after his return from Athens,
when he presented a copy to a friend as ‘a souvenir from me,
which is also of the holy Basil’ (Ep. 115). Recent work has
shown just how fragile are the foundations of the traditional
view, that the collection was actually compiled by Gregory
and Basil.37 Gregory himself introduces the book to its

37
M. Harl, Origène: Philocalie, 1–20, Sur les Écritures, SC 302 (Paris,
1983), 19–41; É. Junod, ‘Basile de Césarée et Grégoire de Nazianze
sont-ils les compilateurs de la Philocalie d’Origène? Réexamen de la lettre
115 de Grégoire’, in Mémorial Jean Gribomont (Rome, 1988), 349–60.
38 Neil McLynn
recipient as ‘the Philocalia of Origen’, implying that he at
least believed that Origen himself had edited the work. So,
if he can offer the volume, without further explanation, as
a memento not only of himself but also of Basil, this is
probably because it bore physical traces—subscriptions or
scholia—of their use of it.38 And once we see the two friends
as users of the text (with Gregory, presumably, as its owner)
rather than as its editors, far the most likely context for such
joint use is during the five years they spent together in
Athens.39 Their subsequent periods of collaboration were
much shorter, and were devoted to projects for which the
Philocalia would have offered little help; besides, Origen’s
works were readily available to Basil’s circle in Cappadocia
and Pontus, so Gregory had little reason to bring the com-
pendium when he visited his friend.40 On the other hand,
Gregory had arrived in Athens via the two cities between
which Origen had divided his career.41 It is attractive to
suppose that he acquired the volume in one of these, as a
portable digest of the great master’s teachings and a useful
aid to the reflections of the serious student.42
The two great themes of the work—biblical hermeneutics
and the problem of free will—would be particularly relevant
to aspiring Christian students; for these of course relate
closely to the issues which would most disturb Augustine as

38
Cf. Palladius, Historia Lausiaca 64, reporting autograph comments
by Origen; annotations in Basil’s hand were allegedly recognizable in
a codex preserved in the library of Caesarea: George Syncellus, Chron.,
ed. B. G. Niebuhr (Bonn, 1829), 382.
39
J. N. Steenson, ‘The Date of the Philocalia’, in R. P. C. Hanson and
H. Crouzel (eds.), Origeniana Tertia (Rome, 1988), 245–52, also argues for
an Athenian context, but still supposes that the pair compiled the work
rather than studied it.
40
Junod, ‘Basile de Césarée et Grégoire de Nazianze’, 360, offers no
arguments for his assertion that Basil and Gregory used the book ‘autour
des années 360’.
41
Before sailing to Athens Gregory studied rhetoric at Palestinian
Caesarea (Or. 7. 6), then briefly at Alexandria (De Vita Sua 128–9).
42
McGuckin, Saint Gregory of Nazianzus, 60–2, emphasizes the
similarities between Prohairesius and Origen, and Steenson, ‘The Date’,
248, also suggests him as a possible Origenist influence: but the fact
remains that Gregory himself provides no hint of this.
 Augustine and Gregory Nazianzen 39
43
a student in Carthage. And the central chapters, providing
material for replies to pagan criticism, would meanwhile
have been especially useful in mid-fourth-century Athens,
a place still dominated by its pagan monuments, ‘richer
in wicked wealth than the rest of Greece’, where young
men were easily ‘swept along by its panegyrists and patrons’
(Or. 43. 21). The Philocalia was less a textbook than a
series of encouraging intimations of the possibility of an
authentically Christian paideia. So one extract (Phil. 12)
exhorts the reader ‘not to faint’ as he tackles the bleaker parts
of Joshua; the next (Phil. 13) consists of Origen’s letter to a
student named Gregory (whom Basil, one suspects, would
never have recognized as his own grandmother’s teacher)
which provides an explicit manifesto for the Christian
student, who should ‘despoil the Egyptians’ by acquiring
pagan learning as a means to Christian ends, and should
meanwhile apply himself to ‘the reading of the divine scrip-
tures’ with the serious and prayerful application that would
allow, finally, ‘participation with God’. The next extract
(Phil. 14) illustrates the techniques necessary by analysing
the names and predicates in a passage of Genesis; then
comes a ‘reply to the Greek philosophers who disparage
the poverty of style of the Holy Scriptures’ (Phil. 15).44 The
organization of the compendium helps give a sense of a do-
it-yourself religious formation, where intensive study of the
Sacred Scriptures was assisted by a restricted number of
well-thumbed secondary texts.
Here again there are echoes of Augustine’s religious
development at Carthage. He too depended upon the piece-
meal collation of assorted primary and secondary sources: in
his case, however, there was no Christian source to teach him
(as the Philocalia taught its readers) to approach the Bible as
a single musical instrument, perfect and harmonized (Phil.
6). Instead, we might see the Manichees and their books
representing for him what Gregory and the Philocalia did

43
Augustine sets out his problems at Conf. 3. 7. 12. The rebarbative
complexities of the biblical text are dealt with at Phil. 1–15; the problem
of evil, Augustine’s particular difficulty, is treated at Phil. 26 as part of the
broader subject of free will.
44
Cf. Augustine, Conf. 3. 5. 9.
40 Neil McLynn
for Basil. And what mattered, ultimately, was the balance
between the two parallel reading programmes. Augustine
thus presents his decisive break with the Manichees as a
diptych. Faustus, the most authoritative expositor of the
Manichaean texts, could not answer Augustine’s questions
about them; at the same time he deferred to Augustine’s
literary expertise, reading with him such classical texts as
he himself desired or Augustine deemed suitable (Conf. 5. 7.
12–13).

4. c a r g o e s o f c u lt u r e
Departures from Athens were attended with as much
ceremony as were arrivals. A fragmentary speech shows
Himerius sending off a pupil with ornate instructions
(reinforced by historical and mythological exempla) about
the future ‘nurture of his mind’, and the maintenance of
the proper interplay between character and eloquence.45
And Gregory describes vividly the day when he and Basil
prepared to depart, ‘the farewell speeches, the processions,
the calls to remain, the groans, the embraces, the tears’ (Or.
43. 24); it was ‘a time for embraces and sorrowful words’, as
he put it elsewhere, ‘for farewell speeches and kindling of
memory’ (De Vita Sua 242–4).
By contrast, Augustine left school without any apparent
fanfare. Indeed, we never see him leave at all. At the end
of Book 3 of the Confessions Augustine is sharing a house-
hold, somewhat uneasily, with his widowed mother (Conf.
3. 11. 19): although the episode probably still belongs to
the time when he was enrolled as a student at Carthage,
biographers have readily transferred it to Thagaste.46 The
progress from student to teacher is invisible. Augustine’s
own narrative seems designed to blur the transition in fact,
combining the bulk of his student career with his early years
as a teacher into the ‘nine years’, from age 19 to 28, that he
spent in thrall to the Manichees.47 This structure serves

45
Himerius, Or. 10.
46
O’Donnell, Augustine: Confessions, ii. 198–9.
47
Conf. 3. 11. 20; 4. 1. 1; 5. 6. 10. Cf. O’Donnell, Augustine: Confessions,
ii. 297–8.
 Augustine and Gregory Nazianzen 41
Augustine’s artistic and apologetic purposes, but also reflects
an aspect of paideia that deserves more attention than it
is conventionally given. For it says something significant
about late antique education that we never see Augustine
‘graduate’.
The overlap in Augustine’s account between studying and
teaching is not merely incidental. Book 4 of the Confessions
covers his early experiences as a teacher: the incidents of the
book are set during ‘those years’ when he ‘was teaching the
art of rhetoric’ (Conf. 4. 4. 2); a section in the middle is
located more specifically ‘at the time when I had first begun
to teach in the town where I was born’ (4. 4. 7). Yet it is in
this book that we have the most vivid glimpses of Augustine
behaving as we might expect a student to do, competing for a
crown in a literary competition (4. 2. 3–3. 5) and enjoying, in
the company of friends, ‘talking and laughing together and
kindly giving way to each other’s wishes, reading elegantly
written books together, sharing jokes and being serious at the
same time, disagreeing occasionally but without rancour’,
and, like Gregory and Basil at Athens, ‘teaching one another,
and learning from one another’ (4. 8. 13). Above all, during
his interlude at Thagaste he found ‘a friend dear to me
through the pursuits we shared’, a former schoolfriend with
whom he resumed the literary pursuits of childhood (4. 4. 7).
Their bond was founded upon this ‘heated enthusiasm of
shared pursuits’, and Augustine lavishes upon it the clichés
of student romance that Gregory had used for himself and
Basil: they were Pylades and Orestes, the friend was ‘half my
soul’ (4. 6. 11). This was a friendship between ‘adulescentes’,
which began when Augustine was 21: about the same age,
that is, as Gregory was when Basil came to Athens.48
The warmth with which Augustine describes this friend-
ship at Thagaste throws into sharp relief the frigidity of his

48
Augustine’s friend was ‘coaevum nobis et conflorentem flore
adulescentiae’ (Conf. 4. 4. 7). Gregory’s birth is most plausibly dated to
c.329/330, Basil’s arrival in Athens to 349/350: see respectively P. Gallay,
La Vie de saint Grégoire de Nazianze (Lyon, 1944), 25–7, and P. J. Fedwick,
‘A Chronology of the Life and Works of Basil of Caesarea’, in P. J.
Fedwick (ed.), Basil of Caesarea: Christian, Humanist, Ascetic (Toronto,
1981), i. 3–19, at 6.
42 Neil McLynn
earlier account of Carthage, where no distinct friends
emerge from the lurid shadows, and even eroticism remains
bleakly anonymous. But our readiness to draw biographical
inferences from this—that Augustine’s initial isolation at
Carthage was real, that he was an awkward provincial, out of
his depth in the big city and therefore easy prey for the
well-organized Manichees49—may well be misplaced. There
are artistic reasons for the emphasis upon this friendship at
Thagaste, which both gives significance to an otherwise
potentially problematic interlude and provides a suitably
heart-driven reason for the subsequent return to Carthage.
And the evocation of cultivated friendship that follows, as
quoted earlier, refers not to a change in circumstances but to
a resumption of existing ties. The Thagaste episode thus
punctuates a period that Augustine treats as a continuum. It
does not imply that Augustine the student had failed to form
the usual bonds. Some might indeed have endured: although
the wealthy Carthaginian Nebridius is first introduced as a
‘truly good and truly chaste adulescens’ (Conf. 4. 3. 6), after
Augustine’s return to the city, there is a reasonable chance
that the relationship had first been forged in the classroom.50
There are similar tricks of perspective in Gregory’s account
of his own youthful friendships. His three or four years at
Athens before Basil’s arrival are greatly foreshortened; and
only incidentally do we discover that he had made previous
experiments in philia.51
Nor again is Augustine exceptional in continuing the
habits of the schoolroom in later life.52 Here too there might

49
Stock, Augustine the Reader, 43–4.
50
The only clue here is that when introducing Alypius and Nebridius at
Conf. 6. 7. 11 Augustine introduces the former as his one-time student,
‘me minor natu’, an odd choice of expression if the same applied to
Nebridius. Assumptions about the relationship have been easily made:
when discussing Augustine’s correspondence with Nebridius, Stock
(Augustine the Reader, 128) makes the latter successively a ‘friend’,
‘protégé’, and ‘student’.
51
Gregory’s failure to claim direct knowledge of Athanasius in
Or. 21 puts his move from Alexandria to Athens before the bishop’s return
from exile in 346. At Alexandria he shared with Philagrius a table and ‘the
delights of lovely comradeship . . . sweated labour over literature, teachers
in common’: Ep. 30.
52
See in general Cribiore, Gymnastics of the Mind, 238–44.
 Augustine and Gregory Nazianzen 43
be a parallel in Gregory and Basil’s (ultimately abortive)
plans to live together in Cappadocia as they had in Athens.
Theirs was to be an ascetic, ‘philosophical’ cell rather than
a Manichaean one, but would otherwise have looked little
different from Augustine’s association with his friend in
Thagaste. Nor would either pair have seemed eccentric.
After Eunapius was suddenly recalled to Sardis by his
parents from his five-year spell in Athens in order to take up,
aged 18, a teaching post, he would spend the afternoons
receiving instruction in higher matters from the philosopher
Chrysanthius, his former teacher.53 The readiness with
which habits of study were thus continued reflects, in
part, the unsystematic diversity—and consequent open-
endedness—of the educational process itself. The schools
could hardly look to a determinate end-product when their
students might stay for anything between a year and a
decade: a sample of some fifty-seven pupils of Libanius at
Antioch shows seventeen staying only a year and eighteen for
two, while nine remained with him for three years, three for
four, five for five, and four for six or more.54 The vaunted
uniformity of paideia was therefore inevitably a mirage, the
more fragile because differences in cultural level were liable
to be exposed in the ruthless zero-sum contests of public
life.55 The leisure habits acquired in the schools—the shared
application to texts and to literary conversation, as well as
more boisterous avocations—were, in this context, of genu-
ine social significance, as a source of much-needed coher-
ence and an opportunity for constructive collaboration.
Augustine and his Thagaste friend (who presumably had
lacked the benefit of polish in the Carthaginian schools)
could consort as equals over their books; Gregory did not
play the Athenian demigod with his Alexandria-trained

53
Eunapius, VS 493, 502–3 (ed. Wright, 512, 552): see the useful
discussion by T. M. Banchich, ‘Eunapius in Athens’, Phoenix 50 (1996),
304–11. For Chrysanthius, see Penella, Greek Philosophers, 75–8.
54
Petit, Les Étudiants de Libanius, 63–4.
55
Brown, Power and Persuasion, 39–40, offers a memorable sketch of
the role of rhetorical culture in ‘the patient recreation, generation after
generation, of the “collective memory” of the urban upper class’. It
remains to delineate more fully the uses of this memory.
44 Neil McLynn
friend Philagrius when expounding a text to him.56 Similarly,
where Eunapius presents himself merely as Chrysanthius’
pupil, observers will doubtless have recognized a partner-
ship.
But what was the relationship between the unsystematic
activity conducted within the schools and the formal cate-
gories represented by the disciplines? Once again, Gregory
suggests an approach to such questions. In his eulogy on
Basil he uses disciplinary terminology to sum up his friend’s
attainments. What form of paideusis, he asks (Or. 43. 23),
did Basil not traverse? It quickly emerges that he had in
fact mastered eight separate disciplines; but in listing these
Gregory is invoking a cultural ideal, rather than describing
a syllabus. He begins by declaring Basil’s supremacy in
rhetoric, grammar, and philosophy: the ordering of these
subjects, which allows a sequence of progressively longer
and more complex questions, most certainly does not reflect
the course of Basil’s actual studies.57 Then come three
matching sciences: astronomy, geometry, and the ‘relations
of numbers’. Basil apparently ‘took’ as much from these as
was necessary to fortify himself against the cleverness of
experts, but Gregory’s prime concern is for symmetry and a
neatly turned epigram: there is no need to suppose that Basil
was taught these subjects specifically at Athens. The next
item, moreover, puts the matter beyond reasonable doubt.
Medical proficiency was thrust upon Basil by his recurrent ill
health, ‘the fruit of philosophy and of zealous application’.
‘Starting from this’—his own self-inflicted needs—Basil
achieved ‘possession of the art’, in a properly philosophical
manner, which is to say that the skill was self-taught.
Gregory introduces medicine, that is, in order to enhance his
panegyric, and provide an appropriate flourish with which to

56
Greg. Naz., Ep. 34, mentioning that he had played the exegete only at
Philagrius’ request, and insisting to his friend that, in another respect,
‘you had your teacher as a student’.
57
The inversion of the logical—and chronological—order of grammar
and rhetoric (for which see, e.g., Augustine, Ord. 2. 12. 35–13. 38) reflects
the moral problems attached to the latter: Gregory insists that Basil’s
character was ‘not like the rhetors’ (Or. 43. 23).
 Augustine and Gregory Nazianzen 45
58
demonstrate Basil’s omnicompetence. And then he caps it
with yet another ‘discipline’ as his climax: all Basil’s other
attainments counted for little beside his paideusis in ēthos,
which made nonsense of the mythological examples of recti-
tude. This is not a reference to advanced studies in moral
philosophy, but a conflation (the more striking for its
smoothness) of specific academic disciplines with their
ethical context. Paideusis was a metaphor: Basil at Athens
was a ship taking on its load, and when he had as full a cargo
as could be crammed into his hold, it was necessary for him
to sail for home (§24). No conceptual scheme was available to
allow more precise quantification of academic attainments.
Much of this terminology, moreover, is recycled: for when
burying his brother, Caesarius, over a decade previously,
Gregory had paid similar tribute to his academic prowess
(Or. 7. 6–8). Here the case is different, for Caesarius had
studied at Alexandria and gone on to win renown as a
doctor—and this was a period when (as an anecdote of
Augustine’s confirms) a doctor’s professional authority
was assured if he could claim to have been trained at
Alexandria.59 But although medicine is again the capstone of
the disciplines, its pre-eminence is surprisingly modest—it
receives less prominence, indeed, in Caesarius’ case than in
Basil’s. Like Basil, Caesarius is awarded two sets of three
paideuseis, this time all scientific—having established his
prowess in geometry, astronomy, and ‘the discipline that
is dangerous to others’ (astrology), Gregory dares anyone to
challenge his primacy in ‘numbers, calculations and divine
medicine’. If medicine is then characterized more fully
than the parts of arithmetic (the division here is presumably
for the sake of symmetry), it still receives less space than
does the conceit that spells out Caesarius’ supremacy. This
itemization of specific excellences, moreover, rounds off an

58
Bernardi, Grégoire de Nazianze: Discours 42–43, 176 n. 2, tries to
solve the problem by supposing that Basil was taught the subject at
Constantinople.
59
For Alexandrian doctors, see Ammianus Marcellinus, Res Gestae 22.
16. 18, with V. Nutton, ‘Ammianus and Alexandria’, Clio Medica 7 (1972),
165–72; Penella, Greek Philosophers, 109–17. Augustine, Civ. 22. 8, shows
the authority of Alexandrian medicine in practice.
46 Neil McLynn
account that had defined Caesarius’ scholarly prowess in
terms of his social relations—his fidelity to teachers, his
affection for his classmates, his wise choice of companions
from among his fellow citizens (especially from among his
fellow Nazianzenes),60 and the recognition he received
from the civic leaders. Like Basil, Caesarius was a great
merchantman, whose schooling consisted of the accumula-
tion of a varied cargo. Nothing is said here to suggest that
the manifest differed from Basil’s. If Gregory is silent about
the literary disciplines, he does not intend to suggest that
Caesarius had not studied these, but is rather presenting
(somewhat awkwardly) his brother’s qualifications as com-
plementary to his own.61 This rhetorical device reflects a real
enough situation, for the two brothers had in fact returned
home to Nazianzus after finishing their studies together.
They were a ‘partnership’: one might envisage them (and
similarly, albeit in different circumstances, Augustine and
his friend at Thagaste) each acting as guarantor for the
other’s attainments. There was, that is, a public aspect to
privately shared absorption in ‘studia’. At Sardis, for
example, Eunapius and Chrysanthius impressed their part-
nership in philosophy upon their fellow citizens by taking
walks, absorbed in conversation, along ‘the public streets’ of
the city.62
Even when Gregory announced a decisive break with
secular pursuits, moreover, the habits of student associations
endured. Nowhere is this more apparent than in his cautious
negotiations with Basil over their project of a joint ascetic
initiative—what has often been criticized as indecision or
evasiveness on Gregory’s part is better understood as the
continued operation, but in a less controlled environment, of

60
The same term πατρ is used twice immediately afterwards, at Or. 7. 9,
specifically in reference to Nazianzus. The estate of Gregory’s friend
Philagrius, who would have studied with Caesarius also (Anthologia
Graeca 8. 100), may well have been in the city territory of Nazianzus.
61
Gregory mentions his own parallel pursuit of his ‘passion for
rhetoric’ at Palestinian Caesarea at Or. 7. 6; he makes no reference here to
his own stint in Alexandria. At Anth. Gr. 8. 91 he credits Caesarius with
mastery of six disciplines, including grammar and (as the summit) ‘the
might of rhetoric’.
62
Eunapius, VS 502 (ed. Wright, 550).
 Augustine and Gregory Nazianzen 47
the same collaborative impulses and competitive pressures
that had formerly been contained within Prohairesius’
school.63 Here again, the friends’ relationship remains on
display: the letters in which Gregory successively mocks and
hymns the monastic ‘thinkery’ that Basil had established
in Pontus are intended for wider circulation, to define the
project (and Gregory’s part in it) within a peer group.64 Nor
could there be more striking testimony to Gregory’s con-
tinued engagement with the world of the schools than his
formal publication of these letters, along with some of
Basil’s replies, over two decades later—as part of a collection
compiled on behalf of a young protégé who was at the time
studying with the rhetors of Caesarea.65
Augustine’s venture at Cassiciacum, too, was an assert-
ively private withdrawal that was deliberately, in the series of
pamphlets he produced there, laid open to the public gaze.
And although his great project of reformatting the liberal
disciplines into vehicles for managing philosophical ascent
might at first seem to have as little direct relation to con-
temporary educational practice as Basil’s retirement to
Pontus, both stand in a similar relationship to the spiritual-
izing, idealizing impulses behind classical education.66 For
the same desire to work ‘through the corporeal to the
incorporeal’ infused the intense friendships between young
men that were encouraged by the schools; nor should we
suppose that Gregory and Basil were the only couple at
Athens (or elsewhere) to conceive their studies as a prepara-
tion for ‘escape from the world’. And Augustine’s plan to

63
In Ep. 1 Gregory cites the need to tend his parents as his excuse for
not joining Basil; in Ep. 2 he offers the counter-proposal that Basil should
join him at Nazianzus. For general background, see McLynn, ‘A Self-
Made Holy Man: The Case of Gregory Nazianzen’, JECS 6 (1998),
463–83, at 467–8.
64
Greg. Naz., Epp. 4–6, replying serially to Basil, Ep. 14; the fact that
Gregory refers directly to Basil’s letter in the first of his replies, despite
having paid an extended visit to his friend in Pontus before writing it, is
clear evidence that this correspondence is constructed as a dialogue for the
benefit of readers.
65
For the context and organization of the letter collection, see
McLynn, ‘Gregory Nazianzen’s Basil’, 183–90.
66
The point is well made, in relation to Basil, by P. Rousseau, Basil of
Caesarea (Berkeley and Los Angeles, 1994), 57–8, 70–2.
48 Neil McLynn
transform the ‘disciplines’ from metaphors into a set of pre-
cisely targeted methodologies might be seen as an exercise in
just such escapology. In launching his grandiose programme,
he was finally announcing a decisive break with the day-
to-day untidiness of the schoolroom. However, the ‘discipli-
narum libri’ should no more be regarded as educational
textbooks than should Gregory’s letter collection. Rather,
in publicly setting himself so monumental a challenge,
Augustine was arguably embarking on an exercise in self-
definition—an exercise that can be connected with the
uncertainty of his status in Milan in autumn 386, after his
resignation from his rhetorical post. It does not detract from
the seriousness of his project thus to interpret it as, in large
part, an exhibition:67 for a central theme of this essay has
been to emphasize that displays of philological prowess were
very serious matters indeed. Rather, we might note how
completely Augustine’s great disciplinary endeavour falls
into abeyance when he leaves Milan (and the literary circles
he was seeking to impress there) in 387. Only now, perhaps,
when he departs for his native town and what he expects will
be permanent immersion there, does Augustine ‘graduate’.
67
In this respect the alternatives offered by O’Donnell, Augustine:
Confessions, ii. 274, are unduly restrictive: to avoid reducing Augustine
to a ‘worn-out pedant exhibiting his expertise’, he makes the project ‘a
serious and original contribution to philosophical literature’.
 3
The Duty of a Teacher: Liminality and
disciplina in Augustine’s De Ordine
C at h e r i n e C o n y b e a r e

In his rural retreat in the autumn of 386, driven by ill

health from his pinnacle of secular achievement, the chair of
rhetoric at Milan, and waiting for baptism to Christian-
ity, Augustine was poised— intellectually and culturally—
between two very different courses of life. His writings of
the period enact a complex and constantly shifting negotia-
tion between these apparent poles, between different inter-
pretations or aspects of the role they might play in choosing
and pursuing a good life. Needless to say, the negotiation
remains inconclusive: Augustine’s intellectual journey draws
him swiftly away from the assumption that firm answers may
ever be found in human inquiry. But the way in which it is
conducted is of great interest.
Augustine at Cassiciacum was self-consciously on the
threshold of something new, and he set out to make that
sense of liminality public with the composition of his dia-
logues there—for, as he explained in De Ordine, ‘quite a few
people suddenly convert to a wonderfully good life, and until
they draw attention to it with some more conspicuous
actions, everyone believes them to be the same sort of people
as they used to be’. 1
Augustine’s sense of liminality carries over into the
literary production from Cassiciacum; and this has been
insufficiently remarked. These are precisely the ‘more

1
‘non pauci se subito ad bonam uitam miramque conuertunt, et donec
aliquibus clarioribus factis innotescant, quales erant, esse creduntur’ (Ord.
2. 10. 29). I use the edition of W. M. Green, in CCSL 29; all translations
of this and other texts, bar that in n. 12, are my own.
50 Catherine Conybeare
conspicuous actions’ through which he announces—or
begins to—that he is now a different sort of person. Why did
Augustine elect—for these ‘actions’ which were personally,
if not yet historically, significant—to write in the form of the
philosophical dialogue? (I make the assumption here that
it was indeed a matter of choice, and that these dialogues are
not simply transcriptions of actual conversations.2) Though
some attempts had been made to ‘Christianize’ the dialogue,
the fact remains that in Latin, at any rate, it was still pre-
dominantly associated with the secular teaching which
Augustine was seeking to revise.3 Not only did the dialogue
form look back chronologically and intellectually to the pre-
Christian world of Cicero, Augustine’s principal model, but
Cicero himself enshrined a generic retrospection, both with
his claims for his interpretation of Greek philosophy and,
more immediately, with his frequent choice of participants
such as Scipio or Cato—august exemplars, but deceased.4
Seen in this light, Augustine’s choice of young, living
interlocutors is just as significant as the more frequently
remarked inclusion of his mother.5 It brings to the fore a

2
I do not wish to rehash here the long debate about the historicity of
the dialogues. Joanne McWilliam provides a welcome corrective to a
rather sentimental interpretative tradition which sees Augustine as (a)
transparent reporter or (b) victim of his genre, with her reading of each of
the characters as of primarily symbolic significance (‘The Cassiciacum
Autobiography’, StPatr 18/4 (1990), 14–43); she gives references for the
‘historicity’ debate, p. 16 nn. 5 and 6.
3
For surveys of the dialogue in the Christian tradition, see B. R. Voss,
Der Dialog in der frühchristlichen Literatur, Studia et Testimonia
Antiqua 9 (Munich, 1970); M. Hoffmann, Der Dialog bei den christlichen
Schriftstellern der ersten vier Jahrhunderte (Berlin, 1966).
4
For a recent insistence on the importance of Cicero for these dia-
logues, see M. P. Foley, ‘Cicero, Augustine, and the Philosophical Roots of
the Cassiciacum Dialogues’, REAug 45 (1999), 51–77; Seth Lerer is inter-
esting on the topic in ch. 1 of Boethius and the Dialogue: Literary Method
in ‘The Consolation of Philosophy’ (Princeton, 1985). The testimonia for
Augustine’s reading of Cicero are systematically studied in M. Testard,
Saint Augustin et Cicéron, 2 vols. (Paris, 1958).
5
See, amid a plethora of references, E. Lamirande, ‘Quand Monique, la
mère d’Augustin, prend la parole’, in A. Zumkeller (ed.), Signum Pietatis:
Festgabe für Cornelius Petrus Mayer OSA (Würzburg, 1989), 3–19;
R. Holte, ‘Monica, “the Philosopher”’ , Augustinus 39 (1994), 293–316,
esp. 306–8.
 The Duty of a Teacher 51
sense of philosophy in the making, as a living tradition
subject to revision and reconsideration, rather than as some-
thing primarily concerned with translation and reinter-
pretation. The Cassiciacum dialogues enact a participatory,
and forward-looking, pedagogical practice: this is essential to
our main theme here.
Equally important, the genre is one which emphasizes
constantly, in conventions and content, the tension between
speaking and writing. Whatever one’s position in the histor-
icity debate, it must be conceded that Augustine is at
immense pains to make it appear that the Cassiciacum dia-
logues actually took place. Participants wander in and out.
Their reactions and facial expressions are remarked upon.
They play for time, they repeat themselves, they express
reluctance to participate, they even (famously) burst into
song. Locations are commented on; so are breaks for meals.
Above all, Augustine dwells on the recording presence of the
pen (almost invariably the pen personified, not the ‘notarius’
(secretary) himself).6 We might expect this from the pro-
grammatic statement at the beginning of the Soliloquia:
Ratio: Right, suppose you’ve discovered something: to what will
you entrust it, so that you can move on to other things?
Augustinus: To memory, I suppose.
Ratio: And is memory so powerful that she can preserve every-
thing satisfactorily, once you’ve thought it through?
Augustinus: It’s difficult—or rather, impossible.
Ratio: Well then, there must be writing.7

6
See, e.g., C. Acad. 1. 1. 4: ‘Adhibito itaque notario, ne aurae laborem
nostrum discerperent, nihil perire permisi’ (‘Having employed a secretary,
lest the breezes disperse our hard work, I allowed nothing to be lost’);
compare Ord. 1. 2. 5: ‘disserebamus inter nos, quaecumque videbantur
utilia adhibito sane stilo, quo cuncta exciperentur, quod videbam
conducere valetudini meae’ (‘we discussed amongst ourselves whatever
seemed to the purpose; obviously, a pen was employed to catch everything—
I took care to hire it for my health’). The mention of the notarius in the
first passage is exceptional.
7
Sol. 1. 1. 1: ‘Ratio: Ecce, fac te invenisse aliquid; cui conmendabis, ut
pergas ad alia?
Augustinus: Memoriae scilicet.
R: Tantane illa est, ut excogitata omnia bene servet?
A: Difficile est, immo non potest.
R: Ergo scribendum est.’
52 Catherine Conybeare
‘Ergo scribendum est’, Reason concludes; not ‘I, you, or we
have to write’, but: ‘there must be writing’.8 This, then,
is why the (apparently impersonal) writing down of the
Cassiciacum dialogues is called for: to circumvent the pitiful
inadequacies of memory. But, as we shall see, this utterly
conventional causa scribendi is gradually interrogated and
undermined as the dialogues proceed, and the liminal space
between the spoken and the written word in the dialogue
form contributes to the interrogation. The relationship
between memory and writing proves to be crucial to
Augustine’s philosophical—or should one say theological?9—
revisionism.
So, we have two types of liminality already in play: the
biographically liminal state in which Augustine found him-
self at Cassiciacum and the liminality between writing and
speech in the genre through which he chose to make it
public. But both are merely contexts—if crucial ones—for
Augustine’s repeated exploration of liminal concepts in the
dialogues. Time and again, he sets up old dichotomies, only
to challenge them by interrogating the space between them:
it is in this shifting intermediary space that he uses con-
cepts which begin to displace, or at least call into question, a
simple binary schematic. As we shall see, this was to prove
essential to his moving away from the dualisms passed down
to him by both his Manichaean and his Neoplatonic
intellectual inheritance.
A passage from the second book of De Ordine illustrates
especially richly Augustine’s internalization of these various
liminalities, and I propose examining it closely. Aptly
enough, it concerns the nature and status of memory,
‘memoria’—and thus the heart of the purported reason for
writing down the dialogues in the first place. As the passage

8
One would normally expect an agent, expressed with a dative (pro)-
noun, as part of a construction like this: my slightly strained translation
attempts to capture the significance of that omission.
9
Note the observation of John Rist: ‘to call Augustine a philosopher
rather than a theologian is not merely to admit a distinction which he
would not have accepted; it is to propose a distinction which he did not
know’: Augustine: Ancient Thought Baptized (Cambridge, 1994), 5. My
argument below should demonstrate my whole-hearted agreement with
this view.
 The Duty of a Teacher 53
opens, Licentius is suggesting to Augustine a view of the
location and qualities of ‘memoria’:
‘I think that memory lives in this subordinate part of the mind
(anima). The wise mind (anima) uses it as if it were a slave, giving
orders to it and imposing the confines of the law on it once it’s
tame and submissive, so that while [the memory-slave] is using
those senses for the things which are no longer necessary to
the wise mind, only to itself, it should not dare to raise itself up or
lord it over its master, nor to use the things which relate to it
indiscriminately and extravagantly. It is transient things which
have the power to relate to that extremely inferior part. What is
memory needed for, if not for transient and—as it were—fugitive
things? After all, that wise man embraces and revels in God, who is
everlasting: one doesn’t hope that he exists, or fear that he doesn’t;
rather, he is constantly present, by the very fact that makes him
truly exist. Immovable and self-contained, he oversees the
slave’s purse, as it were, such that the memory, like a frugal and
punctilious servant, might use it well and keep it carefully.’
While I was pondering this opinion admiringly, I remembered
that on another occasion, in his hearing, I had said a short version
of that very thing. I laughed at him and said, ‘Thank that slave
of yours, Licentius: if he hadn’t served you up something from
his private purse, you might not have had anything to supply. If
memory relates to that part which has admitted that it is ruled by
a good mind as if it were a serving-wench, it is she who has just
helped you—mark my words—to say this. Now, before I go back
to the original schedule, don’t you think that a wise man needs
memory at least for this sort of thing—that is, for honourable and
necessary sorts of teaching?’
‘Why,’ he said, ‘does he need memory, when he actually keeps
everything of his own present at hand? Especially in the case of
the sense which relates to what is before our eyes, we don’t call
memory to our aid. So since a wise man has everything in front
of the inner eyes of his understanding, in other words, he con-
templates God himself—with whom is everything which the
understanding sees and possesses—in utter immobility: why, I
repeat, does he need memory? . . .’
I said, ‘Can that wise man desert his own? When he’s driving
this body, in which he keeps that servant [memory] tied down
by his law, does he by some arrangement abandon the duty of
bestowing favours on whoever he can, and especially that duty
which is most vociferously demanded of him, of teaching wisdom
itself? When he does this, he often prepares something to expound
and investigate from a pre-arranged plan, which is necessarily lost
54 Catherine Conybeare
if he doesn’t commit it to memory. So, either you’ll deny that
benevolent duties are characteristic of a wise man, or you’ll
acknowledge that some things belonging to the wise man are kept
in his memory. Or does he perhaps entrust some necessary part of
his own possessions to himself, for that servant [memory] to keep
safe, not on his own account, indeed, but on that of his associates,
such that [the memory-slave], as a sensible person and a product of
his master’s excellent teaching, would certainly not hang on to it if
[his master] hadn’t ordered [the memory] to keep it in order to lead
the foolish to wisdom?’
He said, ‘I don’t think anything at all is entrusted to it [memory]
by the wise man, if he really is constantly focused upon God,
whether silent or speaking with others; but that well-trained
slave carefully preserves what he may at intervals suggest to his
master when he’s disputing and perform as his own welcome duty
for that most just man, beneath whose power he realizes that
he lives, and he [the memory-slave] does this not, as it were, by
the use of ratio, but under the organization of that highest law and
order.’
‘I make no objection at present to your reasoning,’ I said, ‘so that
we may pursue further what we’ve undertaken.’10

10
Ord. 2. 2. 6–7: ‘[Inquit Licentius:] In qua parte [animae] subiecta
etiam ipsam memoriam puto habitare. Vtitur ergo hac sapiens quasi servo,
ut haec ei iubeat easque iam domito atque substrato metas legis inponat,
ut dum istis sensibus utitur propter illa, quae iam non sapienti sed sibi
sunt necessaria, non se audeat extollere nec superbire domino nec his ipsis,
quae ad se pertinent, passim atque immoderate uti. Ad illam enim
vilissimam partem possunt ea pertinere, quae praetereunt. Quibus autem
est memoria necessaria nisi praetereuntibus et quasi fugientibus rebus?
Ille igitur sapiens amplectitur deum eoque perfruitur, qui semper manet
nec expectatur, ut sit, nec metuitur, ne desit, sed eo ipso, quo vere est,
semper est praesens. Curat autem immobilis et in se manens servi sui
quodam modo peculium, ut eo tamquam frugi et diligens famulus bene
utatur parceque custodiat.
[7.] Quam sententiam eius cum admiratione considerans recordatus
sum id ipsum aliquando me breviter illo audiente dixisse. Tum arridens:
Gratias age, inquam, Licenti, huic servo tuo, qui tibi nisi aliquid de
peculio suo ministraret, nunc fortasse quod promeres non haberes. Nam si
ad eam partem memoria pertinet, quae se velut famulam bonae menti
regendam concedit, ipsa nunc adiutus es, mihi crede, ut hoc diceres. Ergo
antequam ad illum ordinem redeam, nonne tibi videtur vel propter
talia, id est propter honestas ac necessarias disciplinas, memoria opus esse
sapienti?—Quid, inquit, memoria opus est, cum omnes suas res praesentes
habeat ac teneat? Non enim vel in ipso sensu ad id, quod ante oculos
nostros est, in auxilium vocamus memoriam. Sapienti igitur ante illos
 The Duty of a Teacher 55
Licentius begins by associating ‘memoria’ uncompli-
catedly with the transient. We use it to remember fluctuating
things; and it is itself unreliable, as we learnt in the passage
at the beginning of the Soliloquia cited above. It belongs in
a part of the ‘anima’, but a ‘vilissima pars’, an ‘extremely
inferior part’—the part, precisely, that deals with the transi-
ent. Only permanent and immobile things, however, can be
with God: only the higher part of the soul, not tainted by
sense perception, can be called wise.
This is an utterly conventional view of the role and place
of ‘memoria’, which goes back at least to Plotinus (‘memory
is of things that have happened and passed away’11). Behind
Licentius’ interjections throughout this passage lies a figure
very like Plotinus’ upward-looking, solitary sage: ‘What . . .
if one does not depart at all from one’s contemplation of [the
eternal] but stays in its company, wondering at its nature,
and able to do so by a natural power (physis) that never fails?

interiores intellectus oculos habenti omnia, id est deum ipsum fixe

immobiliterque intuenti, cum quo sunt omnia, quae intellectus videt ac
possidet, quid opus est quaeso memoria? . . . [Inquam:] . . . numquidnam
sapiens iste suos potest deserere aut ullo pacto, cum hoc corpus agit, in
quo istum famulum sua lege devinctum tenet, relinquit officium beneficia
tribuendi quibus potest et maxime, quod ab eo vehementissime flagitatur,
sapientiam ipsam docendi? Quod cum facit, ut congrue doceat minusque
ineptus sit, praeparat saepe aliquid, quod ex dispositione eloquatur ac
disputet, quod nisi memoriae commendaverit, pereat necesse est. Ergo aut
officia benevolentiae negabis esse sapientis aut confiteberis res aliquas
sapientis memoria custodiri. An fortasse aliquid suarum rerum non
propter se quidem sed propter suos, sibi tamen necessarium commendat
servandum illi famulo, ut ille tamquam sobrius et ex optima domini disci-
plina non quidem custodiat, nisi quod propter stultos ad sapientiam per-
ducendos sed quod ei tamen ille custodiendum imperarit?—Nec omnino
huic, inquit, commendari quicquam arbitror a sapiente, si quidem ille deo
semper infixus est sive tacitus sive cum hominibus loquens; sed ille servus
iam bene institutus diligenter servat, quod interdum disputanti domino
suggerat et ei tamquam iustissimo gratum faciat officium suum, sub cuius
se videt potestate vivere, et hoc facit non quasi ratiocinando sed summa
illa lege summoque ordine praescribente.—Nihil, inquam, nunc resisto
rationibus tuis, ut quod suscepimus potius peragatur.’
11
Enneads 4. 4. 6. Klaus Winkler has traced closely the relationship
between Plotinus’ thought and Augustine’s in this precise passage of Ord.:
see ‘La Théorie augustinienne de la mémoire à son point de départ’, in
Augustinus Magister, 3 vols. (Paris, 1955), i. 511–19.
56 Catherine Conybeare
Surely one would be . . . oneself on the move towards eter-
nity and never falling away from it at all, that one might be like
it and eternal, contemplating eternity and the eternal by the
eternal in oneself.’12 To be wise is to associate oneself with
the eternal and immutable; memory must be base, for it deals
with ephemera.
However, Augustine manages quietly to problematize this
straightforward vision—this polarization of the eternal and
the mutable—by a sequence of discursive moves.
The first occurs in the speech of Licentius himself, when
the referent of the adjective ‘sapiens’ (‘wise’) slips back to
being the (whole) wise man, not just the part of the ‘anima’
of which wisdom can be predicated (that association has
been made explicit in the sentence before this extract
begins). ‘After all,’ says Licentius, ‘that wise man’— ‘ ille’,
not ‘illa’, ‘sapiens’—
‘ embraces and revels in God, who is
everlasting.’ What are the implications of this? To parody
Augustine’s own ‘ars disputandi’, is the wise man then to be
identified only with the wise part of his ‘anima’? Or does he
in some way contain the whole ‘anima’—the parts which deal
with impermanence as well as the immobile parts? And in
that case, how can he still be called wise? Plotinus dismisses
this potential problem in his treatise on blessedness (Peri
Eudaimonias), insisting that the true sage is completely
detached from the inferior parts of his soul: only the parts
which pertain to blessedness are truly his.13 But Augustine,
as we shall see, is not satisfied with this separation. He
sees very clearly that what sounds plausible on the meta-
physical level—that the spheres of attention of the mind

12
Enn. 3. 7. 5; translation of A. H. Armstrong. The exact nature
and degree of Augustine’s knowledge of Plotinus—or, for that matter,
Porphyry—is not a question which I wish to address here: it is clear to me
that he knew, by some route, the general tenets which I am discussing,
and had engaged with them deeply. The interested reader may note the
aporetic comments of Pierre Hadot in Marius Victorinus: Recherches sur sa
vie et ses œuvres (Paris, 1971), 206–10; for the conviction that Augustine
had read the Enneads, see, e.g., A. Solignac, ‘Réminiscences plotiniennes
et porphyriennes dans le début du de ordine de saint Augustin’, Archives de
Philosophie 20 (1957), 446–65.
13
Enn. 1. 4. 4; see Winkler, ‘La Théorie augustinienne de la mémoire’,
513–14.
 The Duty of a Teacher 57
should be divided—begins to unravel when confronted with
embodiment.
The next move in this potential unravelling is also placed
in Licentius’ mouth, as he develops the image of the wise
man using his memory like a slave: ‘Immovable and self-
contained, he oversees the slave’s purse, as it were, such that
the memory, like a frugal and punctilious servant, might use
it well and keep it carefully.’ ‘Métaphore étrange et curieuse’,
observes Winkler darkly.14 ‘Peculium’, the term translated
as ‘purse’, refers expressly to money or property disbursed
by the head of a household (here ‘dominus’) to one of his
children or slaves, to use as their own; but the privilege of
controlling that property may be revoked by the ‘dominus’ at
any time. The image seems apt when Licentius introduces it,
not least because it formalizes his separation of the spheres
of attention: the wise man remains fixed on the immutable;
his slave, memory, dispenses individual memories from his
own ‘peculium’, which the wise man need not have anything
to do with. But in fact, this metaphor will fatally complicate
Licentius’ simplistic vision. Who, after all, was responsible
for establishing the ‘peculium’ in the first place?
Augustine’s gentle mockery of Licentius—advising him
to thank his memory-slave for helping him out—punctuates
the exchange here. It serves to highlight Augustine’s own
shift in attitude: from admiration to the realization that his
pupil is parroting his own former teaching. His realization
is articulated not just through his laughter but through a
(temporary) change of the vocabulary in which the exchange
is couched: where Licentius has consistently been using
‘servus’, ‘sapiens’, and ‘anima’, Augustine uses ‘famula’
(‘serving-wench’), ‘bona’, and ‘mens’. Augustine thus, within
the internal logic of the dialogue, stages a debate in which he
is effectively arguing with himself and his ‘Plotinian’
opinions—given fictive utterance by Licentius. It is not sur-
prising that the ensuing debate forms the site of a crucial
theoretical revision.
The Augustine within the dialogue counters Licentius’
polarities with an extraordinarily significant move: the
claim that the wise man needs memory ‘for honourable and
14
Winkler, ‘La Théorie augustinienne de la mémoire’, 512.
58 Catherine Conybeare
necessary sorts of teaching’ (‘propter honestas ac necessarias
disciplinas’). The association of memory and teaching had
already been placed in Alypius’ mouth in Contra Academicos:
‘If I don’t forget to mention anything, I shall rejoice in my
memory as well as your teaching (doctrina).’15 But here
Augustine engages far more closely with the combination,
and explores its implications, not for the pupil’s but for the
teacher’s memory. The abstract division of the ‘anima’ with
which Licentius opened has been replaced by the figure of an
actual wise man who, we expect, will act in accordance with
the dictates of wisdom—and it will be part of being wise to
attempt to communicate his wisdom to others.
It is precisely, then, the concept of ‘disciplina’ which
prises open the liminal space in Licentius’ dichotomous
construction, and shows how unrealistic that construction is.
If the wise man has to teach, and needs memory to do so,
how can we neatly divide off memory as dealing only with
the transient?
What does Augustine understand by ‘disciplina’ here?
Pacioni, who provides a running commentary on this
passage, paraphrases, ‘memory is indispensable to the wise
man for learning the liberal disciplines’;16 but this surely
attests to the retroactive pressure exerted by the renowned
discussion of the ‘liberal disciplines’ towards the end of the
second book of De Ordine (as well as an over-rigid inter-
pretation of ‘disciplina’ as concerned only with learning).
There is no suggestion here that ‘disciplina’ might contain
the systematic agenda which is to be set out later: the ‘dis-
cipline(s)’ are not necessarily ‘liberal’. Augustine had sup-
plied a simple explanation of ‘disciplina’ in the Soliloquia:
‘ “discipline” gets its name from [the process of] learning’
(‘disciplina . . . a discendo dicta est’),17 and goes on to say
that ‘a discipline cannot be a discipline unless it teaches true

15
C. Acad. 2. 6. 14: ‘Si enim nihil me fugerit, gratabor cum doctrinae tuae
tum etiam memoriae meae.’
16
‘che la memoria sia indispensabile al sapiente per apprendere le dis-
cipline liberali’: V. Pacioni, L’Unità teoretica del ‘de ordine’ di S. Agostino
(Rome, 1996), 182.
17
Sol. 2. 11. 20; an identical definition in the sermon dating from c.400,
De Disciplina Christiana 1 (PL 40.669–78), stemming from Varro (see
Pollmann’s essay in this volume).
 The Duty of a Teacher 59
18
things’ (‘ . . . nisi vera doceat’). We should note that there
is, in Latin, a fundamental ambiguity in the term ‘disciplina’
(which is reflected in the English ‘discipline’, but not in the
verbs ‘to teach’ or ‘to learn’19). It may denote both learning
(as a fait accompli), and, by extension, the things learnt;
but it may also emphasize the process by which things are
learnt, and hence mean both ‘teaching’ (with the focus on the
person generating the process of learning) and ‘discipline’
(with the focus on the recipient of learning, and the tech-
niques of apprehension which she or he uses).20 Here, it is
precisely the reciprocity of teaching and learning—along
with the importance of its subject-matter being ‘true’—that
is being captured. The point is not that the wise man should
engage with the system of liberal disciplines, but that he
should teach at all. This is where Augustine parts com-
pany decisively with Neoplatonic thought.21 The wise man,
instead of occupying a solipsistic contemplative space, is
suddenly crucially defined by his relationship with his
pupils, with whom he learns through teaching. ‘Disciplina’
precisely negotiates the divide between the metaphysical and
engagement with the corporeal world.22 Now, surely, we

18
‘conlegimus disciplinam, nisi vera doceat, disciplinam esse non
posse’: Sol. 2. 11. 20.
19
Other than in the archaism, now a vulgarism, ‘I’ll learn him’, mean-
ing ‘I’ll teach him a lesson’.
20
Jean Leclercq’s entry under ‘disciplina’ in the Dictionnaire de Spi-
ritualité provides an excellent introduction to the classical background and
Christian development of the term. There are two extended studies of
the term in the early writings of the Church: O. Mauch, Der lateinische
Begriff ‘disciplina’: Eine Wortuntersuchung (Freiburg, Switzerland, 1941);
and W. Dürig, ‘Disciplina: Eine Studie zum Bedeutungsumfang des
Wortes in der Sprache der Liturgie und der Väter’, Sacris Erudiri 4
(1952), 245–79. Each is generally useful, but neither deals explicitly with
this passage of De Ordine.
21
See Winkler, ‘La Théorie augustinienne de la mémoire’, 517: ‘Plotin,
dans sa vie personnelle, prenait au sérieux sa tâche comme maître et ses
devoirs envers ceux qui avaient été recommandés à ses soins, mais dans son
système le rôle des vertus sociales est loin d’être important.’
22
Henri-Irénée Marrou makes a very similar observation of a slightly
later passage in De Ordine: ‘Le passage du sens païen au sens chrétien
[du mot disciplina] est bien marqué dans un texte de saint Augustin [Ord.
2. 8. 25] qui date d’une période où sa pensée était encore tout imprégnée
d’une atmosphère philosophique: disciplina, c’est la sagesse envisagée non
60 Catherine Conybeare
begin to see why Augustine has chosen for his dialogues
a group of ‘young, living interlocutors’, and why he has
chosen to launch his Christian career in dialogue form at all:
it is by enacting the wise man’s commitment to ‘disciplina’
that he will begin publicly to move beyond Neoplatonism.
Moreover, these ‘disciplinae’ are ‘honourable and neces-
sary’ (‘honestas ac necessarias’). The adjectives are not
glossed explicitly here; but the ‘honourable’ must refer both
to the virtue of the wise man and to the truth (as we learnt
from the Soliloquia) of what he teaches. The ‘necessary’ is
rather more complicated. The word ‘necessarius’ occurs four
times in this short extract, along with one instance of
‘necesse’. Each side uses it to enforce agreement by obviating
debate about its associated noun or concept—by making the
implicit claim that the link is natural.23 Licentius has used
the word twice already, once of the (indeterminate) things
which are ‘necessary’ to the memory, but not to the wise
part of the mind; once, explicitly, of the ‘transient and fugi-
tive things’ for which memory is ‘necessary’. It therefore
becomes an important part of the verbal strategy for suggest-
ing that the division between the transient and the eternal
is obvious and straightforward. Here, however, Augustine
challenges that division by stating that teaching is, for the
wise man, ‘necessary’. The shift from mind to man
(‘illa sapiens’ to ‘ille sapiens’), which has already occurred
in Licentius’ speech, is essential to this move (it would be
hard to claim that teaching was necessary to the wise mind
alone). The adjective ‘necessarius’ is tossed in at the end of
Augustine’s brief speech, with no explicit justification; but
it is on this that the complication of ‘memoria’ hinges, and
hence the whole challenge to the dissociation of the
embodied from the divine.

seulement sous son aspect théorique, mais encore dans ses conséquences
pratiques; elle implique une règle de vie’: Bulletin du Cange 9 (1934), 18;
my emphasis.
23
Compare Gillian Beer’s observations on ‘naturally’: ‘Argument has
already been prejudged in that word. Communality is being lined up
behind the speaker’: ‘Representing Women: Re-presenting the Past’, in
C. Belsey and J. Moore (eds.), The Feminist Reader, 2nd edn. (Basingstoke,
1997), 83.
 The Duty of a Teacher 61
Licentius counters by reasserting his original position:
he continues to try to dissociate ‘memoria’ from the
immutable—to divorce it firmly from its liminal status and
put it back in the world of fugitive things, ‘res fugientes’.
Why does the wise man need memory, he asks, when he
keeps ‘all his own things present to him’ (‘omnes suas res
praesentes’)? Or should we translate, ‘his entire reality’—the
divine things which he contemplates and which are the only
things that count? Either way, Augustine calls into question
the idea that this should, or could, be the wise man’s only
concern—because of his obligation to teach, to relate to his
fellow human beings. In the section that I have omitted,
Licentius goes on to protest at some length that he him-
self is not wise—apparently to defy Augustine’s joking
intervention and to imply that his own use of his memory’s
‘peculium’ cannot be included as an argument against him.24
But Augustine—here as in the Cassiciacum dialogues as a
whole—is more concerned to include than to demarcate: he
responds briskly, ‘I’m not wise either’ (‘nec me’); but returns
to his insistence on the importance of ‘disciplina’.
At this point, Augustine inserts a further problematizing
move when he observes that the wise man’s preparations for
disputation ‘are necessarily lost unless he commits them to
memory’.25 Once again, ‘necesse’ signals a remarkable leap
in the argument. Hitherto in the dialogues, ‘memoria’ has
been related to the perceived problem of impermanence,
which is ‘solved’ by writing: so with the exchange in the
Soliloquia, ‘ergo scribendum est’. But here, the role which
‘memoria’ plays is precisely the opposite: it guarantees that
the elements of disputation—or, for that matter, of teaching
more generally—will not be lost; and hence it participates
in the permanence of the divine. Perhaps this picks up
Augustine’s desire, expressed earlier in De Ordine, that the
boys should move ‘praeter codices’ (‘beyond books’), and

24
This is the ingenious interpretation of Winkler, ‘La Théorie
augustinienne de la mémoire’.
25
‘quod nisi memoriae commendaverit, pereat necesse est’. Note the
indicative mood: there is no question of claiming (with the subjunctive)
that the preparations might be lost; they will, necessarily, be so.
62 Catherine Conybeare
26
‘apud sese habitare’ (‘remain self-contained’)—which is
followed in the first instance by Licentius’ extravagant
conversion experience, but for which we may see another,
more convincing, possibility here.27
We also see, in the reinstatement of ‘memoria’, Augustine’s
continued tacit insistence on the importance of situating
his philosophy in embodied human experience: in actual
people thinking and teaching. It is just about possible to
remove the personal aspect from the statement of ‘Ratio’ in
the Soliloquia, ‘ergo scribendum est’, for which we have
already suggested the translation ‘there must be writing’.
Writing may be displaced on to the impersonal agency of
‘the pen’, or an appeal to ‘the record’—as Augustine does
repeatedly in the Cassiciacum dialogues. But how could one
eliminate the personal from the parallel statement, ‘ergo
reminiscendum est’? What meaning can we attach to ‘there
must be remembering’, without a mind, or a person, to do it?
There can be no side-stepping of personal involvement—
unless we somehow separate the mind from the body, things
eternal from things transient; the possibility of which is
precisely the locus of the dispute between Augustine and
Licentius here.
Now, as if to underline that move, we revert to the vexed
question of who has established, and who controls, the
‘peculium’ of Licentius’ simile.
‘Does [the wise man] perhaps entrust some necessary part of his
own possessions to himself, for that servant [memory] to keep safe,

26
On ‘secum’ see J. Doignon, ‘Problèmes textuels et modèles littéraires
dans le livre I du De Ordine de saint Augustin’, REAug 24 (1978), 71–86.
27
Could it be correct to read the ‘extravagant conversion experience’ as
Licentius—as it were—jumping the gun and rushing into a simplistic
solution? Various aspects of the account, not least its position in the logic
of the dialogue, lend themselves to this interpretation; and Augustine
has repeatedly portrayed himself doing the same thing in Sol., leaping to
an apparently obvious conclusion and considering the matter closed—only
to have the inquiry opened up for him once again in the next minute. If
the Licentius of the dialogues bears any relation at all to the historical
Licentius, the conversion clearly didn’t ‘take’: a full decade later,
Augustine is writing to Paulinus of Nola to request his help with the
incorrigibly pagan Licentius (Augustine, Ep. 27; the result is Paulinus,
Ep. 8, largely a poem for Licentius).
 The Duty of a Teacher 63
not on his own account, indeed, but on that of his associates, such
that [the memory-slave], as a sensible person and a product of his
master’s excellent teaching, would certainly not hang on to it if
[his master] hadn’t ordered [the memory] to keep it in order to lead
the foolish to wisdom?’28
Again, the adjective ‘necessarius’ is crucial: the contents of
the ‘peculium’ are indeed necessary to the wise man— ‘ non
propter se . . . sed propter suos’ (‘not on his own account but
on that of his associates’): here is the tell-tale responsibility
of the wise man to those around him which both Plotinus
and Licentius wish to elide. The memory-slave is fully a
part of the wise man, his master: the wise man controls, uses,
and engages with his ‘memoria’. How is the relationship
between the two—master and slave, metaphysical and cor-
poreal, eternal and transient—configured? As ‘ex optima
domini disciplina’: as the product, or by the example, of the
excellent teaching of the master. It is ‘disciplina’, once again,
which both complicates and bridges the liminal space
between dichotomies.29
This is not quite the end of the discussion of ‘memoria’.
Licentius tries one last time to salvage his simile and his
separation of the divine from the corporeal. He tries once
more to give an account of the actions of the memory-slave
which separates the slave’s concerns utterly from those of the
wise man. The wise man is ‘deo semper infixus’ (‘always
focused on God’), whatever else he is (or seems to be) doing;
his slave is responsible, on his own initiative, for administer-
ing the ‘peculium’, ‘not, as it were, by the use of ratio, but
under the organization of that highest law and order’.
Augustine responds with a self-referential shift, which alerts
us to the (already striking) use of the verb ‘ratiocinando’ (‘by
the use of ratio’): ‘I make no objection at present to your
reasoning (rationibus tuis).’

28
Most of the subjects in this passage are expressed simply with pro-
nouns, which I have expanded with the words in square brackets. Though
the Latin is a little convoluted, the referents of the individual pronouns
are not, I think, ambiguous.
29
This also, we may note, lays the groundwork for a paradigm in which
Christ’s teaching is instrumental in negotiating the divide between human
and divine—and human teaching, as part of the imitatio Christi, likewise;
but Augustine does not make this move here.
64 Catherine Conybeare
Several things are going on in this invocation of ‘ratio’.
Contrary to what one might expect from her relatively
unproblematic appearance as interlocutor in the Soliloquia,
‘ratio’ is repeatedly under explicit or implicit interrogation
in these dialogues: ‘ratio’ forms, indeed—perhaps surpris-
ingly—another of Augustine’s ‘liminal’ concepts, though we
shall not explore its role in detail here. Understandably,
given its lengthy philosophical lineage, Augustine is very
attached to the notion that humankind is distinguished from
the beasts—and hence, more godlike—by its use of ‘ratio’,
the faculty of reason. Later in De Ordine, he seems to be
drawing directly on Cicero when he writes: ‘by wise men in
the old days, man was defined in this manner: man is a
rational, mortal animal.’30 In the lines here, however, we see
how Licentius’ personification of ‘memoria’ as a slave has
taken on a rather suspect life of its own, for he has to back-
track suddenly: the memory-slave gives prompts to the wise
man (as if the metaphorical ‘servus’ were a literal ‘animal
rationale’!)—but not, Licentius hastily adds, using reason
(‘ratiocinando’—like a real person), but simply as part of the
way in which things are ordered. Augustine seems to per-
ceive—surely rightly—that both this invocation of ‘ratio’
and its summary dismissal are extremely problematic.
‘Ratio’ is traditionally what makes people more divine: why
then might Licentius associate it with ‘memoria’, which (he
has insisted) deals with the transient and corporeal? But
what, in that case, of the implication that the divine order
of things, ‘summa illa lege summoque ordine’, does not
somehow include the use of reason?
In his response, Augustine reminds Licentius lightly of
his own power to reason, and defers further discussion. He
does not press home his advantage explicitly; but the fact
is that the new Augustine, the Augustine moving towards
baptism and an incarnational theology, has roundly defeated
the old Augustine in the person of Licentius. Licentius has
been caught out by his own metaphor, and has ended by

30
Ord. 2. 11. 31: ‘ipse homo a ueteribus sapientibus ita definitus est:
homo est animal rationale mortale.’ Compare Cicero, Academica 2. 7. 21:
‘si homo est, animal est mortale, rationis particeps’ (‘if he is a man, he is a
mortal animal, partaking of reason’).
 The Duty of a Teacher 65
implying that the divine is indeed inseparable from the
embodied, the permanent from the transient, and wisdom
from despised ‘memoria’ and actual human affairs.
Augustine’s message of integration is crucial to his
incipient understanding of the significance of Christ’s
incarnation; he moves towards that message through his
appreciation of the liminal space between dichotomies—an
appreciation perhaps enhanced by his own biographically
liminal situation, and enacted in his choice of genre in which
to communicate his message. The idea that the divine
and the human should—could—have nothing to do with
each other, that they express mutually exclusive concepts
with no space for further possibilities, was then, and remains,
a tempting and in many ways comfortable option. How-
ever, if my reading of this passage of De Ordine is
correct, Augustine was already, in 386, seeing that so rigid
a separation of spheres ought to be impossible within
Christianity, and was tentatively exploring techniques for
mediating between or reframing altogether the concepts at
stake. His insistence here on the importance of ‘disciplina’,
of the need to communicate godlike wisdom to others, and
on the reciprocities involved, is an extraordinarily significant
move towards that message of integration.31
31
These ideas are developed more fully in my book, The Irrational
Augustine (Oxford, forthcoming). My thinking in general on the integra-
tion of the divine with embodied experience owes most to the work of
Grace Jantzen, especially Becoming Divine: Towards a Feminist Philosophy
of Religion (Manchester, 1998).
 pa r t i i
Disciplinarum libri: The Canon in Question
 4
Augustine’s Disciplines: Silent diutius
Musae Varronis?
D a nuta R. S h a n z e r

1. i n t r o d u c t i o n
A canon of seven liberal arts was already dominant by the
time of Cassiodorus and Isidore of Seville.1 Martianus
Capella also consciously aimed for and achieved a heptad
by deliberately omitting medicine and architecture. When
and how did this canon of seven emerge? Its formation has
often been ascribed to Martianus, who himself was working
with a canon of nine, attributed to Varro, that included
medicine and architecture in addition to the usual seven.
Such were the conclusions of Ritschl’s classic treatment, in
which he outlined the problem of the origin of the heptad
of liberal arts and attempted to reconstruct the contents of
the various ‘disciplinarum libri’, making use of many late
antique sources.2 But the historical problem of the number
and identity of the liberal arts pre-dates Martianus and
Cassiodorus in the late fifth to sixth centuries. Augustine
My gratitude to Charles Brittain and to Peter Lebrecht Schmidt for
their suggestions and help. I would like to dedicate this chapter to
my beloved friend and mentor and favourite Varronian, the late Harry
Jocelyn, who saw it in its infantia (long ago—silent diutius!) and made it the
object of his invariably incisive criticism.
1
In Book 2 of his Institutiones Divinarum et Saecularium Litterarum
Cassiodorus works with what would become the usual unproblematized
medieval cycle of seven. See now J. W. Halporn and M. Vessey,
Cassiodorus: ‘Institutions of Divine and Secular Learning’, Translated
Texts for Historians (Liverpool, 2004), 64 f. [to be corrected in light of the
present essay— Eds.]. Isidore, Origines 1. 2: ‘Disciplinae liberalium artium
septem sunt . . .’ (‘The disciplines that constitute the liberal arts are
seven . . .’).
2
F. Ritschl, ‘De M. Terentii Varronis disciplinarum libris com-
mentarius’, in his Opuscula Philologica, iii (Leipzig, 1877), 352–402, at
352–4.
70 Danuta R. Shanzer
proves a crucial witness, whose own sources and contribu-
tion to the historiography of the ‘artes’ must be disentangled
if we are ever to write the cultural history of the later Roman
Empire and the transition to the early Middle Ages. The
present essay seeks at once to clarify Augustine’s intellectual
biography through source criticism and to elucidate the
literary-intellectual choices confronting later Roman and
early medieval encyclopaedists.
This is the story of the engagement—in all senses of the
word—of a great Christian thinker, Augustine, with a great
pagan scholar, educator, and man of letters, Marcus Terentius
Varro. Their point of intersection is the nature and teaching
of the ‘disciplinae’ or liberal arts. The contested ground is
not merely the canon and substance of the ‘disciplinae’, but
also the literary form in which they were presented by the
authors of textbooks. It is a tale of two lost works, Varro’s
and Augustine’s Disciplinarum Libri, a tale of repression and
suppression that ends in a surprising historical vindication
of the pagan scholar.
While other aspects of Augustine’s life and thought may
be heavily documented, his ‘disciplinae’, unfortunately, are
not. Innumerable questions surround them. Was he working
from or with a canon? If so, what canon was it? We do not
know for certain how many or what his ‘disciplinae’ were,3
or why astronomy was omitted from the list given in the
Retractationes, although it had been included in De Ordine
and De Quantitate Animae.4 The omission of any explicit
allusion to arithmetic in De Ordine has likewise exercised
scholars.5 Augustine’s sources remain largely obscure. Then
there are the questions of intellectual biography raised by his
attempt to write books on the disciplines in Milan in 387, the

3
The lack of a definite article in Latin causes difficulties, e.g., in the
interpretation of ‘de aliis vero quinque’ in Retr. 1. 5. 6 below.
4
H.-I. Marrou, Saint Augustin et la fin de la culture antique, 2nd edn.
(Paris, 1949), 248–50.
5
Ibid. 192 n. 5 suggests that Augustine as a philosopher may have felt
compelled to include philosophy and to telescope the arts of the quad-
rivium into three for a total of seven. I. Hadot, Arts libéraux et philosophie
dans la pensée antique (Paris, 1984), 122, thinks that arithmetic finds an
implicit place at the end of the sequence of mathematical disciplines, even
though it is not mentioned here but instead at Ord. 2. 18. 47.
 Augustine’s Disciplines 71
omission of this project from the narrative of the Confes-
sions,6 and its ultimate dismissal in the Retractationes in 426/7.
There he writes:
Per idem tempus quo Mediolani fui baptismum percepturus,
etiam disciplinarum libros conatus sum scribere, interrogans eos qui
mecum erant, atque ab huiusmodi studiis non abhorrebant; per
corporalia cupiens ad incorporalia quibusdam quasi passibus
certis vel pervenire vel ducere. Sed earum solum de grammatica
librum absolvere potui, quem postea de armario nostro perdidi: et
de musica sex volumina; quantum attinet ad eam partem quae
rythmus vocatur. Sed eosdem sex libros iam baptizatus, iamque ex
Italia regressus in Africam scripsi; inchoaveram quippe tantum-
modo istam apud Mediolanum disciplinam. De aliis vero quinque
disciplinis illic similiter inchoatis; de dialectica, de rhetorica, de
geometria, de arithmetica, de philosophia, sola principia remanse-
runt, quae tamen etiam ipsa perdidimus: sed haberi ab aliquibus
existimo. (Retr. 1. 6)
During the same time that I was at Milan intending to receive
baptism, I also tried to write books about the disciplines, question-
ing those who were with me and who did not shudder at studies
of this sort. I desired to arrive myself or to lead others through
corporeal things to incorporeal things by certain definite steps, as it
were. But of these disciplines, I was able to finish only the book
about grammar, which I subsequently lost from my bookcase and
six books about music, to the extent that it concerns the subject
that is called rhythm. But I wrote these same six books when I had
already been baptized and had already returned to Africa from
Italy. I had only begun that discipline at Milan. But of five [or ‘the
five’] other disciplines that I had similarly embarked upon there,
namely dialectic, rhetoric, geometry, arithmetic, and philosophy,
only the beginnings remain. Even these I also lost, but I believe
that others have them.
Our primary evidence on this subject consists of the enu-
merations of the ‘disciplinae’ in De Ordine, De Quantitate
Animae, and Retractationes, together with De Musica and

6
Conf. 9. 6. 14 would be the expected place. There is a dismissal of
‘omnes libri artium quas liberales vocant’ (‘all the books of the arts that
they call “liberal”’ ) at Conf. 4. 17. 30 with a partial enumeration. It
is interesting that music alone (which Augustine claimed only to have
begun in Milan (Retr. 1. 6)) is alluded to indirectly in Conf. 9. 6. 14 and
Conf. 10. 32. 49–50; in the latter two passages it is ‘saved’ as liturgical
music.
72 Danuta R. Shanzer
Table 4.1. Augustine’s canon of the liberal arts (period of projected
‘disciplinarum libri’)

Ord. 2. 12. 36– Ord. 2. 4. 13– Quant. An. Retr. 1. 6

15. 43 5. 14 33. 72

grammar 2. 12. 36 (grammar) (grammar) grammatica

dialetic 2. 13. 38 disputationes vis ratiocinandi dialectica
rhetoric 2. 13. 38 oratio eloquentia rhetorica
[missing] numerorum numerandi arithmetica
necessitates disciplina
music 2. 14. 39 musica modulandi musica
peritia
geometry 2. 15. 42 geometria dimetiendi geometria
subtilitates
astronomy 2. 15. 42 astrorum (astronomy/ [missing]
motus astrology/
divination)
philosophia philosophia

suggestive writings around it, such as Licentius’ hexameter

poem to Augustine and the Letter to Memor of 408/9.7 Data
for the earlier period, including the retrospective notice
of the Retractationes, are summarized in Table 4.1 (after
Marrou8). Already in the first, partial redaction of De

7
Augustine, Ep. 101.
8
Marrou, Saint Augustin et la fin de la culture antique, 189. I have taken
the liberty of editing his analysis of Quant. An. 33. 72 (the third ‘gradus’).
Its list is of the ‘everything but the kitchen sink’ type, including more than
the ‘liberal’ arts but omitting astronomy and possibly also grammar:
‘inventiones tot signorum in litteris, in verbis, in gestu, in cuiuscemodi
sono’ (probably grammar); ‘vim ratiocinandi et excogitandi’ (dialectic,
literally ‘the strength of reasoning and analysing’); ‘fluvios eloquentiae’
(rhetoric, ‘rivers of eloquence’); ‘carminum varietates’ (metrics, ‘different
types of verse’); ‘modulandi peritiam’ (music, ‘skill in making melodies’);
‘dimetiendi subtilitatem’ (geometry, ‘subtlety in measuring’); ‘numerandi
disciplinam’ (arithmetic, ‘the discipline of counting’); ‘praeteritorum ac
futurorum ex praesentibus conjecturam’ (divination or possibly astrology,
‘conjecturing about the past and the future from present events’).
O. Schissel von Fleschenberg, Marinos von Neapolis und die neuplato-
nischen Tugendgrade (Athens, 1928), 84–5, sees in the passage an originally
Aristotelian tripartite division of the intellect, with the Tugendgrade
starting at the fourth step.
 Augustine’s Disciplines 73
9
Doctrina Christiana most of the ‘disciplinae’ either carry
health warnings or are said not to need teaching per se.10 But
they were no doubt in trouble with Augustine well before
that.
Discussion must start with the second book of De Ordine
and its sources. Since Ritschl’s time, and before then, Varro
has been the prime putative source and the issue. After all,
Augustine reproduced the title of Varro’s encyclopaedia, the
Disciplinarum Libri, in his account of his own work on the
disciplines in the Retractationes (above). Harald Hagendahl
supported this view: ‘In all probability this work [Varro’s
Disciplinarum Libri] was the basis of Augustine’s encyclo-
pedia on the artes, although it cannot be demonstrated in
detail owing to our defective knowledge of both.’11 Dyroff
had interpreted matters slightly differently, but without
much reducing Varro’s role. He derived De Ordine 2. 12.
35 ff. from a neo-Pythagorean source mediated by Varro:
Augustine mentioned Varro twice and praised Pythagoras.12
Ritschl’s thesis, that Varro’s lost Disciplinarum Libri were the
source both for Augustine’s cycle of the liberal arts and for
later Latin encyclopaedists such as Martianus Capella and
Cassiodorus was not seriously challenged until 1984, when
Ilsetraut Hadot published her Arts libéraux et philosophie
dans la pensée antique. Varro’s Disciplinarum Libri and their
survival were an issue again. One of the aims of Hadot’s
book was to establish that the cycle of the seven liberal arts
came into being, not during the fifth century a d in the Latin
West, as had been supposed, by a subtraction of two ‘artes’

9
Begun in 396, finished in 426. See Doct. Chr. 2. 29. 46 for medicine
contrasted with ‘superstitio’; 2. 31. 49 for arithmetic; 2. 31. 48 ff. for
dialectic; 3. 29. 40 for grammar; 4. 3. 5 for rhetoric.
10
Hadot, Arts libéraux et philosophie, 136.
11
H. Hagendahl, Augustine and the Latin Classics, 2 vols. (Göteborg,
1967), i. 267. Cf. ii. 590.
12
A. Dyroff, ‘Über Form und Begriffsgehalt der augustinischen Schrift
De Ordine’, in M. Grabmann and J. Mausbach (eds.), Aurelius Augusti-
nus: Festschrift der Görresgesellschaft zum 1500. Todestage des Heiligen
Augustinus (Cologne, 1930), 15–62. Dyroff completely rejected a Neo-
platonic source: ‘Vor vielem sicher ist, daß in De Ordine sich nicht die
mindeste sichere Spur von Neuplatonismus vorfindet, obwohl genug
Gelegenheit dazu war. . . . In der Trilogie von 386 knüpft er an Cicero und
Varro an . . .’ (p. 47).
74 Danuta R. Shanzer
from Varro’s canon of nine, but considerably earlier in Greek
Neoplatonic literary circles, and that it was transmitted
to the West through Augustine. She sought to undermine
Varro’s importance as a source for the seven-plus-two, by
demolishing the ‘thèse de Ritschl’.13 She argues (1) that
Augustine did not use the Disciplinarum Libri, of whose
shape and content we can in any case know little or nothing,
and (2) that his descriptions of the ‘artes’ in De Ordine,
because they owe much to Neoplatonism, must be derived
from a source later than the mid-first century b c. Her
programme has been characterized as a ‘generoso sforzo di
esorcizzare l’ombra del grande erudito romano’, namely
Varro.14
Before we enter into the detail of Hadot’s study, it may be
helpful to outline the course that my own argument will take.
We begin with a problem: the documentation of Augustine’s
lost work on the liberal arts, the identification of its
source(s), and the reasons behind his gradual disenchant-
ment with the ‘disciplinae’. The traditional opinion, as we
have seen, is that he worked from Varro, but this has been
challenged by Hadot. The second, negative section of my
essay involves detailed criticism of her position. Augustine,
I shall argue, did indeed confront and use the Disciplinarum
Libri. His picture of the ‘artes’ in De Ordine is Platonic (rather
than necessarily Neoplatonic), and there is independent evi-
dence to justify the ascription of such imagery to Varro.
Unlike Hadot, I believe that we can know something of
Varro’s Disciplinarum Libri, that a shadow outline of this
work can be drawn through comparative reconstruction
from later authors and the indirect tradition. In the third,
more positive section of my essay I bring previously un-
appreciated fifth-century evidence to bear on the problem of
its literary form. Finally, in a fourth section, I offer a revised
narrative of Augustine’s engagement with the ‘disciplinae’ and
of Varro’s role in the formation of the medieval canon.

13
Hadot, Arts libéraux et philosophie, 156 ff.
14
U. Pizzani, ‘L’Enciclopedia agostiniana e i suoi problemi’, in Atti del
Congresso internazionale su S. Agostino nel XVI centenario della conversione,
Roma 15–20 settembre 1986, Studia Ephemeridis ‘Augustinianum’ 24
(Rome, 1987), 331–61, at 337.
 Augustine’s Disciplines 75
We shall be able to confirm that the manifestly pagan form
of the Disciplinarum Libri, complete with ανοδο (‘journey
upwards’) and talking Muses, was ultimately unacceptable
to Augustine. In De Ordine, at an early stage in his work on
the ‘disciplinae’, Varro’s influence was close to the surface,
absorbed, but faintly visible. By the time of De Doctrina
Christiana, however, the Roman scholar and his personified
‘artes’ had become too compromised and dangerous. Some of
the African bishop’s squeamishness about personifications
can also be detected in the response of various late antique
Gallic authors to what is almost certainly the fingerprint
of the Disciplinarum Libri. But Varro was ultimately to win,
for in the late fifth century Martianus Capella cheerfully—
though not, as we shall see, unselfconsciously—mixed per-
sonifications, talking females, and a journey to heaven
into the heady allegorical brew of De Nuptiis Philologiae et
Mercurii. And it was his personified liberal arts that would
fill the medieval imagination. Varro’s ghost, one might
caution, should not be too lightly or prematurely laid.

2. h a d ot’s att e m pt e d r e f utat i o n

o f r i t s c h l’s t h e s i s
De Ordine is the linchpin of Hadot’s theory. It was written in
Latin, it had an ascent motif, an apparent canon of seven,
and a strong whiff of philosophy. According to her, it reflects
the establishment of the Neoplatonic canon of seven ‘artes’
as propaedeutic to philosophy, and transmitted it to the
West. It cannot have anything to do with a first-century-b c
encyclopaedist.
Hadot effectively dissects Dyroff’s argument for neo-
Pythagoreanism,15 but her main error (an inappropriate
application of Occam’s razor)16 already appears in her criti-
cism of him: ‘Instead of assuming a mosaic of influences,
we must first look for the most economical explanation.’17

15
Hadot, Arts libéraux et philosophie, 132–5.
16
Pizzani, ‘L’Enciclopedia agostiniana’, 341.
17
Hadot, Arts libéraux et philosophie, 135: ‘Au lieu de supposer une
mosaïque d’influences, il faut d’abord chercher l’explication la plus
économique.’
76 Danuta R. Shanzer
For her, the source for Augustine is ‘incontestably’ Neo-
platonic:18 the tenor of the passage on the ‘artes’ in Book 2 of
De Ordine and the words ‘redeundum’, ‘fugiendum’, ‘pro-
gressus’, and ‘regressus’ at 2. 11. 31 all point to Porphyry’s
lost De Regressu Animae.19
Granted, the flavour of Augustine’s treatment of the
liberal arts in De Ordine 2. 12. 35–15. 43 is clearly philo-
sophical and Platonic,20 what with the dominance of number,
the progression from sensible to intelligible, the imagery of
ascent, and many other details, such as the eye of the soul
(Republic 7. 533d).21 The arts are seven, with a semi-ellipsis
of arithmetic, in the order: grammar, dialectic, rhetoric,
music, geometry, astronomy. But in no case is the philo-
sophical material linked in any organic way to the liberal

18
Hadot, Arts libéraux et philosophie, 132.
19
Ibid. 103 and also 106, citing a weak parallel (not about the liberal
arts) from Civ. 10. 29. Ord. 2. 11. 31: ‘Hic genere posito quod animal
dictum est, videmus additas duas differentias, quibus credo admonendus
erat homo, et quo sibi redeundum esset, et unde fugiendum. Nam ut
progressus animae usque ad mortalia lapsus est; ita regressus esse in
rationem debet.’ (‘Once a genus that is called “animal” has been
established, we see that two differentiae have been added, by which I
believe man was supposed to be admonished both whither he ought to
return and whence he ought to flee. For just as the progression of the
soul to mortal things is a slide downwards, so its return ought to be
towards reason.’)
20
Hadot, Arts libéraux et philosophie, 110, shows the parallels between
Augustine’s section on grammar and the Philebus.
21
See in particular Ord. 2. 14. 39: ‘Hinc se illa ratio ad ipsarum rerum
divinarum beatissimam contemplationem rapere voluit. Sed ne de alto
caderet, quaesivit gradus, atque ipsa sibi viam per suas possessiones
ordinemque molita est. Desiderabat enim pulchritudinem, quam sola et
simplex posset sine istis oculis intueri; impediebatur a sensibus. Itaque in
eos ipsos paululum aciem torsit, qui veritatem sese habere clamantes,
festinantem ad alia pergere, importuno strepitu revocabant.’ (‘From here
that Reason wished to snatch herself away towards most blessed contem-
plation of the divine things themselves. But lest she fall from on high, she
sought steps, and she herself built a path and an order for herself through
her own possessions. For she longed for the sort of beauty that by itself in
its simple form can be seen without these [mortal] eyes, but she was
impeded by the senses. And so she turned her gaze for a while towards
those who by proclaiming that they have the truth call back with
importunate noise the one who hastens to reach other things.’)
 Augustine’s Disciplines 77
22
arts. More importantly, there is no evidence that De
Regressu contained anything about the ‘artes’ at all.23 What
we have here are vignettes of the matter of the ‘discipli-
nae’—e.g., allusions to vowels, continuants, and stops for
grammar24—with philosophizing material around them. One
can mark where the liberal art starts and the philosophy
begins.
When one writes a narrative commentary on this passage,
as Hadot does, the names of many texts and authors pepper
the page:
2. 11. 30: distinguendi et conectendi potestas (‘able to dis-
tinguish and connect’): cf. Phaedrus 266b.25
2. 11. 31: homo est animal rationale mortale (‘man is a
rational and mortal animal’): ‘Cette définition est en effet très
ancienne, mais l’interprétation qui en est faite est essentiel-
lement néoplatonicienne.’26
2. 11. 31: rationale vs. rationabile (‘capable of reasoning’ vs.
‘reasonable’).27
2. 12. 36: Plato, Philebus 18B6–D12.28
2. 13. 38: tripartition of dialectic: Porphyry’s Isagoge (ed.
Busse, 1. 5–6), not Antiochus of Ascalon (ed. Witt, 36).29

22
Hadot, Arts libéraux et philosophie, 128–9, identifies Neoplatonic
material, but none of it can be traced to Neoplatonic texts that discuss the
‘artes’. Her statement that ‘tout le contexte se rapportant à la discipline
grammaticale . . . selon tout vraisemblance, se fonde sur le même texte
néoplatonicien que toute l’argumentation concernant le retour de l’âme
dans sa patrie, effectué par des moyens intellectuels’ (p. 109 n. 33) has no
basis in fact.
23
See the fragments in J. Bidez, Vie de Porphyre, le philosophe néo-
platonicien (Hildesheim and New York, 1980), 27*–44*. Porphyry, Vita
Pythagorae 46–7, which describes the purification of the soul, does indeed
mention the place of the µαθµατα (‘disciplines’) in the process, but only
en passant.
24
Ord. 2. 12. 36.
25
Hadot, Arts libéraux et philosophie, 105.
26
Ibid. 106.
27
Ibid. 107.
28
Ibid. 110, following P. Courcelle, Les Lettres grecques en Occident
de Macrobe à Cassiodore, 2nd edn. (Paris, 1948), 156; see now Pizzani,
‘L’Enciclopedia agostiniana’, 343, for Cicero, De Re Publica 3. 2. 3.
29
See Hadot, Arts libéraux et philosophie, 115–16.
78 Danuta R. Shanzer
2. 14. 41: Reason wove the verbal arts with number: Proclus,
In Prim. Euclid. Elem. Libr. Comm (ed. Friedlein, 24. 21–7)
says that grammar and rhetoric qua sciences are founded on
mathematics.30
2. 14. 39: progress away from the sensible to the intelligible:
Plotinus, Enneads 1. 6; Porphyry, Vita Pythagorae 46–7.31
But none of these proposed sources is ‘hard’. Many are
fuzzy, if not fuzzier (e.g., Proclus, who is much later than
Augustine). Indeed, ‘sources’ is too imprecise a term in this
case, where one may have to distinguish between materials
that Augustine could have used directly (none extant for
the matter in hand!) and ultimate sources that could be far
earlier. Even if the latter are clearer, they are often so widely
disseminated that pinning down a distinctive avenue of
transmission is impossible. Most items Hadot cites are not
clear sources, but similia. The philosophical material is
largely colourless and cannot be pinned down to any one
author or transmitter. Hadot has not proved her unitary
source.32 The philosophizing ‘stuff’ could be Augustine’s,
and the liberal arts someone else’s, or indeed various per-
sons’. Whenever one works on Augustine’s sources, where
they are extant, ‘mosaics’ and contamination are the name of
the game.33
We now proceed to some of the particulars of Hadot’s case
against Ritschl.
a. Varro Dismissed Prematurely
Let Platonists worry about the Platonic material. Our
immediate concern is with Varro, whom Hadot aims to
banish from De Ordine. There is a series of quick points to be
made in Varro’s defence. First, his name (and it is the only
author’s name in the passage other than Cicero’s34) frames
the account of the ‘disciplinae’ and philosophy in Ord. 2. 12.
30
Hadot, Arts libéraux et philosophie, 117.
31
Ibid. 118 says that it comes from De Regressu Animae.
32
Pizzani, ‘L’Enciclopedia agostiniana’, 337; J. Doignon, Dialogues
philosophiques: De Ordine (= Oeuvres de Saint Augustin 4/2) (Paris, 1997),
29–30.
33
Doignon, Dialogues philosophiques: De Ordine, 30: ‘l’éclectisme
d’Augustin’.
34
Ord. 2. 7. 45.
 Augustine’s Disciplines 79
35 and Ord. 2. 20. 54. ‘Quant aux citations de Varron, je les
35

tiens pour occasionnelles’,36 is Hadot’s response. But why

introduce names gratuitously? Second, there is a bona fide
fragment of Varro’s De Grammatica at Ord. 2. 12. 37.37 Third,
there is what is very probably a Varronian division of sound
in Ord. 2. 14. 39, which also appears in a Varronian context
in De Doctrina Christiana.38 These secure attributions by

35
Ord. 2. 12. 35: ‘Quibus duobus repertis nata est illa librariorum et
calculonum professio velut quaedam grammaticae infantia, quam Varro
litterationem vocat; Graece autem quomodo appelletur, non satis in
praesentia recolo.’ (‘Once these two had been discovered, there was born
that business of scribes and accountants, an infant stage of grammar as it
were, that Varro called litteratio; how it is called in Greek I don’t really
remember at the moment.’) Ord. 2. 20. 54: ‘Res enim multum necessaria
mihi prorsus exciderat, quam in illo viro—si quid litteris memoriae
mandatis credendum est; quamvis Varroni quis non credat?—mirari et
paene cotidianis, ut scis, ecferre laudibus soleo . . .’ (‘A very important fact
that I am wont to admire in that man [sc. Pythagoras] and to praise almost
every day entirely slipped my mind—if one should believe in words
entrusted to memory at all, although who would not believe Varro?’)
36
Hadot, Arts libéraux et philosophie, 133.
37
‘Poterat iam perfecta esse grammatica sed, quia ipso nomine profiteri
se litteras clamat unde etiam Latine litteratura dicitur factum est, ut
quidquid dignum memoria litteris mandaretur, ad eam necessario per-
tineret.’ (‘Grammar was already able to be perfect, but because by her very
name she proclaims to profess letters—whence even in Latin she is called
litteratura—it came about that whatever thing worthy of record was
entrusted to writing necessarily fell under her purview.’) Cf. Martianus
Capella, De Nuptiis Philologiae et Mercurii 3, para. 229, in the numbering
of Kopp’s edition (hereafter ‘K’).
38
See Ord. 2. 14. 39: ‘Intellexit nihil aliud ad aurium iudicium perti-
nere quam sonum eumque esse triplicem, aut in voce animantis aut in eo,
quod flatus in organis faceret, aut in eo, quod pulsu ederetur.’ (‘She under-
stood that nothing other than sound pertained to the judgement of the
ears and that it was threefold, consisting in the voice of a living being or in
the sound that the breath makes in instruments or in the sound that is
created by striking.’) Cf. Doct. Chr. 2. 17. 27: ‘Aut enim voce editur, sicuti
eorum est qui faucibus sine organo canunt; aut flatu, sicut tubarum et
tibiarum; aut pulsu, sicut in citharis et tympanis, et quibuslibet aliis quae
percutiendo canora sunt.’ (‘For it is either produced by the voice, such as
of those who sing with their throats without an instrument; or by blowing,
as is the case with trumpets or pipes; or by striking, as is the case with
citharae and drums, and any other instruments that are tuneful when
struck.’) Here the context is clearly Varronian, though downplayed by
Hadot, Arts libéraux et philosophie, 120. She does the same with the allu-
sions to Varro at Ord. 2. 20. 54 in connection with Pythagoras (pp. 130, 133).
80 Danuta R. Shanzer
name and the likely citations of Varronian material pertain-
ing to the ‘disciplinae’ in this passage cannot be dismissed.
Not one of Hadot’s Neoplatonic ‘sources’, let alone the
shadowy De Regressu, can be as well authenticated or shown
to be vestigially present in this way.

b. Literary Frame and Colouring of Augustine’s Account of

the ‘Disciplinae’
Hadot argues that the motifs of abstraction from the cor-
poreal and the ascent have to be Neoplatonic, not Varronian.
In 1991 I translated and commented on Licentius’ Carmen
ad Augustinum 1–15:
Arcanum Varronis iter scrutando profundi
Mens hebet adversamque fugit conterrita lucem.
Nec mirum; iacet omnis enim mea cura legendi
Te non dante manum et consurgere sola veretur.
Nam simul ut perplexa viri compendia tanti
Volvere suasit amor sacrosque attingere sensus,
Quis numerum dedit ille tonos mundumque Tonanti
Disseruit canere et pariles agitare choreas,
Implicuit varia nostrum caligine pectus
Induxit animo rerum violentia nubem.
Inde figurarum positas in pulvere formas
Posco amens aliasque graves offendo tenebras:
Ad summam astrorum causas clarosque meatus,
Obscuros quorum ille situs per nubila monstrat.
Sic iacui nutans . . .
While investigating profound Varro’s secret path, my mind is
dulled, and terrified flees the light it encounters. This is no sur-
prise, for all my interest in the task of reading lies inert if you
fail to lend me a helping hand: it is too insecure to rise on its
own. For as soon as love persuades me to unroll the twisted
compendia of the great man and to venture to approach their
sacred significance—by which proportions he gave the tones of
numbers and taught that the universe sings for the Thunderer
and quickens regular dances—the strenuousness of the subject-
matter enfolds my breast in shifting darkness and has brought a
cloud over my spirit. Thence in my distress I ask for forms of
figures drawn in the sand, and collide with other heavy
shadows—in short, the causes of the stars and their shining
paths, whose invisible location he points out through the clouds.
Thus I lay wavering . . .
 Augustine’s Disciplines 81
I concluded that it drew extensively on De Ordine and con-
tinued the dialogue, that its Varro was indeed Varro of Reate,
that three mathematical ‘disciplinae’ are alluded to, and
finally that the ‘arcanum Varronis iter’ was a compendium
that featured at least music, geometry, and astronomy, as well
as imagery of an upwards journey towards light.39 Evidence
from Licentius suggests that Varro’s work on at least three
mathematical ‘disciplinae’ featured a journey and an ascent.
Thus, pace Hadot, not all ascents had to be Neoplatonic.40
Indeed, it is most surprising not to see Plato’s Cave from
Republic 7 mentioned by her, a text that had all the requisite
elements: an ascent, the terror, and bedazzlement at the
light41 (compare line 2 of Licentius’ poem), a discussion of
all four mathematical disciplines (and dialectic) in which
their ability to turn the faculties upwards is emphasized,42
and even a journey.43 Varro would certainly have known
fundamental texts of Plato.44
39
D. Shanzer, ‘Licentius’s Verse Epistle to Augustine’, REAug 37
(1991), 110–43.
40
Hadot, Arts libéraux et philosophie, 118, assumes that the ‘gradus’
imagery refers to the Neoplatonic flight of Daedalus. This is simply
wrong. Ascent by ‘gradus’ is not the same as flight. The same applies to
Retr. 1. 5. 6: ‘quibusdam quasi passibus certis’ (‘by certain definite steps as
it were’) and Ep. 101. 3: ‘quibusdam quasi gradatis itineribus’ (‘by certain,
as it were, gradated progressions’). W. Hübner, ‘Die “artes liberales” im
zweiten Buch von De Ordine’, in Charisteria Augustiniana Iosepho Oroz
Reta dicata (Madrid, 1994), 317–44, at 323, characterizes the word
‘gradus’ as Varronian. He is, however, prepared to concede a rather mixed
flight and climbing metaphor to Augustine. For subsequent development
of ‘gradus’ imagery in Augustine, see Pollmann below, 225–31.
41
Rep. 516c and e. Augustine clearly had access to the idea indirectly;
see Sol. 2. 25. 1.
42
Rep. 522c, 525d, 526e, 527b, 529a, 531c, 531d (for the arts’ relation-
ship with one another); for dialectic see 532a, 533d.
43
Rep. 532e.
44
He had studied with Antiochus of Ascalon in Athens at the Academy
between 84 and 82. See Cicero, Ep. ad Atticum 13. 19. 3; Academica Pos-
teriora 1. 3. 12. Also F. Della Corte, Varrone: Il terzo gran lume romano
(Genoa, 1954), 173 n. 10. W. Burkert, ‘Cicero als Platoniker und Skep-
tiker’, Gymnasium 72 (1972), 175–200, at 198, notes that Cicero never
mentions the simile of the Cave. But there is no doubt that he had direct
access to the Republic. See A. A. Long, ‘Cicero’s Plato and Aristotle’, in
J. F. G. Powell (ed.), Cicero the Philosopher (Oxford, 1995), 37–61, at 43–5.
Also T. B. deGraff, ‘Plato in Cicero’, Classical Philology 35 (1940),
143–55, at 146 and 150 nn. 66 and 67 for Gyges and Er.
82 Danuta R. Shanzer
A work of the late fifth century, De Statu Animae of Clau-
dianus Mamertus, likewise alludes to Varro in a context
where three mathematical disciplines appear, and speaks of a
progress from the visible to the invisible:
Marcus Varro, sui saeculi peritissimus et teste Tullio omnium sine
dubitatione doctissimus, quid in musicis, quid 〈in arithmeticis〉,
quid in geometricis, quid in philosophoumenon45 libris divina
quadam disputatione contendit, nisi ut a visibilibus ad invisibilia,
a localibus ad inlocalia, a corporeis ad incorporea miris aeternae
artis modis abstrahat animum et in corpora, hoc est in adversa sibi
dilapsum sui compotem faciat?46
What does Marcus Varro, the most skilled of his age, and on
Cicero’s testimony, without doubt, the most learned, seek to gain
by an almost divine disputation in his musical writings, 〈his
arithmetical writings〉, his geometrical ones, in the books of
philosophical matters, if it not be to draw away the soul in certain
wondrous ways that belong to an eternal art from the visible to
the invisible, from what has place to what has no place, from the
corporeal to the incorporeal and to make it master of itself, even
though it has slipped into bodies, that is to say into things that are
adverse to it[s nature].
It has been argued (by Bömer,47 Theiler,48 and D’Ales-
sandro49) that this passage is a mosaic of purely Neoplatonic
or Augustinian reminiscences. But again, why would
Claudianus gratuitously drag in Varro in the context of the
disciplines, if the name were not significant?50 The allusion

45
Ritschl, ‘De M. Terentii Varronis disciplinarum libris commen-
tarius’, 371, hesitated to connect this with De Dialectica, and suggested De
Philosophia instead.
46
Claudianus Mamertus, De Statu Animae 2. 8 (CSEL 11.130.2).
47
F. Bömer, Der lateinische Neuplatonismus und Neupythagoreismus und
Claudianus Mamertus in Sprache und Philosophie (Leipzig, 1936), 158,
believes that he had no access to Varro.
48
W. Theiler, ‘Porphyrios und Augustin’, in Forschungen zum Neu-
platonismus (Berlin, 1966), 160–251, at 165 n. 15, warns that the influence
of Varro should not be overestimated and that Claudianus is not using
Varronian terminology.
49
P. D’ Alessandro, ‘Agostino, Claudiano Mamerto, Cassiodoro e i Dis-
ciplinarum libri di Varrone’, in MOUSA: Scritti in onore di Giuseppe
Morelli (Bologna, 1997), 357–70, at 368.
50
Pizzani, ‘L’Enciclopedia agostiniana’, 338.
 Augustine’s Disciplines 83
would be pointless. As for the Varronian origin of the ascent
motif: nihil obstat.51
No one, to my knowledge, has cited the next chapter (2. 9)
of De Statu Animae:
Aurelius Augustinus et acumine ingenii et rerum multitudine et
operis mole veluti quidam Chrysippus argumentandi virtute aut
Zenon sensuum subtilitate aut Varro noster voluminum magni-
tudine, et qui profecto talis natura adtentione disciplinis exstiterit,
ut non immerito ab istis corporalibus nostri saeculi Epicureis aut
Cynicis spiritalis sophista dissenserit, libro ad Hieronymum de
origine animae sic pronuntiat . . .52
Aurelius Augustinus by the sharpness of his native wit, the multi-
tude of subjects, and the bulk of his writing, like a certain Chry-
sippus for his strength in argument, a Zeno for the subtlety of his
senses, or our Varro through the size of his œuvre, and one who
indeed was such in nature, inclination, and training in the liberal
arts that not without reason did he, a spiritual wise man, disagree
with those corporeal Epicureans or Cynics of our age, in his book
to Jerome about the origin of the soul, so pronounces . . .
Here Augustine is triply compared to a learned trio of
Chrysippus, Zeno, and Varro, with respect to his ‘acumen’
and ‘virtus argumentandi’, his ‘multitudo rerum’ and ‘sub-
tilitas’, his ‘operis moles’ and ‘voluminum magnitudo’. In
the line-up of the final triad of ‘natura’, ‘adtentione’, and
‘disciplinis’, the last item, ‘disciplinae’, is appropriately
associated with Varro. Would Claudianus drag in Varro
as an irrelevance twice?53 We may well need to revise our
customary interpretation of Sidonius’ comparison of Varro
and Augustine: not the pagan antiquarian and his Christian

51
Della Corte, Varrone, 239 n. 4, goes further, comparing the passage to
Retr. 1. 6 and concluding that the idea of the ascent must be Varronian.
52
CSEL 11.133.10 f.
53
Hadot, Arts libéraux et philosophie, 166: ‘Nous voyons donc que le fait
qu’un auteur tardif cite Varron et même une oeuvre précise de celui-ci, ne
garantit d’aucune manière que l’auteur ait eu une connaissance directe de
cette oeuvre.’ She then cites Fontaine on proximate and ultimate sources.
She misunderstands the application of the principle. Fontaine is not
saying that later writers cite authors out the blue: he is saying that some-
one like Macrobius will steal material on Republican Latin from Gellius
and cite only the ultimate source. If a later author cites Varro, Varro is still
the point. The question is how, if at all, the allusion was mediated.
84 Danuta R. Shanzer
opponent of the City of God, but perhaps two en-
cyclopaedists.54
The order of Varro’s disciplines may be unclear, and
citations with precise attributions non-existent, but many
more shadowy traces can be discerned by comparative
reconstruction. Using this method, one can draw some
conclusions not only about what the work comprised, but
also about its literary shape.
c. Number of Books and of ‘disciplinae’ in Varro’s
Disciplinarum Libri (also known as Musae)
Hadot claims that we do not know whether the nine books
of Varro’s Disciplinarum Libri each contained one of the
nine different disciplines.55 She fails, however, to note
Ritschl’s discussion of this very question56 and the testimony
of Vitruvius57 that nine disciplines were treated: ‘item
Terentius Varro de novem disciplinis, unum de architectura’
(‘likewise Terentius Varro on 〈the?〉 nine disciplines, one on
architecture’). With clear evidence for nine disciplines and
also for nine books, the default assumption must be that
one book was devoted to each discipline.58 But there may be
new evidence for the distribution of the ‘disciplinae’ over
nine books. To unearth it, we have to travel back in time
and do some research into the conventions of titulature of
ancient miscellaneous, encyclopaedic, and multi-book works,
specifically those that were known as ‘Muses’.
Μου̃σαι (‘Muses’) may have been the title of a work of
Heraclitus.59 Unfortunately we know nothing of its format,
but in the Greek world the nine books of Herodotus were

54
See Sidonius Apollinaris, Ep. 2. 9. 5: ‘Similes scientiae viri, hinc
Augustinus hinc Varro lectitabantur’ (‘Men of similar knowledge, on one
side Varro on the other Augustine, were read’).
55
Hadot, Arts libéraux et philosophie, 157.
56
Ritschl, ‘De M. Terentii Varronis disciplinarum libris commen-
tarius’, 356.
57
Vitruvius, De Architectura 7, praef. 14.
58
More than ‘quelque probabilité’ as Hadot, Arts libéraux et philoso-
phie, 157, suggests.
59
See A. Ernout and L. Robin, Lucrèce: De Rerum natura, 2nd
edn. (Paris, 1962), on Lucretius 1. 657, citing Plato, Sophistes 242d
and Diogenes Laertius 9. 12.
 Augustine’s Disciplines 85
60
also known as ‘the Muses’. F. Oomes, however, in his TLL
article on ‘musa’ drew attention to various passages in which
he suggested that ‘Musae’ in Latin was the title of a mis-
cellaneous work.61 Gellius had attested such a piece by
Aurelius Opilius in Noctes Atticae 1. 25. 17 and in the preface
to the same work.62 So too did Suetonius, De Grammaticis
6. 2: ‘composuitque variae eruditionis aliquot volumina, ex
quibus novem unius corporis. Quae quia scriptores ac poetas
sub clientela Musarum iudicaret, non absurde et fecisse et
inscripsisse se ait ex numero divarum et appellatione.’ (‘And
he wrote a number of books reflecting different sorts of
learning, of which nine [were joined] in one volume. These,
because he judged writers and poets to be part of the clientela
of the Muses, quite reasonably he said that he had written
and titled after the number and the names of the goddesses.’)
Among the passages mentioned by Oomes was Cicero,
Academica Posteriora 1. 2: ‘Silent enim diutius Musae
Varronis quam solebant, nec tamen istum cessare sed celare
quae scribat existimo.’ (‘The Muses of Varro are quiet for
longer than they are wont, but I do not think that he has
stopped, but that he is hiding what he writes.’) W. Buchwald
took Oomes’ argument further and suggested that Cicero
had alluded to a (then unfinished) work of Varro’s known to
have been nine books long, the Disciplinarum Libri.63 They
were the ‘magnum opus in manibus’, not De Lingua Latina.64
60
Lucian, Herodot. 1: αλλ αγωνιστν  Ολυµπων κα παρεχεν εαυτν
α˛δων τὰ στορα κα κηλ!ν τοὺ παρ"ντα αχρι του̃ Μούσα κληθ#ναι
τὰ ββλου αυτου̃, %νν&α κα αυτὰ ο'σα (‘But he presented himself as a
champion of the Olympians, singing his histories and beguiling also those
present with the result that his books were called the Muses, since they too
were nine.’)
61
F. Oomes, ‘musa’, TLL 8. 1692. 7.
62
‘Aurelius Opilius in primo librorum quos “Musarum” inscripsit’;
Noctes Atticae, praef. 5: ‘Namque alii “Musarum” inscripserunt’.
63
W. Buchwald, ‘Musae Varronis’, Museum Helveticum 23 (1966),
215–17.
64
See Della Corte, Varrone, 263, for their dating in 47/5. The Dis-
ciplinarum Libri come after them in Jerome’s catalogue, and Della Corte
dates the completion of Book 8 to 33 (presumably following Ritschl, ‘De
M. Terentii Varronis disciplinarum libris commentarius’, 400, who sees
Pliny, Naturalis Historia 29. 4. 65 as an allusion to Book 8) and of Book 9
possibly to somewhat later, in 31. He rightly notes that work on the
Disciplinarum Libri must have been started considerably earlier.
86 Danuta R. Shanzer
To the discussion he added Arnobius, Adversus Nationes
6.1: ‘ridet temporibus priscis Persas fluvios coluisse,
memorialia ut indicant scripta . . . pro Marte Romanos
hastam, Varronis ut indicant Musae.’ (‘He derides the fact
that the Persians worshipped rivers in ancient times, as the
written records indicate . . . and that the Romans wor-
shipped a spear representing Mars, as the Muses of Varro
indicate.’) This passage has a close parallel in the work of
Clement of Alexandria, who also attributes the cult of the
Mars spear to a there unspecified work of Varro.65
Buchwald’s suggestion that the Disciplinarum Libri may
also have been known as the ‘Musae’ was rejected by H.
Dahlmann.66 He thought that there was no evidence that the
Arnobius testimonium depended on Clement, and that there
was no need for the two to go back simultaneously to an
authentic work of Varro’s. Both testimonia, according to him,
were dependent on an anonymous intermediate source that
had not specified in which work the Varronian information
appeared. This origin was supposed to explain the disparity
between Musae Varronis and Ουάρρων ( σνγγραφεύ. Hence
Dahlmann concluded that the expression ‘Musae Varronis’
in the Academica was not authentic evidence for a title of a
work: it merely meant ‘literarische Produktion’.67
Dahlmann’s argument is not convincing, and various
points should be made in response. ‘Musae’, as ‘literary pro-
duction’, is not flat or neutral, but poetic, coy, precious,
or teasing. Arnobius has no such attitude to his source, nor
was Varro’s work verse, as was the ‘musa Lucretia’ (note the
singular of the noun and the adjective rather than the plural
of the noun with a genitive) to which Arnobius once
alluded.68 Had Arnobius sought to undermine Varro, he
might have spoken of his ‘fabulae’. The tone of the allusion

65
Clement of Alexandria, Protrepticus 4. 46: Ουάρρων ( συγγραφεύ.
66
H. Dahlmann, ‘Silent diutius Musae Varronis quam solebant’, in
J. Collart (ed.), Varron: Grammaire antique et stylistique latine (Paris,
1978), 85–9. Also more recently by Hübner, ‘Die “artes liberales” im
zweiten Buch von De Ordine’, 337, though he wistfully characterizes
Buchwald’s suggestion as ‘verlockend’.
67
Dahlmann, ‘Silent diutius Musae Varronis quam solebant’, 86.
68
Arnobius, Adversus Nationes 3. 10, using the singular.
 Augustine’s Disciplines 87
69
is neutral, and ‘Musae’ is most probably a title. Further-
more, had Arnobius’ source not specified what work some
prose citation came from, he would hardly have spon-
taneously attributed it to ‘Varronian Muses’. Συγγραφεύ, the
term used by Clement, is the regular way of designating a
fact-collector or prose-writer. It is far more probable that the
source, whatever it was, attributed the quotation to the
Musae Varronis, as a title, and that Clement simply dropped
the title of the work altogether.
Dahlmann likewise sought to dispose of the phrase
‘Musae Varronis’ in Cicero’s Academica by arguing that it
referred not to the title of a work, but to Varro’s ‘literarische
Schriftstellerei’.70 The word occurs in a speech of Atticus’.
Dahlmann sees the tone of the passage as ‘erfüllt von
Höflichkeit und Bewunderung’.71 This is possible.72 But a
more probing reading might see here a double entendre, where
the Muses are both Varro’s inspiration and even more aptly
the patronesses of the books also known as the Musae.73
Atticus might have been making a gentle joke by personify-
ing Varro’s learned ladies. A little later the books of Varro
become hosts for Cicero.74

69
Arnobius uses ‘indicare’ both for people and for books as sources, see
Adversus Nationes 6. 11: ‘memorialia ut indicant scripta’; 2. 71: ‘ut indicat
supputatio’; 5. 18: ‘Sextus Clodius indicat’; 6. 22: ‘Posidippus . . . indicat’.
70
Dahlmann, ‘Silent diutius Musae Varronis quam solebant’, 88.
71
Ibid. 89.
72
Cicero regularly used ‘Musae’ in a figurative sense. See De Finibus
5. 49: ‘variis avido satiatus pectore Musis’ (‘the eager breast satiated by the
Muses’—his own translation of Homer); Ep. ad Familiares 16. 10. 2 (to
Tiro): ‘Tu Musis nostris para ut operas reddas’ (‘Prepare to pay attention
to our Muses’). But there are also uses that stay close to personification,
e.g., Fam. 1. 9. 23: ‘nam me ab orationibus diungo fere referoque ad
mansuetiores Musas, quae me maxime sicut iam a prima adulescentia
delectarunt’ (‘For I almost unyoke myself from speeches and take myself
back to the more gentle Muses that have given me the greatest pleasure
just as they already did in my first youth’). For the voice of the Muses, see
Orator 62: ‘Xenophontis voce Musas quasi locutas ferunt’ (‘They say that
the Muses spoke as it were in the voice of Xenophon’).
73
See Appendix below for a fuller discussion of the passage.
74
Academica Posteriora 1. 3. 9: ‘nam nos . . . tamquam hospites tui libri
quasi domum reduxerunt’ (‘For your books have led us like guests back
home’).
88 Danuta R. Shanzer
To summarize: there are good reasons for believing
that Varro wrote something whose title, official or unofficial,
was Musae, in nine books, each corresponding to one of the
‘disciplinae’. The evidence for such a view has long been
generally known, even if differently interpreted. Other
sources may yet enable us to say more about Varro’s lost
work.

3. n e w e v i d e n c e f o r va r r o’s m u s a e
Sidonius Apollinaris wrote a letter in 471 to his friend Nym-
phidius. In it he gave a laudatory account of Claudianus
Mamertus’ De Statu Animae (see above) while demanding
the immediate return of his own copy from Nymphidius. He
speaks of Claudianus as ‘revealing that what they call the
nine Muses are disciplines not women’ (‘novem quas vocant
Musas disciplinas aperiens esse, non feminas’).
Namque in paginis eius vigilax lector inveniet veriora nomina
Camenarum, quae propriam de se sibi pariunt nuncupationem.
Illic enim et grammatica dividit et oratoria declamat et arithmetica
numerat et geometria metitur et musica ponderat75 et dialectica
disputat et astrologia praenoscit et architectonica struit76 et metrica
modulatur. (Ep. 5.2)
For in his pages the attentive reader will find the more authentic

75
ponderat makes no sense as a verb for music. It is almost certainly
corrupt. If one compares a text such as the Anonymus de Constitutione
Mundi (PL 90.908C): ‘Geometria est immobilis magnitudinis doctrina
ponderatrix’ (‘Geometry is the subject that measures stationary mass’) or
the anonymous distich (H. Walther, Alphabetisches Verzeichnis der Versan-
fänge mittellateinischer Dichtungen (Göttingen 1969), no. 7273): ‘Gram
loquitur, Dia vera docet, Rhet verba colorat / Mus canit, Ar numerat, Geo
ponderat, Ast colit astra’ (‘Grammar speaks, Dialectic teaches the truth,
Rhetoric colors words, Music signs, Arithmetic counts, Geometry weighs,
Astronomy worships the stars’), it seems possible that a complex corrup-
tion left music with geometry’s verb. ‘Metitur’ may well be the obvious
and later corrective interpolation.
76
Architecture would have been useful for Ritschl, ‘De M. Terentii
Varronis disciplinarum libris commentarius’, 364. He reconstructed the
presence of architecture from her deliberate exclusion by Martianus (ibid.
367) and from Vitruvius, De Architectura 7, praef. 14. This passage but-
tresses his analysis.
 Augustine’s Disciplines 89
names of the Camenae, who provide their own titles for them-
selves.77 For there grammar draws distinctions, oratory declaims,
arithmetic counts, geometry measures, music weighs [sic], dialectic
disputes, astrology foretells, architecture builds, and metrics
modulates.
Care in interpretation is demanded. Someone, not Claudi-
anus, had personified the ‘disciplinae’ as Muses—that is,
women—not sober disciplines. But in Claudianus’ work are
found the real disciplines, each performing the action
appropriate to her sphere. This list of ‘disciplinae’ has noth-
ing to do with the actual content of De Statu Animae.78
Its purpose is panegyric; it is an inflated way of calling
Claudianus’ work learned.79 None the less, one may deduce
something of Sidonius’ canon of the liberal arts from it, and
it may be possible to determine whence it came. The first
sentence assembles a number of key concepts: the number
nine, the Muses, and a list of liberal arts, thereby proving the
connection between ‘Musae’ and ‘disciplinae’. But who was
the author of the false Muses? When this testimonium is
combined with that of Arnobius and Cicero, an obvious
answer emerges. Varro wrote something called the Musae.
The Muses are usually nine, and Varro is known to have used
this number.80 Sidonius provides sound evidence that in
some unspecified work the ‘Musae’ were associated with the
‘disciplinae’. Since Sidonius is recalling a canon from
memory, we need not be unduly bothered by the apparent

77
A. Loyen, Sidoine Apollinaire: Lettres, 3 vols. (Paris: 1960–70),
ii. 175: ‘dénomination approprié à leur personnalité’.
78
Sidonius had almost certainly not even read it. See Ep. 4. 3 for his
belated thanks.
79
Sidonius, when chided for not thanking Claudianus for the dedi-
cation of De Statu Animae, would eventually lay it on even thicker in
Ep. 4. 3. 5.
80
See Arnobius, Adversus Nationes 3. 38: ‘Novensiles Piso deos
esse credit novem in Sabinis apud Trebiam constitutos, hos Granius
Musas putat consensum adcommodans Aelio, novenarium numerum tradit
Varro, quod in movendis rebus potentissimus semper habeatur et
maximus’. (‘Piso thinks the Novensiles are nine gods who dwell in
Trebia among the Sabines, but Granius thinks them the Muses in agree-
ment with Aelius [Stilo], and Varro ascribes the number nine, because
it is always considered the greatest and most powerful in moving
things.’)
90 Danuta R. Shanzer
omission of Medicina and the substitution of Metrica (as
distinct from Grammatica or Musica).81
a. Varro as Common Source for the Personified ‘Disciplinae’
of Martianus Capella and Fifth-Century Gallic Writers
Ariston of Chios and Aristippus unflatteringly compared the
arts to Penelope’s (i.e., Philosophy’s) handmaidens; sub-
sequently Martianus Capella drew lively personifications of
the liberal arts.82 Otherwise there is no literary evidence for
personification of the figures. But neither the Hellenistic
writers nor Martianus had nine arts; nor did they identify
them with the Muses.83 So neither Ariston or Aristippus nor

81
This must be a pure slip, since metrics is never a separate liberal art,
but part of grammar or of music: i.e., the cross-over point between the
verbal and the mathematical ‘artes’ (as, e.g., in Augustine, Martianus
Capella, Cassiodorus, and Isidore). See M. Simon, ‘Zur Abhängigkeit
spätrömischer Enzyklopädien der Artes liberales von Varro’s Discipli-
narum libri’, Philologus 110 (1966), 88–101, at 97.
82
The ‘artes’ personified as handmaidens of Philosophy are attested for
Ariston of Chios by Stobaeus 3 (ed. Hense, 246.1): (Αρστων ( Χο τοὺ
περ τὰ %γκύκλια µαθµατα πονουµ&νου, αµελου̃ντα δ. φιλοσοφα,
/λεγεν (µοου ε0ναι το µνηστ#ρσι τ# Πηνελ"πη ο2 αποτυγχάνοντε
%κενη περ τὰ θεραπανα %γνοντο (‘Ariston of Chios said that those
who work on the liberal arts while neglecting philosophy are like the
suitors of Penelope, who, failing to gain her, consorted with the maids.’)
For Aristippus see Diogenes Laertius 2. 79 (students of the %γκύκλια
παιδεύµατα who stopped short of studying philosophy were Penelope’s
suitors who won the maids, but missed the mistress), noted by F. Kühnert,
Allgemeinbildung und Fachbildung in der Antike (Berlin, 1961), 6. L. M. de
Rijk, ‘ “Enkyklios Paideia”: A Study of Its Original Meaning’, Vivarium 3
(1965), 24–93, at 79–81, attributes the authentic form of the mot to Bion
of Borysthenes, as cited by pseudo-Plutarch, De Liberis Educandis 7D (10)
(ed. Bernardakis, 15. 26–16. 4).
83
Some of Martianus’ Muses’ songs derive colouring from the ‘artes’.
See 2. 118K for Urania and astronomy; 2.120K for Polymnia and music
and geometry; 2. 122K for Clio and rhetoric, dialectic, and grammar;
Hadot, Arts libéraux et philosophie, 146–7, for arithmetic in Erato’s song,
if ‘rationibus’ means ‘calculs’. There may also be hints of a connection
between the material of the ‘artes’ and ‘disciplinae’ in De Nuptiis138K, where
both the arts and the disciplines collect the books that Philologia vomits
up. Plutarch, Quaestiones Conviviales 746J alludes to Polyhymnia as a
Muse for Geometry. As H. Kees, ‘Musai’, in RE xvi.1 (1933), 684, points
out, the common pairing of Muses and arts is unknown in classical
Greece. See ibid. 729–30 for a table of arts (though not liberal arts) associ-
ated with different Muses.
 Augustine’s Disciplines 91
De Nuptiis Philologiae et Mercurii lies behind the allusion in
Sidonius’ letter to Nymphidius. Yet Claudianus Mamertus,
another Gallic writer of c.470 and the subject of Sidonius’
letter, also mentions personified ‘disciplinae’ and speaks
bitterly of the reception accorded the learned ladies, in his
letter to Sapaudus:
Video os Romanum non modo neglegentiae, sed pudori esse
Romanis, grammaticam uti quandam barbaram barbarismi et solo-
ecismi pugno et calce propelli, dialecticen tamquam Amazonem
stricto decertaturam gladio formidari, rhetoricam acsi grandem
dominam in angusto non recipi, musicen vero et geometricam atque
arithmeticam tres quasi furias despui, posthinc philosophiam
[atque] uti quoddam ominosum bestiale numerari.84
I see that the Roman tongue is not just something to neglect for
Romans, but something to be ashamed of. Grammar is being
repelled, as if she were some sort of she-barbarian, by the fist and
kick of barbarism and solecism; Dialectic is feared as if she were an
Amazon about to fight with drawn sword; Rhetoric as if she were
an over-sized mistress cannot be received in a tight space; Music,
Geometry, and Arithmetic are despised as if they were three
Furies; Philosophy next [and] is counted as something ominous
and bestial.
The passage clearly exploits and exaggerates the per-
sonifications ironically. In Claudianus’ time the liberal
arts are rejected as if they were the barbarous invaders of
the author’s native Gaul. Claudianus was not working
from Martianus,85 yet a number of features of Martianus’

84
Claudianus Mamertus, Ep. ad Sapaudum (CSEL 11.204.22 f.). The
text is clearly corrupt. One either excises ‘atque’ or supplies ‘astrologiam’
or ‘astronomiam’, a science that, coupled with philosophy in its natural
aspect, could well have been regarded as ‘ominosum’; cf., e.g., Augustine,
Ord. 2. 15. 42: ‘. . . astrologiam genuit, magnum religiosis argumentum
tormentumque curiosis’ (‘. . . generated astrology, a great proof to the
religious and torture to the [excessively] curious’). Charles Brittain,
however, has suggested to me that ‘philosophy’ (as distinct from the
propaedeutic dialectic) is correct, and that Claudianus addresses its poor
reception in his own times.
85
De Nuptiis was written in Africa c.470–80. See D. Shanzer, A
Philosophical and Literary Commentary on Martianus Capella’s De Nuptiis
Philologiae et Mercurii Liber 1 (Berkeley and Los Angeles, 1986), 1–28. It
is unattested in Gaul before the time of Gregory of Tours (Decem Libri
Historiarum 10. 18).
92 Danuta R. Shanzer
descriptions of the ‘artes’ could seem to have been parodied
by Claudianus. Dialectic, for example, says of herself:
Ni Varronis mei inter Latiares glorias celebrati mihi eruditio
industriaque suppeteret, possem femina Doricae nationis apud
Romuleae vocis examina aut admodum rudis aut satis barbara
reperiri. Quippe . . . Marci Terenti prima me in Latinam vocem
pellexit industria ac fandi possibilitatem per scholas Ausonas
comparavit.86
If the erudition and industry of my Varro were not available
among the glories of Latium, I, a woman of the Doric nation,
would be found rather uncultivated or indeed quite barbarous
among the throngs of those who speak Romulus’ tongue. To be
sure . . . the industry of Marcus Terentius first coaxed me to
Latin speech and provided the chance to speak in the Ausonian
schools.
Martianus’ work reflects a period when there was a tripartite
division of Greek, Roman, and barbarian, and when an edu-
cated woman from Greece might be rejected at Rome as an
undesirable alien. His Dialectic says that, although she is a
Greek, had it not been for the efforts of Varro, she might have
been considered a ‘barbara’ in Rome. To this we should
compare Claudianus’ Dialectica, a creature of barbaric and
Amazonian proportions, and his Grammatica, who is
repelled as if she were a ‘barbara’ by the punch and kick of
solecism. Martianus’ Rhetorica is a lofty and excessively
self-confident individual, ‘quaedam sublimissimi corporis
ac fiduciae grandioris’,87 who kisses people vulgarly and
audibly: ‘eiusque verticem deosculata cum sonitu (nihil enim
silens, ac si cuperet, faciebat), sororum se consortio societa-
tique permiscuit’ (‘Having kissed her head with a smack for,
quite intentionally, she never did anything silently, she
joined the company of the sisters’).88 She thus fits Claudi-
anus’ teasing soubriquet of ‘grandis domina’ (‘large mis-
tress’) perfectly. Martianus’ Arithmetica, with her rapidly
signing fingers and shining rays, is mistaken for a new

86
3. 335K (ed. Willis, 109. 1–6).
87
5. 426K (ed. Willis, 148. 12–13).
88
5. 565K (ed. Willis, 200. 12–14).
 Augustine’s Disciplines 93
89
Hydra. His Music (in jest) is a dangerous Bacchant, ready
for sparagmos.90
The two Gallic writers of the later fifth century knew
some source that treated ‘artes’ (nine and seven91) and
included some sort of allusion to the arts as Muses (Sidonius)
and arts and female personifications, big ladies and bar-
barians (Claudianus). Martianus has seven ‘disciplinae’, but
made a deliberate departure from a canon of nine that was
known to him. Martianus’ ‘disciplinae’ are portrayed as large
and voluble, and occasionally comic or monstrous women; at
least one was a ‘grandis domina’, and another was treated as
a ‘barbara’. Dialectica refers to Varro jokingly as ‘Varro
meus’ (‘my Varro’).92 Astronomy alludes to him as ‘quidam
Romanorum non per omnia ignarus mei’ (‘A certain one
among the Romans who is not entirely ignorant of me’).93
These similarities are unlikely to be fortuitous. None of
these fifth-century authors is dependent on another: they
reflect a common source. Martianus had Varro Menippeus in
mind when he composed De Nuptiis.94 He also frequently
alludes to Varro in the portions of the work devoted to the
‘disciplinae’.95 In the introduction to the book on Dialectic,
for example, Varro is mentioned twice.96 Both Sidonius and

89
7. 729K (ed. Willis, 261. 15–19). The motif is Menippean.
90
See 3. 326K (ed. Willis, 105. 12): ‘Musices impetu, cuius praevertis
officium, discerperis’ (‘You will be torn to pieces by an attack from Music,
whose duties you anticipate’).
91
Claudianus cites seven, not nine. This need not mean that he did not
know nine ‘artes’. Ready acceptance of the humbler and more practical
medicine and architecture disqualified them from this list.
92
Martianus 4. 335K (ed. Willis, 109. 1).
93
8. 817K (ed. Willis, 310. 12). As D’Alessandro, ‘Agostino, Claudiano
Mamerto, Cassiodoro’, 364, rightly observes, this passage and Cassio-
dorus, Institutiones 2. 7. 2 guarantee the existence of a Varronian treatise
on astronomy.
94
Shanzer, A Philosophical and Literary Commentary on Martianus
Capella’s De Nuptiis Philologiae et Mercurii Liber 1, 29–44.
95
4. 335K (ed. Willis, 109. 1–4); 6. 578K (203. 18); 6. 639K (222. 5)
and 6. 662K (234. 7), mediated by Pliny; 8. 817K (310. 12); 9. 928K
(356. 12), through an intermediate source.
96
4. 335K (ed. Willis, 109. 1–4).
94 Danuta R. Shanzer
Claudianus knew of Varro,97 and both allude to him in con-
texts that clearly involve the mathematical ‘disciplinae’ and
philosophy.98
Even though there is some trace of a doctrine of the ‘dis-
ciplinae’ personified as Muses in Maximus of Tyre,99 his
world was a Greek-speaking one, and the lost Disciplinarum
Libri is the obvious candidate for the Gallic writers. While
we may know nothing for certain of the literary form of the
Disciplinarum Libri (the direct citations and allusions are
unrevealing),100 the work of Martianus and these other late
97
For explicit allusions to Varro in Sidonius, see Carm. 2. 90; Carm. 14,
praef. 2–3; Carm. 23. 151 and Ep. 2. 9. 5 to Donidius; Ep. 4. 3. 1 to
Claudianus; Ep. 4. 10 to Felix (mediated by Jerome); Ep. 8. 6. 18 to
Namatius. For Claudianus, see De Statu Animae 2. 9 (CSEL 11.133.12);
Ep. 2 (CSEL 11.206.1).
98
De Statu Animae 2. 8 (CSEL 11.130.2), discussed at length above;
ibid. 2. 9 (CSEL 11.133.12). For Sidonius, see Carm. 14, praef.: ‘Audacter
affirmo, musicam et astrologiam, quae sunt infra arithmeticam conse-
quentia membra philosophiae, nullatenus posse sine hisce nominibus
indicari. Quae si quispiam ut Graeca, sicut sunt, et peregrina verba con-
tempserit, noverit sibi aut super [semper Buecheler] eiuscemodi artis
mentione supersedendum, aut nihil omnino se, aut certe non ad assem,
Latiari lingua hinc posse disserere. Quod si aliquis secus atque assero,
rem se habere censuerit, do quidem absens obtrectatoribus manus; sed
noverint sententiam meam discrepantia sentientes sine Marco Varrone,
sine Sereno, non Septimio, sed Sammonico, sine Censorino, qui de die
natali volumen illustre confecit, non posse damnari.’ (‘I boldly affirm
that music and astrology which are the limbs of philosophy that follow
arithmetic [in importance] cannot be discussed without these names. If
anyone despises them as Greek and foreign (as indeed they are), let him be
aware that he must either always steer clear of the mention of that sort of
art or else know that he cannot say anything at all or certainly not anything
thoroughly in Latin about this. If anyone thinks that the matter is other
than as I claim, I give in to these complainers, to be sure, since I’m absent;
but let them know that even though they maintain a different point of
view, mine cannot be condemned without [condemning] Marcus Varro,
Serenus [not Septimius, but Sammonicus], and Censorinus who wrote an
outstanding work on the natal day.’)
99
Maximus of Tyre 10. 9 (ed. Koniaris, 123. 7): του̃το αρα κα ο3
ποιητα τν Μνηµοσύνην α4νττονται Μουσ!ν µητ&ρα, Μούσα µ.ν τὰ %πιστµα
5νοµάζοντε, 7γάθεον χορν κα /ργον ∆ι", υπ Μνηµοσύνη δ. γεννωµ&να
κα συνταττοµ&να. (‘This also the poets riddling designate as Memory, the
mother of the Muses, and they name the Muses the sciences, a most holy
chorus and a work of Zeus, born and subordinated to Memory.’)
100
See Ritschl, ‘De M. Terentii Varronis disciplinarum libris com-
mentarius’, passim.
 Augustine’s Disciplines 95
Latin authors enable us to hazard some guesses about it. One
could reconstruct a work in nine books, containing one dis-
cipline per book, an equation of each discipline with a Muse,
and talking personifications of these figures. Varro’s long-
winded learned disciplines were the Muses with which
Atticus teased him. To each ‘disciplina’/‘musa’ he may like-
wise have been ‘Varro meus’. The passage from Sidonius
cited above provides another hint, aside from the passage
in the Academica of Cicero, that Varro’s work on the liberal
arts in nine books may have been cited familiarly as the
‘Musae’.101 Further support may be found in Fulgentius,
Mythologiae 1. 15: ‘Nos vero novem Musae doctrinae atque
scientiae dicimus modos’ (‘But we say that the nine Muses
are types of learning and knowledge’).102
Now Book 3 of Martianus, De Grammatica, begins with an
interchange between Muse (Satura) and poet. The text of
the poem is corrupt, but some sense can be elicited. At the
end of Book 2 Martianus had claimed that he was about to
drop the mythological dress of his ‘fabula’.103 Here, however,
the Muse wants ‘amicta fictis commenta ferre’ (‘to produce
arguments cloaked in literary fictions’). The poet protests
that ‘artes vera fantes’ (‘arts who speak true things’) will
prepare the writings of the following books. Satura jokingly
insists that the Arts must be dressed for decency. The
poet protests that they will do their teaching in disembodied
form (‘asomato in profatu’, ‘in bodiless utterance’). The
Muse insists that nothing can be made ready except by a
‘figminis figura’, a personification.104 Martianus gives in,
and each of the following books sports a personified
‘disciplina’.
The purpose of this poem has always been obscure,
although it is clearly programmatic. It has long been
suggested that Martianus deliberately confronted Varro and

101
Martianus’ encyclopaedia was referred to as the ‘Philologia’ by its
own author. See M. De Nonno, ‘Un nuovo testo di Marziano Capella:
la metrica’, Rivista di Filologia 118 (1990), 129–44.
102
Ed. Helm, 25. 18.
103
2. 220K (ed. Willis, 57. 24 ff.).
104
3. 221–2K (ed. Willis, 58. 11; 59. 2, 7–9, 17, 21).
96 Danuta R. Shanzer
his canon of nine in Book 9, where he excludes medicine and
architecture.105 If Varro’s encyclopaedia featured personified
disciplines in fancy dress, then the opening of Book 3 may
represent an abortive attempt to depart from Varronian
practice. In the final metrum of his encyclopaedia Martianus
admitted that he had mingled Muses and Gods and made the
‘disciplinae’ babble in crudely personified forms: ‘immiscuit
/ Musas deosque, disciplinas cyclicas / Garrire agresti cruda
finxit plasmate.’106 Nothing more precise can, or should, be
said now;107 but, given this new possibility of the presence of
Muses and personifications in the title and, no doubt, the
text of Varro’s Disciplinarum Libri, estimates of Martianus’

105
9. 891K (ed. Willis, 339. 4): ‘Sed quoniam his mortalium rerum cura
terrenorumque sollertia est, nec cum aethere quidquam habent superisque
confine, non incongrue si fastidio respuuntur, in senatu caelite reticebunt,
ab ipsa deinceps virgine explorandae discussius.’ (‘But because they have
skills pertaining to, and care for, mortal and terrestrial matters, and have
nothing in common with heaven or the gods, it is not inappropriate that, if
they are rejected with distaste, they fall silent in the celestial senate, but
subsequently be explored in greater detail by the virgin herself.’)
106
9. 998K (ed. Willis, 385. 4).
107
There are numerous texts that could add tesserae to the mosaic: e.g.,
Apuleius, Florida 20 (ed. Helm, 40. 19) on the intoxicating and healing
draughts from the ‘Musarum creterra’ (‘crater of the Muses’), which
include ‘litteratoris rudimento’, ‘grammatica’, ‘rhetorica’, ‘poeticam
commentam’, ‘geometria’, ‘musica’, ‘dialectica’, and ‘philosophia’; also
Ausonius, Griphus Ternarii Numeri, ep. (ed. Green, 112. 34–9): ‘quam multa
enim de ternario sciens neglexi: tempora et personas, genera et gradus,
novem naturalia metra cum trimetris, totam grammaticam et musicam
librosque medicinae, ter maximum Hermen et amatorem primum philosophiae
Varronisque numeros, et quidquid profanum vulgus ignorat’ (‘. . . all of
grammar and music and the books of medicine, thrice-great Hermes and
the first lover of philosophy and the numbers of Varro . . .’). Green has no
comment on the meaningfulness of the triad of ‘grammatica, musica,
medicina’. It could be shorthand for the 3 × 3 liberal arts; Varro’s name
and the mention of medicine point in that direction. Note that Griphus
30–3 contains an allusion to the Varronian story of the origin of the nine
(versus three) Muses: ‘Et lyrici vates numero sunt Mnemonidarum / Tris
solas quondam tenuit quas dextera Phoebi / Sed Citheron totiens ternas ex
aere sacravit / Religione patrum, qui sex sprevisse timebant.’ (‘The lyric
poets are the same in number as the daughters of Mnemosyne, which
three alone the right hand of Phoebus once held, but Citheron consecrated
three apiece of bronze, according to ancestral religion, for they feared to
have spurned six.’)
 Augustine’s Disciplines 97
originality may have to be revised. In the final hymn, sung
108

by Thalia, the lines ‘vos disciplinas omnes / ac nos sacrate

Musas’ (‘consecrate all the disciplines and us too, the
Muses’),109 one of the first hints that the work will deal with
the ‘artes’, may be a deliberate departure from Martianus’
primary source. And, as already noted, the programmatic
final metrum may contain a similar allusion.
b. The Place of the Practical on the Road Upwards
But what of the progress from corporeal to incorporeal, the
upward road that I suggested was already present in Varro
and reflected both in Augustine110 and in Claudianus
Mamertus? How can it be reconciled with the presence
of the all-too-terrestrial medicine and architecture in
Martianus? Kühnert had already raised this objection to
exclude the ascent metaphor from Varro.111 Medicine was in
Book 8,112 and architecture perhaps in Book 9 (as suggested
by the evidence of Martianus). Be that as it may, the
‘arcanum iter’ upwards can still work. All the passages that
discuss abstraction from the corporeal to the incorporeal
concern the quadrivium or its parts, and these alone are
mentioned in Licentius, Claudianus, and in De Ordine.113
The voyage may well have begun on earth, ascended, and

108
It would be interesting and probably worthwhile to pursue the
connections between Muses and liberal arts in visual evidence as well.
109
2. 126K (ed. Willis, 40. 18–19).
110
Passim in Ord. 2. 14. 39 ff. Note Ord. 2. 20. 53: ‘cum et vitae regulas
et scientiae non tam itinera quam ipsos campos ac liquida aequora . . .
et breviter et ita plane significasti’ (‘Since you have briefly and so straight-
forwardly laid out both the rules of life and not so much roads of
knowledge as the very fields and liquid seas themselves’).
111
Kühnert, Allgemeinbildung und Fachbildung in der Antike, 365.
112
Nonius Marcellus: ‘Varro disciplinarum lib. VIII’, ‘Varro dis-
ciplinae lib. VIII’—about ‘lusciosi’ and ‘portulaca’ respectively (ed.
Lindsay, 135. 9–11, 551. 13).
113
Ord. 2. 14. 39: ‘Sed ne de alto caderet, quaesivit gradus atque ipsa
sibi viam per suas possessiones molita est’ (‘But lest she fall from on high,
she sought steps and she herself built a path through her possessions’).
This initiates the mathematical disciplines. See also Augustine, Ep. 101. 3.
98 Danuta R. Shanzer
then returned to earth.114 Plato’s educands had to go back
down to the Cave, after all.115 Martianus’ explanation of the
omission of medicine and architecture—‘quoniam his
mortalium rerum cura terrenorumque sollertia est, nec cum
aethere quicquam habent superisque confine’ (‘Because their
care is for mortal things and their skill pertains to the earthly,
nor do they have anything in common with heaven or with
the gods’)—may even reflect Varro’s own strictures.116
Augustine’s canon seems to consist of just seven arts.117
But medicine and architecture are none the less present in
the margins, next to one another, just before the series of
seven. De Ordine 2. 11. 32, a discussion of ‘ratio’ in sight and
hearing as opposed to taste, touch, and smell, mentions
proportion in the prescriptions administered by the medicus
(and the food seasoned by the cook!) as well as proportion in
building (‘ratio’ and the architect). It is possible that these
two arts, mentioned here cursorily, may be contained in the
first part of Augustine’s tripartite classification in De Ordine
2. 12. 35: ‘unum est in factis ad aliquem finem relatis,
alterum in dicendo [the trivium], tertium in delectando
[the quadrivium]’ (‘One consists in deeds applied to some
end, the other in speaking, and the third in delighting’).

c. Contents, Canons, and Caveats

Lists are incomplete and lists differ. Our Gallic authors
diverge, each from the other and each from himself as well as
from Ritschl’s reconstruction of the De Novem Disciplinis,

114
For a Muse grounded, see Martianus at 1. 28K (ed. Willis, 13. 8–10):
‘sola vero, quod vector eius cycnus impatiens oneris atque etiam sub-
volandi alumna stagna petierat, Thalia derelicta in ipso florentis campi
ubere residebat.’ (‘Thalia alone, because the swan who was carrying
her, intolerant of the burden and also of flying upwards had sought its
ancestral swamp, was left alone sitting on the fertile surface of a flowering
field.’) The joke is still somewhat obscure.
115
Rep. 539e–540a.
116
9. 891K (ed. Willis, 339. 4). Hadot, Arts libéraux et philosophie, 150,
is engaging in special pleading when she claims that Martianus is imitating
Macrobius, Saturnalia 7. 15. 14–15, not Varro, in excluding medicine and
architecture. Macrobius mentions only medicine.
117
The heptad was unquestionably significant. He returned to it in
Quant. An. 33. 70–6 and Doct. Chr. 2. 7. 9. See the discussion by Pollmann
in this volume, 225–7.
 Augustine’s Disciplines 99
which, according to him, contained the trivium, the quad-
rivium, medicine, and architecture (see Table 4.2).118
Medicine is omitted from all passages but one. Is this an
impediment to discerning a Varronian canon of nine in
the background? No, for none of the passages is a formal
and exclusive citation of all of the ‘artes’. Completeness
was unlikely. Many students of canons and canonical lists
fall victim to the fallacy of demanding a higher degree of
precision or exhaustiveness from their sources than those
very sources required or were intended to have.119 When one
deals with sources on the ‘artes’, two crucial factors have to
be taken into account. The first is the author’s purpose
in listing them. Sidonius, Ep. 5. 2 is aiming at a number,
nine. Claudianus does not say he is, and Claudianus is not
interested in the trivium in De Statu Animae. In Ep. 5. 2
Sidonius provided a drink from a fire hydrant and was bound
to get disorganized in the midst of forty-five proper names.
The second factor is human error. Most people have trouble
rattling off the members of any given canonical group,
without some thought and either fingers or a piece of paper.
I submit that both Sidonius and Claudianus were describ-
ing—in their own ways—a single unitary elephant (see
J. G. Saxe, ‘The Blind Men and the Elephant’).
The Disciplinarum Libri120 were nine books long.121 The
general content of some books was reconstructed by Ritschl,
deconstructed by Hadot, and reconstructed in the present

118
The order in which these ‘artes’ appeared was reconstructed by
Ritschl, ‘De M. Terentii Varronis disciplinarum libris commentarius’,
352–402. Only in the case of arithmetic (Book 5) and medicine (Book 8)
can we be sure of the position of the discipline within the whole (ibid.
368).
119
This is the point made by Marrou, Saint Augustin et la fin de la
culture antique, 218: ‘Les discordances sont plus spécieuses que réelles. Un
bon nombre s’expliquent par le caractère même des textes d’où ces listes
sont tirées; c’étaient souvent des exemples jetés au hasard sous la plume,
non des énumérations méthodiques et exhaustives.’
120
Title: Aulus Gellius, Noctes Atticae 10. 1 .6 and 18. 15. 2; Nonius
Marcellus (ed. Lindsay) 135. 10 and 551. 15; Cassiodorus, Institutiones
2. 3. 2.
121
Number of books: Vitruvius, De Architectura 7, praef. 14;
Cassiodorus, Institutiones 2. 3. 2; Isidore, Origines 2. 23.
100 Danuta R. Shanzer
Table 4.2. The Gallic canons compared

Claudianus, Sidonius, Ep. Claudianus, Sidonius to

Stat. An. 2. 8 5. 2 (not in Ep. ad Claudianus,
sequence) Sapaudum Ep. 4. 3 (singuli
artifices)

in musicis musica musice cum Orpheo

plectrum
〈in arithmeticis〉 arithmetica arithmetica
geometria geometrica cum
Archimede
radium cum
Euclide
mensuras
in philoso- dialectica dialectice Socrates, Plato,
phoumenon Aristotle
libris [ =
dialectic?]
astrologia 〈astrologia〉 [?]
grammatica grammatica
architectonica cum Vitruvio
perpendiculum
oratoria rhetorica Aeschines,
Demosthenes,
Hortensius,
etc.
metrica [a cum Aescula-
clear error] pio baculum
medicina [?]
in philoso- philosophia
phoumenon
libris [?]

essay (Table 4.3). But establishing order is difficult, because

in economical ‘need to know’ citations the ‘artes’ appear as,
for example, ‘Varro in geometriae volumine’,122 or ‘Varro in
libro quem de astrologia conscripsit’, without indication of
what book the art occupied, or even whether the work formed
part of the Disciplinarum Libri. There are also potential

122
Cassiodorus, Institutiones 2. 7. 3, 4 (ed. Mynors, 155. 11, 157. 12).
 Augustine’s Disciplines 101
Table 4.3. Reconstructions of Varro’s disciplinae

Ritschl’s Hadot’s Martianus’ Shanzer’s

Varro Varro allusions to Varro
Varro, and (external
parallels evidence)

grammar
(358) (162–3) no 3. 229 parallels Augustine
Isidore evidence for its Ord. 2. 12. 35 Sidonius
Martianus place in the Claudianus
Augustine Disciplinarum Martianus
Libri
dialectic
(356) (160, 163–4, mentions Varro Augustine
Martianus 166) no 4. 335 Sidonius
evidence for its (misascribed by Claudianus
existence in the Della Corte Martianus
Disciplinarum (241–2) to
Libri Rhetoric)
rhetoric
(357) (172) no Augustine
Priscian evidence for its Sidonius
existence in the Claudianus
Disciplinarum Martianus
Libri
geometry
(359)
Cassiodorus (173) mentions Varro Augustine
Martianus mentioned by 6. 578, 6. 639, Licentius
Cassiodorus, and 6. 662 Claudianus
but not (mediated by Martianus
necessarily as Pliny)
part of the
Disciplinarum
Libri
arithmetic
(362–3) (159, 175) no
Incertus de trace at all [?] Augustine
grammatica anywhere Claudianus
Augustine Martianus
Nonius
Gellius
Claudianus continued
102 Danuta R. Shanzer
Table 4.3. Continued

Ritschl’s Hadot’s Martianus’ Shanzer’s

Varro Varro allusions to Varro
Varro, and (external
parallels evidence)

astronomy
(361) (167) argument mentions Varro Augustine
Cassiodorus, from Martianus 8. 817 Licentius
etc. is ex silentio Sidonius
(175) no Martianus
evidence Cassiodorus,
Inst. 2. 7. 2
music
(361) no (159, 175) mentions Varro Licentius
explicit 9. 928 (mediated Claudianus
testimonium by an Martianus
Claudianus intervening
source)
medicine (366–8)
Martianus present and omitted (see Nonius
Nonius correct 9. 891) 〈Augustine〉
Pliny 〈Sidonius〉
Isidore
architecture
(364–6) present and omitted (see Vitruvius
correct 9. 891) Martianus
〈Augustine〉
Sidonius
Martianus

problems with Roman numerals. Positivists work themselves

up into a lather over what may be a false reading, a dropped
or added ‘I’.123 Much of our evidence for what book a given
discipline occupied is subject to that sort of corruption. We
cannot afford to forget this.

123
Hadot, Arts libéraux et philosophie, 159, criticizes Ritschl (a bit
unfairly) for this sort of emendation. See Ritschl, ‘De M. Terentii
Varronis disciplinarum libris commentarius’, 370, for musical material
‘in tertio [sic] disciplinarum’ (‘in the third book of the disciplines’).
 Augustine’s Disciplines 103
Then there is the matter of order. Scholars typically make
arguments of the type, ‘X can’t be following Y, because
his order is different’. These are unlikely to be universally
valid. Order could be varied for philosophical or literary
purposes.124 Was sight a higher or lower sense than
hearing?125 Should arithmetic come at the beginning (as
foundation),126 the middle (sacred hebdomad),127 or the end
of the quadrivium (number in abstracto)?128 Should music
come first,129 or second in the quadrivium because it linked
the ears and sound inherent in the appreciation of the arts of
the trivium with the number underlying the quadrivium?130
Or should Music come last to sing an epithalamium?131 Or
did the mind progress ‘per aspera ad astra’?132 More ‘give’
(and more room for the requirements of aesthetic and
moral agendas) is required in our rules for analysis. All these
factors make this sort of source criticism difficult.

124
Hadot, Arts libéraux et philosophie, 169, makes the point well.
125
Augustine, Ord. 2. 14. 41 found the objects of hearing, sounds, lesser
things because they were ‘sensibiles res’ that passed on as opposed to the
objects that the mind sees. Authoritative sources include Aristotle, Meta-
physics A1; Protrepticus Fr. 7. Alain de Lille, Anticlaudianus 5. 258 made a
different decision: only the horse of hearing accompanies Prudentia and
Theologia into the higher regions of heaven (reflecting a Pauline hierarchy
of the senses suggested by Romans 10: 17: ‘fides ex auditu’ (‘faith comes
from hearing’)).
126
Ritschl, ‘De M. Terentii Varronis disciplinarum libris com-
mentarius’, 369, incorrectly calls this a later innovation. Hübner, ‘Die
“artes liberales” im zweiten Buch von De Ordine’, 341, traces its initial
position back to Plato.
127
As in Book 7 of De Nuptiis.
128
As implicitly in Augustine, Ord. See 2. 15. 43: ‘in his ergo omnibus
disciplinis occurrebat ei omnia numerosa’ (‘therefore in all these
disciplines she encountered entirely things pertaining to number’).
129
Through its control of metrics (shared with grammar) it linked
the verbal and mathematical arts. It seems to be first in Augustine,
Ord. 2. 14. 39.
130
As could be the implication of Augustine, Ord. 2. 14. 39–41. Ritschl,
‘De M. Terentii Varronis disciplinarum libris commentarius’, 369, rightly
points out that its position is the most variable among the arts of the
quadrivium.
131
As in Book 9 of De Nuptiis.
132
As in the conclusion of Book 2 of Cassiodorus’ Institutiones
(ed. Mynors, 158. 3).
104 Danuta R. Shanzer

4. au g u st i n e a n d t h e m u s e s
Non enim audiendi sunt errores gentilium superstitionum, qui
novem Musas Iovis et Memoriae filias esse finxerunt. Refellit eos
Varro, quo nescio utrum apud eos quisquam talium rerum doctior vel
curiosior esse possit. Dicit enim civitatem nescio quam, non enim
nomen recolo, locasse apud tres artifices terna simulacra Musarum,
quod in templo Apollinis donum poneret, ut quisquis artificum
pulchriora formasset, ab illo potissimum electa emeret. Itaque con-
tigisse ut opera sua quoque illi artifices aeque pulchra explicarent,
et placuisse civitati omnes novem, atque omnes esse emptas, ut in
Apollinis templo dedicarentur; quibus postea dicit Hesiodum
poetam imposuisse vocabula. Non ergo Iupiter novem Musas
genuit, sed tres fabri ternas creaverunt. Tres autem non propterea
illa civitas locaverat, quia in somnis eas viderat, aut tot se cuiusq-
uam illorum oculis demonstraverant; sed quia facile erat animad-
vertere omnem sonum, quae materies cantilenarum est, triformem
esse natura. Aut enim voce editur, sicuti eorum est qui faucibus
sine organo canunt; aut flatu, sicut tubarum et tibiarum; aut pulsu,
sicut in citharis et tympanis, et quibuslibet aliis quae percutiendo
canora sunt. Sed sive ita se habeat, quod Varro retulit, sive non ita,
nos tamen non propter superstitionem profanorum debemus musicam
fugere, si quid utile ad intellegendas sanctas. scripturas rapere
potuerimus . . . (Doct. Chr. 2. 16. 27)
We must not pay heed to the errors of the pagan superstitions that
pretended that the nine Muses were the daughters of Jove and
Memory. Varro refutes them, and I doubt that anyone among them
could be more learned in such matters or more curious about them.
He said that a certain city, whose name I do not recall, had con-
tracted out to three craftsmen three images of the Muses apiece to
dedicate in the temple of Apollo as a gift, on the understanding
that whichever of the craftsmen made the more beautiful images,
from him by preference it would buy the chosen artwork. And so it
came about that the artisans themselves too displayed equally
beautiful statues, and the city liked all nine of them, and all were
bought to be dedicated in Apollo’s temple. He says that the poet
Hesiod ascribed names to them. Therefore it was not Jupiter who
begat the nine Muses, but three artisans created three a piece. That
city did not dedicate three Muses because it had seen them in
dreams or because so many had presented themselves to the sight
of one of them, but because it is easy to see that all sound, which is
the material of songs, is threefold in nature. Either it is produced
by voices, as is the case with those who sing with their throats
without an organ, or by blowing, as of trumpets or flutes, or by
 Augustine’s Disciplines 105
striking, as with citharae and drums, or whatever other instruments
give tongue when struck. Whether this is the case, as Varro tells it,
or not, we all the same should not flee music on account of the
superstition of the profane, as long as we can snatch something
useful for understanding sacred scripture thence . . .
In De Doctrina Christiana 2. 16. 26 Augustine saves one
liberal art, music. Immediately (in the passage quoted above)
he argues against the ‘errors of pagan superstition’ that con-
sider the Muses the daughters of Jupiter and Memory.
Varro, the most learned and most ‘curiosus’ in such matters,
refuted them. Augustine narrates a rationalizing Varronian
account of the reproduction of statues of the (originally)
three Muses. Three statues were reproduced thrice and
thus became nine. The Muses are three not because they
appeared in dreams or in person, but because sound’s
classification is tripartite: vocal, blown, or struck.133 Varro
reappears at the end of the passage.
In Retractationes 1. 3. 2 Augustine regretted that he had
‘attributed a great deal to the liberal arts’ and that he
had said, albeit in jest, that the Muses were goddesses:
Verum et in his libris displicet mihi saepe interpositum fortunae
vocabulum . . . et quod multum tribui liberalibus disciplinis, quas
multi sancti multum nesciunt; quidam etiam qui sciunt eas sancti
non sunt; et quod Musas quasi aliquas deas quamvis iocando
commemoravi.
Indeed in these books too I regret that I frequently inserted the
name of Fortune . . . and that I attributed a great deal to the liberal
arts of which many holy people are very ignorant and even some
who know them [well] are not holy and that I mentioned the Muses
as if they were some sort of goddesses—even if only in jest.
The passage he was alluding to was De Ordine 2. 14. 41:
Et quoniam illud quod mens videt semper est praesens et
inmortale adprobatur—cuius generis numeri apparebant—sonus
autem, quia sensibilis res est, praeterfluit in praeteritum tempus

133
See Ord. 2. 14. 39 for the identical triple classification of sound.
Hadot, Arts libéraux et philosophie, 120, notes that it appears in Doct. Chr.
in ‘un contexte varronien, ce qui ne prouve pas nécessairement que ce soit
Varron qui ait inventé cette division. Varron l’a probablement tirée d’une
œuvre grecque.’
106 Danuta R. Shanzer
inprimiturque memoriae, rationabili mendacio iam poetis favente
ratione Iovis et Memoriae filias Musas esse confictum est. Unde ista
disciplina sensus intellectusque particeps musicae nomen invenit.
And because what the mind sees is always present and is con-
sidered immortal—numbers seemed to be of this sort—but sound,
because it is a thing apprehended by the senses, flows by into time
past and is imprinted on memory, by a reasonable falsehood, with
reason on the side of the poets, it was pretended that the Muses
were the daughters of Memory. From this that discipline, which
partakes both of intellection and sense perception, found the name
‘music’.
He thus rejected his statement of 386, where he had called
the falsehood ‘reasonable’ (rationabile) and consistent with
‘reason’ (ratione), by claiming to have been joking. De-
mythologization and demystification of the Muses show
that Augustine was disquieted by them.134 In De Doctrina
Christiana, I would guess, he sneakily tried to refute Varro
from the latter’s own works. Varro had discussed the evolu-
tion of the canon from three to nine; yet he is cited by
Augustine to prove that the Muses are three in number (not
nine) and not goddesses, even though there is independent
evidence that Varro assigned a ‘novenarius numerus’ or
ninefold number to them.135 Augustine would never mention
them by name again.136
Varro was a rationalizer, an antiquarian, a student of the
gods, and a devotee of personifications. It was the form and
fancy dress, not just the matter, of his Disciplinarum Libri
that almost certainly disquieted Augustine. For, unlike
Apollonius of Tyana,137 Augustine had little time for literary

134
Or could it be the number of occasions on which he teased Licentius
about his Muses? See Ord. 1. 3. 6: ‘Nam video tibi Musam tuam lumen ad
lucubrandum accendisse’ (‘For I see that your muse has lit the lamp for
burning the midnight oil’); 1. 8. 24: ‘Vade ergo interim ad illas Musas’ (‘Go
in the meantime to those Muses’).
135
See the text of Arnobius cited above, n. 80.
136
Confirmed by a search of the Chadwyck–Healy electronic Patrologia
Latina database.
137
See the charming anecdote about Apollonius and the Mesopotamian
customs-officer at Philostratus, Vita Apollonii 1. 20, where he ‘declares’
his accompanying ladies Sophrosyne, Dikaiosyne, etc. They are not slave-
girls, he says, but mistresses (despoinai).
 Augustine’s Disciplines 107
Alexandrianism, where books and disciplines could become
ladies.138 However, the Platonic upward journey through the
mathematical disciplines, as outlined in the Disciplinarum
Libri, was a powerful and attractive concept.139 In charac-
teristic fashion Augustine internalized and adapted it.140
He also made use of some of Varro’s disciplinary material.141
He largely suppressed the personifications of the ‘artes’—
but not entirely successfully.142 Consider the implicit per-
sonification of grammar in De Ordine 2. 12. 36: ‘Poterat iam
perfecta esse grammatica sed, quia ipso nomine profiteri se
litteras clamat unde etiam Latine litteratura dicitur factum
est, ut quidquid dignum memoria litteris mandaretur, ad
eam necessario pertineret.’143 But an even more notable
personification dominates this passage of De Ordine, Ratio
herself, with whom Augustine would soon converse in the
Soliloquia. He ensures, however, that she is not a visible
person, let alone a female one, but a disembodied voice. Even
whether this is an inner or an outer voice is unclear: ‘ait mihi
subito sive ego ipse, sive aliusquis 〈sive〉 extrinsecus sive
intrinsecus . . .’144 (‘Some one said to me suddenly, whether
I myself or someone else, be it outside or inside . . .’). In De

138
See the Callimachean ‘programmatic’ passages: (1) Aetia, prol.
11–12: ‘of two, that Mimnermus was sweet . . . the big (µεγάλη) woman
did not teach’, often thought to refer to his ‘big’ elegy, named after his
mistress Nanno; (2) Fr. 398, on Antimachus’ Lyde: ‘Lyde, a fat (παχύ)
book and not clear’.
139
Hübner, ‘Die “artes liberales” im zweiten Buch von De Ordine’,
322, rightly notes that the upward journey begins with the mathematical
disciplines.
140
Ord. 2. 14. 39; 2. 15. 42: ‘terram caelumque collustrans’ (‘wandering
over earth and heaven’); 2. 15. 43: ‘Hic se multum erexit multumque
praesumpsit’ (‘At this point she raises herself a great deal and takes much
upon herself’); 2. 20. 53: ‘scientiae non tam itinera quam ipsos campos ac
liquida aequora’ (‘not so much roads of knowledge as very fields and liquid
seas’).
141
See Doignon, Dialogues philosophiques: De Ordine, 363–78, for a
terse and recent summary.
142
See also Doct. Chr. 2. 25. 39 on statues and 3. 6. 11 on Neptune.
Hübner, ‘Die “artes liberales” im zweiten Buch von De Ordine’,
312, notes that there is partial personification in Ord. that lends it ‘eine
Art romanhafte Handlung’.
143
Translated above, n. 37.
144
Sol. 1. 1.
108 Danuta R. Shanzer
Ordine 2. 16. 45 we can see Augustine consciously playing
with the idea of the ‘corpus’ and the ‘anima’ of grammar (or
Grammar): Monica grasps its soul while leaving its body to
the ‘diserti’ (‘eloquent’).145 Varro must have been among the
latter.
In De Ordine are signs both of the present and of the
future. Full assent is given to grammar and dialectic, but
almost no attention is paid to rhetoric, perhaps because of
the author’s own conflict about its morality and the post he
had given up.146 Astrology is already designated as danger-
ous.147 Geometry appears in the Soliloquia.148 Grammar and
dialectic are reconsidered in Soliloquia 2. 11. 19–20, but
the literary duties of the grammarian have already become
suspect; that trend would continue in the diatribes of Book 1
of the Confessions.149 The unquestioning acceptance in the
Soliloquia that the face of ‘Veritas’ could be reached through
education in the disciplines and that the disciplines could be
recovered by (Platonic) reminiscence150 had to be carefully
retracted in 427.151 In De Doctrina Christiana, it is true,

145
‘Sed tu contemptis istis vel puerilibus rebus vel ad te non pertinenti-
bus ita grammaticae paene divinam vim naturamque cognosces, ut eius
animam tenuisse, corpus disertis reliquisse videaris.’ (‘But you, having
rejected those things that are either childish or have nothing to do with
you, will so know the almost divine power of grammar that you seem to
hold its soul, while leaving its body to the eloquent.’)
146
See Conf. 9. 2. 3–4, 4. 7. It is not entirely clear whether Augustine
resigned officially before the date of De Ordine (November 386).
147
Ord. 2. 15. 42.
148
Sol. 1. 9. 3; 2. 32. 8.
149
On the presentation of the ‘artes’ in Conf. see Burton’s essay below.
150
Sol. 2. 35. 1: ‘Tales sunt, qui bene disciplinis liberalibus eruditi;
siquidem illas sine dubio in se oblivione obrutas errunt discendo et
quodammodo refodiunt, nec tamen contenti sunt nec se tenent, donec
totam faciem veritatis, cuius quidam in illis artibus splendor iam sub-
rutilat, latissime atque plenissime intueantur.’ (‘They are like this who are
well educated in the liberal arts; if they err in learning them because they
are no doubt buried in oblivion inside them, and somehow dig them out,
they are nonetheless not content nor do they contain themselves until they
gaze most extensively and exhaustively upon the whole face of the truth,
whose splendor already glows somewhat more gently in those arts.’)
151
Retr. 1. 4. 4: ‘Item quodam loco dixi, quod disciplinis liberalibus
eruditi, sine dubio in se illas oblivione obrutas eruunt discendo, et quodam
modo refodiunt. Sed hoc quoque improbo: credibilius est enim, propterea
 Augustine’s Disciplines 109
Augustine saved music, the one liberal art that he continued
to pursue in Africa and returned to with longing and disquiet
in the Confessions.152 But by the time of the Letter to Memor
in 408/9 it too had largely fallen from favour, as had the
so-called liberal arts.153

This essay has concentrated on the younger Augustine

and his attitude to the disciplines, to suggest that Hadot’s
dismissal of Varro’s Disciplinarum Libri as a source was pre-
mature. The tortured relationship of the bishop of Hippo to
the scholar of Reate that we see in the City of God almost
certainly ante-dates that work by more than twenty years. I
hope to have illuminated that relationship by reading closely
and to some extent between the lines. It may be possible to

vera respondere de quibusdam disciplinis, etiam imperitos earum, quando

bene interrogantur, quia praesens est eis, quantum id capere possunt,
lumen rationis aeternae, ubi haec immutabilia vera conspiciunt; non quia
ea noverant aliquando, et obliti sunt, quod Platoni, vel talibus visum est.’
(‘Likewise I said somewhere that those learned in the liberal arts doubtless
dig them up if they are buried in oblivion within them and somehow
excavate them. But this too I reject, for it is more believable that even
those who are unskilled in the disciplines can give correct answers about
some of them when they are questioned skillfully, because the light of
eternal reason is present to them to the extent that they can understand it,
when they see these immutable truths. It is not the case that they knew
them once and forgot them, as Plato and such people thought.’)
152
Pizzani, ‘L’Enciclopedia agostiniana’, 356, notes that De Musica is
too long to form part of an encyclopaedia, but has developed a life of its
own. Music had a place in Conf. 9. 6. 14–7. 15 and 10. 33. 49.
153
Ep. 101 to Memor (408/9) 1: ‘non quia nolui, sed quia non potui . . .
per nostrum ministerium non litteris illis, quas variarum servi libidinum
liberales vocant’ (‘Not because I was unwilling, but because I could not
. . . through my offices, not those letters that the slaves of various lusts call
“liberal” ’); 3: ‘Verum quia in omnibus rerum motibus quid numeri
valeant facilius consideratur in vocibus, eaque consideratio quibusdam
quasi gradatis itineribus nititur ad superna intima veritatis, in quibus viis
ostendit se sapientia hilariter et in omni providentia occurrit amantibus’
(‘It is true that in all movements of things it will more easily be seen what
power numbers have in the case of voices, and this examination strives
towards the secret upper regions of truth by certain successive paths as it
were; in these paths wisdom shows herself cheerfully and meets those who
love her in all providence’). He sends Book 6 of De Musica, but not the
first five, which he does not deem worthy of Julian.
110 Danuta R. Shanzer
break up a bedevilling log-jam in Varronian source criticism
by granting that lists are rarely of lapidary perfection, but
have their purposes in their own contexts, that numerals can
be corrupted, and that enumerative texts need not invariably
fetishize their predecessors’ order and hierarchy.
As for Augustine himself, it is crucial not to forget that
all his writings were occasional. They respond to social
forces and texts ‘out there’ in his largely lost environment.
We must read between the lines and examine tergiversations
and temporizings, and wonder whether his aversion to the
disciplines after an initial attempt to use them as a cure for
the ‘plagae opinionum’154 (‘wounds consisting of opinions’)
was partly internal and whether the mathematical ones were
beyond him. But there were no doubt external factors too.
These may have included Ambrose155 or fundamentalist
friends.156 Augustine may even have foreseen his eventual
position in De Doctrina Christiana. But above all there was
the intimidating realization that all the Bible lay ahead of
him for exegesis, and that there would never be world
enough and time. These factors can help us to understand
what happened between November 386 and April of 387
to take the bloom off Augustine’s disciplines and cause
him essentially to abandon the project. But ultimately his
personal intellectual crises and qualms about secular learn-
ing and female personifications had little effect. Within less
than 150 years a Christian Boethius would hold a prison
colloquy with a superbly personified and fully visualized
Philosophy.157 And the seven liberal arts, both personified
and canonized, would reign throughout the Middle Ages.

154
Ord. 1. 1. 3: ‘Quod hi tantum adsequuntur, qui plagas quasdam
opinionum, quae vitae cotidianae cursus infligit, aut solitudine inurunt aut
liberalibus medicant disciplinis.’ (‘These alone pursue this who either
impress indelibly through solitude or heal with the liberal disciplines
certain wounds [consisting] of opinions that the course of daily life inflicts
upon them.’)
155
For whose attitude, see P. Brown, Augustine of Hippo: A Biography
(London, 1967), 112, and the essay by Heßbrüggen-Walter in this volume.
156
Augustine may hint at the existence of this adverse body of opinion
when he mentions those he did consult when working on the ‘artes’, who
‘did not shudder at these sorts of disciplines’, in Retr. 1. 5. 6.
157
Contrast Augustine’s treatment of Ratio in Sol. 1. 1.
 Augustine’s Disciplines 111
Appendix: Cicero and the ‘Muses’ of Varro
Atticus: ‘Omitte ista quae nec percunctari nec audire sine molestia
possumus, quaeso,’ inquit, ‘et quaere potius ecquid ipse novi.
Silent enim diutius Musae Varronis quam solebant, nec tamen
istum cessare sed celare quae scribat existimo.’ ‘Minime vero,’
inquit ille, ‘intemperantis enim arbitror esse scribere quod
occultari velit; sed habeo magnum opus in manibus, quae iam pri-
dem; ad hunc enim ipsum’ (me autem dicebat) ‘quaedam institui,
quae et sunt magna sane et limantur a me politius.’ (Cicero,
Academica 1. 1–2)
Then Atticus said, ‘Please leave out those things that we cannot ask
about or hear without annoyance and inquire rather whether there
is anything new with him [sc. Varro]. For Varro’s Muses are silent
for rather longer than they used to be, but I do not think that he is
delaying, but hiding158 what he writes.’159 ‘Absolutely not,’ says he
[sc. Varro], ‘for I consider it the act of an injudicious man to write
what he might want to remain hidden. But I have a major work on
my hands, 〈and have had it〉160 for some time. For I have set to work
on some writings addressed to this man himself (he meant me
[sc. Cicero]) that are both a major enterprise and are being filed
and polished by me.’
Atticus had suggested that Cicero dedicate a work to Varro
in 54 (Ep. ad Atticum 4. 16. 2). Cicero did not have an appro-
priate item and delayed doing so till he decided to introduce
Varro into the second edition of the Academica (Att. 13. 12.
3). At that time (45 b c) Varro had already promised him the
dedication of an unspecified work two years earlier.161 It is
generally assumed that the work promised by Varro was
Book 5 of De Lingua Latina which he did indeed eventually

158
The point is humorously picked up by Cicero at Academica 1. 3:
‘nihil enim eius modi celare possumus’ (‘For we cannot conceal anything
of that sort’).
159
Or better still, given the subjunctive, ‘what sort of thing he is
writing’.
160
‘Quae’, however, is plural.
161
Cicero, Ep. ad Atticum 13. 12. 3: ‘Iam Varro mihi denuntiaverat
magnam sane et gravem προσφ9νησιν. Biennium praeteriit cum ille
καλλιπδη adsiduo cursu cubitum nullum processerit.’ (‘Already Varro
had announced to me the dedication of a very major and serious work.
Two years have gone by, but that Kallipides, though running persistently,
has not progressed [even a] yard.’)
112 Danuta R. Shanzer
dedicate to Cicero. It is also assumed that De Lingua Latina
is the work under discussion at the opening of the
Academica, the one Atticus teases Varro about.162
If, however, the badinage is only about De Lingua Latina,
then there is a need to explain why he could be said to
be ‘hiding’ it (‘celare’), given that Books 2–4 were out and
dedicated to Publius Septimius (De Lingua Latina 5. 1).163 If
two works are at issue, however, the problems are solved.
Atticus teases Varro about the Musae, and Varro is depicted
by Cicero as responding with the excuse that his ‘magnum
opus in manibus’ (the Disciplinarum Libri) is causing the
delay.
162
L. Straume-Zimmermann, F. Broemser, and O. Gigon, Marcus
Tullius Cicero: Hortensius, Lucullus, Academici Libri (Darmstadt, 1990),
454–5.
163
H. Dahlmann, ‘Marcus Terentius Varro’, in RE Supplementband vi
(1935), 1172–277 at 1204: ‘Daß Varro die Bücher II–IV an Septimius
schon vor 47, ehe er die Absicht hatte, an Cicero ein Werk zu richten,
publiziert hat, möchte ich deswegen für wahrscheinlich halten, weil er
sonst der Einheitlichkeit wegen das ganze Werk unter Ciceros Namen
gestellt haben würde.’
 5
Divination and the Disciplines of
Knowledge according to Augustine
W i l l i a m E. K l i n g s h i r n

Nam si singulas disciplinas percipere magnum est,

quanto maius omnis.
Cicero, De Natura Deorum1

1. i n t r o d u c t i o n
Shortly before his ordination as bishop, probably in 394,
Augustine composed, in addition to Contra Adimantum,
several popular sermons against the Manichaean Adimantus,
whose Disputationes against the Law and the Prophets had
recently come to his attention.2 Possidius lists five ‘tractatus’
in this series, of which three survive: Sermons 1, 50, and 12.3
The last (and probably latest) of these defends a passage
from the book of Job: ‘Behold came angels into the sight of

I am grateful to the American Council of Learned Societies for a fellow-

ship supporting the research on which this chapter is based.
1
‘For if it is a daunting task to grasp individual disciplines, how much
more so to grasp all of them!’ (1. 11).
2
Retr. 1. 22. 1. For the career and writings of Addas (Latinized
as Adimantus), traditionally one of Mani’s three early disciples, see
G. S. Gasparro, ‘Addas-Adimantus unus ex discipulis Manichaei: For the
History of Manichaeism in the West’, in R. E. Emmerick (ed.), Studia
Manichaica, iv: Internationaler Kongreβ zum Manichäismus, Berlin,
14.–18. Juli 1997, Berichte und Abhandlungen, Berlin-Brandenburgische
Akademie der Wissenschaften, Sonderband 4 (Berlin, 2000), 546–59.
3
Possidius iv. 29–33, ed. A. Wilmart, Miscellanea Agostininiana, ii:
Studi Agostiniani (Rome, 1931), 167. On the chronology, see F. Cavallera,
‘Notes chronologiques et hagiographiques sur quelques sermons de saint
Augustin’, Bulletin de Littérature Ecclésiastique 31 (1930), 21–30, at 21–3.
114 William E. Klingshirn
God, and the devil in their midst. And God said to the devil,
“Where do you come from?” And the devil replied, “Circling
the whole world, I have arrived”’ (Job 1: 6).4 In this sermon,
Augustine is concerned to point out how God could speak
to the devil without the devil’s actually seeing him. His aim
is to defend the passage from Adimantus’ charge that it
contradicts the gospel saying that ‘Blessed are those who
are pure of heart, for they shall see God’ (Matt. 5: 8).5 To
do this, Augustine launches into what amounts to a com-
prehensive survey of all the modes of divine communication
with humankind.
Now there are many ways in which God speaks with us. At times
he speaks through an instrument, as through a codex of the divine
scriptures (per codicem divinarum scripturarum). He speaks through
a heavenly body, as he spoke to the Magi through a star (per
stellam). . . . He speaks through a lot (per sortem), just as he spoke
concerning the choice of Matthias in place of Judas. He speaks
through the human soul (per animam humanam), as through a
prophet. He speaks through an angel (per angelum), just as we
understand him to have spoken to certain patriarchs and prophets
and apostles. He speaks through a creature (per . . . creaturam)
endowed with speech and sound, just as we read and believe of
voices produced from the heavens, although no one is visible to the
eye. Finally God speaks directly to a person, not externally
through the ears or eyes, but internally in the soul, and not in one
way only, but in dreams (in somnis) . . . or with a person’s spirit
lifted up (spiritu hominis assumpto), which the Greeks call ecstasis,
. . . or in the mind itself (in ipsa mente), when each person under-
stands his majesty and his will.6

4
Serm. 12. 1: ‘Ecce venerunt angeli in conspectum dei, et diabolus in
medio eorum. Et deus ait diabolo: Unde venis? Qui respondens dixit:
Circumiens totum orbem adveni.’
5
Ibid.: ‘Beati qui puro sunt corde, quia ipsi deum videbunt’ (CCSL 41.
165).
6
Serm. 12. 4: ‘Multi autem modi sunt, quibus nobiscum loquitur
deus. Loquitur aliquando per aliquod instrumentum, sicut per codicem
divinarum scripturarum. Loquitur per aliquod elementum mundi, sicut
per stellam magis locutus est. . . . Loquitur per sortem, sicut de Mathia
in locum Iudae ordinando, locutus est. Loquitur per animam humanam,
sicut per prophetam. Loquitur per angelum, sicut patriarcharum et
prophetarum et apostolorum quibusdam locutum esse accipimus.
Loquitur per aliquam vocalem sonantemque creaturam, sicut de caelo
 Divination and the Disciplines 115
This list is consistent with Augustine’s keen and tolerant
interest in divination in the period between his conversion
and the writing of the Confessions. In addition to the
expected forms of divine communication (angels and
prophets), it also mentions several kinds that can clearly
be categorized as divinatory (although Augustine would
never have used the word).7 These include divination by
lot, by the stars, by ecstasies and dreams, and by the direct
mental transfer that he believed could occur in certain cases.
Augustine included these types of divination because, like
many other Christians, he believed that under the right con-
ditions they could legitimately convey God’s knowledge.
Other types, such as those involving animal sacrifice or the
consultation of demons, he deliberately left off the list,
for the same reasons that he gave in Confessions 4. 2. 3 for
refusing to employ a haruspex.
It is notable that, along with objectionable kinds of divina-
tion, Augustine also omitted diviners from his catalogue.
Not that such figures are entirely absent from his early
writings: half a book of the Contra Academicos is taken up
with his description of the ‘hariolus’ Albicerius; astrologers
are featured in the Confessions; and various kinds of diviners
are attacked in the sermons. But, unlike divination, which
could in certain forms and under certain circumstances
channel God’s knowledge, diviners, Augustine believed,
had no access of their own to offer. In the worst cases, for
instance the haruspex in the Confessions, that was because
they performed sacrifices or trafficked with demons. But
even if diviners supplied harmless knowledge, Augustine
believed it was either of a trivial kind, like Albicerius’, or, if
more significant, the result of a divinely inspired conjuction

voces factas, cum oculis nullus videretur, legimus et tenemus. Ipsi denique
homini, non extrinsecus per aures eius aut oculos, sed intus in animo non
uno modo deus loquitur, sed aut in somnis . . . aut spiritu hominis
assumpto, quam graeci extasin vocant, . . . aut in ipsa mente, cum quisque
maiestatem vel voluntatem intellegit’ (CCSL 41.167–8). For the date, see
P.-P. Verbraken, Études critiques sur les sermons authentiques de Saint
Augustin, Instrumenta Patristica 12 (The Hague, 1976), 55.
7
See the article of J. den Boeft in AL ii. 518–19, s.v. ‘Divinatio’.
116 William E. Klingshirn
between the diviner’s chance words and thoughts already in
the inquirer’s mind (Conf. 4. 3. 5). Such for instance was his
explanation in Confessions 7. 6. 9 of the apparent accuracy of
some astrologers’ predictions.8
The connection between Augustine’s opinion of diviners
and his understanding of the disciplines lies just below the
surface of this reasoning. For the kind of diviner against
whom Augustine’s polemic was directed was the diviner
who operated, according to Cicero’s (and ultimately Plato’s)
distinction, by skill (‘ars’) rather than inspiration (‘natura’).9
By ceaselessly attacking such figures in his sermons and
other writings, Augustine aligned himself with existing
Christian teaching, philosophical opinion, and imperial
legislation. But, more significantly for our purposes, he
also implicitly rejected a long Roman tradition of treating
skilled diviners as possessors of a ‘disciplina’, an organized
system of interpreting signs and obtaining knowledge
that was not innate or inspired, but could be learned and
applied.
Why Augustine did this is, in retrospect, not particularly
puzzling. First, in the ‘economy of knowledge’ in which
divination operated, too much emphasis on technical sys-
tems of interpretation threatened to place skill before virtue.
Already evident in Plato’s hostility to artificial divination,
this view repeatedly resurfaces in the Platonic tradition.
Another problem for Augustine and other Platonists was
that technical divination, whether it emphasized skill or
virtue, seemed to overvalue human agency and thus leave no
room for divine autonomy or, in Christian terms, grace. A
third problem was that the successful workings of divination
were widely attributed to daimones, whom Christians con-
demned as demons. But Augustine’s mature views on these
matters, as expressed most fully in City of God, took a long
time to develop. And in the fifteen years from conversion

8
See further W. E. Klingshirn, ‘The Figure of Albicerius the Diviner
in Augustine’s Contra Academicos’, StPatr 38 (2001), 219–23.
9
Plato, Phaedrus 244C–E; Cicero, De Divinatione 1. 6. 12. See, in
general, A. Bouché-Leclercq, Histoire de la divination dans l’Antiquité,
i (Paris, 1879), 29–91.
 Divination and the Disciplines 117
to the Confessions, his changing treatment of the disciplines
can be used as a lens through which to view his continuing
fascination with divination. It is especially interesting to
observe Augustine using techniques from the disciplines
he accepted, particularly grammatical learning and the
dialectical arts of division and classification, against the
disciplines he rejected.10

2. d i v i nat i o n a s a d i s c i p l i n e
For almost half a millennium before Augustine, the over-
whelmingly positive connotations of the word ‘disciplina’
had appealed to Romans who wished to portray artificial
divination as a respectable science. Its core meanings—
‘learning’ and ‘what or how a learner has learned’ (clearly
displayed in the unsyncopated form ‘discipulina’11)—
associated it not only with teaching (‘doctrina’), knowledge
(‘scientia’), and skilled practice (‘ars’), but also with severity,
manners, moral formation, and good order.12 Used of the
philosophical schools and of training in grammar, music, and
other skills, the word brought to mind the social recognition
and organization of the higher arts, and even more re-
assuringly the stabilizing, traditional authority of a teacher,
without whom no one could apprehend a discipline.13 It was,
accordingly, for Cicero the right word to describe not only

10
On the grammarian as ‘a man of distinctions’, see R. A. Kaster,
Guardians of Language: The Grammarian and Society in Late Antiquity
(Berkeley and Los Angeles, 1988), 19.
11
e.g., in the entry for ‘Tages’ in Festus, De Verborum Significatu:
‘Tages nomine, geni filius, nepos Iovis, puer dicitur discipulinam
haruspicii dedisse duodecim populis Etruriae.’
12
The fundamental studies of the word are by A. Gudeman, TLL v. 1
(1915), cols. 1316–26, s.v. ‘disciplina’, and O. Mauch, Der lateinische
Begriff Disciplina: Eine Wortuntersuchung (Freiburg, Switzerland, 1941).
See also H.-I. Marrou, ‘ “Doctrina” et “Disciplina” dans la langue des
Pères de l’Église’, Archivum Latinitatis Medii Aevi 9 (1934), 5–25, at
10–12. See also, in this volume, the definition offered by Catherine
Conybeare, above 59.
13
Augustine, Util. Cred. 17. 35: ‘et si unaquaeque disciplina, quamq-
uam vilis et facilis, ut percipi possit, doctorem aut magistrum requirit’
(CSEL 25[1].46).
118 William E. Klingshirn
haruspicy, suitably denominated the ‘Etrusca disciplina’,15
14

but also augury, its Roman counterpart.16

Cicero declined to pay exactly the same compliment to
astrology, but in the first book of De Divinatione, expressing
views favourable to divination, he did call it, synonymously,
an ‘ars’ and a ‘scientia’ (1. 1. 2).17 Later in the same book he
associated astrologers with haruspices, to whom he explicitly
assigned a ‘disciplina’ (1. 19. 36). Further, in a letter of 46
b c, he associated astrologers with augurs, likewise credited
in the text with a ‘disciplina’.18 But Cicero went no further.
He did list ‘astrologia’ along with the liberal arts of music,
geometry, grammar, and rhetoric in De Oratore (1. 42. 187),
but by this he meant what we would call astronomy.19 Varro
did the same in his De Astrologia,20 possibly one of the nine
books of his Disciplinae.21
After Cicero and Varro, ‘disciplina’ became a standard
Latin term for the teaching of the divinatory sciences.22 This
usage reached its high point in fourth-century pagan circles.
As before, ‘disciplina’ continued to be used of individual
14
‘haruspicum disciplina’, ‘haruspicinae disciplina’: De Divinatione
1. 2. 3, 41. 91; 2. 23. 50; see also 2. 12. 28, 38. 80.
15
C. O. Thulin, RE vi. 1 (1907), cols. 725–30, s.v. ‘Etrusca disciplina’.
16
‘Atqui et nostrorum augurum et Etruscorum disciplinam . . . res ipsa
probavit’: De Natura Deorum 2. 10. See also De Divinatione 1. 17. 33,
47. 105; 2. 35. 74; De Legibus 2. 8. 20, 13. 32–3; De Natura Deorum 2. 9.
17
On the related meanings of ‘disciplina’, ‘ars’, and ‘scientia’, see
G. Schrimpf, ‘Disciplina’, in Joachim Ritter (ed.), Historisches Wörterbuch
der Philosophie, ii (Basel and Stuttgart, 1972), cols. 256–9.
18
Fam. 6. 6. 7 (= 234. 7, ed. Shackleton-Bailey).
19
For the difference, see Cicero, De Divinatione 2. 42. 88, and
more generally, W. Hübner, Die Begriffe ‘Astrologie’ und ‘Astronomie’ in
der Antike, Akademie der Wissenschaften und der Literatur, Mainz,
Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse, 1989,
no. 7 (Stuttgart, 1990).
20
Cassiodorus, Institutiones Divinarum et Saecularium Litterarum
2. 7. 2.
21
F. Ritschl, ‘De M. Terentii Varronis disciplinarum libris com-
mentarius’, in his Opuscula Philologica, iii (Leipzig, 1877), 361; I. Hadot,
Arts libéraux et philosophie dans la pensée antique (Paris, 1984), 175. See
now the essay by Shanzer above, esp. 93.
22
For haruspicy, see Thulin, RE vi. 1, col. 725; for augury, see
J. Linderski, ‘The Augural Law’, ANRW 16.3 (1986), 2146–312, at
2237–40; and for astrology, see Vitruvius, De Architectura 1. 1. 16; 9. 2. 1,
6. 2; Apuleius, De Dogmate Platonis 1. 3.
 Divination and the Disciplines 119
23 24
forms of divination: augury, haruspicy, and, at least once,
astrology.25 But the term also began to be used of divinatory
systems in general. In his description of the state of learn-
ing in contemporary Alexandria, for instance, Ammianus
Marcellinus lists among ‘the masters of the disciplines’
(‘disciplinarum magistri’) those skilled in geometry, music,
astronomy, and mathematics, as well as those who possess
the knowledge (‘scientia’) that ‘shows the ways of the fates’
(22. 16. 17). In a digression on the Persian magi, he tells us
that in addition to what they learned from Zoroaster, they
handed on to their offspring the ‘disciplines of predicting
the future’ (‘disciplinis praesentiendi futura’, 23. 6. 33). In
his account of the treasonous divination session performed
by the diviners Patricius and Hilarius, he represents Hilarius
as saying that the ring they used for the ceremony was
‘consecrated by secret disciplines’ (‘anulum . . . mysticis
disciplinis initiatum’), that is by learned divinatory rituals
(29. 1. 31). Finally, in the long analysis of divination that
justifies Julian’s skill in it, Ammianus explains that humans
receive the ‘gifts of divining’ (‘munera divinandi’) from what
they obtain ‘through the various disciplines’ (‘per disciplinas
varias’, 21. 1.8). It was therefore fitting for a philosopher like
Julian, ‘both learned and zealous for all knowledge’ (‘erudito
et studioso cognitionum omnium’, 21. 1. 7), to possess
significant divinatory abilities.
Such all-encompassing knowledge, potent both in divina-
tion and in philosophical wisdom, had become a common-
place by the fourth century. Consider Servius’ explanation
of the phrase ‘hominum divumque interpres’, used of the
Etruscan diviner Asilas in Aeneid 10. 175:
23
e.g., in commentaries on the Aeneid by Servius (5. 530) and Servius
Danielis (2. 692–3; 3. 60, 90, 359; 4. 453, 462; 5. 530). See also Macrobius,
Sat. 1. 24: ‘Apud poetam nostrum, inquit, tantam scientiam iuris
auguralis invenio, ut, si aliarum disciplinarum doctrina destitueretur, haec
illum vel sola professio sublimaret’ (“In our poet”, he said, “I find so great
a knowledge of the augural law that if the teaching of the other disciplines
were left behind, the profession of this knowledge, even by itself, would
ennoble him”).
24
Servius on Aen. 1. 733; 4. 166; Servius Danielis on Aen. 1. 2, 422;
Ammianus Marcellinus 21. 1. 10.
25
Firmicus Maternus, Mathesis 2. 30. 14. See also Arnobius, Adversus
Nationes 1. 5: ‘at Chaldaeorum ex reconditis disciplinis’.
120 William E. Klingshirn
An interpres is an intermediary (medium). For the interpreter
(interpretator) both of the gods and of the humans to whom he
indicates divine thoughts (divinas mentes) is called an interpres.
What Nigidius Figulus says must also be noted, that these arts
(artes) are so interconnected that one cannot exist without another.
For this reason Virgil assigns the knowledge (scientia) of all the
arts of divining (omnium divinandi artium) to those whom he
wishes to prove perfect.26

The perfect diviner of this ideal was like the diviner

described in a horoscope by Firmicus Maternus: ‘knowing
much and inquiring thoroughly into the secrets of all the
disciplines’ (‘multa sciens et omnium disciplinarum secreta
perquirens’).27 For many, Virgil himself was the supreme
example of such comprehensive disciplinary knowledge.28
Macrobius, for instance, describes him in the Commentary on
the Dream of Scipio as ‘lacking in no discipline’ (‘nullius
disciplinae expers’, 1. 6. 44) and ‘most skilled in all the dis-
ciplines’ (‘disciplinarum omnium peritissimus’, 1. 15. 12).29
It is thus with the highest praise for Vettius Agorius Praetex-
tatus that Aconia Fabia Paulina is presented on their joint
epitaph as saying, ‘Releasing me pure and chaste from the
fate of death by the good of your disciplinae, husband, you
lead me into temples and dedicate your servant to the gods.’30
The most systematic association between the liberal and
the divinatory disciplines can be found in the Marriage

26
‘ “Interpres” medium est: nam et deorum interpretator, et hominum,
quibus divinas indicat mentes, interpres vocatur. Et notandum quod ait
Nigidius Figulus, has artes ita inter sese esse coniunctas, ut alterum sine
altero esse non possit: unde his quos perfectos vult probare Vergilius,
omnium divinandi artium praestat scientiam’: G. Thilo and H. Hagen
(eds.), Servii Grammatici qui Feruntur in Vergilii Carmina Commentarii,
ii (Leipzig, 1883), 408.
27
Mathesis 5. 1. 16; cf. 5. 1. 17: ‘multarum artium ac disciplinarum
scius’.
28
J. den Boeft, ‘Nullius Disciplinae Expers: Virgil’s Authority in (Late)
Antiquity’, in L. V. Rutgers et al. (eds.), The Use of Sacred Books in the
Ancient World (Leuven, 1998), 175–86.
29
Similar language is found in Servius Danielis: ‘omnium discipli-
narum scientiam’ (on Aen. 1. 305) and ‘disciplinis omnibus eruditam’ (on
Aen. 8. 314).
30
‘Tu me, marite, disciplinarum bono / puram ac pudicam sorte mortis
eximens / in templa ducis ac famulam divis dicas’ (ILS 1259. 22–4).
 Divination and the Disciplines 121
31
of Philology and Mercury of Martianus Capella. Here
the liberal disciplines, allegorized as maidens raised in
Mercury’s household (1. 36K), are balanced by the seven
mantic disciplines, allegorized as maidens reared by Philolo-
gy’s mother Phronesis (9. 892K). Unlike the maidens repre-
senting the liberal disciplines, the virgins raised by Phronesis
are never subjected to a full examination. Thus we learn the
specialties of only four of them (9. 894–6K); the other three
are a mysterious trinity.32 Genethliace represents Astrology;
Symbolice is concerned with the interpretation of omens and
portents; Oeonistice stands for augury through birds; and the
last sister interprets Jupiter’s lightning through haruspicy.33
Since the interpretation of omens and portents in Roman
religion was the usual province of haruspices (although
other kinds of diviners could also be employed), the four
divinatory specialties named here are in fact equivalent to
the same three we have seen characterized elsewhere as
‘disciplinae’.
Compared with the liberal disciplines, all seven mantic
disciplines are said to be endowed with an equal or greater
erudition (9. 892K). This consists of a knowledge of divine
secrets (‘superum . . . pectorum arcana’) and of inter-
pretations of the hidden divine will (‘voluntatis abditae
interpretamenta’, 9. 893K). They alone are said to be able
to mediate between humans and gods (‘nam inter divina

31
On the work and its late fifth-century date, see D. Shanzer, A
Philosophical and Literary Commentary on Martianus Capella’s De
Nuptiis Philologiae et Mercurii Liber 1 (Berkeley and Los Angeles, 1986),
1–28. Paragraph numbers below are from the edition of Kopp.
32
In his commentary on Martianus Capella, Remigius of Auxerre
states that Augustine names all seven mantic disciplines: ‘Augustinus
in libro De Civitate Dei earum nomina dicit’: C. E. Lutz (ed.), Remigii
Autissiodorensis Commentum in Martianum Capellam, Libri III–IX
(Leiden, 1965), 298. I am not aware of any single passage in which
Augustine does this, although he does list many types of divination at
various points, including astrology (Civ. 5. 7) and augury, haruspicy,
inspired prophecy, and dreams (Civ. 8. 16).
33
For these identifications, see L. Cristante, Martiani Capellae De
Nuptiis Philologiae et Mercurii Liber IX (Padua, 1987), 196–202, and
L. Lenaz, Martiani Capellae De Nuptiis Philologiae et Mercurii Liber
Secundus (Padua, 1975), 110–11.
122 William E. Klingshirn
humanaque discidia solae semper interiunxere colloquia’,
9. 893K), and when shamefully expelled from the earth
(‘terris indecenter expulsas’, 9. 898K)—a telling indication
of Martianus’ late date34—they will be fittingly received
in the stars. Raised by Phronesis in the same household as
Philology and constituting a part of her dowry, they can
therefore be understood in the allegory to have played a
significant formative role in her education. Accordingly, the
Muses Terpsichore and Euterpe praise Philology for her
knowledge of the arts of divination (2. 124–5K).
Already in the fourth century, the full conjunction of dis-
ciplinary and divinatory knowledge that Martianus Capella
was later to bestow upon Philology was crystallized into
the set phrase ‘scientiae disciplinis’ (‘by the disciplines of
knowledge’). Arnobius of Sicca, writing before 310, pro-
vides our earliest example. Addressing pagans who claim to
possess ‘sapientia’ and ‘intelligentia’, he demands to know
from where they have received so much wisdom (‘tantum
sapientiae’) or mental perception (‘acuminis et vivacitatis
tantum’), or from what disciplines of knowledge (‘scientiae
disciplinis’) they have obtained so much understanding
(‘tantum cordis’) or derived so much divination (‘divina-
tionis tantum’).35 Beneath Arnobius’ sarcasm, we glimpse
the pagan claim that the ‘disciplines of knowledge’ were the
source of understanding and divination. A more direct
example comes from the introduction to Book 5 of Firmicus
Maternus’ Mathesis, datable to the 330s. Addressing his
dedicatee Mavortius, the astrologer promises that a capable
mind prepared by the reading of the four previous books and
illuminated by the stars ‘will arrive at the divine secrets of
this knowledge (divina istius scientiae secreta) and, initiated
in all the disciplines of this knowledge (initiatum omnibus
istius scientiae disciplinis), will receive the whole order of

34
Shanzer, A Philosophical and Literary Commentary, 21.
35
Arnobius, Adversus Nationes 2. 6: ‘Unde quaeso est vobis tantum
sapientiae traditum, unde acuminis et vivacitatis tantum vel ex quibus
scientiae disciplinis tantum cordis adsumere, divinationis tantum potuistis
haurire?’ On this sense of ‘cor’, see also Adv. Nat. 2. 12: ‘ac dum vestris
fiditis cordibus et quod typhus est sapientiam vocatis.’
 Divination and the Disciplines 123
divine interpretation (totum ordinem divinae interpreta-
tionis)’. The result will be the ability ‘both to discover and
explain human fates, (fata hominum et invenire pariter et
explicare)’.36
Christians who used the phrase ‘scientiae disciplinis’
avoided its divinatory connotations by distinguishing the
knowledge produced by the disciplines from true wisdom.
Commenting on the proverb ‘Dilige sapientiam et disci-
plinam’ (Prov. 1: 2), Ambrose of Milan wrote that ‘wisdom is
what flows from understanding, and discipline is moulded
into a certain natural shape by the virtues and strengthened
as a mental condition by the disciplines of knowledge’.37 The
biographers of Caesarius of Arles sharpened this distinction
in the mid-sixth century when they described a futile
attempt by the grammarian and rhetor Julianus Pomerius
to polish their hero’s ‘monastic simplicity’ (‘monasterialis
. . . simplicitas’) by the ‘disciplines of secular knowledge’
(‘saecularis scientiae disciplinis’).38 Another Christian
approach was to emphasize the distinction between secular
knowledge and true wisdom by using the phrase ‘sapientiae
disciplinis’ (‘by the disciplines of wisdom’).39

36
Firmicus Maternus, Mathesis 5, praef. 1: ‘Si itaque capax ingenium
et flagrantis animi desiderio commotum praecedentes hauserit libros et
sit oportuna stellarum radiatione conceptum, ad divina istius scientiae
secreta perveniet, et, initiatum omnibus istius scientiae disciplinis, totum
ordinem divinae interpretationis accipiet, ut possit fata hominum et
invenire pariter et explicare.’
37
Ambrose, Explanatio Psalmorum xii, Ps. 43: 2: ‘Sapientia est quae
sensu scaturit, disciplina in habitudinem quandam naturae coalita forma
virtutum et mentis confirmata sententia scientiae disciplinis’ (CSEL 64.
260).
38
Vita Caesarii 1. 9, ed. G. Morin, Sancti Caesarii Episcopi Arelatensis
Opera Omnia nunc primum in unum collecta, ii (Maredsous, 1942), 300.
For the deliberate use here of ‘secular’ in place of ‘liberal’, see Augustine,
Civ. 6. 2 (CCSL 47.167): ‘ut in omni eruditione, quam nos saecularem,
illi autem liberalem vocant.’ The credentials of Julianus Pomerius are
given in his entry in Kaster’s prosopography of grammatici (Guardians of
Language, 342–3, no. 124).
39
The phrase is a favourite of Ambrose’s. See his De Isaac vel Anima 1.
2: ‘et hauriret purae sapientiae disciplinas’; 3. 9: ‘et interioris sapientiae
disciplinas’ (CSEL 32[1].642, 648); De Iacob et Vita Beata 1. 1. 4: ‘et
sapientiae instrueret disciplinis’; 1. 5. 17: ‘sapientiae disciplinis’
(CSEL 32[2].6, 15); De Ioseph 13. 75: ‘sapientiae multiplicis disciplinas’
124 William E. Klingshirn
Augustine did not use these phrases himself, but he
repeatedly problematized the assumptions that stood behind
them. What were the disciplines that gave access to know-
ledge or wisdom? Could they teach something that was not
good or true? What precisely did they teach? And, in the
matter of divination, why could they not transmit the truest
and most secret knowledge of all—the knowledge of past,
present, and future—that is, God’s knowledge?

3. d i v i nat i o n i n t h e c a s s i c i a c u m d i a lo g u e s
In Augustine’s early writings, the liberal disciplines them-
selves and people well educated in them seem capable
of almost anything.40 There cannot be a bad discipline,41 dis-
ciplines cannot teach what is false,42 and whatever can be
known is known through a discipline.43 This high valuation
of the disciplines set strict criteria for inclusion, which the
liberal disciplines could meet and other disciplines could
not. Broader criteria might have included other disciplines.
Apuleius’ criteria were so broad that even robbers could
have a ‘disciplina’ (Metamorphoses 4. 9, 18)—a standard
no less telling for all its irony. Magic was a discipline for him
too (Met. 2. 20; 3. 18), and sometimes a very wicked one
(Met. 3. 16: ‘maleficae disciplinae perinfames sumus’;

(CSEL 32[2].117); Explanatio Psalmorum xii, Ps. 36: 64: ‘quae in nullo
discrepauit a sapientiae disciplinis’ (CSEL 64.123). See also Jerome,
Ep. 99. 2: ‘et de intimis sapientiae disciplinis’; Ep. 100. 10: ‘et sapientiae
disciplinis suas animas dedicarunt’ (CSEL 55.212, 222); and Rufinus’
translation of Origen, In Leviticum Homiliae, hom. 9. 2: ‘si qui mentem
suam adornaverit sapientiae disciplinis’ (SC 287.80).
40
Ord. 1. 8. 24: ‘nam eruditio disciplinarum liberalium modesta sane
atque succincta et alacriores et perseverantiores et comtiores exhibet ama-
tores amplectendae veritati, ut ardentius appetant et constantius insequan-
tur et inhaereant postremo dulcius, quae vocatur, Licenti, beata vita’ ‘For
instruction in the liberal disciplines—measured, of course, and kept con-
cise—makes those who love to embrace truth more eager, persistent, and
elegant, so that they seek more passionately, pursue more steadfastly, and
in the end cling more delightfully to what is called, Licentius, the happy
life’.
41
Lib. Arb. 1. 1. 2 (CCSL 29.212).
42
Sol. 2. 11. 20–1 (CSEL 89.71–4).
43
Imm. An. 1. 1 (CSEL 89.102).
 Divination and the Disciplines 125
Met. 9. 29: ‘armis facinerosae disciplinae suae’). For
Firmicus Maternus (after his conversion), pagan signs and
symbols were the devil’s ‘disciplina’ (De Errore Profanarum
Religionum 18. 1: ‘quae illis in istorum sacrilegiorum
coetibus diaboli tradidit disciplina’).
By valorizing the disciplines as he did, Augustine gave up
the opportunity to depict artificial divination as a wicked or
diabolic discipline. But he also gained the ability to recast
divination as an ‘anti-discipline’ that could then be con-
trasted with real disciplines like grammar and music. The
story of the diviner Albicerius fills precisely this role in Book
1 of Contra Academicos.44 All the interlocutors in the dia-
logue agree that his knowledge, such as it was, had nothing
to do with intellectual training. Completely lacking gram-
matical schooling (‘qui grammatici scholam vix transiens
vidisset aliquando’, 1. 6. 18) and living the kind of sensual
and depraved life that only someone with no humane
‘scientia’ could lead (‘quae si Albicerius ille didicisset, num-
quam . . . tam luxuriose deformiterque vixisset’), Albicerius
possessed none of the most honorable disciplines (‘honestis-
simas disciplinas’, 20). His ability to recite lines of poetry
from other people’s minds was something that even the most
uneducated (‘imperitissimis’) could manage—with the help
of demons (‘ab huius aeris animalibus quibusdam vilissimis,
quos daemonas vocant’, 20). For he could neither teach
the metres themselves (‘ipsa metra’) nor make up his own
verses (‘versus proprios’, 21). Albicerius’ low intellectual
profile and the demonic inspiration that stood in the place
of real learning thus allow Augustine, in the character of
Flaccianus, to recommend the disciplines over demonic
divination:
For that most learned man (vir ille doctissimus) used to ask those
who admired such things whether Albicerius could teach grammar
or music or geometry. But who could know him and not admit that
he was entirely unskilled (imperitissimum) in all of these matters?
For this reason he used to mightily urge that those who had learned

44
On the subject of Albicerius’ appearance in Book 1, I have benefited
from conversations with Karin Schlapbach and from her recently pub-
lished commentary, Augustin: Contra Academicos (vel De Academicis)
Buch 1, Patristische Texte und Studien 58 (Berlin and New York, 2003).
126 William E. Klingshirn
such things (qui talia didicissent) not hesitate to prefer their own
souls to that divination (illi divinationi) and take care to instruct
and advise their own mind in those disciplines in which it was
possible to leap across and fly over the airy nature of invisible
spirits. (1. 7. 21)
Standing behind this contest between demonic divination
and the soul’s own learning is a complicated Neoplatonic
debate over the proper means to ascend to God: more
specifically, by what combination of virtue, knowledge,
ritual, and skill this might be accomplished.45 In such a
context, if one had to make a choice, an emphasis on the
disciplines preferred virtue and knowledge, an emphasis on
divination, ritual and skill. The case of Albicerius makes
Augustine’s choice clear from the beginning, and prepares
the way for the treatment of the disciplines not only in the
remaining two books of Contra Academicos, but also, more
prominently, in De Ordine.
In this dialogue, divination is repeatedly presented as the
wrong means for acquiring knowledge. As an anti-discipline,
it serves to underscore the point that the divinely ordered
universe is best understood by the orderly acquisition of the
liberal disciplines. So, early in Book 1, Licentius jokingly
asks why, if superstitious people (‘superstitiosi’) can take
auguries from mice (‘de muribus augurari’), he cannot inter-
pret Augustine’s intervention as an oracular voice (1. 3. 9,
with 1. 5. 14), something that would be consistent with the
whole scheme of divine order. Mice were considered sacred
to Apollo,46 and these musings accordingly elicited from
Augustine the Virgilian tag ‘Sic Pater ille deus faciat, sic
altus Apollo!’,47 suitably Christianized by the change of the
genitive plural ‘deum’ (‘of the gods’) to ‘deus’ (Aen. 10. 875:
‘Sic pater ille deum faciat, sic altus Apollo’). This reference

45
The subject is vast. A good starting-point is P. Athanassiadi,
‘Dreams, Theurgy and Freelance Divination: The Testimony of
Iamblichus’, JRS 83 (1993), 115–30.
46
M. Tulli Ciceronis De Divinatione, ed. A. S. Pease, University of
Illinois Studies in Language and Literature 6 (1920), 438; repr. Darmstadt
(1963), 276.
47
I take this reading from J. Doignon, ‘Problèmes textuels et modèles
littéraires dans le livre I du De ordine de saint Augustin’, REAug 24 (1978),
71–86, at 79–82.
 Divination and the Disciplines 127
to Apollo prompts Augustine to recall Aeneid 3. 88–9 (‘Quem
sequimur? Quove ire iubes? Ubi ponere sedes? / Da, pater,
augurium, atque animis illabere nostris’), which he then
Christianizes by transferring Apollo’s divinatory powers to
God the Father, ‘who now sends an augurium and slips into
our minds’. The ‘vates’ of truth, Augustine concludes, are
not those inspired by Apollo, but those who can be wise
(‘sapientes’, 1. 4. 10). Artificial divination (in the form of
augury by mice) is rejected, while natural divination is shown
to be none other than the wise person’s search for truth.
Licentius extends this line of reasoning by (naïvely) won-
dering what would have happened if some great ‘vates’ or
astrologer (‘Chaldeus’) had predicted the subject of the dia-
logue long before its committal to writing. He is uncertain
whether any diviner has actually done so, either on his own
or in response to an inquirer (‘a consultore coactus’48), but
supposes that if a chain of unknown causes actually links one
event to the next, then, theoretically at least, this hidden
order could be uncovered. Along with the events themselves,
so too could the words that describe them in a book be pre-
dicted beforehand (1. 5. 14). Whatever the merits of the
argument, it is immediately shut down by Augustine’s
observation that his young interlocutor does not realize
how much has been said against divination (‘contra divina-
tionem’) and by what sorts of men (1. 6. 15). Again, though
not as bluntly as in Contra Academicos, the skills of diviners
are shown to be relevant to the problem at hand, but at the
same time to be inappropriate for investigating it.
In Book 2, as Augustine approaches his description of the
liberal disciplines, which belong to the category of learning
by reason, he pauses briefly at the category of learning by
authority, which can be human or divine. Here he draws
renewed attention to the shortcomings of divination. Divine
authority is more secure than human authority, but students
should beware of what masquerades as divine authority. In
anti-demonic language strongly reminiscent of Contra Aca-
demicos, Augustine warns inquirers against ‘the wondrous

48
The phrase recalls C. Acad. 1. 7. 21: ‘vel coactus a quopiam consulto-
rum’. The similarity is noted by Schlapbach, Augustin: Contra Academi-
cos, 186.
128 William E. Klingshirn
deception of airy creatures’ (‘aeriorum animalium mira fal-
lacia’), who deceive curious, acquisitive, and fearful souls
‘through certain divinations and not a few interpretations of
matters pertaining to the bodily senses’ (‘per rerum ad istos
sensus corporis pertinentium quasdam divinationes nonnul-
lasque sententias’, 2. 9. 27). The message is clear. Not only
does demonic divination have nothing to do with the reason
from which the disciplines issue; it also has nothing to do
with authority, either divine or human. In the framework of
learning Augustine sets up in De Ordine, divination has no
place. When we arrive at the description of astronomy
(‘astrologia’) in the list of the liberal disciplines, no question
of divination arises. Augustine merely nods at the disci-
pline’s ambiguity, ‘a great proof for the pious, and a tor-
ment for the curious’ (‘magnum religiosis argumentum
tormentumque curiosis’, 2. 15. 42).

4. d i v i nat i o n a n d s u pe r st i t i o n i n
d e d o c t r i na c h r i st i a na
In the aftermath of the Cassiciacum dialogues, Augustine
retreated from his earlier valuation of the liberal disciplines,
eventually admitting that he had attached too much weight
to them (‘et quod multum tribui liberalibus disciplinis’, Retr.
1. 3. 2). By the time he had completed the second book of
De Doctrina Christiana, he no longer argued that the liberal
disciplines were the key to all wisdom: at best, they could
serve as useful tools for interpreting the unknown signs of
Scripture. The narrower function Augustine now assigned
to the liberal disciplines also had the effect of further
hardening their division from other disciplines of know-
ledge. Thus, in the synopsis of pagan learning that occupies
the second half of Book 2, the liberal disciplines belong
to the category of divinely instituted ‘doctrinae’, whereas
divination is placed in the category of human institutions.
The liberal disciplines are further distinguished from
divination by the fact that they stand at the top of their genus
(because they pertain to reason rather than the bodily
senses), whereas in its genus divination stands at the bottom,
since it is a form, indeed the very paradigm, of ‘supersti-
tion’. To appreciate the full complexity of this categoriza-
 Divination and the Disciplines 129
tion, we need to examine the role that divination plays as an
anti-discipline in Book 2 of De Doctrina Christiana.
Book 2 is centrally about the knowledge (‘scientia’) needed
by the interpreter of Scripture, which Augustine uses a
semiotic framework to discuss.49 The book focuses mostly on
unknown signs, leaving the subject of ambiguous signs for
Book 3. It deals with both literal and metaphorical signs,
and with the knowledge of ‘languages’ (2. 11. 16, 16. 23) and
‘things’ (2. 13. 20, 16. 23–4) needed to understand these. But
while a knowledge of languages and things sufficed for many
problems in Scripture, Augustine also thought that inter-
preters needed an understanding of the larger frameworks of
meaning in which signs made sense. Such a ‘knowledge of
signs’ (2. 13. 20), however, was dangerous, both intrinsically,
because of its tendency to generate excessive pride (2. 13. 20,
38. 57, 42. 63), and extrinsically, because of its connections
to pagan learning. At least in part, Augustine’s classification
of the forms of knowledge was intended to protect his
readers from these dangers.
He prefaces his formal classification with a preliminary
discussion of mathematics (2. 16. 25) and music (2. 16. 26),
disciplines closely related to one another.50 Their function
here is to establish the absence of ‘superstition’ as the first
condition of legitimacy for all human learning: ‘Indeed,
we find both number and music mentioned with respect in
several places in the holy scriptures. But we must not listen
to the fictions of pagan superstition, which have represented
the nine Muses as the daughters of Jupiter and Memory’
(2. 16. 26–17.27).51 The question of how many Muses there
were and who their parents were combines number and
49
On its structure, see G. A. Press, ‘The Content and Argument of
Augustine’s De Doctrina Christiana’, Augustiniana 31 (1981), 165–82, at
173–7, and K. Pollmann, Doctrina Christiana: Untersuchungen zu den
Anfängen der christlichen Hermeneutik unter besonderer Berücksichtigung
von Augustinus, De doctrina christiana (Freiburg, Switzerland, 1996),
esp. 147–55.
50
Doct. Chr. 4. 20. 41: ‘illa musica disciplina, ubi numerus iste
plenissime discitur’.
51
Trans. R. P. H. Green, Augustine: De Doctrina Christiana (Oxford,
1995). On the contrast between the treatment of the Muses here and in Ord.
2. 14. 41, see S. MacCormack, The Shadows of Poetry: Vergil in the Mind
of Augustine (Berkeley and Los Angeles, 1998), 63–4, and above, 105–9.
130 William E. Klingshirn
music into a single problem, which Augustine attempts to
solve—inconclusively, as it turns out—by an amusing story
from Varro. ‘But whether Varro’s story is true or not,’ he
continues, ‘we should not avoid music because of the associ-
ated pagan superstitions if there is a possibility of gleaning
from it something of value for understanding holy scripture’
(2. 18. 28, trans. Green).
After these programmatic comments, Augustine embarks
on a more systematic treatment of the problem, based on a
distinction between the ‘doctrinae’ that humans establish
themselves and those that they take notice of, whether
already enacted under divine administration or divinely
inscribed in nature (2. 19. 29).52 In both genera, Augustine
takes pains to distinguish between legitimate and illegitimate
forms of learning based on the absence or presence of
superstition. And although he occasionally refers to other
forms of superstition, such as the ‘magicae artes’ that he
knew almost exclusively from literature (‘quae quidem
commemorare potius quam docere assolent poetae’, 2. 20.
30) and the incantations, symbols, amulets, and other
remedies used by folk healers (2. 20. 30, 23. 36, 29. 45), most
of his examples come from divination, and specifically tech-
nical divination. This makes sense, since Book 2 is about
knowledge and the interpretation of unknown signs, and
technical divination dealt with both,53 though illegitimately,
as Augustine had come to believe.
The first genus concerns humanly devised forms of learn-
ing. Augustine begins with its lower, superstitious branch
(2. 20. 30–24. 37). Both pulled by his theory of signs and
pushed by his need to reject divination as a discipline,
Augustine defines superstition as a pact of signification
made between humans and demons. The practical effect of
this definition is to radically alter the field of signs inter-

52
On the underlying distinction between nomos and physis that this
represents, see C. Schäublin, ‘De Doctrina Christiana: A Classic of
Western Culture?’, in D. W. H. Arnold and P. Bright (eds.), De Doctrina
Christiana: A Classic of Western Culture (Notre Dame, Ind., and London,
1995), 47–67, at 50–8.
53
H. R. Seeliger, ‘Aberglaube, Wissenschaft und die Rolle der historica
narratio in Augustins Doctrina Christiana’, Wissenschaft und Weisheit 43
(1980), 148–155, at 151.
 Divination and the Disciplines 131
preted by diviners. Instead of dealing with signs inscribed
in nature by divine providence, as theories favourable to
technical divination envisioned,54 Augustine’s diviners deal
with signs sent by fallen angels, whose fugitive meanings are
based only on convention. Divinatory knowledge thus
emerges as an anti-type of disciplinary knowledge, since it
deals not with what God has established in creation but
merely with what humans and demons with the same inten-
tions have agreed upon.55 While it was not necessarily false
(2. 23. 35), such knowledge was deceptive, illusory, and of
course very dangerous (2. 23. 36).
In this section Augustine concentrates on only three kinds
of technical diviners: augurs, haruspices, and astrologers.56
As we have seen, these were the only kinds of diviners
considered in the Latin tradition to practice a discipline.
Other types attacked by Augustine elsewhere, such as lot
diviners or dream-interpreters, were not thought by anyone
to be so endowed, and so are of no concern to him here.
The discussion is carefully arranged to give prominence to
astrology, Augustine’s favourite divinatory science. After a
brief mention of haruspices, augurs, healing practices, and
other superstitions (2. 20. 30), astrologers are attacked by
traditional arguments in a single continuous section (2. 21.
31–22. 34). This is followed by a more general discussion of
demonic involvement in divinatory signs (2. 23. 35–24. 37),
in which further references are made to haruspices (by the
portents they explained, 2. 23. 36) and augurs (2. 24. 37).
Augustine’s main interest in this section lies in explaining
how superstitious and especially divinatory knowledge rests
on human and demonic conventions and intentions, and not
in directly contrasting this knowledge with disciplinary
knowledge. Indeed, he only once explicitly compares any of
the superstitious ‘doctrinae’ to any of the disciplines. ‘To
this category (ad hoc genus) also belong all the amulets and
54
Cicero, De Divinatione 1. 52. 118: ‘sed ita a principio inchoatum esse
mundum ut certis rebus certa signa praecurrerent’.
55
On the role of intentionality, see R. A. Markus, ‘Augustine on Magic:
A Neglected Semiotic Theory’, in his Signs and Meanings: World and Text
in Ancient Christianity (Liverpool, 1996), 135–9.
56
Only the biblical examples he cites (2. 23. 35) belong to the category
of natural divination.
132 William E. Klingshirn
remedies that the discipline of physicians (medicorum . . . dis-
ciplina) likewise condemns’ (2. 20. 30). It is only later in the
book that the full meaning of this statement emerges, in a
passage contrasting the ‘natural’ power of true medicine
with the ‘conventional’ power of superstitious healing
(2. 29. 45). At this point, ‘the discipline of physicians’ may
be taken to stand for all the true disciplines, which are found
in what God, not humans, have established, and which must
therefore be opposed to superstitious knowledge.
Following his discussion of the superstitious branch of
humanly established learning, Augustine moved to its
higher, non-superstitious branch (2. 25. 38–26. 40), which by
definition included no divination. Its first subdivision
(‘superflua luxuriosaque’) consists mainly of the arts, like
dancing, painting, and literature, that represent something
else, whose signification is based on convention (2. 25. 38).
Its second subdivision (‘commoda . . . et necessaria’) con-
tains the social and economic conventions on which civilized
life depends (2. 25. 39–40) and also human sign systems like
writing, language, and shorthand symbols (2. 26. 40). As a
means of representing rational thought, these sign systems
make a smooth transition to the second genus, which consists
of divinely established ‘doctrinae’.
This genus is divided into two branches, the lower per-
taining to the bodily senses (2. 27. 41–30. 47), the higher to
rational thought (2. 31. 48–38. 56). In his discussion of the
bodily senses, Augustine returns to the question of supersti-
tion. He draws explicit contrasts between ‘doctrinae’ in this
category and superstitions that he had identified earlier in
the book. Following the order of those superstitions, his com-
parisons served three purposes. The most immediate was to
differentiate learning based on the senses from any kind of
superstition, which always threatened when bodies were
involved. This differentiation served in the second place to
define legitimate ‘doctrinae’ more exactly by comparing them
with illegitimate ‘doctrinae’. The third purpose was to dis-
place divination from the category it would otherwise have
occupied. This is illustrated in Martianus Capella, where the
mantic disciplines are associated with Philology, who repre-
sents human learning, and the liberal arts with Mercury, who
represents divine reason. As Ilsetraut Hadot explains, while
 Divination and the Disciplines 133
the liberal arts ‘exist independently of human realities, in the
very realm of Ideas’, the mantic arts, although sacred, ‘are
closely tied to the human condition and its needs’.57
Augustine begins his comparisons with history. As a narra-
tive of what has been done under God’s order and adminis-
tration, history is located in the genus of divinely instituted
‘doctrinae’ (2. 28. 44). To explain this counterintuitive cate-
gorization, Augustine differentiates between the narration
of ‘what has been done’ (‘facta’) and the narration of ‘what
must be done’ (‘facienda’). In keeping with his purposes, he
cites divination as an example of the latter. ‘History narrates
what has been done in a faithful and useful way, whereas the
books of haruspices and similar writings set out to teach
things that must be done or observed, with the audacity of
a counselor and not the reliability of a witness.’58 The
link between ‘the books of haruspices and similar writings’
mentioned here and ‘the books of haruspices and augurs’
mentioned earlier (2. 20. 30) is unmistakable, and serves to
introduce a train of similar comparisons.
In the next passage, Augustine commends a knowledge of
natural history, but warns readers against putting it to super-
stitious uses (2. 29. 45). He again points back to an earlier
passage, in this case on superstitious ‘remedia’ (2. 20. 30).
‘For we have separately treated that category too, already
distinguished from this legitimate and unrestricted kind’
(‘nam et illud genus iam distinctum ab hoc licito et libero
separavimus’). Links between the two sections include the
mention of incantations, magical symbols, and amulets, and
the contrast, here made explicit, between natural and con-
ventional healing.
Knowledge of the stars comes next (2. 29. 46). In this
section, again deploying highly traditional arguments,59

57
‘La loi des nombres et de leurs rapports mutuels, qui forme la
base des sept disciplines cycliques, existe indépendamment des réalités
humaines, au niveau même des Idées. Par contre, les arts mantiques et la
théurgie sont des arts sacrés, il est vrai, mais liés étroitement à la condition
et aux besoins humains’: Hadot, Arts libéraux et philosophie, 151.
58
‘Historia facta narrat fideliter atque utiliter, libri autem haruspicum
et quaeque similes literrae facienda vel observanda intendunt docere,
monitoris audacia, non indicis fide’ (2. 28. 44, trans. adapted from Green).
59
e.g., Cicero, De Divinatione 2. 42. 88 ff.
134 William E. Klingshirn
Augustine differentiates the predictions of astronomers
from those of astrologers. Both use their knowledge of signs
to reason from the known to the unknown, but astronomers
focus their predictions on the stars themselves, in contrast
to astrologers, who try to draw conclusions about human
deeds and outcomes (‘in nostra facta et eventa’, 2. 29. 46).
The passage contains obvious echoes of Augustine’s earlier
description of astrologers, whose attempts to predict ‘both
our actions and the outcomes of those actions’ (‘vel actiones
nostras vel actionum eventa’, 2. 21. 32) lead them into super-
stitious error (2. 22. 33). Such comparisons between the pre-
dictions made by diviners and those made by other skilled
professionals are often found in discussions of divination. In
addition to helping to distinguish between astronomers and
astrologers, they also help to assign ‘the other arts’ to their
proper category (2. 30. 47).
In this subdivision, Augustine gathers together the arts
of manufacture (construction, carpentry, pottery making),
service (medicine, agriculture, sailing), and action (dancing,
running, wrestling). Some of these arts have been mentioned
in other categories: dancing was classified as a superfluous
human institution at 2. 25. 38, and medicine and agriculture
were said to be liable to superstitious use at 2. 29. 45. What
links all these arts here, says Augustine, is that their practi-
tioners use past experiences to make conjectures about future
ones (‘harum ergo cunctarum artium de praeteritis experi-
menta faciunt etiam futura conici’, 2. 30. 47). At first, this
seems like an odd point to find in common among such
diverse arts. But the divinatory context explains it. For the
three arts of service mentioned by Augustine—medicine,
agriculture, and sailing—are among those whose predictions
were traditionally contrasted with divinatory predictions.
As Cicero’s brother explains in De Divinatione: ‘There are
many things foreseen by physicians, ship captains, and also
by farmers, but I do not call the predictions of any of them
divination.’60
60
Cicero, De Divinatione 1. 50. 112: ‘Multa medici, multa guberna-
tores, agricolae etiam multa praesentiunt, sed nullam eorum divinationem
voco’ (trans. adapted from W. A. Falconer (Loeb Classical Library, 1923)).
For further references, see Pease (ed.), M. Tulli Ciceronis De Divinatione,
ad loc.
 Divination and the Disciplines 135
Having argued and re-argued the inferior status of tech-
nical divination, Augustine moves to the only category in
which he believed the true disciplines could be found, those
divinely instituted ‘doctrinae’ that pertain to reason (2. 31.
48–38. 56). Like the higher branch of humanly instituted
‘doctrinae’, this branch also makes no mention of divination.
Rather, it discusses the two parts of dialectic, disputation
and logic (2. 31. 48–35. 53), rhetoric (2. 36. 54–37. 55), and
mathematics (2. 38. 56). Mentioned together prominently
at the beginning of the section (2. 31. 48) and in the first
summary that follows (2. 39. 58–9), only ‘disputatio’ and
‘numerus’ are termed disciplines, but they may be taken to
stand for others. ‘Disputatio’ certainly stands for the other
disciplines of language, and ‘numerus’ for the mathematical
disciplines, or at least music. Astronomy is mentioned in
this category, but only in a reprise of divine institutions per-
taining to the senses (2. 32. 50); geometry is not mentioned.
Bracketed by renewed admonitions to avoid pride (2. 38. 57,
42. 63), two summaries close the book. These take the form
of a package of advice for young people and an allegorical
explanation of the spoils of Egypt. Although superstition is
mentioned, no hint of divination occurs here.

5. t e c h n i c a l a n d nat u r a l d i v i nat i o n
in the confessions
In Book 2 of De Doctrina Christiana Augustine attempted to
demonstrate both the (diminished) value of the disciplines
for acquiring knowledge and their superiority over technical
divination. He decisively followed through on these efforts
in the Confessions, where the kind of knowledge that he
represents himself as most earnestly seeking is supplied
neither by the disciplines, which play only a minor role,
nor by technical divination, which he rejects in the form of
astrology, but by God’s mysterious providence. This is
channelled through significant persons, books, events, and
two forms of natural divination: his mother’s dreams and his
own biblical lot divination.
To illustrate how small a role the liberal disciplines play in
Augustine’s search for knowledge in the Confessions, we need
only examine the three occurrences of ‘disciplina’ in the
136 William E. Klingshirn
work. In each case, the word is associated with a failure to
achieve true knowledge. In Book 1, Augustine complains
that people are more offended by the mispronunciation of
the word for person ‘contra disciplinam grammaticam’ than
by hatred of an actual person (1. 18. 29). And in Book 5,
Faustus the Manichaean, reputedly endowed, like some
Neoplatonic sage, with the mastery of all knowledge
(‘honestarum omnium doctrinarum peritissimus et apprime
disciplinis liberalibus eruditus’, 5. 3. 3), disappointingly
turns out to be unqualified in any of the liberal disciplines
except grammar (‘hominem expertem liberalium discipli-
narum nisi grammaticae’, 5. 6. 11).
Of course, Augustine’s high expectations of Faustus’
learning, as judged by the standards of his own education,
do concede some value to the liberal disciplines, as do other
episodes in the work. But technical divination holds no
similar place in the Confessions. This is demonstrated above
all by the example of astrology, to which Augustine ulti-
mately assigned no standing even as an ‘ars’, let alone as a
‘disciplina’. This realization was brought home to him by the
arguments and experiences of well-educated men: the pro-
consul Vindicianus (‘vir sagax medicinae artis peritissimus’,
4. 3. 5) and the aristocrats Nebridius (4. 3. 6)61 and Firminus
(‘liberaliter institutus et excultus eloquio’, 7. 6. 8–10). As a
result, Augustine concluded that the true predictions
made by astrologers are produced not by art but by chance
(‘non arte . . . sed sorte’); likewise, their false predictions
come about not from lack of skill (‘artis inperitia’) but from
random error (‘sortis mendacio’, 7. 6. 9).
The same logic applied of course to predictions made
by lot, particularly to lots drawn from inspired books, as
Vindicianus pointed out (4. 3. 5). The pagan definition he
gives of ‘sors’ as ‘a power . . . everywhere diffused in the
nature of things’ (‘vim sortis . . . in rerum natura usque-
quaque diffusam’) was easily transformed by Augustine into
a definition of divine grace (7. 6. 10).62 This move allowed
Augustine to portray the central divinatory event in the

61
For his education, see Kaster, Guardians of Language, 314–15.
62
B. Bruning, ‘De l’astrologie à la grâce’, in Collectanea Augustiniana:
Mélanges T. J. Van Bavel, ii (Leuven, 1990), 575–643.
 Divination and the Disciplines 137
Confessions, his consultation of a codex of Paul’s Letters, as a
form not of divination by art, but of divination by nature.
The third important form of divination in the Confessions,
Monica’s dreams, already belonged to this category, and
so Augustine could portray both his own and Monica’s div-
inatory experiences as effusions of divine revelation rather
than technical divinatory consultations. Thus is natural
divination shown to be consistent with God’s providence
and Augustine’s search for knowledge.
Ultimately, of course, Augustine and Monica move
beyond even this. In their Neoplatonic ascent (9. 10. 24) and
subsequent reflection on it (9. 10. 25), their souls rise not
only above the stars and hence above astrology,63 but also
beyond bibliomancy and oneiromancy.
If all dreams and visions in the imagination are excluded, if
all language and every sign and everything transitory is silent . . .
then he alone would speak not through them but through himself.
We would hear his word (verbum eius), not through the tongue
of the flesh (per linguam carnis), nor through the voice of an angel
(per vocem angeli), nor through the sound of thunder (per sonitum
nubis), nor through the obscurity of a simile (per aenigma
similitudinis). Him who is in these things we love we would hear
himself without them. (9. 10. 25)64
Carried by this experience beyond their previous modes of
natural divination—Monica beyond her ‘dreams and visions’
and Augustine beyond ‘all language and every sign’—mother
and son move for an instant not simply beyond the need for
divination, but more profoundly beyond the need for any
kind of mediation. In the world which Augustine created for
his readers in the Confessions, this was as far as one could get
from the ‘disciplines of knowledge’. Of course, outside
his world the same kind of ascent was promised to readers of
Porphyry, Iamblichus, and other Neoplatonists by those very
disciplines. In this respect, it is worth noting that Augustine
only uses the word ‘disciplina’ in connection with any kind

63
L. C. Ferrari, ‘Augustine and Astrology’, Laval Théologique et
Philosophique 33 (1977), 241–51, at 242.
64
Translation adapted from H. Chadwick, Saint Augustine: Confessions
(Oxford, 1991).
138 William E. Klingshirn
of divination in three passages from Book 10 of the City of
God, where it is theurgy that earns the distinction (‘apparere
theurgian esse . . . disciplinam’, 10. 9; ‘per nescio quam
theurgicam disciplinam’, 10. 10; ‘theurgica disciplina’,
10. 27). But by the time he wrote about theurgy in the City
of God, it was safe for Augustine to apply the theurgists’ own
word to their practice, for he stood on the far side of a shift
in his thinking about divination, which we can clearly see
in Book 10 of the Confessions. There, in a preview of his
later arguments against theurgy, he dismissed the claims of
practitioners whose reliance on their own ‘doctrina’ and on
tactical alliances with demons left them unable to return to
God (10. 42. 67). They needed a mediator, he said, but their
mediator was false. As both divine and human, only Christ
could truly serve this purpose (68).
By this move, Augustine not only attacked theurgy and
Neoplatonism; he also permanently reconfigured for himself
the problem of human access to divine knowledge, and thus
the problem of divination and the disciplines of knowledge.
For if it was in Christ that ‘all treasures of wisdom and
knowledge are hidden’ (‘in quo sunt omnes thesauri
sapientiae et scientiae absconditi’, Col. 2: 3), as he wrote at
Confessions 10. 43. 70, then it was only through Christ that
these treasures would be accessible to humankind. Thus,
Augustine’s concept of ‘the centrality . . . of Christ as
mediator of wisdom’, which began to emerge in his thinking
in the mid-390s, marked an important break not only in
his thinking about wisdom, knowledge, philosophy, and
Christianity, as Carol Harrison has recently argued,65 but
also in his thinking about divination and the disciplines of
knowledge. Indeed, as we can see in retrospect, it was an
emphasis on Christ’s role as mediator that made biblical
learning, biblical disciplines, and biblical divination so
appealing to Augustine in De Doctrina Christiana 2 and
Confessions 8. Although such an emphasis is not entirely
absent from earlier writings (e.g., C. Acad. 2. 1. 1), its still
predominantly Platonic deployment there allowed him to

65
C. Harrison, Augustine: Christian Truth and Fractured Humanity
(Oxford, 2000), 36–8, quotation at 37.
 Divination and the Disciplines 139
take a more traditional approach to problems of divination
and the disciplines in the later 380s and early 390s.66
But these two stages of Augustine’s early thought are also
linked by another theme, operating in both periods, which
Christ’s mediation only throws into greater prominence.
This is the enormous gap between God and man that calls
for divination (and mediation) in the first place. To discuss it
briefly, and so to conclude, it will be instructive to examine
Augustine’s correction of a passage in De Immortalitate
Animae in which he had stated that there was no knowledge
that was not contained in a discipline (‘Omne autem quod
scit animus, in sese habet; nec ullam rem scientia com-
plectitur, nisi quae ad aliquam pertineat disciplinam’, 1. 1).
He had been thinking only of human souls, he conceded in
his retractatio: ‘it did not occur to me that God does not learn
disciplines, and yet he possesses a knowledge of all things
that also includes a foreknowledge of future things’ (‘nec
venit in mentem Deum non discere disciplinas et habere
omnium rerum scientiam in qua etiam praescientia est
futurorum’, Retr. 1. 5. 2). God in fact was the only being
whose knowledge (and foreknowledge) was not learned.
Some of that knowledge he widely shared, according to De
Doctrina Christiana, by means of his creation itself and the
divinely instituted ‘doctrinae’ and ‘disciplinae’ that humans
investigated, discovered, and in one way or another learned
(2. 27. 41). But foreknowledge or, to put it more broadly,
divinatory knowledge, God kept more carefully guarded. As
essentially secret knowledge, it belonged to God’s hidden
providence, and like God’s power constituted one of the
great asymmetries separating the Creator from his creation.
It might be snatched away prematurely by demons,67
opaquely hinted at in the Holy Scriptures, or openly con-
veyed to humans by special modes of communication, but
it could never be entrusted to any professional class of
diviners or system of interpretation. Augustine makes this
point in many ways throughout his early writings, not
least remarkably by scrupulously avoiding any association

66
For another assessment of the changing balance in Augustine’s
thought, see the discussion by Conybeare above.
67
Tertullian, Apologeticum 22. 9 (CCSL 1.129).
140 William E. Klingshirn
between divination and the disciplines. By such consistent
effort, made possible by his own disciplined grammatical
and dialectical training, he aimed to maintain the proper
boundaries, in power and knowledge, between heaven and
earth. As we can see in the sermon with which we began
(Serm. 12. 12, CCSL 41.174), these were boundaries that
Augustine believed could only be transgressed by Christ, the
‘power and wisdom of God’ (1 Cor. 1: 24), a mediation that
guaranteed their perdurance to the end of time.
 6
The Vocabulary of the Liberal Arts
in Augustine’s Confessions
Ph i l i p Bu r to n

Following his conversion back to Christianity in the spring

of 386, Augustine embarked on a project to produce hand-
books of grammar, dialectic, rhetoric, music, geometry,
arithmetic, and philosophy—
a version of what is known to
us as the ‘canon’ of an education in the liberal arts. 1 We
know this because he tells us so in the first book of his
Retractationes (1. 6):
Per idem tempus quo Mediolani fui, baptismum percepturus,
etiam disciplinarum libros conatus sum scribere, interrogans eos
qui mecum erant atque ab huiusmodi studiis non abhorrebant,
per corporalia cupiens ad incorporalia quibusdam quasi passibus
certis vel pervenire vel ducere. Sed earum solum de grammatica
librum absolvere potui, quem postea de armario nostro perdidi,
et de musica sex volumina, quantum attinet ad eam partem quae
rhythmus vocatur. Sed eosdem sex libros iam baptizatus, iamque
ex Italia regressus in Africam scripsi; inchoaveram quippe tantum-
modo istam apud Mediolanum disciplinam. De aliis vero quinque
disciplinis illic similiter inchoatis, de dialectica, de rhetorica, de
geometria, de arithmetica, de philosophia, sola principia reman-
serunt. Quae tamen etiam ipsa perdidimus; sed haberi ab aliquibus
existimo.
Throughout that time, when I was at Milan waiting to receive
baptism, I tried also to compose books on the arts. I questioned

My particular thanks to Gillian Clark and Bob Coleman for their

encouragement and helpful comments.
1
The larger question of the origins of the ‘canon’ will not be con-
sidered here. It is not entirely clear that Augustine even knew of a
canon, though it seems a reasonable assumption. See the Introduction and
Shanzer’s essay in this volume.
142 Philip Burton
those of my companions who were not averse to such studies, in my
desire to arrive or to lead others by certain steps through things
material to things immaterial. But of these arts I was able to com-
plete only my book on grammar, which later I lost from my book-
case, and six books on music, touching the part known as ‘rhythm’.
But these six books I wrote after my baptism and my return from
Italy to Africa; at Milan I had only begun that art. Of the other five
arts I had likewise begun there—dialectic, rhetoric, geometry,
arithmetic, and philosophy—only the beginnings have remained,
and even these I have lost, though I think some people have them.
It has often been remarked that Augustine’s interest in the
liberal arts suggests a distinctly ‘intellectual’ understanding
of Christianity. Certainly he had been influenced to part
from the Manichees by his discovery of incompatibilities
between their cosmology and the laws of mathematics and
geometry. His other early post-conversion works, such as the
Soliloquia and De Magistro, are philosophical dialogues in
form. The very verb he uses in this passage to describe his
technique, ‘interrogare’, has a long history as a technical term
of rhetoric and dialectic.2 Augustine is clearly optimistic
about the possibility of taking the traditional view of the arts
as a pathway from the physical to the incorporeal world and
combining it within his own Christian Platonism.
Around 391 work on the project was abandoned. Augustine
did not simply lose interest in Christian applications of the
arts, as can be seen from De Doctrina Christiana, begun
around 395–7. None the less, his later writings do betray an
increasing pessimism about the possibility, or desirability, of
a Christianization of the traditional curriculum. I suggest
that we can see the beginnings of this pessimism if we look at
the vocabulary which Augustine uses to describe the arts
in another work, closely contemporary with De Doctrina

2
For its use as a general term for dialectical reasoning, see, e.g.,
Cicero, De Fato 28: ‘sic . . . interrogant: si fatum tibi est ex hoc morbo
convalescere, sive tu medicum adhibueris sive non adhibueris, con-
valesces’ (TLL vii(1). 2272.71 ff.), and for the specific sense ‘questioning
requiring yes-or-no answer’, TLL vii(1). 2266.28 ff. Augustine is aware of
both senses: Doct. Chr. 4. 10. 25: ‘in collocutionibus [i.e., philosophical
dialogues] est cuique interrogandi potestas’; ibid. 3. 3. 6: ‘ad perconta-
tionem multa responderi possunt, ad interrogationem vero aut non aut
etiam’.
 The Vocabulary of the Liberal Arts 143
3
Christiana, namely the Confessions. In doing so, we shall
also have occasion to note some ways in which he com-
bines classical ‘disciplina’ with his own practice of biblical
interpretation.4
The Greek terminology we have seen Augustine using in
the Retractationes goes back at least to the first century
before the Christian era. ‘I shall do my best,’ says Cicero
in the Academica, ‘to speak Latin—except in the case of
words such as “philosophy” or “rhetoric” or “physics” or
“dialectic”, which along with many others custom now
uses as if they were Latin words.’5 Again in De Finibus
he writes that ‘we follow the example of our ancestors in
using “philosophy”, along with “rhetoric”, “dialectic”,
“grammar”, “geometry”, and “music”, even though they
could be expressed in Latin, as being our own’.6 The
one notable absentee from this list, ‘arithmetica’, makes
its first appearance in Vitruvius, and is common from the
first century a d onwards; the Elder Pliny, for instance,
describes the sculptor Pamphilus as ‘learned in all subjects,
especially arithmetic and geometry’.7 By Late Antiquity,
this Greek terminology was clearly the norm. Martianus
Capella uses a scheme based on ‘grammatica, dialectica,
rhetorica, geometria, astronomia, arithmetica, harmonia’;

3
The terminology of De Doctrina Christiana will be mentioned mar-
ginally for purposes of comparison. The range of disciplines covered in
this work is wider than that of the Confessions, and the intended audience
may be more specialized. This question would deserve analysis of its own,
the results of which may differ from those offered here.
4
As background, note especially H.-I. Marrou, Saint Augustin et la fin
de la culture antique, 4th edn. (Paris, 1958), 561–83, who first discusses the
technical senses of ‘scientia’ and ‘sapientia’, then surveys the evidence
relating to Augustine’s manuals on different ‘artes’.
5
Cicero, Academica 1. 7. 25: ‘nisi in huiuscemodi verbis, ut philos-
ophiam aut rhetoricam aut physicam aut dialecticam appellem, quibus ut
aliis multis consuetudo iam utitur pro latinis’.
6
Cicero, De Finibus 3. 2. 5: ‘Quamquam ea verba quibus instituto
veterum utimur pro latinis, ut ipsa philosophia, ut rhetorica, dialectica,
grammatica, geometria, musica, quamquam latine ea dici poterant, tamen
quoniam usu percepta sunt, nostra ducamus.’
7
Pliny, Historia Naturalis 35. 76: ‘omnibus litteris eruditus, praecipue
arithmetica et geometria’.
144 Philip Burton
Cassiodorus has ‘musica’ for ‘harmonia’, but is otherwise the
same.8
What were the alternatives? It is a commonplace of loan-
word studies that when a technical vocabulary is borrowed
by one language from another, there is usually a period of
experimentation, during which various translations are tried
out; usually, the loan-words emerge as winners.9 But the
picture is not always so simple. Where the loan-words tri-
umph, they often retain a foreign flavour—as for instance
with ‘philosophia’ and ‘rhetorica’, marked out as foreign by
phonology, orthography, and general cultural association. As
for the translations, where these were already current in the
language, they may retain a flavour of the word they were
used to translate. Where they were novel coinages, the mere
fact that we often know them only from citations by later
authors proves that they did not vanish entirely.
One very important stylistic consequence follows from
this. All stylistics rests on the twin notions of substitutability
and markedness.10 The principle of substitutability assumes
that it is possible to say fundamentally the same thing in
more than one way. At the lexical level, this substitutability
is what we call synonymy. The very concept of synonymy
is a major stumbling-block for many semanticists. But if we
are to find examples of it anywhere, surely it is in cases of
technical translation, where we know that one word is
used precisely to stand in for another one. The principle of
markedness assumes that where two words or constructions

8
For further detail and references, see Shanzer’s essay in this volume,
esp. 98–102.
9
For this phenomenon in the fields of philosophy and rhetoric, see
R. G. G. Coleman, ‘The Formation of Specialized Vocabularies in
Philosophy, Grammar, and Rhetoric: Winners and Losers’, in M. Lavency
and D. Longrée (eds.), Actes du Ve Colloque de Linguistique latine
(Louvain, 1989), 77–89; for medicine, D. R. Langslow, Medical Latin in
the Roman Empire (Oxford, 2000), 127–30, with caveats and qualifications;
for biblical translation, P. Burton, The Old Latin Gospels: A Study of their
Texts and Language (Oxford, 2000), 147–8.
10
In this paragraph I draw upon terminology and approaches familiar
from linguistics. For definitions and examples, see D. Crystal, A Dic-
tionary of Linguistics and Phonetics, 3rd edn. (Oxford, 1991), 211–12,
340–1); J. Lyons, Introduction to Theoretical Linguistics (Cambridge,
1968), 427–8, 451.
 The Vocabulary of the Liberal Arts 145
may be substituted for each other, the less common of the
two is more marked, and so more interesting from a stylistic
point of view. Marked terms may contrast with unmarked
ones either by dialect (‘faucet’ versus ‘tap’), sociolectally
(‘napkin’ versus ‘serviette’), or chronologically (‘record-
player’ versus ‘gramophone’). In each pair, both terms have
the same referend, but one of the two terms would strike
many English-speakers as unusual or forced. It is the inter-
play between marked and unmarked terms in the Confessions
which concerns us now.
We shall first consider ‘grammar’ and ‘rhetoric’. Through-
out the Confessions Augustine is ambivalent about human
language. The initial acquisition of language he enjoys, but it
leaves him deeper in what he calls ‘the stormy fellowship of
human life’ (1. 8. 13). And grammar as it is taught in schools
merely reinforces this obsession with human values rather
than divine ones: ‘How carefully,’ he exclaims, ‘the children
of men observe the rules governing letters and syllables that
they have received from previous speakers, while ignoring
the eternal rules that lead to everlasting salvation, which they
have received from you!’ (1. 18. 29). At the heart of his cri-
tique is the Platonic belief that the conventional educational
system is a fraud. Teachers and parents are ‘grammar-
mongers and their customers’ (‘venditores grammaticae vel
emptores’, 1. 13. 23; cf. 4. 2. 2: ‘docebam . . . rhetoricam, et
. . . loquacitatem . . . vendebam’), but in fact knowledge
cannot be imparted. Moreover, the ‘artes’ of grammar and
rhetoric are in any case concerned with fictions and plausi-
bilities rather than with truth; the better educated your
grammarian, the more likely he is to know that Aeneas did
not, in fact, come to Carthage (1. 13. 22).

grammar
The usual word for grammar in the Confessions is ‘gram-
matica’, and usually it has bad connotations.11 It is Greek
11
For a less rhetorical presentation of Augustine’s views on grammar,
see Doct. Chr. 3. 29–40. There he declares that it is not his intention to teach
the various literary tropes, in case he gives the impression of teaching the
‘ars grammatica’; useful as they are for the understanding of Scripture,
they are also familiar to those with no formal training in grammar.
146 Philip Burton
‘grammatica’ that Augustine is forced against his will to
study, and ‘grammatica’ which teaches men to believe that a
dropped aitch is worse than a miscarriage of justice.12 Faus-
tus the Manichee knows only ‘grammatica’ of all the liberal
arts—and has only an average knowledge of that.13
What were the alternatives here? From an early date, the
loan-word ‘grammatica’ had been used alongside ‘litterae’,
itself a translation of τὰ γράµµατα (an etymology known to
Augustine), as well as ‘cognitio litterarum’ and similar peri-
phrases. Cicero himself, although he allowed ‘grammatica’
honorary Latin citizenship, preferred these expressions. And
Augustine too uses them in the Confessions, alongside
‘grammatica’. So, for instance, where he criticizes ‘gram-
matica’ for being concerned with merely human traditions,
he goes on to observe that the zealous prosecutor’s awareness
that he is doing to others what he would not have done to
himself is as sharply engraved in his being as his knowledge
of ‘letters’ (‘non est interior litterarum scientia quam scripta
conscientia’, 1. 18. 29). Elsewhere, ‘litterae’ in their broader
sense may be good or bad: the Scriptures are ‘sacrae litterae’
(12. 31. 42), but literalism is characteristic of Manichee
exegetics; Augustine is freed only through Ambrose’s
reiteration of the principle that ‘the letter kills, but the Spirit
gives life’ (2 Cor. 3: 6, quoted at 6. 4. 6). The term ‘litterae’,
then, is not the positive counterpart of an intrinsically
negative ‘grammatica’; in itself neutral, it may be used in
favourable or unfavourable contexts. A basic knowledge
of literacy is necessary, but not sufficient, for a true reading
of the Scriptures.14

12
See Conf. 1. 18. 29 ff., where ‘grammatica’ consorts with other such
shady terms as ‘syllaba’, ‘soloecismus’, and ‘barbarismus’; it is ironic that
the parameters of good Latin can be defined only by using Greek words.
For constellation of Greek words, compare Conf. 3. 3. 6, where ‘schola
rhetoris’ is followed up with ‘typhus’, ‘diabolicus’, and ‘daemonium’.
13
Conf. 5. 6. 11: ‘expertus sum hominem expertem liberalium
disciplinarum nisi grammaticae atque eius ipsius usitato modo’.
14
This would seem to imply that the Christian exegete needed at least
some basic, conventional education. Note that in Doct. Chr. praef. 4
Augustine side-steps even that minimal requirement, with his story of the
barbarian Christian slave who learnt his letters after a three-day prayer
session; the absence of any human agency (‘nullo docente homine’) is
 The Vocabulary of the Liberal Arts 147
Similarly ambivalent is Augustine’s other word for
grammar: namely, ‘litteratura’. This term, as Augustine
knew, went back at least to Varro, but was already obsolete in
Seneca’s day.15 Yet he uses it three times in the Confessions,
twice in connection with another archaic translation,
‘oratoria’ for ‘rhetorica’. Again, this translation has a good
republican pedigree—one thinks of Cicero’s dialogue, the
Partitiones Oratoriae—but again, this time thanks to Quin-
tilian, we know that it was obsolete by the first century a d.16

stressed. It has to be borne in mind, however, that Augustine sees this

example as an exception.
15
Augustine, Ord. 2. 12. 35: ‘Quibus [sc. litteris et numeris] duobus
repertis nata est illa librariorum et calculonum professio, quaedam gram-
maticae infantia, quam Varro litterationem vocat’ (‘When these two arts
[i.e. literacy and numeracy] had been discovered, there came into being
the profession of the book-man and the bean-counter; this was, so to
speak, the infancy of the art of grammar, which Varro calls litteratio’);
Martianus Capella, De Nuptiis Mercuriae et Philologiae 3. 229 (ed. Kopp):
Γραµµατικ dicor in Graecia, quod γραµµ linea et γράµµατα litterae
nuncupentur . . . hincque mihi Romulus Litteraturae nomen ascripsit,
quamvis infantem Litterationem voluerit nuncupare, sicut apud Graecos
Γραµµατιστικ primitus vocitabar’ (‘Grammatikē am I called in Greece,
as a pen-stroke is called grammē and letters are called grammata . . . for this
reason Romulus gave me the name of Litteratura, though in my infancy he
chose to call me Litteratio, just as I was originally called grammatistikē
among the Greeks’). The distinction between ‘litteratio’ (basic literacy)
and ‘litteratura’ (literary studies) has been painstakingly reconstructed by
R. Kaster, C. Suetonius Tranquillus: De Grammaticis et Rhetoribus
(Oxford, 1995), 86–93. For ‘litteratura’ in both senses, and its obsoles-
cence, see Seneca, Ep. 88. 20: ‘prima illa, ut antiqui vocabant, litteratura,
per quam pueris elementa traduntur’ (‘the earliest litteratura (as the
ancients called it), the one through which children are taught their
letters’).
16
Compare Cicero’s use of the phrase ‘facultas oratoria’ at De Inven-
tione 1. 6–7; De Oratore 1. 245. For the formation, we might compare
Ovid’s Ars Amatoria, clearly a parody of a technical handbook: M. Janka,
Ovid ‘Ars Amatoria’, Buch 2: Kommentar (Heidelberg, 1997), 31–4. For
the obsolescence of the term, see Quintilian, Institutio Oratoria 2. 14.
1–4: ‘Rhetoricen in latinum transferentes tum oratoriam, tum oratricem
nominaverunt . . . et haec interpretatio non minus dura est quam illa
Plauti essentia atque queentia’ (‘Latin translators sometimes called rhet-
oric oratoria, sometimes oratrix . . . a translation no less harsh than those
famous words of Plautus’, essentia and queentia’). Quintilian’s own title
(if manuscripts are to be trusted) is, of course, evidence of the currency of
the adjective.
148 Philip Burton
The three passages repay attention. The first is Augustine’s
description of his days at sixth-form college in Madaura,
where he had gone ‘for the sake of acquiring literary and
oratorical expertise’ (‘litteraturae atque oratoriae percipien-
diae gratia’, 2. 3. 5). It is hard to know what, if anything, to
read into this, though one is tempted to think that Augustine
is using the archaic expression to describe the provincial
schooling available in Madaura in the 360s.17 His second
use is ambivalent rather than neutral: Simplicianus tells
Augustine how Marius Victorinus had taken the emperor
Julian’s prohibition of Christians teaching ‘litteratura’ and
‘oratoria’ as the cue for his professional retirement (8. 5. 10).
Here Augustine’s attitude is more complex. Victorinus’
decision would be one of the factors influencing Augustine’s
own retirement, but Augustine does not approve of Julian’s
edict; elsewhere he equates it with persecution.18 His focus
in this passage is not on the evils of the conventional edu-
cational system, but on the uses to which it may be put. More
clearly positive is the third instance of ‘litteratura’, in a key
passage for our present study. In Book 10 of the Confessions,
Augustine discusses the arts, not as they are taught in school,
but in a more Platonic sense, as things accessible only to
the intellect: ‘as for the nature of grammar and the art of
dialectic (nam quid sit litteratura, quid peritia disputandi), how
many categories there are . . . [A]ll these things are in my
memory, without my having kept an image and left the
reality outside’ (10. 9. 16). Grammar in this more positive
sense may fairly be described as ‘litteratura’. In the same
passage Augustine refers also to rhetoric, to which we now
turn.

17
Madaura, a centre of Romanization in North Africa in the second
century a d, seems to have undergone something of a renovation pro-
gramme around and shortly after Augustine’s time there. C. Lepelley,
Les Cités de l’Afrique romaine au Bas-Empire, 2 vols. (Paris, 1979–81),
ii. 128–33, lists a series of public works on buildings ‘incuria paene ad
interitum redactae’ (‘all but reduced to destruction through neglect’) or
‘tot retro annis ruinarum labe deformes’ (‘besmirched with the stain of
delapidation going back over many years’).
18
Civ. 18. 52: ‘an ipse non est ecclesiam persecutus, qui christianos
liberales litteras docere vetuit?’ (‘What of him who forbade Christians to
teach the liberal arts? Did he not persecute the Church?’).
 The Vocabulary of the Liberal Arts 149

r h e to r i c
The Confessions contain seven instances of ‘rhetorica’, all
referring to Augustine’s pre-conversion career. The one
example we have encountered (4. 2. 2) speaks for all:
‘Throughout those years I taught the art of rhetoric (artem
rhetoricam), and taught the knack of domineering speech,
being myself dominated by my own cupidity.’19 But in
the key catalogue of the arts as they exist in the memory,
Augustine again avoids this term, asking instead how it is
that we remember rhetoric (‘numquid sicut meminimus
eloquentiam?’, 10. 21. 30). There is every reason to think
that he is here using the word ‘eloquentia’ in its technical
sense, which goes back at least to Cicero, is solidly attested
in Quintilian,20 and most memorably of all appears in
Eumolpus’ introduction of himself in Petronius’ Satyricon:
‘Where is dialectic? Where astronomy? Where the well-
tended path of philosophy? Who has ever gone to a temple
to perform his vow, having attained unto rhetoric?’ (‘ubi est
19
Note the description of rhetoric as an ‘art’, found also at 5. 12. 22;
in Doct. Chr. 4. 60 Augustine explicitly endorses Cicero’s view that
eloquence is not so much invented as discovered. Three of the other seven
instances of ‘rhetorica’ co-occur with ‘profiteri’ or ‘professio’, to be con-
trasted with the ‘confessio’ of the Christian Augustine. Note also how his
own enterprise in Doct. Chr. is carefully paraphrased as a ‘means of
presentation’ (‘modus proferendi’, 1. 1. 1 and elsewhere), even though the
closeness to rhetoric is obvious already from 2. 37. 55: ‘sed haec pars [i.e.
rhetorica or eloquentia] discitur, magis ut proferamus ea quae intellecta
sunt’.
20
Cicero does draw a loose distinction between untutored eloquence
and rhetoric proper. See, e.g., De Inventione 1. 5: ‘Civilis quaedam ratio
est, quae multis et magnis rebus constat. Eius quaedam magna et ampla
pars est artificiosa eloquentia, quam rhetoricam vocant’ (‘Political science
consists of many important things. An important and noble part of it is
eloquence refined by art, or “rhetoric”, as it is called’), or De Oratore 1.
167: ‘non defuit . . . patronis . . . eloquentia neque dicendi ratio, sed iuris
civilis scientia’ (‘Those lawyers lacked neither eloquence nor a scientific
approach to speaking, but rather a knowledge of law’). Quintilian, how-
ever, uses ‘eloquentia’ as a straightforward translation: Institutio Oratoria
2. 16. 2: ‘Sequitur quaestio, an utilis rhetorice. Nam quidam vehementer
in eam invehi solent . . .: eloquentiam esse, quae poenis eripiat scelestos’
(‘The question follows, whether rhetoric is a good thing. Some people
habitually criticize it harshly, saying that it is eloquence which rescues the
wicked from their punishment’).
150 Philip Burton
dialectica? ubi astronomia? ubi sapientiae cultissima via?
quis umquam venit in templum et votum fecit, si ad elo-
quentiam pervenisset?’, 88). Elsewhere in the Confessions,
Augustine’s use of the term ‘eloquentia’ is similar to his use
of ‘litterae’ for ‘grammatica’. It is found both in pejorative
contexts, as when he studies ‘with criminal intent’ the clas-
sics of rhetoric (‘libri eloquentiae’) at Carthage (3. 4. 7), but
also in reference to the Scriptures: Moses is notable for his
‘facultas eloquendi’ (12. 26. 36), but never for his ‘facultas
rhetorica’.
The ‘loq’- root appears also in another key passage of the
Confessions. Having described his university education in
Book 3, Augustine sums it up by telling God he was taught
nothing: ‘Whatever I learnt of the arts of speech and logic,
or of the measurement of figures or music or mathematics
(de arte loquendi et disserendi . . . de dimensionibus figurarum et
musicis et de numeris), I learnt with no great difficulty—and
not through any human tradition. You know, O Lord my
God, that my ready understanding and critical acumen also
are your gifts’ (4. 16. 30). We should note in passing that this
summary, like the Platonic catalogue of the arts as they truly
are in Book 10, is couched almost completely in Latin terms.
Only ‘musica’ is given a Greek name, on this its one appear-
ance in the Confessions.21 (We shall return below to another
use, or abuse, of the ‘loq’- root.)
The other possible Latin translation of ‘rhetorica’ to be
considered here is ‘facundia’. This is not a canonical
translation in the way that ‘eloquentia’ is, but there is
some evidence that in earlier Latin it could bear this
sense. Sallust uses it of the primitive rhetorical skills of the

21
In De Musica Augustine repeatedly uses the phrase ‘peritia’ or ‘ars
(bene) modulandi’, also found in Cassiodorus’ Institutiones; the first
attestation of this formula is in the third-century grammarian Censorinus
(10. 3). The phrase is not used in the Confessions, though it is notable that
Augustine favours church music only if it is done ‘cum convenientissima
modulatione’ (Conf. 10. 33. 50). We should, therefore, be prepared to hear
echoes of music in all ‘mod’- words in the Confessions: ‘modus’, ‘moderari’,
‘omnimodus’, ‘multimodus’, and so on. And as Klingshirn notes above,
129, we should remember that music is an arithmetical discipline, and so
falls under the broader rubric of ‘numerus’.
 The Vocabulary of the Liberal Arts 151
22
second-century tribune Memmius, and Ovid of the
Scythians’ first faltering steps on the road to eloquence.23
The Sallustian tag is recycled by Augustine in reference
to his dedication of De Pulchro et Apto to Hierius, famous
for his ‘Graeca facundia’ (4. 14. 21); it is used also in the
description of his early trips to hear Ambrose preaching in
the basilica at Milan, where he is interested only in the style
of Ambrose’s preaching, not its content (‘quasi explorans
eius facundiam’, 5. 23. 13). Again, it is difficult to generalize
on the basis of two examples. It appears, however, that this
is an old-fashioned translation, not common in classical
Latin, and that Augustine reserves it for rather unfavourable
contexts; he is clearly critical of his pre-conversion
preoccupation with rhetoric.

dialectic
We have seen how Augustine’s attitude to the ‘artes’ under-
goes a major change between his conversion and the later
decades of his life. Nowhere is this more visible than in his
use of the term ‘dialectica’.24 In his first post-conversion
work, the Contra Academicos of 386, he observes that ‘by
adding dialectic, which is either the essence or the sine qua
non of philosophy, Plato is said to have perfected that art’
(‘Plato . . . subiungens . . . dialecticam, quae aut ipsa esset
aut sine qua sapientia omnino esse non posset, perfectam
dicitur composuisse philosophiae disciplinam’, 3. 18. 37). By

22
Sallust, Iugurtha 30. 4: ‘quoniam ea tempestate Romae Memmi
facundia clara pollensque fuit’ (‘since at that time Memmius’ speech-craft
was famous and prepollent at Rome’).
23
Ovid, Tristia 2. 273: ‘discitur innocuas ut agat facundia causas’ (‘Men
study speech-craft, to defend the just’).
24
On matters relating to dialectic in Late Antiquity, including
Manichee disputations, Christian deployment of the practice, and ultim-
ately the growing preference for authority based on hierarchy, tradition,
assent to formulas, and unanimous or at least univocal acclamation, see
R. Lim, Public Disputation, Power and Social Order in Late Antiquity
(Berkeley and Los Angeles, 1995). An additional perspective is provided
by Heßbrüggen-Walter in this volume.
152 Philip Burton
the last decade of his life, he is not so charitable:25 the first
disciples, he writes in the City of God, ‘were uneducated, not
versed in grammar, not armed with dialectic, not puffed up
with rhetoric’ (‘impolitos, non peritos grammatica, non
armatos dialectica, non rhetorica inflatos’, 22. 5)—three
Greek words together. His works against Julian of Eclanum,
also from his last decade, are full of sneers at Julian’s pre-
tensions as a ‘dialecticus’: where, he asks at one point, is
Julian’s vaunted grasp of Aristotle and the other tricks of the
dialectical trade (‘ubi est acumen tuum, quo tibi videris
categorias Aristotelis assecutus, et aliam dialecticae artis
astutiam’, Contra Iulianum 3. 2. 7)?
But more than half of the 120 or so examples of ‘dialec-
tica’ in Augustine’s writings appear in a very different
work—the polemic Against Cresconius the Donatist Gram-
marian, written in 405–6, and so chronologically quite close
to the Confessions. Cresconius had accused Augustine of
being a ‘dialecticus’. Augustine, however, is unrepentant;
dialectic, he says, is ‘nothing other than the art of disputa-
tion’. A ‘dialecticus’ is merely what in Latin is called a
‘disputator’, and, he says, ‘to criticize under a Greek name
what you have no choice but to praise under a Latin name
is nothing other than to impose upon the uneducated,
and to insult the educated’ (‘Improbrare ergo in vocabulo
graeco, quod approbare cogeris in latino, quid est aliud
quam indoctis praetentare fallaciam, doctis facere iniu-
riam?’, 1. 14. 17). For good measure, Augustine adds that the
question-and-answer sessions we find Jesus practising in the
Gospels were themselves an example of ‘dialectica’. A
shrewd point. But Augustine’s rather shrill defence of
‘dialectica’ may lead us to suspect that Cresconius had
touched a sore point. Augustine had, in fact, hardly used the
word ‘dialectica’ in his published works in the twenty years
since Contra Academicos.
Certainly he does not use it in the Confessions, but he
has a range of alternatives. One of his alternatives, ‘peritia

25
The turning-point can be clearly observed at Doct. Chr. 2. 31. 48–
35. 53, where he cautions against an unbridled use of the ‘disciplina dis-
putationis’, also called ‘scientia definiendi, dividendi atque partiendi’ or
‘regulae conexionum’; see also Pollmann below, 221.
 The Vocabulary of the Liberal Arts 153
disputandi’, we have encountered already in his catalogues
of the arts as they exist in the mind (10. 9. 16, 4. 16. 30). The
translation is canonical; Cicero in the Orator links the ‘dis-
putandi ratio’ with ‘praecepta disserendi’ as Latin versions
of ; διαλεκτικ26 Here, as in other cases, preference for
Latin words has the important consequence of enabling
Augustine to link the ‘artes’ to specific biblical passages,
where the same words are used in a non-technical sense. The
supreme dialectician, for him, is God; it is God who teaches
us, by forcing us to confront the inconsistencies and false
values in purely human conventions. The link is provided by
Isaiah 1: 18, cited at Conf. 13. 19. 24: ‘venite, disputemus,
dicit dominus.’ ‘Come let us reckon up [or do dialectic], says
the Lord’—a passage to which we shall return later.
Another important verb is ‘sermocinari’;27 not, perhaps,
so familiar a rendering of διαλ&γοµαι as ‘disputare’ or ‘dis-
serere’, but a rendering none the less. Both the Auctor ad
Herennium and Quintilian use the term ‘sermocinatio’ to
refer to the rhetorical figure known as διάλογο, in which
individuals referred to in a speech are made to converse in
language appropriate to their character,28 and ‘sermo’ or
‘sermocinatio’ can also be used of philosophical dialogue,
whether Xenophontic, Platonic, or Ciceronian.29 Indeed it is

26
Cicero, Orator 32. 113–14: ‘disputandi ratio et loquendi dialecticum
sit . . . ipse Aristoteles tradidit praecepta plurima disserendi, et postea
qui dialectici dicuntur spinosiora multa pepererunt’ (‘The scientific
approach to argument and speech [is] dialectic . . . Aristotle himself left
many teachings on the art of debate, and the later “dialecticians” produced
many teachings more thorny still’).
27
See also Conybeare in this volume, 50–1.
28
Rhetorica ad Herennium 4. 52. 65: ‘Sermocinatio est cum alicui
personae sermo attribuitur et is exponitur cum ratione dignitatis’ (‘Ser-
mocinatio is when a speech is ascribed to a character and is set forth in a
way proportionate to their status’); Quintilian, Institutio Oratoria 9. 2. 31:
‘sermones hominum assimulatos dicere διαλ"γου malunt, quod latino-
rum quidam dixerunt sermocinationem’ (‘Fictional speeches by indi-
viduals are called “dialogues”; some Latin-speakers have called this
phenomenon sermocinatio’).
29
Cicero, De Officiis 1. 37. 134–5: ‘Sit ergo hic sermo, in quo Socratici
maxime excellunt, levis minimeque pertinax; insit in eo lepos . . . habentur
autem plerumque sermones aut de domesticis negotiis aut de re publica
154 Philip Burton
easy to see how ‘sermocinari’, a middle verb with a similar
basic meaning to διαλ&γοµαι, could easily take on the sense
‘to engage in dialectic’. Augustine himself is aware of this
possibility. Cresconius the Donatist grammarian is pum-
melled mercilessly: if, Augustine asks, ‘neither our Lord nor
his interlocutors conducted dialectic in the course of their
“sermocinatio”, then what in your opinion is dialectic?’30 Of
the five instances of ‘sermocinari’ in the Confessions, all can,
at a push, be classed under the heading of ‘dialectic’ or ‘intel-
lectual discourse’. On one occasion, Augustine fuses the
double sense of ‘sermo’ as a rendering both of ; (µιλα,
‘homily’ or ‘sermon’ in the English sense, and of διάλογο:
Monica goes to church to hear not the preacher, but ‘you,
[O God], in your homilies’ (‘te [Deus] in tuis sermonibus’,
5. 9. 17). I would suggest that the notion of ‘dialectic’
is strongly present at this point. Monica has often been
viewed as a classic instance of the untaught sage, and indeed
Augustine refers to God teaching her in the ‘inner schola’ or

aut de artium studiis atque doctrina’ (‘The sort of conversation in which

the Socratics excel should be light and not too single-minded; it should
have a humorous element . . . these conversations are usually about either
household management or politics or the pursuit and study of the arts’);
De Amicitia 1. 24. 38: ‘[Scaevola] exposuit nobis sermonem Laelii
de amicitia . . . eius disputationis sententias memoriae mandavi . . . ut
tamquam a praesentibus haberi sermo videretur’ (‘[Scaevola] recounted to
us the conversation on friendship which Laelius held . . . I learnt by heart
many of phrases from that discussion . . . so that the conversation should
seem to be conducted by parties actually present’); Horace, Odes 3. 21.
9–10: ‘non ille, quamquam Socraticis madet / sermonibus, te neglegit
horridus’ (‘Drenched as he is in Socratic dialogues, he’s not wild and
woolly; he won’t overlook you’).
30
C. Cresc. 1. 19. 23: ‘video quid fortasse dicturus sis, nec illos [sc.
captiosi interrogatores domini] nec illum in ea sermocinatione dialectice
egisse. Si ergo nec illi qui captiose atque invidiose sermocinantur, nec
illi qui tales eorum responsione convincunt, dialectice agunt, dic nobis
tandem quid sit dialectica?’ (‘I think perhaps I see what you’re going to
say, that neither [Jesus] himself nor those [who tried to catch him
out with questions] practised dialectic in the course of their conversation.
If, then, dialectic is practised neither by those who conduct conversations
out of envy and a desire to catch people out, nor by those who refute
such people with an answer, don’t keep us in suspense—what is
dialectic?’).
 The Vocabulary of the Liberal Arts 155
lecture-theatre of her heart. At two other points in the
later books of the Confessions, he uses ‘sermocinari’ again to
refer to God as supreme dialectician, calling on those who
can to ‘listen to [God’s] inner dialectic’ (‘audiat te intus
sermocinantem qui potest’, 11. 9. 11), and beseeching God
to ‘console me, and reason with me’ (‘tu me alloquere, tu
mihi sermocinare’, 12. 10. 10). For reasons that will become
clearer in a moment, this special use of philosophical
language in the later books is entirely appropriate.

geometry
We have already noted the appearance of geometry as
‘dimensio figurarum’ in the list of the ‘artes’ as they really
are, given in Book 4. In the catalogue of the arts in Book 10
we find a similar translation, ‘dimensionum rationes’
(10. 12. 19). Various other phrases also occur, referring to the
measurement of shapes, of the earth, or of the cosmos more
generally. Thus we find ‘dimensiones figurarum’ (4. 19. 30),
and references to ‘measuring the distances between the stars
and seeking the fulcra of the earth’ (‘neque . . . siderum
intervalla dimetimur vel terrae libramenta quaerimus’, 10.
16. 24). The one expression we never find is ‘geometria’
itself—even though it is one of Cicero’s honorary Latin
words. Augustine’s alternatives are, in fact, reminiscent of
some of the expressions Cicero himself tried and dropped.31
But again, his translations both play up the radical sense
of the words and allow him to link the arts to the Scriptures:
God, the Great Geometer, has ‘arranged all things in
measure and number and weight’, and the humblest believer

31
Cicero, Disputationes Tusculanae 1. 5: ‘at nos metiendi ratio-
cinandique utilitate huius artis [sc. geometriae] terminavimus modum’
(‘We have restricted the boundaries of this art to the practical purpose
of measuring and calculating’); De Senectute 49. 6: ‘mori videbamus in
studio dimetiendi paene caeli atque terrae C. Galum’ (‘We watched as
Gaius Galus died in the pursuit of measuring all but heaven and earth’).
Cf. Horace, Odes 1. 28. 1–2: ‘te maris et terrae numeroque carentis hare-
nae / mensorem cohibent, Archyta’ (‘You, Archytas, measurer of the sea
and the land and the countless sand . . . are circumscribed within a little
dust; what slight reward!’).
156 Philip Burton
has a truer grasp of the source of geometry than the finest
infidel professional.32

arithmetic
Turning to mathematics, we may not be surprised to find
that the word ‘arithmetica’, like ‘geometria’, is entirely
avoided in the Confessions, even though it had been current
in educated Latin for centuries. Instead, Augustine uses
‘numerus’.33 This is the regular translation in Republican
Latin—Cicero uses it twice in the context of Platonic
philosophy34—and, like ‘disputare’, it enables Augustine to
link the liberal arts to the Scriptures. The Confessions fam-
ously open with a reference to ‘numerus’, cited from Psalm
147: 5: ‘Magnus es, domine, et laudabilis valde. Magna
virtus tua, et sapientiae tuae non est numerus.’ Likewise,
many of the forty instances of ‘numerus’ and ‘numerare’ in
the Confessions appear in biblical citations. So, for instance,
Augustine can invoke Matthew 10: 20, ‘to you the hairs of
our head are numbered’ (‘tu vero, cui numerati sunt
capilli nostri’, 1. 12. 19), or the statement that Christ was
‘numbered among us’ (‘et numeratus est inter nos’, 5. 3. 5,

32
Conf. 5. 4. 7, citing Wisdom 11: 21: ‘fidelis homo . . . dubitare stultum
est quin utique melior sit quam mensor caeli et numerator siderum et
pensor elementorum et neglegens tui, qui omnia in mensura et numero et
pondere disposuisti’ (‘It is foolish to doubt that the believer is better off
than the measurer of the sky and the counter of the stars and the weigher
of the elements who has no regard for you, who have arranged everything
according to measure and number and weight’).
33
Compare Doct. Chr. 2. 38. 56, where Augustine uses ‘numeri
disciplina’ (as elsewhere just ‘numerus’ or ‘numeri’) of the unchanging
laws of mathematics, with relevance also for geometry and music, but
distinguished from the arbitrary ‘numerus’ of metre.
34
Cicero, De Re Publica 1. 10. 16: ‘cuius [sc. Platonis] in libris multis
locis ita loquitur Socrates, ut etiam cum de moribus, de virtutibus,
denique de re publica disputet, numeros tamen et geometriam et harmo-
niam studeat Pythagorae more coniungere’ (‘There are many passages in
[Plato’s] works where Socrates speaks in such a way that even when
discussing ethics, moral qualities, or politics, he betrays a desire to link
them to mathematics and geometry and harmony, Pythagoras-style’); cf.
De Finibus 5. 29. 87: ‘cur Plato Aegyptum peragravit ut a sacerdotibus
barbaris numeros et caelestia acciperet’ (‘why Plato scoured Egypt to learn
mathematics and astronomy from the barbarian priests’).
 The Vocabulary of the Liberal Arts 157
citing Isaiah 53: 12, Mark 15: 28)—this last in a passage
concerned explicitly with the relationship between math-
ematics and theology. By choosing ‘numeri’ rather than
‘arithmetica’, Augustine is able to portray God as the Great
Mathematician.

ph i lo s o ph y
Augustine famously encounters philosophy at the age of 18,
through reading Cicero’s lost protreptic Hortensius. Wisdom
is with God; and the Hortensius teaches him that ‘the love
of wisdom has a Greek noun: philosophy’ (‘amor autem
sapientiae nomen graecum habet philosophiam’, 3. 4. 7–8).
It would be easy to say that ‘amor sapientiae’ is Augustine’s,
or Cicero’s, gloss on ‘philosophia’. But Augustine does not
present us with the Greek term, then with a Latin equiva-
lent, or vice versa. Rather, ‘philosophia’ is presented as
simply the ‘Greek noun’ to describe the love of wisdom, in
the same way that ‘amor sapientiae’ is the corresponding
Latin noun phrase. True philosophy, he suggests, is above
and beyond linguistic particularism.35 It is true that this ini-
tial, positive evaluation is swiftly followed by a health warn-
ing, from the Apostle Paul: ‘See that none deceive you
through philosophy and empty distraction, according to
human tradition, according to the elements of this world’
(3. 4. 8, citing Col. 2: 8). But the reference here is to
‘bad’ philosophy, seen as characterized by atheist material-
ism and an obsession with intellectual pedigrees. This sort
of philosophy is always ‘philosophia’, its practitioners
‘philosophi’. The Latin terms are generally reserved for
‘good’ philosophy. In his catalogue of the arts in the memory
in Book 10, Augustine talks of philosophy—that is, in this
case, ethics—in purely descriptive terms. He speaks of the

35
Once again, we find a more dispassionate treatment of this discipline
in Doct. Chr. (2. 40. 60). The nuances of his introductory phrase ‘philos-
ophi autem qui vocantur’ (‘those who are called philosophers’) are hard to
read. Like the use of quotation marks in English, the expression ‘qui
vocantur’ might imply a rejection of the claim of (all) philosophers to be
friends of the truth; but such expressions are often used with Greek words,
even familiar ones, with little more implication than the practice of putt-
ing (say) German loan-words in italics.
158 Philip Burton
‘affectiones’ or ‘perturbationes animi’, but uses nothing that
we can call even a semi-technical translation of φιλοσοφα.
Elsewhere, Augustine typically uses some locution involving
‘sapientia’. For the verb, he avoids the pejorative ‘philos-
ophari’ in favour of the more positive calques ‘sapientiam
colere’ (6. 12. 20) or ‘diligere’ (3. 4. 8), or simply ‘sapere’.
For the noun, we find ‘studium sapientiae’ (4. 14. 21), or just
plain ‘sapientia’.
‘Sapientia’ is, once again, a familiar rendering of ‘philos-
ophia’ from Republican times onward. But, as in the other
cases already observed, Augustine blends this traditional
discourse of intellectual activity with his new biblical
wisdom. Sixteen of the sixty instances of ‘sapientia’ in the
Confessions appear in biblical citations, beginning with the
prominent citation of Psalm 147: ‘sapientiae tuae non est
numerus’ (1. 1. 1). ‘Sapientia’ is the quality par excellence of
God; quoting Paul again, Augustine even identifies it
with Christ: ‘The Only-Begotten himself has become our
wisdom and righteousness’ (‘ipse autem unigenitus factus est
nobis sapientia et iustitia’, 5. 3. 5, citing 1 Cor. 1: 24, 30).
And Augustine concludes his exegesis of the creation story
with the statement that God, the Supreme Philosopher,
teaches us all these meanings through a process of dialectic
(‘haec nobis disputas, sapientissime deus noster’, 13. 18. 23).

Having examined Augustine’s practice for each of the

‘artes’, we should now inquire into the principles lying
behind it. It is true that Latin writers often intersperse
Greek and Latin technical vocabulary, especially in lists;36
but that does not seem to be an adequate explanation
here. Why are established words so regularly reserved for
pejorative contexts, or else ignored altogether? It is true that
Latin writers frequently use clusters of Greek terms in order
to distance themselves from their subject-matter. But
Augustine himself explicitly warns Cresconius against criti-
cizing in Greek what he approves in Latin, and in fact a high

36
Compare, for instance, Petronius’ catalogue of ‘dialectica, astrono-
mia, sapientia, eloquentia’ (Satyricon 88, quoted above) with Quintilian’s
juxtaposition of ‘grammatice, musice, philosophia [praecepta sapientiae]’
with ‘ratio siderum’ and ‘eloquentia’ (Inst. Or. 1. 4. 4–5).
 The Vocabulary of the Liberal Arts 159
percentage of the Greek words in the Confessions are there
precisely because of their occurrence in biblical Latin. We
may grant that, up to a point, Greek words are used by
Augustine to indicate disapproval. But that principle by
itself will not account for the distribution of liberal-arts
terms that we find in the Confessions.
The clue to Augustine’s usage is surely the wider semantic
resonance of his translations. Our original concept of substi-
tutability may be stretched to breaking-point here. Perhaps
we should think of Greek words as having not a defined set
of familiar Latin translations, but a penumbra of various
translations. The extent of this penumbra is necessarily
uncertain, and it may be worth while considering some
borderline examples. Take ‘loquacitas’. We have seen this
word juxtaposed to ‘rhetorica’ (‘docebam in illis annis
rhetoricam, et victoriosam loquacitatem victus cupidine
vendebam’, 4. 2. 2). Augustine uses it also when speaking of
his retirement from teaching rhetoric or as a decision to
‘withdraw the service of my tongue from the speech-market’
(‘subtrahere ministerium linguae meae nundinis loquacita-
tis’, 9. 2. 2), shortly afterwards expressed as ‘gaining release
from the rhetorical profession’ (‘[solvi] a professione rhe-
torica’, 9. 4. 7). These juxtapositions in themselves suggest
an equivalence between the terms. Underlying them there
may be an extension of the analogical principle by which
many Latin translations are formed: <&ω: loquor :: ρ́ητορικ:
x, where x = ars loquendi/eloquentia/loquacitas. Augustine is
not above making such loaded analogies explicitly: compare
µυθ": fabula :: µυθικ": x, where x = fabulosus (Civ. 6. 5).37
If we accept that ‘loquacitas’ might be an equivalent to
‘rhetorica’ in some contexts, we might draw the analogy
for ourselves elsewhere; Monica, for instance, notably avoids
the ‘loquacitas’ of old women (Conf. 5. 9. 17), primarily

37
Disingenuously presented as the unfortunate outcome of the limited
resources of Latin. The passage in full reads ‘latine si usus admitteret,
genus quod [Varro] primum posuit, fabulare appellaremus. Sed fabulosum
dicamus; a fabulis eum mythicon dictum’. For this loaded use of the
pejorative -‘osus’ suffix, we may compare Conf. 4. 1. 1: ‘contentiosa
carmina et agonem coronarum faenearum’, where there seems to be an
implicit analogy αγ9ν: contentio :: αγωνιστικ": contentiosus.
160 Philip Burton
‘garrulity’, but perhaps, for Augustine, qualitatively no
different from empty rhetoric—pleasurable and persuasive
speech, but void of substance? And should we not also
think of the smooth-talking Manichees, also typically
loquacious?
We may also see the penumbra principle at work if we
consider again the case of dialectic. Both διάλογο and ‘dis-
putatio’ broaden their meaning in later usage to mean not
just dialogue but any intellectual disquisition, with or with-
out other interlocutors; a dialogue may thus be a monologue.
This is not, in itself, a problem for Augustine: in the
Soliloquies, he openly acknowledges that dialogue may occur
within one and the same individual.38 So in the Confessions
Ambrose’s sermons at Milan are described as disputations,
as is Augustine’s own self-consolatory disquisition on the
death of Monica. None the less, Augustine is never quite at
ease with purely monologic dialectic: the term ‘disputatio’ is
also used of Faustus the Manichee’s public talks, where
Augustine is dismayed to find that no time is allowed for
question and answer.39 In this respect, he shares a wide-

38
Augustine, Sol. 2. 7. 14: ‘Ridiculum est si te pudet, quasi non ob id
ipsum elegerimus huiusmodi sermocinationes . . . soliloquia vocari et
inscribi volo . . . cum enim neque melius quaeri veritas possit quam inter-
rogando et respondendo, et vix quisquam inveniatur quem non pudeat
convinci disputantem . . . placuit a me ipso interrogatum mihique
respondentem, deo adiuvante, quaerer’ (‘It is ridiculous for you to be
embarrassed, as if it were not for just this reason that we chose this form of
conversation . . . I would have them called and entitled the Soliloquies . . .
since truth cannot better be sought than by question-and-answer tech-
niques, and since it is hard to find anyone who is not embarrassed if shown
to be wrong in the course of a debate, I decided to seek truth by asking
myself and replying to myself, with the help of God’). Note the various
Latin terms for dialectic here: ‘sermocinatio’, ‘interrogare’, ‘disputare’.
39
Conf. 5. 6. 11: ‘aviditas mea . . . delectabatur quidem motu affectuque
[Fausti] disputantis . . . sed moleste habebam quod in coetu audientium
non sinerer ingerere illi . . . conferendo familiariter et accipiendo ac red-
dendo sermonem’ (‘my greed . . . was well satisfied with Faustus’ attitude
and outlook as a dialectician . . . but I found it regrettable that from
my position in the general mass of the audience I was not allowed to press
him on any points . . . in intimate conversation, trading words with
each other’). See also M. Vessey, ‘Conference and Confession: Literary
Pragmatics in Augustine’s “Apologia contra Hieronymum” [i.e. Confes-
sions]’, JECS 1 (1993), 173–201.
 The Vocabulary of the Liberal Arts 161
spread late antique concern that dialectic could easily
become mere verbal point-scoring, rather than an open
search for truth. Some Latin terms for public debate—
‘altercatio’, ‘certamen’—virtually invite this understanding.
Augustine notably avoids them in the Confessions.
We do, however, often find him using ‘colloquium’, ‘col-
loqui’, or sometimes ‘loqui cum’—not canonical translations
of διάλογο, but I think translations just the same. The
‘loq’- root is common in the translation literature as a render-
ing of the Greek λογ-; as witness Augustine’s own coinage
‘soliloquium’ for µον"λογο.40 And like his compatriot
Cyprian of Carthage he may have known the text ‘venite,
disputemus, dicit dominus’, as already cited above (Isaiah
1: 18 = Conf. 13. 19. 24) in the form ‘venite, colloquamini’.
Indeed, we need only see a few examples to realize that he is
using it in the sense ‘disputatio’. The sort of relaxed philo-
sophical dialogue Augustine hopes to have with Faustus
is a ‘colloquium’ (5. 6. 10). His question-and-answer dispu-
tation on the drunken beggar in Milan is described in similar
terms (‘locutus sum cum amicis qui mecum erant’, 6. 6. 9).
So too are the philosophical discussions he enjoys with
Monica before her death in Ostia (‘colloquebamur ergo soli’,
9. 10. 23; ‘cum quibusdam amicis meis materna fiducia
colloquebatur’, 9. 11. 28). Once we accept that ‘colloquium’
may have this force of ‘philosophical dialogue’, we may
sense its meaning in other, less expected contexts. For
example, Monica’s ‘amica colloquia’ (9. 9. 19) to her fellow
housewives on how to avoid getting a beating counts as
dialectic in two senses: specifically, because household
management is one of the oldest themes in the dialectical
tradition, and generally, because any homely chat with a
serious moral outcome may claim to be an inheritor of the
original Socratic technique.
Our penumbra, therefore, is hazy around the edges; we
might expect to find the liberal arts in yet more passages of
the Confessions, in some disguise or other. We might also
need to look harder at apparently everyday words which

40
For further examples from the translation literature, see Burton, Old
Latin Gospels, 131.
162 Philip Burton
also carry some intellectual or philosophical charge. One
example will suffice. We began our inquiry by looking at the
dialogical format of Augustine’s early manuals of the arts,
and in particular the word ‘interrogare’, ‘to pose a philo-
sophical question (requiring a yes-or-no answer)’. Once we
are alerted to this technical sense, our reading of the Confes-
sions is enriched at various points. Augustine interrogates,
among others, his hypothetical grammarian (1. 13. 22), his
soul (4. 4. 9), and himself (6. 6. 9); he is himself interrogated
by Nebridius (9. 3. 6). Book 10 is a lengthy ‘interrogatio’,
first of the created world, then on the senses, on the nature of
God; Books 11 and 12 present scriptural exegesis as dialectic
(11. 12. 14; 12. 4. 6). The language of dialectic runs deep in
the Confessions.
We have touched on the ways in which Augustine exploits
biblical language to support his argument, and this question
deserves some further consideration before we finish.
Augustine’s detection of references to the liberal arts in the
Book of Genesis or the prophecies of Isaiah strike us as far-
fetched, perhaps even an abuse of Scripture. On this reading,
the question of how the Bible became the focus of Western
culture could become a depressing one: the story of how
intellectuals abandoned any search for first principles in
favour of an ever more elaborate (and implausible) exegesis
of a sacred text. It is true that Augustine refers repeatedly
in the Confessions to the ‘auctoritas’ of Scripture, and takes
its inerrancy as axiomatic. We may congratulate him on his
‘creative engagement’ with Scripture, or some such quality—
patronizingly, if it is to be silently understood that we would
never venture to be quite so creative ourselves. More profit-
ably, we might investigate the intellectual framework within
which he describes his commitment to Scripture. This task
cannot be pursued any further here; yet it is clear that it
must involve two things: first, an investigation into the ter-
minology of ‘auctoritas’ and its intellectual pedigree; and
second, a consideration of the importance of providence
in Augustine’s scriptural exegesis. For if his discovery of
dialectic, geometry, and so forth in the Scriptures is (on one
level) the merest accident of translation, it is also clear that
from Augustine’s perspective these translations may them-
selves be seen as providential. The assertion of eternal
 The Vocabulary of the Liberal Arts 163
providence as manifested in the authority of the Scriptures
is rounded off with a Ciceronian phrase, ‘administratio
rerum humanarum’ (Conf. 6. 5. 7; compare De Natura
Deorum 2. 1. 3). If Augustine’s commitment to the Scrip-
tures is ultimately an act of faith, it is at least one that he
attempts to rationalize within the recognized intellectual
schemes of antiquity.41
To conclude: in avoiding the traditional terminology of
the arts, Augustine was attempting to revalorize them.42 His
alternative terms are a mixture of established and familiar
translations (such as ‘disputatio’, ‘numeri’, ‘figurae’), revived
archaisms (‘oratoria’, ‘litteratura’), and apparently novel
usages of his own (‘colloquium’). But in his distribution
of these terms, Augustine goes beyond mere variatio. By
avoiding Greek in his discussion of the higher arts, and by
exploiting the nuances of his various Latin translations, he
compels the willing reader to consider these ‘disciplinae’ as
something other—and more—than a knack or a trade or even
a profession that one learns at school or university. It is

41
On this matter, see the discussion by J. J. O’Donnell, Augustine:
Confessions, 3 vols. (Oxford, 1992), ii. 353–4, with references.
42
Conf. 10. 12. 19: ‘Audivi sonos verborum quibus significantur [sc.
numerorum dimensionumque rationes] cum de his disseritur; sed illi alli,
istae autem aliae sunt; nam illi aliter graece, aliter latine sonant; istae vero
nec graecae nec latinae sunt, nec aliud eloquiorum genus’ (‘I have
heard the sounds by which [mathematics and geometry] are signified when
they are discussed; but the sounds are one thing, and mathematics and
geometry are another. The sounds are different in Greek and Latin, but
mathematics and geometry are not in Greek or Latin, nor in any other type
of speech’). Cf. Conf. 10. 20. 29: ‘Nam hoc [sc. nomen vitae beatae] cum
latine audit Graecus, non delectatur, quia ignorat quid dictum sit; nos
autem delectamur, sicut etiam ille si graece hoc audierit; quoniam res ipsa
nec graeca nec latina est’ (‘When a Greek-speaker hears the Latin
word [“happiness”], he gets no pleasure from it, since he doesn’t know
what has been spoken. We, however, do get pleasure, just as he would if he
heard it spoken in Greek. This is because the thing itself is neither Greek
nor Latin’). In this respect, Augustine is closer than he would like to admit
to Porphyry of Tyre. For Porphyry, ‘the real barbarians are those who
cannot, or will not, speak the language of the fatherland’: G. Clark,
‘Translate into Greek: Porphyry of Tyre on the New Barbarians’, in
R. Miles (ed.), Constructing Identities in Late Antiquity (London, 1999),
112–32, at 130.
164 Philip Burton
perhaps unfortunate that he did not press further in this
direction. For however we may now judge his intentions, it is
clear from the later history of the ‘disciplinae’ that the
demystifying terminology of the Confessions had only a
limited impact.43
43
See Shanzer’s chapter above, esp. 104–10.
 pa r t i i i
Doctrina christiana: Beyond the Disciplines
 7
The Grammarian’s Spoils:
De Doctrina Christiana and the
Contexts of Literary Education
Cat h e r i n e M. Ch i n

Knowledge of the circumstances of Augustine’s education

in Roman North Africa is, of course, indispensable to the
writing of Augustine’s history, knowledge of the circum-
stances of Augustine’s own writing of De Doctrina Chris-
tiana no less important for an understanding of that work’s
place in Augustine’s thought. In this essay, however, I would
like to contextualize Augustine slightly differently: rather
than focusing on Augustine’s specific setting in the intel-
lectual history of the later Roman Empire, I would like
to explore the ways in which the seminal De Doctrina
Christiana is productive of two larger ideological contexts,
into which Augustine places the task of reading: namely,
the contexts of ‘paganism’ and ‘Christianity’. Analysis of
Augustine’s attitudes toward pre-Christian Roman literature
has often been formulated in terms of either the ‘conflict’
between Augustine’s Christianity and such literature, or the
‘accommodation’ of pagan thought in his overall Christian
scheme.1 Both approaches presuppose that the opposition of
two already existing categories, ‘Christianity’ and ‘pagan-

I would like to thank Andrew S. Jacobs and Elizabeth A. Clark, whose

critical suggestions and advice have immeasurably improved this chapter.
1
e.g., H.-I. Marrou, Saint Augustin et la fin de la culture antique (Paris,
1938; 2nd edn. with ‘Retractatio’, 1949); H. Hagendahl, Augustine and the
Latin Classics (Göteborg, 1967); P. Courcelle, Les Confessions de saint
Augustin dans la tradition littéraire: Antécédents et posterité (Paris,
1963), pt. 1; S. MacCormack, Shadows of Poetry: Virgil in the Mind of
Augustine (Berkeley and Los Angeles, 1998).
168 Catherine M. Chin
ism’, forms the background to Augustine’s thought. Much
scholarship in late ancient studies, however, has shown that,
in Robert Markus’s felicitous phrase, late Roman paganism,
at least, ‘existed only in the minds, and, increasingly, the
speech habits, of Christians’.2 The production of paganism
in late ancient speech habits is, I think, central to under-
standing Augustine’s own formulation of paganism and, by
extension, to understanding how Christianity emerges as
paganism’s polar opposite in Augustine’s writing. De Doct-
rina Christiana is, I would like to argue, one of the textual
moments in which speech habits produced both paganism
and Christianity. Specifically, I shall argue that Augustine in
De Doctrina Christiana uses the decontextualizing and dis-
locating techniques of ancient grammatical writing to pro-
duce the opposing concepts of Christianity and paganism,
and to locate the educated Christian subject in relation to
them.

1. au g u st i n e o n s i g n s a n d t h i n g s :
d i s lo c at i o n a n d t r i a n g u lat i o n
Although the ‘tractatio scripturarum’ that forms the subject
of De Doctrina Christiana is often studied in relation to
classical rhetorical ‘tractatio’,3 the ‘treatment’ of Scripture in

2
R. Markus, The End of Ancient Christianity (Cambridge, 1990), 28.
Earlier works include A. Cameron, ‘Paganism and Literature in Late
Fourth Century Rome’, in M. Fuhrmann (ed.), Christianisme et formes
littéraires de l’Antiquité tardive en Occident (Geneva, 1977), 1–30;
J. J. O’Donnell, ‘Paganus’, Classical Folia 31 (1977), 163–9; and idem,
‘The Demise of Paganism’, Traditio 35 (1979), 45–88. Much of the rele-
vant literature concentrates on the question of a ‘pagan revival’ in the
fourth century; for a recent consideration of the issue, see C. W. Hedrick,
Jr., History and Silence: Purge and Rehabilitation of Memory in Late
Antiquity (Austin, Tex., 2000), esp. ch. 3, ‘Unspeakable Paganism?’.
3
G. A. Press, ‘The Subject and Structure of Augustine’s De Doctrina
Christiana’, AugSt 11 (1980), 99–124; K. Eden, ‘The Rhetorical Tradition
and Augustinian Hermeneutics in De Doctrina Christiana’, Rhetorica 8
(1990), 45–63; M. Scanlon, ‘Augustine and Theology as Rhetoric’, AugSt
25 (1994), 37–50; C. Harrison, ‘The Rhetoric of Scripture and Preaching’,
in R. Dodaro and G. Lawless (eds.), Augustine and his Critics (London,
2000), 214–30.
 The Grammarian’s Spoils 169
Books 1–3 has much in common with late ancient gram-
matical textual analysis, also known as ‘tractatio’.4 The
reading practices which Augustine advocates for the reso-
lution of verbal ambiguity, for example (language study,
appropriate word division, and familiarity with a wide
variety of word usages5), are those developed in the schools
of the ‘grammatici’ for the interpretation of ancient texts.6
Augustine himself explicitly compares his task in De Doc-
trina Christiana to the work of the late Roman ‘litterator’ or
‘grammaticus’:7 ‘Whoever teaches how [the Scriptures]
should be understood is like the expositor of letters, who
teaches how they ought to be read.’8 The grammarian
Diomedes, Augustine’s contemporary,9 defines the task of
grammar simply as ‘the understanding of the poets and the
ready elucidation of writers and historians, and the logic

4
Priscian, for example, analysing the Aeneid, begins his detailed
grammatical discussions of each line with the imperative ‘Tracta singulas
partes’: ed. H. Keil, Grammatici Latini (hereafter GL) iii (Leipzig,
1860) at 1. 9, 2. 44, 3. 67, 4. 84, 5. 93, 6. 109, 7. 135, 8. 157, 9. 169, 10.
185, 11. 198, 12. 210. Cf. M. Irvine, The Making of Textual Culture:
‘Grammatica’ and Literary Theory, 350–1100 (Cambridge, 1994), 178–89.
5
Doct. Chr. 3. 1. 1.
6
Irvine, Making of Textual Culture, 179–83; for other examples of
early Christian use of grammatical techniques, see, e.g., B. Neuschäfer,
Origenes als Philologe (Basel, 1987); F. Young, Biblical Exegesis and the
Formation of Christian Culture (Cambridge, 1997); and C. Schäublin,
Untersuchungen zu Methode und Herkunft der Antiochenischen Exegese
(Cologne, 1974).
7
In ancient usage the Roman ‘litterator’ is sometimes distinguished
as the teacher of elementary literacy, the more advanced linguistic and
literary instruction being reserved for the ‘grammaticus’ proper. However,
as Robert Kaster has noted, there was in practice considerable overlap
between these two teaching professions in Late Antiquity: ‘Notes on
“Primary” and “Secondary” Schools in Late Antiquity’, Transactions
of the American Philological Association 113 (1983), 323–46. For further
discussion of this passage of the prologue, see Pollmann below, 210–11.
8
Praef. 9. All translations from Doct. Chr. are my own; references to
the Latin text are to the edition of R. P. H. Green Augustine: De Doctrina
Christiana (Oxford, 1995), though I have retained only the traditional
numbering.
9
On the dating of Diomedes, see R. A. Kaster, Guardians of Language:
The Grammarian and Society in Late Antiquity (Berkeley and Los Angeles,
1988), 271.
170 Catherine M. Chin
of speaking and writing correctly’.10 Augustine similarly
describes the programme of De Doctrina Christiana as cover-
ing first the correct method of understanding Scripture, and
then the correct method of presenting it (1. 1. 1). That De
Doctrina Christiana shares its basic approach with the
late ancient ‘ars grammatica’ should not, of course, come as
a surprise: Augustine’s discussions of language in other
works also strongly recall the work of fourth- and fifth-
century Latin grammarians, and Augustine’s own early De
Grammatica was, arguably, a work in the same grammatical
tradition as Servius or Donatus.11
The particular placement of grammar in De Doctrina
Christiana, however, is significant. The ‘ars grammatica’
proper, the discussion of verbal signs, does not begin until
Book 2. The discussion of signs in Book 1 seems cursory:
after noting the difference between ‘things’ (‘res’) and ‘signs’
(‘signa’) (1. 2. 2.), Augustine spends the bulk of Book 1 on
‘things’ and the kinds of love the reader ought to bear
them. The relationship between the reader and ‘things’
is, famously, construed as a relationship based on love
(‘amor’) (1. 4. 4), ultimately on the love of God, who is, for
Augustine, the highest ‘thing’: ‘una quaedam summa res’
(1. 5. 5). Augustine thus posits the dislocation of humanity
from God (e.g., 1. 10. 10) as paradigmatic for the relation-
ship between readers and ‘things’ more generally. This dis-
location, in turn, underlies the ‘amor’ that draws humanity
to ‘things’ and to God. In other words, in Book 1 of De

10
GL i. 426. This is, clearly, a broad definition of grammar: Servius,
in his commentary on Donatus, argues that grammar proper is concerned
especially with the eight parts of speech (GL iv. 405), but ‘Sergius’ (on
whom see Kaster, Guardians of Language, 429–30), also commenting on
Donatus, repeats Diomedes’ assertion that the ‘ars grammatica consists
principally in the understanding of the poets and in the logic of speaking
or writing correctly’ (GL iv. 486).
11
Cf. G. Bellissima, ‘Sant’ Agostino grammatico’, in Augustinus
Magister (Paris, 1955), i. 35–42; J. Collart, ‘Saint Augustin grammairien
dans le De magistro’, REAug 17 (1971), 279–92; V. Law, ‘St. Augustine’s
“De grammatica”: Lost or Found?’, RechAug 19 (1984), 155–83; Irvine,
Making of Textual Culture, 169–78; at p. 178 Irvine calls Doct. Chr. ‘a
Christian ars grammatica’.
 The Grammarian’s Spoils 171
Doctrina Christiana, the reader is construed as a subject
fundamentally desiring the enjoyment of God, since Augus-
tine defines the relationship between humanity and ‘things’,
with God as the paradigmatic ‘thing’, as a separation that
engenders ‘amor’ between the two.12
This amorous—indeed, in the Platonic sense, erotic—
relationship forms the background into which grammar is
to be set. Anne Carson, in her discussion of eros in Greek
literature, remarks, with regard to Sappho’s fragment 31:
‘Where eros is lack, its activation calls for three structural
components: lover, beloved, and that which comes between
them.’13 That Augustine conceives of the relationship
between humanity and ‘things’, particularly divine things, as
based on lack, and resulting in ‘amor’, is clear, both in De
Doctrina Christiana and in other works.14 In De Doctrina
Christiana, ‘that which comes between’ readers and ‘things’
is the sign: ‘All teaching is either of things or of signs, but

12
H.-J. Sieben has argued that ‘caritas’ in Doct. Chr. is among the
‘things’ at which Scripture aims: ‘Die “Res” der Bibel: Eine Analyse von
Augustinus, De Doctrina Christiana 1–111’, REAug 21 (1975), 72–90, at
78–9. Although Augustine uses ‘caritas’ more frequently than ‘amor’ in
Book 1, it is, I think, significant that ‘amor’ provides Augustine with his
definition of ‘enjoyment’ at 1. 4. 4, and that this ‘amor’ is directed toward
the Trinity at 1. 5. 5. The conflation of ‘amor’, ‘caritas’, and other such
terms (e.g., ‘dilectio’ at 1. 35. 39) under the heading of ‘love’ may not be
entirely out of order, as Augustine does not always use them as distinct
technical terms: cf. K. Pollmann, Doctrina Christiana: Untersuchungen zu
den Anfängen der Christlichen Hermeneutik unter besonderer Berücksichti-
gung von Augustinus, De doctrina christiana (Fribourg, 1996), 126–7. It is
important here to note, however, that ‘caritas’ and ‘amor’ may not be
completely synonymous for Augustine, since at Doct. Chr. 3. 10. 16, he
describes ‘caritas’ as movement towards ‘enjoyment’, earlier defined as
an instance of ‘amor’. In short, even if ‘caritas’ is an Augustinian ‘res’, the
relation between ‘things’ and readers will still be based on ‘amor’. On the
interpretive function of ‘caritas’ in Doct. Chr., see esp. Pollmann, Doctrina
Christiana, 121–47, and W. S. Babcock, ‘Caritas and Signification in
De Doctrina Christiana 1–3’, in D. W. H. Arnold and P. Bright (eds.), De
Doctrina Christiana: A Classic of Western Culture (Notre Dame, Ind.,
1995), 145–63, at 154–7.
13
A. Carson, Eros the Bittersweet (Princeton, 1986), 16.
14
Most famously, perhaps, at Conf. 10. 27. 38.
172 Catherine M. Chin
things are learned through signs’ (1. 2. 2).15 Moreover, the
pre-eminent form that the sign takes, for Augustine, is the
word, since words are signs ‘whose whole use is to signify.
No one uses words except to signify’ (1. 2. 2). Nouns, verbs,
conjunctions, and the other parts of speech are the things
that, as words, ‘have gained supremacy in signifying’; ‘all
other signs are scant in comparison to words’ (2. 3. 4.) To the
extent that it was the discipline of grammar, in antiquity,
that concerned itself with the functioning of individual
words,16 the third part of Augustine’s erotic triangle is
grammar, the science of signs. Signs ‘come between’ human-
ity and ‘things’, both in the sense of mediating between them
(since things are learned through signs) and in the sense of
perpetuating their disjunction: not only does misunder-
standing the signs of Scripture lead the reader astray (1. 36.
40–1), but the mere existence of scriptural signs under-
scores the separation of the reader from the divine ‘res’
(1. 37. 41–38. 42). In the first book of De Doctrina Christiana,
then, Augustine triangulates:17 having construed the rela-
tionship between humanity and God as essentially dis-
junctive, he supplies the necessary third point of the lover’s
triangle, ‘that which comes between them’, under the guise
of grammar. Grammar occupies the space of dislocation.18

15
On the significance of Augustine’s ‘per’ in ‘res per signa’, see C. P.
Mayer, ‘Res per signa: Der Grundgedanke des Prologs in Augustins
Schrift De doctrina christiana und das Problem seiner Datierung’, REAug
20 (1974), 100–12, at 104.
16
Grammatical ‘tractatio’, as illustrated most prominently in Priscian’s
Partitiones, tended to be either word-by-word analysis of written works
or discussion of individual words in phrase-by-phrase reading; for dis-
cussion of one such analysis, see Kaster, Guardians of Language, ch. 5, on
Servius’ Virgil commentary. On earlier uses of etymology in grammatical
analysis, see M. Amsler, Etymology and Grammatical Discourse in Late
Antiquity and the Early Middle Ages (Amsterdam, 1989), esp. 15–31.
17
Carson, Eros the Bittersweet, 17.
18
For ‘triangulation’ as used here, see ibid. 17. Carson (p. 11) indicates
her debt to Lacan for the description of erotic lack; Lacan’s interest in
structuralist approaches to desire is articulated esp. in ‘The Function and
Field of Speech and Language in Psychoanalysis’, in J. Lacan, Écrits,
trans. A. Sheridan (New York, 1977), 30–113. Lacan locates language, as
the symbolic, in the displacement of ‘the thing’, and suggests (p. 104) that
this displacement is instrumental in producing desire.
 The Grammarian’s Spoils 173
2. au g u st i n e o n s po i l s a n d s c r i pt u r e :
r e c o n t e xt ua l i zat i o n a n d g r a m m a r
There is a further sense, however, in which grammatical
writing in Late Antiquity served as a forum of dislocation
and of decontextualization. One of the tasks of the ‘ars
grammatica’ was to remove verbal signs from their original
signifying contexts in both written texts and spoken
language, and to reconfigure them as signs of linguistic
regularity.19 This shift from what Roland Barthes calls the
‘symbolic’ function of the sign to the ‘paradigmatic’ func-
tion is a fundamentally dislocating and relocating gesture.20
It entails the conceptual dismantling, as it were, of prior
contexts, and the imagination of new contexts into which the
same signs will fit. The process is most visible in the use
of Latin quotations in handbooks of ‘ars grammatica’. For
example, to illustrate alliteration, the grammarians Donatus,
Charisius, and Priscian quote the following line from
Ennius: ‘o Tite tute Tati tante tyranne tulisti’.21 In its new
context in the grammars, the line refers not primarily to any
Titus but to the principle of alliteration itself. The repetition
of the single, now mobile, line in several handbooks sug-
gests, indeed, the persistence of the paradigmatic associ-
ations of the line, superseding its symbolic associations.
Individual words are also used in the handbooks to signify
grammatical points quite different from their everyday
meanings. Donatus, for example, uses the verbs ‘sto’ (‘I
stand’) and ‘curro’ (‘I run’)—on one level antonyms—to
illustrate precisely the same thing: namely, what an active
verb is.22 Notably, it is not the ‘meaning’ of these verbs to
which Donatus appeals, but their morphology, as he defines

19
The appeal to written texts and to common usage generated much
debate in antiquity over the relative authority of ‘auctoritas’ over and
against ‘usus’. On Augustine’s concept of ‘authority’, see K.-H. Lütcke,
Auctoritas bei Augustin (Stuttgart, 1968); and Amsler, Etymology and
Grammatical Discourse, 100–8.
20
R. Barthes, ‘The Imagination of the Sign’, in idem, Critical Essays,
trans. R. Howard (Evanston, Ill. 1972), 205.
21
Ennius, Annales 1. 113; Donatus, Ars Grammatica 3. 4. 5 (GL iv. 398);
Charisius, Ars Grammatica 4. 4 (GL i. 282); Priscian, Partitiones 7. 141
(GL iii. 492).
22
GL iv. 360.
174 Catherine M. Chin
an active verb as one ending in ‘-o’, to which ‘r’ may not be
added: we do not say ‘stor’ or ‘curror’.23 The same approach is
found in grammatical work more broadly: e.g., in Priscian’s
Partitiones, in which the answer to ‘Why [is arma] neuter?’ is
‘Because all nouns ending in “-a” in the plural are without
question neuter’.24 Words are here entered into a signifying
system quite different from the obvious (semantic) one. To
the extent that the purpose of the handbooks is to create a
grammatical metalanguage out of the language already in
use, decontextualization and recontextualization of signs are
practices fundamental to them. Without the possibility of
transferring words from a symbolic signifying context to a
paradigmatic context, grammar as a discipline would be
impossible.
Not only semantic units, but entire units of knowledge,
were the objects of the grammarian’s transferral; in textual
analysis the grammarian was to redistribute, piecemeal, the
knowledge produced in other disciplines. Macrobius’ Satur-
nalia, which in grammatical fashion represents the bulk of
ancient learning as an extended gloss on the writings of
Virgil,25 advocates this attitude toward knowledge in its
preface: ‘Let us gather then from all sources and from them
form one whole, as single numbers combine to form one
number.’26 The same sentiment is expressed in Martianus
Capella’s Marriage of Philology and Mercury, in which the
personified Grammar is held to have authority in poetry,
rhetoric, philosophy, history, mathematics, and music, all of
which contribute to the explication of texts.27 Not only is
the grammarian configured, ideally, as a polymath,28 but

23
GL iv. 360.
24
GL iii. 461.
25
Saturnalia 1. 24. 9–25; but note the difference between Macrobius’
portrayal of grammatical practice as performed by the fictional figure of
Servius and that performed by the historical figure of Servius: Kaster,
Guardians of Language, ch. 5.
26
Saturnalia 1, praef. 8, trans. Percival Vaughan Davies, in Macrobius:
The Saturnalia (New York, 1969).
27
De Nuptiis Philologiae et Mercurii 3. 225–7, 230, 263, 326 (ed.
Kopp).
28
For discussion of this trope in earlier texts, see Kaster, Guardians of
Language, 59–64; Irvine, Making of Textual Culture, 45, 51.
 The Grammarian’s Spoils 175
grammar itself is understood to be the practice of breaking
down larger fields of knowledge into ‘mobile’ decontextual-
ized units.
It is, of course, this conception of knowledge, as some-
thing that can be transferred in discrete fragments from one
context to another, that Augustine invokes in Book 2 of De
Doctrina Christiana, on the spoiling of the Egyptians:
[T]he Egyptians had not only idols and heavy burdens, which the
people of Israel hated and fled, but also vessels, gold and silver
ornaments, and clothes, which that people secretly claimed for a
better use when they left Egypt. (2. 40. 60)
The fourth-century equivalents of these spoils, according to
Augustine, are the ‘liberal disciplines more fitting to be used
for truth’ (ibid.).29 The transference of knowledge from
one arena to another is not merely a matter of sharing
methodology or philosophical assumptions; Augustine has
in mind something rather more ‘literal’, as his remarks
immediately before the Exodus metaphor show:
I think it would be possible for someone who could, and who
wanted to perform some great and beneficial task for the use of
the brothers, to commit to writing the geography, animals,
plants and trees, stones and metals and whatever sorts of unknown
things Scripture mentions, discussing and explaining them.
(2. 39. 59)30

29
The trope of ‘spoiling the Egyptians’ as a metaphor for Christian
‘use’ of the liberal arts is not unique to Augustine; for discussion of the
use of the Exodus metaphor from Marcion on, see C. Gnilka, Chrēsis: die
Methode der Kirchenväter im Umgang mit der Antiken Kultur, i: Der
Begriff des ‘rechten Gebrauchs’ (Basel, 1984), 57 n. 120. At pp. 102–33
Gnilka surveys the use by early Christian and other ancient writers of
the parallel trope of the student as bee—the student, bee-like, is to take the
‘nectar’ of literature and put it to proper, usually philosophical, use—but
without connecting it with the work of ancient grammarians.
30
Augustine may here have in mind the sort of project more commonly
associated with Roman antiquarianism: e.g., the second-century diction-
ary of Festus, De Verborum Significatu, ed. W. M. Lindsay (Leipzig,
1913), which lists meanings particularly of earlier Latin religious terms,
arranged roughly alphabetically. In 419, Augustine himself compiled a
list, the Locutiones in Heptateuchum, not of geographical or botanical
terms from the Bible, but of unidiomatic Latin phrases in the Heptateuch,
as a similar kind of reading aid.
176 Catherine M. Chin
The idea of compiling ‘source-books’ of knowledge that
can be applied to the explication of Scripture suggests a
thoroughgoing dislocation and recontextualization of know-
ledge: Augustine, like his grammarian contemporaries, is
interested in changing the signification of previously exist-
ing signs by ‘literally’ removing them from their symbolic
signifying contexts.31 The ‘tractatio scripturarum’, following
this method of reading, is simply another kind of spoliation;
and Augustine’s spoliating style is that of the late ancient
grammarian.

3. p r e s e n c e , a b s e n c e , a n d g r a m m a r
The configuration of scriptural ‘tractatio’ as spoliation,
another level of dislocation within the already disjunctive
erotic triangle, suggests the ways in which De Doctrina
Christiana addresses larger issues of cultural and religious
identity. As spoliation, grammatical practice not only rejects
‘original’ contexts, it simultaneously evokes them.32 Servius’
and Priscian’s commentaries are not, after all, meant to
occlude Virgil, but to illustrate, and perpetuate, his impor-
tance in later Roman literary education. Analysis of particu-
lar words in grammatical texts provokes repeated reference
to ‘antiqui’ and to usages common ‘apud maiores’.33 Here
the grammarians imagine a historical context into which the
text, removed from its literary context, can be placed. It is,
however, a past that has been homogenized to a great
extent: little distinction is made in the handbooks between
exempla from very different periods in Latin literary history,
from Plautus to Horace. Donatus can introduce nearly any
quotation with the homogenizing ‘ut’; 34Pompeius, more
precise with names, none the less at one point runs through

31
This disjunctive procedure is applied even to the signs of Scripture,
which must be brought into the context of the ‘rule of faith’ in order to be
understood properly: Doct. Chr. 3. 2. 2.
32
Cf. Lacan, ‘Function and Field’, 86: ‘For the function of language is
not to inform but to evoke.’
33
For discussion of the negotiation necessary between the authority
of ‘the ancients’ and the grammarian’s authority, see Kaster, Guardians of
Language, 171–93.
34
e.g., Plautus at GL iv. 393 and Horace at iv.395.
 The Grammarian’s Spoils 177
quotations from Virgil, Persius, Terence, and Plautus in the
space of about twenty lines,35 without distinguishing the
authors by anything other than name. The historical context
created in grammatical work is, then, a particularly broad
one, a past that is notable mostly for being different from the
present.36 The decontextualizing practices of grammar thus
entail the reimagining of a broad ‘originary’ context, an
‘antiquity’, that marks the difference between the ‘auctor’
and the later reader.
Similarly, in Augustine’s metaphor of spoliation, the gold
and silver of the Egyptians may be used by the Israelites in
Egypt, but they are first explicitly marked as ‘Egyptian’.
Hence Augustine’s query: ‘Do we not see with how
much gold and silver and clothing Cyprian, that sweetest
teacher and blessed martyr, was laden when he left Egypt?’
(2. 40. 61). The very visibility of Cyprian’s ‘Egyptian’ goods
argues for the continuing ‘Egyptianness’ of the liberal arts in
Augustine’s scheme. Here an originary context is invoked
again to mark difference, now the difference between
‘Egypt’/‘paganism’ and ‘Israel’/‘Christianity’. Moreover, by
the proposed confinement of the liberal arts to the kinds of
‘source-books’ that Augustine imagines, their marking as
‘pagan’—that is, not ‘Christian’—is, at least in theory, per-
petuated. This simultaneous recontextualization and decon-
textualization of knowledge is the grammatical matrix
within which ‘paganism’ is produced in De Doctrina
Christiana. Augustine’s excursus in Book 2, the list of the
branches of knowledge, and of what from them is to be either
retained or rejected (2. 19–42),37 is as much a programme of

35
He is illustrating the uses of the noun: GL v. 136.3–25.
36
The division of time into ‘then’ and ‘now’ as a hermeneutical tech-
nique in Late Antiquity: E. A. Clark, Reading Renunciation: Asceticism
and Scripture in Early Christianity (Princeton, 1999), 145–52; also Kaster,
Guardians of Language, esp. 183.
37
The role of this list in Doct. Chr. has long been debated; Pollmann,
Doctrina Christiana, 89–108 (as in her essay in the present volume), places
it within the overall structure of the work by reading Doct. Chr. as funda-
mentally dihairetic in structure; see esp. 149–55 on Book 2. L. M. J. Ver-
heijen, however, has argued that the list of pagan studies is a ‘digression’
from the primary argument of Book 2: ‘Le De Doctrina Christiana de
Saint Augustin: Un manuel d’herméneutique et d’expression chrétienne
178 Catherine M. Chin
‘pagan’ education as it may be of Christian.38 Or, to posit
another Augustinian triangulation: between the Christian
reader and the ‘tractatio scripturarum’ must come a third
term, paganism, the ‘doctrinae apud gentes’ (2. 39. 58),
which both mediates and perpetuates the disjunctive rela-
tionship between reader and text.
At the same time, however, Augustine uses the desiring
relationship between Christian reader and sacred text to
reject the ‘original’, ‘pagan’ contexts of his spoils. Augustine
refers to past authorities in the same homogenizing terms as
the grammarians, but does so in order to highlight the need
to assign them to a different place in the Christian scheme.39
As spoils, the liberal arts are there precisely to be recontextu-
alized, moved from a hypothetical ‘Egypt’ to a hypothetical
‘Israel’. Here again is a matrix within which Christianity, as
the opposite of paganism, can be imagined and invoked. If
‘pagans’ are the monolithic ‘gentes’ from whom ‘doctrinae’
are taken, ‘Christians’ are the equally monolithic ‘fratres’ on
whose behalf the ‘gentes’ are despoiled (2. 39. 59). Augustine’s

avec, en II.19.29–42.63, une charte fondamentale pour une culture

chrétienne’, Augustiniana 24 (1974), 10–20. While I would not argue that
the discussion at 2. 19–42 is merely tangential to the rest of Book 2, I agree
with C. Schäublin that Augustine here ‘abruptly shifts his viewpoint’:
‘De Doctrina Christiana: A Classic of Western Culture?’, in Arnold and
Bright (eds.), De Doctrina Christiana: A Classic of Western Culture, 47–67,
at 50. The thoroughness of the ‘review’ of learning in the passage allows
Augustine to conjure up ‘paganism’ precisely by means of the mass of
detail not directly pertinent to his argument (e.g., the list of different
kinds of superstition at 2. 20. 31); cf. Lacan, ‘Function and Field’, 86:
‘what is redundant as far as information is concerned is precisely that
which does duty as resonance in speech.’ [For a contrasting view of the
relevance of Augustine’s treatment of superstitious practices in Book 2,
see Klingshirn’s essay above—Eds.]
38
For the controversy over whether or not Doct. Chr. is a programmatic
guide to Christian ‘education’ or ‘culture’, see E. Kevane, ‘Augustine’s
De Doctrina Christiana: A Treatise on Christian Education’, RechAug 4
(1966), 97–133; Verheijen, ‘Le De Doctrina Christiana de Saint Augustin’;
G. A. Press, ‘Doctrina in Augustine’s De Doctrina Christiana’, Philosophy
and Rhetoric 17 (1984), 98–120, and Schäublin, ‘De Doctrina Christiana:
A Classic of Western Culture?’, 47–52.
39
Cf. Amsler, Etymology and Grammatical Discourse, 102–3.
 The Grammarian’s Spoils 179
rhetorical flourishing of famous Christian names—Cyprian,
Lactantius, Victorinus, Optatus, and Hilary, ‘passing over in
silence those who are still alive, and innumerable Greeks’
(2. 40. 61)—serves less to illustrate appropriate use of the
liberal arts than to present the reader with an imagined
crowd of ‘Israelites’ who have left ‘Egypt’. Augustine
presents Cyprian and Lactantius as standing in for a large,
undifferentiated, and anonymous body of ‘our many good
faithful men’ (ibid.). In the same way that the grammarians’
‘antiqui’ conjure a vague ‘classical world’, Augustine’s
language here conjures an equally vague Christian one.40
The category ‘the people of God’ stands as a structural
parallel to ‘Egyptians’, and, importantly, is presented as
obviously separate: ‘spoils’, as such, must be transferred
from one owner to another. The idea of ‘spoils’ here invokes
the two possible owners of the liberal arts in Augustine’s
scheme: Christianity and paganism.
The production of ‘Christianitas’ as an abstraction is
intimately related to the grammatical tasks of decontextual-
izing and recontextualizing knowledge for the explication of
texts. The metaphor of spoiling the Egyptians implies not
only an ‘Egypt’ and ‘Israelites’, but also a more general
‘Israel’. Augustine ends his use of the metaphor in Book 2
by claiming that ‘the wealth of gold, silver, and clothing
that that people took with them out of Egypt’ was small
compared to ‘that of the riches which it had afterwards
in Jerusalem, as was evident especially during the reign of
Solomon’ (2. 42. 63). Christianity is finally posited as a loca-
tion and a separate political entity, parallel to the ‘Egypt’ of
‘paganism’. The idea of spoliation allows Augustine to
project an independent existence for Christianity, in a way
that a more literal description of late ancient Christians as
inhabitants, and products, of the Roman Empire might

40
‘Lists’ of important figures deployed to evoke ‘a certain abstract
commonality’: P. Cox Miller, ‘ “Differential Networks”: Relics and Other
Fragments in Late Antiquity’, JECS 6 (1998), 113–38, at 134; and eadem,
‘Strategies of Representation in Collective Biography: Constructing the
Subject as Holy’, in T. Hägg and P. Rousseau (eds.), Greek Biography and
Panegyric in Late Antiquity (Berkeley and Los Angeles, 2000), 209–54.
180 Catherine M. Chin
41
not. While on one level Augustine’s use of the story of
Israel leaving Egypt is clearly secondary to the advent of
Christianity, on another, the story becomes a device for pro-
ducing Christianity as a free-standing conceptual entity. It
is the ultimate goal of the Israelites’ departure, and thus the
place of greatest ‘usefulness’ in the larger goal of progress
toward the divine (2. 42. 63). The production of Christianity
as such in Augustine’s speech habits thus occurs through the
metaphor of spoliation, inseparable from the simultaneous
imagination, and appropriation, of ‘doctrinae apud gentes’.42
By creating these parallel locations, the dislocation of
knowledge—the idea of spoliation—invokes the opposed
categories of Christianity and paganism as the two cultural
contexts for late Roman education. Moreover, the two com-
peting ideas are placed within the amorous disjunction
between reader and ‘res’, the disjunction with which
Augustine opens De Doctrina Christiana. At 3. 1. 1, Augus-
tine maintains that the decontextualized ‘doctrinae apud
gentes’ are simply to be counted among the ‘necessary
things’ for the reader of Scripture.43 Similarly, Christianity
is a parallel region of utility in providing the necessary (and
markedly singular) ‘doctrina’ for the Christian reader,
though it is a region, ‘Israel’, whose origins are persistently
marked as ‘Egyptian’. This placement suggests, in turn, that
41
Geographical metaphors used to create religious identity, in a related
context: B. Leyerle, ‘Landscape as Cartography in Early Christian
Pilgrimage Narratives’, Journal of the American Academy of Religion
64 (1996), 119–43; J. Elsner, ‘The Itinerarium Burdigalense: Politics and
Salvation in the Geography of Constantine’s Empire’, JRS 90 (2000),
181–95. G. Frank, The Memory of the Eyes: Pilgrims to Living Saints in
Late Antiquity (Berkeley and Los Angeles, 2000), considers the ways in
which geographical narrative is linked with the ideological productiveness
of group biography
42
Of twenty-four instances of the word ‘christianus’ in Doct. Chr.,
fourteen occur in Book 2, in the course of Augustine’s description of the
branches of learning and how they are to be despoiled for the ‘tractatio
scripturarum’: 2. 12. 17; 2. 16. 24; 2. 18. 28; 2. 23. 36; 2. 25. 38; 2. 25. 40;
2. 29. 45; 2. 35. 53; 2. 39. 59 (twice); 2. 40. 60 (twice); 2. 40. 61; 2. 41. 62.
The other occurences are at praef. 4 and 5 (twice); 1. 14. 13; 1. 30. 32; 3. 8.
12; 4. 1. 1; 4. 7. 11; 4. 14. 31; 4. 31. 64.
43
In this passage, the phrases ‘scientia linguarum’ and ‘cognitio
quarundam rerum necessarium’ denote what Augustine has covered in
Book 2.
 The Grammarian’s Spoils 181
the imaginative constructs of Christianity and paganism
both mediate and perpetuate the separation that is the basis
of the work as a whole.44 In Augustine’s approach to the
‘tractatio scripturarum’, the presence of Christianity is con-
stantly invoked, addressed, and desired; yet Christianity’s
absence, paganism, ‘Egypt’, must continually be called upon
in order for Scripture to be treated at all.

4. c o n c lu s i o n
In De Doctrina Christiana, the two possible contexts in
which reading can be undertaken are Egypt and Israel,
paganism and Christianity, and reading necessitates
movement from one to the other. To the extent that the later
history of Latin reading, and of Latin Christianity, could
continue the rhetorical use of this division,45 the project
visible in De Doctrina Christiana, as in other works of the
same period,46 may be seen to have done its ideological work.
At the same time, however, the ambiguity and tension
involved in the series of triangulations that Augustine
proposes also involve a perpetuation of the paradoxical con-
ditions of the division: the later transmission of classical
literature might then be seen, not simply as following an

44
Carson, Eros the Bittersweet, 117, refers to R. Barthes, A Lover’s Dis-
course, trans. R. Howard (New York, 1978), 15, to describe the evocative
yet paradoxical aspect of such amorous writing: ‘Endlessly I sustain the
discourse of the beloved’s absence; actually a preposterous situation; the
other is absent as a referent, present as allocutory. This singular distortion
generates a kind of insupportable present; I am wedged between two
tenses, that of the reference and that of the allocution: you have gone
(which I lament), you are here (since I am addressing you). Whereupon I
know what the present, that difficult tense, is: a pure portion of anxiety.’
Cf. Lacan, ‘Function and Field’, 65.
45
As in, e.g., Cassiodorus’ separation of ‘divine’ and ‘human’ in the
Institutiones; or, more negatively, the use of ‘Egypt’ in Caesarius of Arles,
Sermo 99 (ed. G. Morin), which compares traditional learning with the
biblical ten plagues. On grammar in Cassiodorus, see Irvine, Making of
Textual Culture, 195–209.
46
For fuller discussion of the trope of ‘separation’ in patristic litera-
ture, and of similar approaches to the question of reading by Christian
writers, see Gnilka, Chrēsis.
182 Catherine M. Chin
independent trajectory of ‘classical scholarship’,47 but as the
ongoing product of the idea of division, necessary for the
equally ongoing articulation of a separate, biblically based
‘Christianity’.48 On such a reading, the dislocating and
recontextualizing practices of grammar materially create the
contexts of literary education, by motivating the scribal pro-
duction of bodies of texts that can represent what Augustine
calls ‘Israel’ and ‘Egypt’.
I would like to return, finally, to Anne Carson’s reading of
Sappho’s fragment 31. In the poem, the man who sits
between the speaker of the poem and the object of her desire
seems ‘equal to the gods’. In contrast, the speaker’s
altogether human body is both quickened and, as she says,
‘almost killed’ by desire.49 Augustine’s amorous triangle in
De Doctrina Christiana places grammar, and its simultaneous
evocations of paganism and Christianity, where they, in turn,
seem ‘equal to the gods’: reified, that is, and authoritative.
The desiring Christian reader, on the other hand, is both
quickened and immobilized by the simultaneous presence
and absence of the object of desire, God, Augustine’s
‘summa res’, whose presence and absence is mediated by
grammar and instantiated through Augustinian grammar’s
productions of Christianity and paganism. The words of De
Doctrina Christiana both suggest the ‘highest thing’ to the
reader and generate, through grammatical spoliation, the
‘godlike’ parallel and interdependent structures of Christi-
anity and paganism. The divine ‘res’ is, as Augustine says,
only available ‘per signa’. I suggest, then, that Augustine’s
famous vacillations between ‘Christianity’ and ‘paganism’ in
literary work, at least as such vacillations can be found in De
Doctrina Christiana, are less a reaction to some ontologically

47
As plotted, notably, by L. D. Reynolds and N. G. Wilson, Scribes and
Scholars: A Guide to the Transmission of Greek and Latin Literature
(Oxford, 1968, and later editions). On medieval uses of classical texts,
primarily for elementary education, see S. Reynolds, Medieval Reading:
Grammar, Rhetoric and the Classical Text (Cambridge, 1996).
48
Reynolds, Medieval Reading, 7–17, discusses the use of the ‘auctores’
as preparation for the reading of Scripture; on the parallel positions of
Virgil and Christian biblical epic in medieval grammars, see Irvine, Mak-
ing of Textual Culture, 364–71.
49
Quoted in Carson, Eros the Bittersweet, 12.
 The Grammarian’s Spoils 183
prior categories of ‘Christian’ and ‘pagan’ than they are the
effect of Augustine’s triangulating manœuvre in the text
itself. The Egypt from which the Christian grammarian
removes intellectual spoils, and the Israel to which the spoils
are removed, are equally conceptual products of the act of
spoliation, a late ancient speech habit in which De Doctrina
Christiana participates.
 8
Augustine’s Critique of Dialectic:
Between Ambrose and the Arians
St e fa n He ß b r Ü g g e n-Wa lt e r

1. i n t r o d u c t i o n
Since the publication of Henri Marrou’s seminal disserta-
tion on Augustine’s education and its connection to the
learned culture of antiquity,1 Augustine’s views on dialectic
have been the object of considerable scholarly attention,
culminating in Jean Pépin’s monograph on the topic in
1976.2 But beyond the indisputable fact that Augustine
was concerned with dialectical problems and had his own
opinions on the philosophical status of the discipline, com-
mentators disagree over how to present the evolution of his
position. Did his attitude towards dialectic change in the
course of his philosophical development? If so, when?
Marrou voices no opinion on the topic, but quotes texts from
all periods of Augustine’s thought indifferently.3 He thus
seems to accept, at least implicitly, that Augustine’s concep-
tion and evaluation of dialectic remained essentially the
same. Pépin believes that a change took place, but only at the
very end of Augustine’s career. According to him, late texts
like Contra Iulianum and the Opus Imperfectum contra
Iulianum indicate a significant shift in Augustine’s view of
dialectic, towards a more sceptical attitude.4 However, as
regards the period between De Ordine and De Doctrina
Christiana, otherwise so decisive for Augustine’s intellectual

1
H.-I. Marrou, Saint Augustin et al fin de la culture antique, 4th edn.
(Paris, 1958).
2
J. Pépin, Augustin et la dialectique (Villanova, Pa., 1976).
3
e.g., Marrou, Saint Augustin et la fin de la culture antique, 245–6.
4
Pépin, Augustin et la dialectique, 243, 254.
 Augustine’s Critique of Dialectic 185
development, he states explicitly that no major change took
place.5 Pépin’s interpretation has recently been challenged,
from both sides. The old thesis that there was no change at
all has been reasserted.6 At the same time, an attempt has
been made to locate the major change at a much earlier stage
of Augustine’s development, in his dialogue De Magistro.7
There is more at stake here than exegetical neatness or
the contour of an intellectual biography. The influence of
Augustine’s view of logic on medieval theology in the West
cannot be overestimated. His thought set the agenda on the
topic for centuries. One example: in order to defend the
application of dialectical procedures to the solution of theo-
logical problems, Peter Abelard quotes both De Ordine and
De Doctrina Christiana.8 The question of what Augustine’s
views on this matter actually were thus commands attention
beyond the circle of strictly ‘Augustinian’ scholarship.
What do we have to know in order to decide whether or
not Augustine was sympathetic towards dialectic and its
use for the solution of theological problems? First of all, we
must sort out the different kinds of evidence for his view of
the discipline. If changes appear in his position over time, we
must then ask how and why they came about. As such
research proceeds, it becomes clear that Augustine was in
fact addressing problems of fundamental philosophical
importance and that he gives us at least the sketch of a
basically sound and reasonable solution. Besides the
question of how Augustine’s outlook may have developed
over the years, which has until now been the main focus of
the debate, other, more interesting questions come to the
fore. Why did Augustine change his view? What led him to
focus on dialectic itself as a topic of philosophical reflection?
Why was it so important for him to formulate an approach to
this particular discipline?
For clarity’s sake, we begin by summarizing our con-
clusions.
5
Ibid. 169.
6
J. Lienhard, ‘Augustine on Dialectic: Defender and Defensive’,
StPatr 33 (1997), 162–6.
7
J. Brachtendorf, ‘The Decline of Dialectic in Augustine’s Early Dia-
logues’, StPatr 37 (2001), 25–30.
8
Abelard, Theologia Summi Boni, ed. Ostlender-Nägeler, 66 ff.
186 Stefan Heßbrüggen-Walter
First, the reality of change. Although the wording of
Augustine’s definitions of dialectic varies only slightly over
the years,9 we are not justified in assuming that successive
definitions continue to designate the same thing. Augustine
does indeed change his view of dialectic, and not only with
regard to its value in a Christian education; in the period
between De Ordine and De Doctrina Christiana he gains vital
new insights, developing a completely new vision of what the
discipline can and cannot achieve, and of why it is worth the
trouble to acquire some knowledge of it. Second, a motive
for change: there is enough circumstantial evidence to
make it probable that the alteration in Augustine’s position
was stimulated not only by disenchantment with his own
earlier views, but also by a desire to combat the ‘heresy’ of
Arianism. The most influential foes of dialectic in his
immediate environment during the period in question were
those—notably Ambrose, bishop of Milan—who, in their
arguments against the Arians, made a point of objecting
that the latter applied dialectical laws and procedures to the
solution of doctrinal questions in Trinitarian theology,
instead of relying on the ‘simple’ truths of faith. Is it not
likely that Augustine was prompted by this diagnosis of a
heretical abuse of dialectic to re-examine his own relation
to the discipline? Third, and finally, an intellectual-
biographical datum: the development of Augustine’s thought
from the Cassiciacum dialogues towards De Doctrina
Christiana entails, among other changes, an important
clarification and enhancement of his view of dialectic, its
scope, function, and knowledge claims.

2. d i a l e c t i c a n d t h e a s c e n t to t h e o n e :
de ordine
Although it is commonly held that after his conversion
Augustine was as much a Neoplatonist as a Christian, the
precise relationship of these two aspects of his thought
remains difficult to grasp. One central tenet of his Neo-
platonism must, however, be introduced here, in order for it

9
Pépin, Augustin et la dialectique, 161.
 Augustine’s Critique of Dialectic 187
to be possible for us to appreciate the role played by dialectic
in this context. It concerns the relationship between human
beings and God. For the Neoplatonist that Augustine was,
our souls are nothing but isolated spiritual substances that
have been separated from the divine All-soul. They long for
a reunion with this ground of their existence. To facilitate
a return to the divine, the soul must turn inwards and find
within itself evidence of former communion with the
All-soul.10 In the second book of his De Ordine, Augustine
gives an account of how this can be accomplished. It is the
seven liberal arts that are supposed to turn the soul towards
its rational essence and make it aware of its capabilities.
Mastering the skills involved in the practice of the liberal
arts thus becomes the essential pre-condition for achieving
the desired metaphysical knowledge.
In the beginning of his exposition in the second book
of De Ordine, Augustine introduces a discipline capable of
delivering knowledge of the divine law that governs the
world and is ‘written in our souls’.11 At 2. 18. 47 this dis-
cipline is called by its proper name, philosophy. Philosophy
enables us to obtain knowledge of the divine law, but such
knowledge cannot be gained simply by reading texts or lis-
tening to a teacher. These and similar activities (comprised
by the term ‘eruditio’, education) must be accompanied by
a life of philosophy.12 And although it is possible in principle

10
e.g., Plotinus, Enneads 5. 1. 2. R. J. O’Connell, St Augustine’s Early
Theory of Man (Cambridge, Mass., 1968), 114, regards this account as an
important source for Augustine.
11
Ord. 2. 8. 25: ‘Haec autem disciplina ipsa dei lex est, quae apud eum
fixa et inconcussa semper manens in sapientes animas quasi transcribitur,
ut tanto se sciant vivere melius tantoque sublimius, quanto et perfectius
eam contemplantur intellegendo et vivendo custodiunt diligentius’ (‘This
discipline is the very law of God, which, while forever remaining fixed and
unshaken with him, is copied (as it were) into the souls of the wise, so that
they may know that their capacity to live better and more excellently
depends on their contemplating this law [or discipline] more perfectly in
their minds and observing it more conscientiously in their lives’).
12
Ord. 2. 18. 47: ‘Haec igitur disciplina eis, qui illam nosse desiderant,
simul geminum ordinem sequi iubet, cuius una pars vitae, altera
eruditionis est’ (‘Therefore this discipline commands those who desire
to know it to follow a double order, one part of which is life, the other
education’).
188 Stefan Heßbrüggen-Walter
to lead a good life by following the authority of faith, the
autonomous philosophical way of life is to be preferred:
There are two necessary ways to learning: authority and reason.
Authority comes first as far as the sequence [of learning] is con-
cerned. Reason comes first as far as the thing itself [i.e., learning] is
concerned. There is a difference between what is put first in the
sequence of our actions and what is rated more desirable.13
All of us learn some of the knowledge that is necessary for
leading a good life by following authorities, but we should
try to gain knowledge by use of reason alone. How then does
the use of reason prepare us for a philosophically inspired
life? As Augustine explains in De Ordine 2. 11. 30, reason is
our ability to discern things and to unite them, presumably
in a concept or judgement. And since the main obstacle to
knowing God and his laws is our infatuation with the senses
and the merely apparent knowledge they deliver, we must
return to ourselves and discern our souls as the divine part of
our nature.
Reason is a movement of our mind which is able to distinguish and
connect the objects of learning. It is extremely rare that human
beings are able to use reason as a guide either to understand God
or the soul itself, be it in ourselves or in others. This is because
those who have moved beyond their soul and busied themselves
with the affairs of the senses must come back to themselves, which
is difficult.14
What makes this theory of self-knowledge so interesting is
that the knowledge of ourselves as souls is not achieved
through mystical insight or contemplation. Instead, we find
this knowledge through an analysis of the products of the
human mind. They themselves are ordered, or, as Augustine
puts it, ‘rationabile’ (2. 11. 31). Actions in general, as well as

13
Ord. 2. 9. 26: ‘Ad discendum item necessario dupliciter ducimur,
auctoritate atque ratione. Tempore auctoritas, re autem ratio prior est.
Aliud est enim, quod in agendo anteponitur, aliud, quod pluris in
appetendo aestimatur.’
14
Ord. 2. 11. 30: ‘Ratio est mentis motio ea, quae discuntur, dis-
tinguendi et conectendi potens, qua duce uti ad deum intellegendum vel
ipsam quae aut in nobis aut usque quaque est animam rarissimum omnino
genus hominum potest non ob aliud, nisi qui in istorum sensuum negotia
progresso redire in semet ipsum cuique difficile est.’
 Augustine’s Critique of Dialectic 189
speech acts in particular, are to be regarded as ‘rationabilia’.
By analysing speech, dialectic helps us to turn the soul away
from the realm of sense towards self-reflection: in its
products, the soul discerns the ability it has to impose order
on the world. Hence dialectic is the pre-condition for all
further acquisition of knowledge. And there is a second
dimension: as soon as the soul knows itself as the lawgiver of
nature, it becomes aware of its own rational nature.
There are three potential difficulties with such a view.15
First, it seems to conflict with the originally theurgic and
mystical ambitions of Neoplatonism. If the return of the
human soul hinges merely on the acquisition of certain
knowledge about the phenomenal world, that return
becomes a quasi-mechanical process devoid of any spiritual
meaning. However, since this is a problem only within
the framework of Neoplatonism and is irrelevant for a dis-
cussion of the role of dialectic in Augustine’s later thought,
it will not be pursued further here.
The second difficulty cannot be so easily dismissed by a
Christian such as Augustine had become. How was the
thesis of the superiority of the philosophical way of life over
trust in the authority of revelation to be reconciled with a
Christian faith? As we shall see, this issue is raised by
Ambrose when he criticizes the application of dialectic to
theological problems. Later, Augustine himself would advert
to it in his Retractationes, as he reflected on his early work. ‘I
am displeased with these books,’ he would write, ‘because
I attributed much to the liberal arts, which were largely
unknown to many of the saints.’16 It cannot be a necessary

15
O’Connell, St Augustine’s Early Theory of Man, 227, describes it as
‘intellectualism . . . exacerbated by Varro’s mathematical semi-rational-
ism’. I. Hadot, Arts libéraux et philosophie dans la pensée antique (Paris,
1984), 156 f., argues that the theory of the disciplines in De Ordine was not
influenced by Varro’s Disciplinarum Libri and regards De Ordine as the
starting-point for the early medieval idea of the seven liberal arts as a
propaedeutic to philosophy. In contrast to O’Connell, she sees De Ordine
as the culmination of a historical development starting in Middle Platon-
ism. (For Augustine’s debt to Varro, see now Shanzer’s essay in this
volume.)
16
Retr. 1. 3: ‘Verum et his libris displicet mihi . . . quod multum tribui
liberalibus disciplinis, quas multi sancti multum nesciunt.’
190 Stefan Heßbrüggen-Walter
pre-condition of Christian salvation to know the liberal arts,
since the Apostles and other saints were not well versed in
them.17
The third difficulty is central to the argument of the
present essay. Augustine conceives of dialectic in De Ordine
as a science capable of generating knowledge that is not
limited to the formal aspects of argument but includes what
may be called ‘ontological’ or ‘material’ truths. As we shall
see, it is precisely this understanding of the power of the
discipline that was apparently taken for granted by the
Arians of Augustine’s day, and fiercely contested by their
‘Catholic’ critics. Dialectic ‘knows the ways of knowing’,
Augustine writes: ‘It alone does not merely want to make us
knowledgeable, but is able to do so as well.’18 Dialectic
achieves this end by two means. First, it directs the soul
towards its nature as spiritual substance. Secondly, it serves
as the foundation of all other sciences, helping us to discern
the structure of the phenomenal world; that is why it can be
called the discipline of disciplines (‘disciplina discipli-
narum’).19 This position, too, is fully in accord with a Neo-
platonist conception of the discipline. Dialectic enables
us to gain insights into the nature of our soul and into the
structure of the outer world. In this sense, it can be said to
generate ontological or material truths.
In what follows, it will be shown that both questions raised
(for a Christian) by the philosophy of dialectic found in De
Ordine—of the relation between dialectic and truths of the

17
Cf. O’Connell, St Augustine’s Early Theory of Man, 227: ‘In the case
of one who is “ignorant” but capable of the learning embodied in the
disciplines, his theory of intellectualist salvation has relatively clear
sailing. But how is Augustine to understand the fact that not all Christians
are apt candidates for such a course of studies, equipped for such an
intellectualist way of “return”?’
18
Ord. 2. 13. 38: ‘Scit scire, sola scientes facere non solum vult sed
etiam potest.’
19
Although Pépin, Augustin et la dialectique, 190, is right to say that
Augustine nowhere explicitly sides with either party in the ancient dispute
as to whether dialectic is an instrument or a part of philosophy, it is
clear that his Neoplatonic conception of the discipline led him in early
years into the Platonist camp that declared dialectic to be a part of
philosophy.
 Augustine’s Critique of Dialectic 191
faith, and of whether dialectic can indeed generate material
truths—had considerable resonance in milieux with which
Augustine was connected. The first question we shall find
recurring in Ambrose’s critique of dialectical method, the
second in the Trinitarian argumentation of two Arian (more
strictly, Anomian) theologians, Aetius and Eunomius.

3. a m b r o s e aga i n st t h e d i a l e c t i c i a n s
Ambrose is not the only church father to express misgivings
about the role of dialectic in theology, but he is of particular
importance to us inasmuch as we can assume that Augustine
was largely acquainted with his thought.20 The relevant
arguments can be found in two texts of Ambrose: namely,
the ‘explanatio’ of Psalm 36 and his De Fide.
In his ‘explanatio’ of Psalm 36, Ambrose uses exegesis of
verse 16 (‘Melius est parum iusto super divitias peccatorum’)
for a polemic against the reputed resources of pagan philo-
sophers. It is better, he says, to be poorly endowed with the
intellectual goods of ancient tradition than to immerse one-
self in sacrilegious discourse that contradicts the revealed
truth of the Scriptures. A person may be ‘rich in words as
the worldly philosophers are’, he writes, ‘when [the latter]
dispute about sacrilegious practices, the movement of the
planets, the star of Jove or Saturn, about the generation of
man, the cult of artificial images (simulacra), or about geom-
etry and dialectic’.21 While it may be no surprise to find that
Ambrose does not regard inquiries into the generation of
human beings or the motion of planets named after Roman
gods as worthy of reflection, his disdain of dialectic (and
geometry) would seem to need some explaining.

20
For new arguments for Augustine’s close engagement with
Ambrose’s thought and writing from an early stage in his Christian
intellectual development, see M. Dulaey, ‘L’Apprentissage de l’exégèse
biblique par Augustin (1): Années 386–389’, RE Aug 48 (2002), 267–95,
and ‘(2): Années 390–392’, RE Aug 49 (2003), 43–84.
21
Ambrose, Explanatio Psalmorum 36: 16: ‘est . . . dives in verbis ut
sunt philosophi istius mundi de sacrilegiis disputantes, de motu siderum,
de stella Iovis ac Saturni, de generationibus hominum, de simulacrorum
cultu, de geometria et dialectica’.
192 Stefan Heßbrüggen-Walter
Hints for an explanation are to be found in De Fide, in a
chapter dedicated to the Arian heresy. Arius had denied the
consubstantiality of the three persons of the Trinity, holding
that the Son had been created by the Father before the
beginning of time in order to be the Father’s agent in the
creation of the world. Ambrose states that according to
the Arians, Father and Son are completely dissimilar, and
cites a radical strain of Arian thought promoted by Aetius
and Eunomius. He then speculates on the reasons for this
marked deviation from what he takes to be a constitutive part
of Christian revelation. He writes:
Now since the heretic says that [God’s nature] is dissimilar, and
because he tries to argue for that in cunning disputations (versutis
disputationibus), it is up to us to state what has been written: ‘See to
it that no one makes a prey of you by philosophy and empty deceit,
according to human tradition, according to the elements of this
world, and not according to Christ. For in him the whole fullness
of the body dwells bodily.’22
The quotation from Colossians 2: 8–9 at the end of the pas-
sage shows that Ambrose believes the Arians to be bewitched
by philosophy. They do not follow the simple truths of the
Scriptures. They believe that they are able to recognize God
rationally, and they develop elaborate arguments in support
of their views, even though St Paul himself had warned
the Church against such presumptiousness. The tool for
their cunning disputations (‘versutae disputationes’) is
dialectic:
All the power of their poisons they repose in dialectical disputa-
tion, which according to the philosophers’ view is defined as
having not the power of asserting but the desire of destroying
(non astruendi vim habere, sed studium destruendi). But it has not

22
Ambrose, De Fide 5. 41: ‘Nunc quoniam haereticus dicit esse dis-
similem, idque versutis disputationibus astruere nititur, dicendum est
nobis quod scriptum est: “Cavete ne quis vos depraedetur per philos-
ophiam, et inanem seductionem secundum traditionem hominum, et
secundum elementa huius mundi, et non secundum Christum: quia in
ipso habitat omnis plenitudo divinitatis corporaliter.”’ Translation of
biblical quotation adapted from NRSV.
 Augustine’s Critique of Dialectic 193
pleased God to save his people by dialectic: ‘The kingdom
of God is in simplicity of faith, not in verbal controversy (1 Cor.
4: 20)’23
According to Ambrose, dialectic—certainly as used by the
Arians—was the ruination of true Christian faith.
How far was this opinion of Arian reasoning justified?
Two main sources are available to us: the Apologia written
by Eunomius and the Syntagmation by his teacher Aetius.24

4. dialectic and the knowledge of god:

eunomius and aetius
The Anomians (from anomoios, ‘dissimilar’) denied any
likeness between Father and Son on the grounds that it is
constitutive of the Father that he be uncreated (agennētos).25
If Father and Son are similar in substance, the Son must
be uncreated as well. This cannot be true, however, because
it endangers God’s singularity; there cannot be two uncre-
ated deities. Consequently, Father and Son cannot be
identical (homoousios) or even similar (homoios) in substance,
because only an uncreated Deity could be similar to the
Father.
For an understanding of the role that the Anomians
may have played in the development of Augustine’s view
on dialectic, two aspects of their thought are particularly
relevant: their theory of language and their adherence to

23
De Fide 5. 42: ‘Omnem enim vim venenorum suorum in dialectica
disputatione constituunt, quae philosophorum sententia definitur non
astruendi vim habere, sed studium destruendi. Sed non in dialectica
complacuit Deo salvum facere populum suum: “Regnum enim Dei in
simplicitate fidei est, non in contentione sermonis.”’
24
The Syntagmation has survived in two recensions, the first contained
in the Panarion of Epiphanius, the second in the Pseudo-Athanasian
Dialogues on the Trinity. I follow the edition and translation provided by
L. R. Wickham, ‘The Syntagmation of Aetius the Anomean’, JTS 19
(1968), 532–69. Eunomius’ Apologia is extant, despite an order given by
the emperor Arcadius in 398 to destroy all books of the Anomians. The
work was probably composed c.360.
25
Wickham, ‘Syntagmation of Aetius’, 537 f.
194 Stefan Heßbrüggen-Walter
Stoic propositional logic.26 The analysis will show that they
agree with Augustine’s De Ordine. Dialectic is a science
capable of producing true insights into the structure of
being, to be derived by application of the rules of Stoic
propositional logic.
The Anomian theory of language rests on what may be
called the principle of ‘meaning–essence identity’: a differ-
ence in the concepts signified by particular terms is in itself
always sufficient justification for assuming that the objects
designated by those terms are substantially different as well.
In everyday use, the terms ‘father’ and ‘son’ signify diverse
concepts or, in Stoic language theory, sēmainomena. This fact
has implications for Trinitarian theology. Eunomius writes:
We state that the Holy Scripture teaches us to believe that the Son
is an offspring, because we do not believe that essence (ousia) is one
thing and the signified concept (sēmainomenon) is another thing
distinct from it: it [essence] is rather the substance (hypostasis)
signified by the name, because the name truly expresses essence.27
The principle of ‘meaning–essence identity’ is fundamental
to the Anomian argument against the orthodox view of the
Trinity. The assumption that the name ‘truly expresses
substance’ leads to the consequence that we gain a true
understanding of the Holy Trinity only if we analyse our
concept of God. That is to say, a complete understanding of
God’s essence is something graspable by his creatures.28

26
The application of dialectical tools in theology did not go undisputed
among Arian theologians. The Anomians were criticized for it not only by
the orthodox but also by supporters of the Arian creed, like Philostorgius:
J. de Ghellinck, ‘Un Aspect de l’opposition entre hellénisme et christian-
isme: L’attitude vis-à-vis de la dialectique dans les débats trinitaires’,
in his Patristique et Moyen Age, iii (Brussels and Paris, 1948), 247–310,
at 270.
27
Eunomius, Apologia 12. 8–12, from the edition by B. Sesboü é et al.,
SC 305.234–99, my translation. Wickham, ‘Syntagmation of Aetius’, 540,
also notes the acceptance of ‘meaning–essence identity’ and links it to a
theory of the divine origin of language. See also J. Daniélou, ‘Eunome
l’Arien et l’exégèse néo-platonicienne du Cratyle’, Revue des Études
Grecques 69 (1956), 412–32.
28
Cf. de Ghellinck, ‘Un Aspect de l’opposition’, 274. Sesboü é et al.,
Basile de Césarée: Contre Eunome, SC 299.16, see in this the culmination
of a whole ancient Greek philosophical tradition.
 Augustine’s Critique of Dialectic 195
Eunomius’ teacher Aetius accepts this principle as well,
as is apparent from section 16 of the Syntagmation, a com-
pendium of thirty-seven concise arguments presented in a
letter to his followers:29
If ingeneracy is revelatory of essence (ousia), it is reasonable for it
to be contrasted with the offspring’s essence. If ‘ingeneracy’ has no
meaning, a fortiori ‘offspring’ reveals nothing. How could non-
entities be contrasted? If, again, the word ‘ingenerate’ is contrasted
with the word ‘generate’, silence following the utterance of the
words, the Christian hope turns out to begin and stop; it is based
on magnificent language but not on what the natures really are,
which is the intended meaning of the names.30
For Aetius, if something ‘has a meaning’ (sēmainei), it is
‘revelatory’ (dēlotikon) of the essence of the signified: a word
is meaningful only if it reveals the essence of its referent. A
denial of this principle would render the contrast between
being ingenerate and being generate meaningless, and would
thus destroy faith—make ‘the Christian hope stop’. Con-
sequently, the Nicene party, which denies the diversity of
substances, cannot be using the respective terms in a
meaningful way. For the words they use to have any mean-
ing, it is necessary to accept the diversity of substances cen-
tral to the Anomian view.
The second feature of dialectical theory prominent in
Anomian argumentation is their application of Stoic propo-
sitional logic. We only can grasp God’s essence by means of a
conceptual analysis if we rely on correct inferential pro-
cedures in order to derive the relevant arguments. An
example from Aetius’ Syntagmation:
If the ingenerate nature is not patient of origination, it is what it is
called, but if it is patient of origination, the affections (pathē) of
origination will be superior to the substance of the Deity.31
Wickham has noted that such arguments cannot be under-
stood as syllogisms in the Aristotelian sense, and classifies

29
For textual history and historical background, see Wickham,
‘Syntagmation of Aetius’, 532–5.
30
Ibid. 546.
31
Ibid.
196 Stefan Heßbrüggen-Walter
them loosely as deductive arguments.32 But in fact the
example can quite easily be analysed as a slightly disguised
Stoic hypothetical syllogism using modus ponens (if p then
q; p; q) and modus tollens (if p then q; not q; not p). If one
adds to the quoted argument one more premiss that was
probably taken for granted by every participant in the dis-
cussion—namely that ‘no affection of the Deity is superior
to it’—then the proof intended by Aetius can be
reconstructed as follows:
1. If the ingenerate nature of the Deity is patient of origi-
nation, its affections are superior to the Deity [premiss]
2. No affections of the Deity are superior to it [premiss]
3. The ingenerate nature of the Deity is not patient of
origination [follows from (1) and (2) by applying modus
tollens]
4. If the nature of the Deity is not patient of origination, it
is ingenerate [premiss]
5. The nature of the Deity is ingenerate [follows from (3)
and (4) by applying modus ponens]
The rules applied here are fundamental axioms of Stoic
logic, used in order to infer theological insights from
self-evident premisses.33
The diagnosis put forward by Ambrose thus appears to
be correct. The Arians endeavour to prove their points by
dialectical argumentation, applying a theory of the relation-
ship between language and the essence of entities, and the
inferential rules of Stoic logic. They believe that it is possible
to generate material truths concerning the ontological struc-
ture of the Trinity by means of unaided reason.
Having offered this account of a possible major context for
Augustine’s thought, we are in a position to reconsider his
presentation of dialectic in De Doctrina Christiana. The view
that he gives of that discipline is more nuanced than that of
either of the other parties, Ambrose or the Anomians.

32
Wickham, ‘Syntagmation of Aetius’, 534. But cf. de Ghellinck, ‘Un
Aspect de l’opposition’, 278 ff.
33
The principles governing these inferences can be traced back to
Chrysippus: M. Frede, Die stoische Logik (Göttingen, 1974), 131. They
were also known to Cicero, as shown by his Topica 54–7.
 Augustine’s Critique of Dialectic 197
Let us now see how Augustine applies both parts of dia-
lectic, the science of definition and the science of inference,
to theological problems in exegesis and systematic theology.

5. dialectic as a formal discipline:

d e d o c t r i na c h r i st i a na
As far as dialectic is concerned, Augustine in Book 2 of De
Doctrina Christiana may be seen pursuing two related aims.
He evidently wants to circumvent the kind of critique that
Ambrose made of the discipline, while at the same time
curbing the pretensions of Arians and others—including
perhaps his own former self—who asserted that dialectic was
a science capable of producing material insights into the
structure of God and his creation.
Before entering into detailed analysis, we should remind
ourselves of the place of dialectic in the system of the dis-
ciplines sketched in De Doctrina Christiana. According to
Augustine, arts and sciences as they can be found in the
‘pagan’ heritage of Late Antiquity fall into two classes: ‘One
consists of things which have been instituted by human
beings, the other consists of things already developed, or
divinely instituted, which have been observed by them’
(2. 19. 29).34 He makes it clear that dialectic belongs in
the second class, consisting of disciplines which may
be appropriated by Christian thinkers without hesitation,
albeit with due precautions. The laws of dialectic are part of
nature, broadly considered, and have been instituted in the
same way as the laws of any other ‘natural’ science:
The validity of syllogisms (veritas conexionum) is not something
instituted by humans, but observed and recorded by them, so that
the subject may be taught or learnt. It is built into the permanent
and divinely instituted system of things (est in rerum ratione
perpetua et divinitus instituta). The historian does not himself
produce the sequence of events which he narrates, and the writer
on topography or zoology or roots or stones does not present things
instituted by humans, and the astronomer who points out the

34
English translation of De Doctrina Christiana by R. P. H. Green,
Saint Augustine: On Christian Teaching (Oxford, 1997), occasionally
slightly modified.
198 Stefan Heßbrüggen-Walter
heavenly bodies and their movements does not point out some-
thing instituted by himself or any other person; likewise the logi-
cian who says that since the consequent is false, the antecedent
must be false, may be saying something perfectly true, but does not
make it true, for he only points out the truth of it. (2. 32. 50)
Logical laws are no mere figment of human imagination.
They are true, and the world conforms to them. (This point
is central to the rebuttal of criticism like Ambrose’s.) They
exist and can be discerned and described in the same way as
entities in other spheres of the created world. Hence any-
body who applies dialectic to particular problems can find
truth, provided two conditions are fulfilled: that person
applies the laws of the science correctly, and the premisses
from which he or she starts are true. Logic thus becomes a
powerful tool for the understanding of Scripture.35 Things
become more complicated, however, if the dialectician,
wishing to deceive an adversary, uses an invalid dialectical
rule in order to construct a sophism, or if the relevant
premisses are not true. ‘One must beware’, Augustine writes,
‘of indulging a passion for wrangling and making a puerile
show of skill in trapping an opponent’ (2. 31. 48). Yet, com-
pared to what we find in Ambrose, the critique offered
here of dialectic is very mild. It concerns only the abuse of
dialectic, not the science itself. And, according to Augustine,
even the falsehood of dialecticians who contrive to deceive
their adversaries is only childish, not ‘venomous’ as Ambrose
would have it.
A paradigmatic instance of such abuse of dialectic is the
sophism: a formally invalid inference sufficiently similar to
correct applications of dialectical rules as to convey the
impression of being valid itself. Augustine gives a fairly
simple example. First premiss: ‘You are not what I am.’
Granted. Next: ‘I am a human being.’ Conclusion: ‘You are
not a human being’ (ibid.). As Pépin has pointed out, this

35
Doct. Chr. 2. 31. 48: ‘disputationis disciplina ad omnia genera
quaestionum, quae in litteris sanctis sunt, penetranda et dissolvenda,
plurimum valet’ (‘The discipline of disputation [i.e. dialectic] is extremely
useful for investigating and resolving all kinds of questions in the holy
writings’).
 Augustine’s Critique of Dialectic 199
36
sophism is part of the Stoic tradition. According to the
testimony of Simplicius,37 in order to solve the sophism,
dialecticians usually noted the ambivalence of the term
‘human being’, used to designate a universal as well as an
individual.38
A second cause of dialectical error is the derivation of
wrong conclusions from wrong premisses. For the modern
reader, this is common knowledge: logic does not create
truth, it only guarantees conservation of truth in inferences.
But those in Augustine’s time who were interested in the
speculative potential of dialectic, like the Arians, or hostile to
the discipline, like Ambrose, apparently had to be reminded
that dialectic was in fact a purely formal science. The
adversaries of dialectic needed to be aware that one may use
false sentences as antecedent or consequent in hypothetical
syllogisms without claiming them as true. Thus the true
inferential connection of propositions that themselves may
be regarded as false can still lead to valid—and thus true—
results. Augustine writes:
There are also such things as valid logical syllogisms based on
false statements (verae conexiones ratiocinationis falsas habentes
sententias), which attack a mistake made by an opponent. But
these are advanced by honest and clever people to embarrass the
person whom they are seeking to attack and make him abandon his
misconception, by showing that if he chooses to stick to it he is
logically compelled to uphold what he condemns. (2. 31. 49)

If the adversary of the dialectician holds false opinions,

one way to convince him of that is to derive apparently
absurd conclusions from them, forcing him to accept that the
premisses are equally wrong. To make this point, Augustine
uses the authority of Scripture: Paul himself uses the
procedure in 1 Corinthians 15. In Augustine’s reading, the

36
Pépin, Augustin et la dialectique, 213, following Diogenes Laertius in
attributing the sophism to Chrysippus; K. Hülser, Die Fragmente zur
Dialektik der Stoiker, iv (Stuttgart and Bad Cannstatt, 1988), 1762 f., is
sceptical.
37
In Aristotle, Categories 105. 7–20 (ed. Hülser, 1247).
38
T. G. Bucher, ‘Zur formalen Logik bei Augustinus’, Freiburger
Zeitschrift für Philosophie und Theologie 29 (1982), 3–45, at 12 f., is
unaware of the Stoic roots of the sophism and its ancient analysis.
200 Stefan Heßbrüggen-Walter
passage is meant to refute those who deny the resurrection of
the dead. ‘The apostle Paul was not advancing true state-
ments when he said, “neither did Christ rise”, and “our
preaching is in vain”, and “your faith is in vain”, and then
other things, which are completely false’ (ibid.). This is the
false thesis that has to be refuted by dialectical means: those
assuming the impossibility of resurrection must also state
that Christ has not risen. Their argument would run along
the following lines, applying modus ponens: ‘If there is no
resurrection of the dead, Christ has not been resurrected.
There is no resurrection’ (which is the way things are put in
1 Cor. 15). But since, according to Paul at least, Christ’s
resurrection can be regarded as a historical fact, it is possible
to apply modus tollens: ‘If there is no resurrection of the
dead, Christ has not been resurrected. Christ has been resur-
rected. There is resurrection of the dead.’ We thus arrive at
the inference that the Christian faith is after all justified.
Both the antecedent and the consequent of the hypothetical
proposition in 1 Corinthians 15 are wrong, but it is none the
less possible to infer a truth from them. Augustine not only
renders the argument, but also gives a logical analysis. He
states the explicit premiss of the adversaries of Christianity
who are the target of Paul’s argumentation, then shows—
though without using the term itself—that in the given
example Paul applies modus tollens.
In the example just given, there obviously comes a
moment when it is necessary to appeal to a truth that can
only be known from revelation and the sacred texts, namely
the ‘fact’ that Christ has risen; no dialectician is in a position
to say whether or not that is true. However, at almost the
same moment in Augustine’s argument another ‘fact’
emerges, or is confirmed: namely, that the rules connecting
truths (or falsehoods) in such a way that their inferential
connections become obvious are not a matter of faith or
revelation; they can be taught in the philosophical schools.39
Augustine concludes: ‘There are, then, valid syllogisms
based not only on true propositions but also on false ones;
it is easy to learn which of them are valid even in schools

39
A point already emphasized by Bucher, ‘Zur formalen Logik bei
Augustinus’, 34.
 Augustine’s Critique of Dialectic 201
outside the church (in scholis illis . . . quae praeter ecclesiam
sunt). But the truth of propositions must be sought in the
church’s holy books (in sanctis libris ecclesiasticis)’ (2. 32. 49).
This is an important insight, which would take the bite out
of Ambrose’s criticism of dialectic. Of course, dialectic is an
art common to both pagans and Christians. But this is not a
reason to despise it. Dialectic is concerned only with the
formal derivation of valid syllogisms, not the material truth
of propositions, which in matters of the faith (according to
Augustine) can only be ascertained by appeal to the revealed
truth of Scripture.
So far in this section we have dealt only with the part
of dialectic that gives rules of valid inference. There is,
however, the second important subject area: definition and
division.40 In this regard too, Augustine stresses that rules
for the correct formation of definitions can be derived
rationally. As an example of them, he reminds his readers
that only those traits that really belong to the thing in ques-
tion should be accepted in a definition. What is valid in this
part of the science is again purely formal, and can even help
to deal with falsehoods; wrong definitions may be grouped
together according to whether they are logically possible or
not. Here is the salient passage from Book 2:
The knowledge of definition, division, and classification (scientia
definiendi, dividendi atque partiendi), though often applied to false
things, is not in itself false; and it was not instituted by man, but
discovered as part of the way things are (in rerum ratione comperta).
For just because it is often applied by poets to their fables and by
false philosophers or heretics (in other words, false Christians)
to the tenets of their misguided systems, that does not make it
wrong to say that in defining or dividing or classifying something
you must not include something irrelevant or leave out something
that is relevant. This is true, even if the things being defined or
classified are not true. (2. 35. 53)
Why is this discussion of definition included in De
Doctrina Christiana? It is true, of course, that the question

40
For an overview of the relevant Stoic positions, see A. A. Long and
D. N. Sedley, The Hellenistic Philosophers, i (Cambridge, 1987), 190–5,
and for fuller documentation, Hülser, Fragmente zur Dialektik der Stoiker,
iv. 714–27 (frr. 621–31).
202 Stefan Heßbrüggen-Walter
of how to define a concept had been part of dialectic from
Plato’s time onwards, and had been integrated into Peri-
patetic and Stoic logic.41 But a more interesting motive than
simple compliance with disciplinary tradition is suggested
by the reference to ‘false philosophers or heretics (in
other words, false Christians)’ in the passage just quoted.
Augustine is no more historically specific here than he is
elsewhere in the treatise. Would it be too far-fetched,
however, to think that he alludes to the Arians, among
others? The principle of ‘meaning–essence identity’ is of
paramount importance for their theological argumentation,
and it certainly belongs to the science of definition. We saw
earlier that Augustine had himself once taken an extremely
ambitious view of the value of dialectic for Christian
philosophy, similar in many respects to that of the Arians.
The claims made for the discipline in De Doctrina Christiana
are much more modest, and clearly subordinate to those
made for the authority of Scripture and the traditional
teachings of the Church. May we assume that the change
in Augustine’s position was to some extent a result of his
reflection on the heretical abuse of dialectical methods?
Corroboration for this inference may be sought in a work
from a later stage of Augustine’s philosophical development,
his De Trinitate. In this work, rather than straightforwardly
refute the principle of ‘meaning–essence identity’, he
chooses instead to demonstrate that its application by the
Arians is misguided. To that end, he applies rules from
the science of definition which he derives from Aristotle’s
Categories, thus meeting the Arians on their own ground.42
Augustine summarizes the Arian argument as follows:

41
Pépin, Augustin et la dialectique, 171 ff.
42
Augustine criticizes Arian theology in Books 5–6 of Trin. M.
Schmaus, Die psychologische Trinitätslehre des Hl. Augustinus (Münster,
1927), 143, and A. Schindler, Wort und Analogie in Augustins Trinitätslehre
(Tübingen, 1965), 151 ff., among others, hold that the primary
target of this critique was Eunomius, an assumption recently questioned
by M. R. Barnes, ‘The Arians of Book V, and the Genre of De Trinitate’,
JTS 44 (1993), 185–95. None of the passages in Trin. 5 have exact
parallels in extant Arian texts, so certainty in this matter is impossible.
However, one problem that Barnes raises for the traditional view can
 Augustine’s Critique of Dialectic 203
The Father is called Father in relation to the Son, and the Son is
called Son in relation to the Father, but ‘unbegotten’ is said in
relation to himself, and ‘begotten’ in relation to himself; and there-
fore, if whatever is said in relation to himself is said according to
substance, while to be unbegotten and to be begotten are different,
then the substance is different.43
To paraphrase: ‘pater’ as well as ‘filius’ is a relative term;
the contrast between ‘ingenitus’ and ‘genitus’ is predicated
of substance (‘ad se ipsum dicitur’). This predication of
substance is in itself a sufficient criterion for substantial
diversity of Father and Son. The first step in Augustine’s
counter-argument consists in an analysis of what it means to
be ‘genitus’. ‘Genitus’ is a relative concept par excellence.
Sons are begotten because they are sons—and they are
sons, because they are begotten. If ‘son’ is a relative term,
‘begotten’ is relative as well; one can always ask ‘begotten
by whom?’44 Now, an Arian theologian might concede that
‘genitus’ is a relative concept. But he would deny that the
same is true of ‘ingenitus’ and assume that it describes a
non-relative property. At this point, Augustine introduces a
rule that is not explicitly contained in Aristotle’s Categories.
Negation of a given term does not change the category it
belongs to. This rule is ‘proven’ in a peculiar way. Augustine
introduces examples from all ten categories, negates them,
and shows that the negation does not change the category of

be answered by arguments presented here. He suggests (p. 192) that if

Eunomius had been a target, Augustine would have been forced to reflect
in Trin. on his own use of tools similar to those deployed by his heretical
adversary. But, as we have seen, the problem of how to apply dialectic
in theological matters had already been addressed in Doct. Chr. Having
suggested there that the tools of dialectic had been misused by heretics,
why should Augustine hesitate to use Aristotle’s Categories correctly in
order to deal with the results of such misuse?
43
Trin. 5. 6. 7: ‘Pater ad filium dicitur et filius ad patrem; ingenitus
autem ad se ipsum et genitus ad se ipsum dicitur. Et ideo si quidquid
ad se ipsum dicitur secundum substantiam dicitur; diversus est autem
ingenitum esse et genitum esse; diversa igitur substantia est.’ Trans.
A. W. Haddon (modified).
44
Ibid.: ‘Ideo quippe filius quia genitus et quia filius utique genitus.
Sicut autem filius ad patrem sic genitus ad genitorem refertur, et sicut
pater ad filium ita genitor ad genitum.’
204 Stefan Heßbrüggen-Walter
45
the predicate. He then concludes that something can only
be ‘ingenitus’ if it has not been produced by a ‘genitor’,
refuting the basic Arian assumption that ‘ingenitus’ as predi-
cated of the Godhead falls under the first category.46
This discussion in De Trinitate not only shows the extent
to which Augustine was aware of Arian theology and its use
of dialectical tools, but also confirms that reflection on topics
in the science of definition constituted for him an important
element of anti-heretical discourse—as we might have
surmised from a passing remark about ‘false Christians’ in
De Doctrina Christiana 2. 35. 53.

6 . c o n c lu s i o n : au g u st i n e a n d d i a l e c t i c
If the arguments presented above are convincing, the dis-
continuity between Augustine’s early conception of dialectic
and his developed view of the discipline as it is presented in
De Doctrina Christiana will have become both more apparent
and more explicable. Should we then think of this change
as marking a ‘decline in dialectic’, if only in Augustine’s
estimation?47 I think not. Such an assessment would be
correct only if dialectic could justifiably be considered a
science capable of generating material insights for Christian
theology. That is in fact exactly the understanding of the
discipline presupposed by theologians who were regarded by
Augustine and others as ‘heretical’. It is not a view that he
shared by the time he began writing De Doctrina Christiana.

45
Trin. 5. 6. 7: The ‘proof’ depends on two auxiliary premisses:
Aristotle’s list of categories is exhaustive, which might easily be granted.
And—more problematic—all predicates belonging to a category behave
identically in this respect; thus it suffices to test one predicate in each
category.
46
Trin. 5. 7. 8: ‘Non ergo receditur a relativo praedicamento cum
ingenitus dicitur. Sicut enim genitus non ad se ipsum dicitur sed quod ex
genitore sit, ita cum dicitur ingenitus non ad se ipsum dicitur sed quod
ex genitore non sit ostenditur’ (‘Thus there is no departure from relative
predication when he is called “unbegotten”. For as “begotten” is not said
in relation to himself but in virtue of the fact that he is from the begetter,
so when he is called “unbegotten” it is not said in relation to himself but in
virtue of the fact that he is shown not to be from the begetter’).
47
See above, n. 7. For Augustine’s attitude in his earlier works, see also
Conybeare’s essay in this volume.
 Augustine’s Critique of Dialectic 205
There, instead, he produces a ‘critique’ of the discipline in a
very modern, almost Kantian sense. He shows what can and
cannot be achieved by dialectical means. He proves that any
overtly optimistic assessment of the potential of dialectic
leads to philosophical and theological error. Such error is
not, however, the fault of the discipline itself—as some of its
more radical modern critics would believe—but of adherents
who fail to recognize its quite limited scope. Dialectic, as
Augustine came to conceive of it, is supposed to secure the
formal, not the material, validity of inference and definition,
being thus (as Kant would put it), not an organon, but a canon
for cognition.
 9
Augustine’s Hermeneutics as a
Universal Discipline!?
Ka r la Po l l m a n n

1. i n t r o d u c t i o n
The sense of abstract terms often varies between domains of
research. For our purpose, ‘hermeneutics’ may be defined as
theoretical reflection on the conditions under which a trans-
mitted text becomes understandable to a contemporary
reader. More generally, hermeneutics is ‘the classical dis-
cipline concerned with the art of understanding texts’, a
theory of the rules for interpretation, whereas ‘exegesis’
refers to the practical application of such rules to a given text
or texts.1
If we accept these definitions of hermeneutics, we shall
not be able to point to a single (extant) work dedicated as a
whole to this discipline before Late Antiquity. In ancient
literature of the ‘classical’ periods, we find only rudimentary
reflections on the ways of interpreting a text correctly, inte-
grated in various practical contexts. So, for example, in the
grammatical tradition we have the fourfold way of dealing

1
H.-G. Gadamer, Hermeneutik, i: Wahrheit und Methode, 6th edn.
(Tübingen, 1990), 169: ‘Die klassische Disziplin, die es mit der Kunst des
Verstehens von Texten zu tun hat, ist die Hermeneutik’ (p. 146 in Engl.
trans. (London, 1975)). See R. Barthes et al., Exégèse et herméneutique
(Paris, 1971) for hermeneutics as ‘savoir de l’appropriation du sens’
(p. 15), as an enriched synonym for ‘interpretation’ (p. 278), and for the
various exegetical methods (p. 280). See also A. C. Thiselton, The Two
Horizons: New Testament Hermeneutics and Philosophical Description with
Special Reference to Bultmann, Gadamer, and Wittgenstein (Grand Rapids,
Mich., 1980), 10–12, and D. F. Wright, ‘Augustine: His Exegesis
and Hermeneutics’, in M. Saebø (ed.), Hebrew Bible/Old Testament: The
History of its Interpretation, i, pt. 1 (Göttingen, 1996), 701–30.
 Augustine’s Hermeneutics 207
with the canonical school authors: ‘lectio’ (correct pronunci-
ation, intonation, and division of words); ‘emendatio’ (text-
ual criticism); ‘enarratio’ (detailed commentary on the text
as now established, including its historical, mythological,
and rhetorical aspects);2 and ‘iudicium’ (judgement of the
authenticity and value of a piece of literature).3 In rhetoric,
there is the ‘status’-system of the rhetorician Hermagoras
(second century b c), the so-called Status Hermagorae; one
section of this contains the Status Legales,4 which give rules
on how to solve specific difficulties in understanding a given,
fixed text—in this case, that of a written law—in order to be
able to apply it correctly to a specific case. Questions of the
interpretation of written laws are also treated in the legal
literature, as for example in the Digest (1. 3 and 50. 17).
In Christian tradition before the end of the fourth century,
hermeneutical reflection is also relatively rare. The only
exception is Origen, who deals in the fourth book of De
Principiis (‘On the Principles of Christian Faith’) with the
topic of how to interpret Holy Scripture. He explains
the method of ‘allegorical’ interpretation, which enables the
reader to detect one or even several hidden senses beneath
the literal meaning of a text. This technique has its ‘pagan’
equivalent in allegorical interpretation of Homer, Virgil, or
Orphic texts (Derveni papyrus),5 which can be traced back to
the sixth century b c.6

2
K. Eden, Hermeneutics and the Rhetorical Tradition: Chapters in the
Ancient Legacy and its Humanist Reception (New Haven and London,
1997), 20–40.
3
See, e.g., Quintilian, Institutio Oratoria 1. 4. 1–3; 1. 8. 1–21.
4
See Cicero, De Inventione 2. 116–54 (ambiguity, letter and intent,
conflict of laws, reasoning by analogy, definition), and Augustine (?), De
Rhetorica 11 (four types of controversy or ‘quaestiones legales’: ‘scriptum
et voluntas’, ‘contentio legum contrariarum’, ‘ambiguitas’, ‘conlectio’).
5
See now A. Laks and G. Most (eds.), Studies on the Derveni Papyrus
(Oxford, 1997), including an edition and translation of the text.
6
See esp. K. Froehlich (trans. and ed.), Biblical Interpretation in
the Early Church (Philadelphia, 1984); P. Rollinson, Classical Theories of
Allegory and Christian Culture (Pittsburgh, 1980); D. Dawson, Allegorical
Readers and Cultural Revision in Ancient Alexandria (Berkeley and Los
Angeles, 1992); Z. Zlatar, The Epic Circle: Allegoresis and the Western Epic
Tradition from Homer to Tasso (Sydney, 1993).
208 Karla Pollmann
It is already clear that a stronger interest in hermeneutical
questions arises when there is a fixed and canonical text to
consider, like Homer, Virgil, written laws, oracles—or the
Bible. For the most part, the obscurity of a text and the
difficulty of understanding it are the results of its age and
the circumstances of its transmission. This factor acquires
additional importance when a text is granted religious or
quasi-religious authority, with the implication that its mes-
sage needs to be understood. The assumption then is that the
text has a meaning—probably hidden—that is relevant to the
contemporary reader.
The manner of dealing with a text is in many cases highly
arbitrary, and also closely connected to socio-cultural norms
and prevailing ideologies. So it seems reasonable to expect
that a theory of interpretation (that is, hermeneutics) will
take account of at least some of the factors just mentioned,
and that it will offer an intellectual framework within which
hermeneutical rules are supposed to operate.7
The first two works to deal exclusively and specifically
with the discipline of hermeneutics are Tyconius’ Liber Regu-
larum and Augustine’s De Doctrina Christiana. Both authors
lived at the end of the fourth century, both were from
North Africa, and both were Christians. There is no clear
evidence as to why these two individuals at this particular
time wrote those works; they themselves give no hints at all.8

7
Cf. Barthes et al., Exégèse et herméneutique, 285, where P. Ricoeur
emphasizes ‘qu’il n’existe pas de méthode innocente; que toute méthode
suppose une théorie du sens qui n’est pas acquise, mais qui est elle-même
problématique’; similarly J. S. Croatto, Biblical Hermeneutics: Toward a
Theory of Reading as the Production of Meaning (Maryknoll, NY, 1987;
original Spanish edn., Buenos Aires, 1984), p. x: ‘there is no such thing as a
nonhermeneutic reading of the Bible’.
8
I have written on this elsewhere: see K. Pollmann, ‘La genesi del-
l’ermeneutica nell’Africa del secolo IV’, in Cristianesimo e Specificità
regionali nel Mediterraneo Latino (sec. IV–VI): XXII Incontro di Studiosi
dell’antichità cristiana, Studia Ephemeridis ‘Augustinianum’ 46 (Rome,
1994), 137–45, and eadem, Doctrina Christiana: Untersuchungen zu den
Anfängen der christlichen Hermeneutik unter besonderer Berücksichtigung
von Augustinus, De doctrina christiana (Fribourg, 1996), 32–65; also
B. Sundkler and C. Stead, A History of the Church in Africa (Cambridge,
2000), 21–30; R. J. Forman, Augustine and the Making of a Christian
Literature (Lewiston, NY, 1995), 131–6.
 Augustine’s Hermeneutics 209
We can only speculate about what led Augustine in 396
to start writing a hermeneutical treatise; contrary to his
practice in many other works, including (perhaps the most
famous instance) the City of God, he nowhere refers to a
specific occasion. That he knew Tyconius’ Liber Regularum
can be seen from the paraphrase of that work that he
includes at the end of Book 3 of De Doctrina Christiana.9
In general, De Doctrina Christiana seems to avoid concrete,
historical allusions that could detract from the universal
validity of its precepts. Nor do we know why Augustine only
finished the work after a break of thirty years, at the end of
his life, in 426/7. It is remarkable, in any case, that he felt
able to complete De Doctrina Christiana without any major
alteration of his original plan, as he himself testifies in his
Retractationes. (It is often stated that the older Augustine
advocated a rather negative or pessimistic anthropology,
depriving humanity of almost all dignity before the grace of
God. Yet he let stand remarks in De Doctrina Christiana
about the human ability to be a temple and a reasonable
prophet of God.10)
In the prologue to De Doctrina Christiana Augustine
defines this work as ‘certain precepts for the treatment of
the Scriptures’ (‘praecepta quaedam tractandarum scrip-
turarum’, prol. 1) and defends it against potential critics.
Then he defines the purpose of the work positively: in the
same way that all who know letters are able to read for them-
selves and do not need anybody else to read to them, so
everybody who accepts the precepts of De Doctrina
Christiana should be able to resolve the obscurities of the
Bible on their own behalf. At least they may avoid the

9
See M. Moreau, I. Bochet, and G. Madec (eds. and trans.), Saint
Augustin: La doctrine chrétienne, BA 11(2) (Paris, 1997), 562–81; Poll-
mann, Doctrina Christiana, 196–215. Less satisfactory is G. Bray, Biblical
Interpretation Past and Present (Downers Grove, Ill., 1996), who calls
Doct. Chr. ‘an amplification of Tyconius, whom [Augustine] regarded as
too simplistic’ (p. 92).
10
Especially in the prologue and at the end of Book 4. G. Lettieri,
L’altro Agostino: Ermeneutica e retorica della grazia dalla crisi alla meta-
morfosi del De doctrina christiana (Rome, 2001), dedicates an extensive
study to the theological differences between the earlier and the later parts
of Doct. Chr.
210 Karla Pollmann
absurdity of depraved meanings (‘aut certe in absurditaem
pravae sententiae non incidat’, prol. 9). Thus, like Tyconius,
Augustine asserts the universal effectiveness of his her-
meneutical rules, but he does so with a subtle limitation,
laying greater emphasis than his African precursor on the
independence of the individual reader and interpreter of the
Bible (e.g., prol. 1 and 9).
Augustine, we have noted, compares his hermeneutics to
the letters of the alphabet by means of which persons are
able to understand a given text without another’s help. This
is what he says:
[T]hose . . . who explain to an audience what they understand in
the scriptures are, as it were, performing the office of reader and
pronouncing letters they know, while those who lay down rules
about how they are to be understood are like the person who
teaches literacy, who gives out the rules, that is, on how to read. So
just as the person who knows how to read does not require another
reader, when he gets hold of a volume, to tell him what is written in
it, in the same way, those who have grasped the rules we are
endeavouring to pass on will retain a knowledge of these rules, like
letters (quasdam regulas velut litteras tenens), when they come
across anything obscure in the holy books, and will not require
another person who understands to uncover for them what is
shrouded in obscurity. (prol. 9)11
Two things are remarkable about this comparison. First,
we might wonder how Augustine can state that a simple
knowledge of letters is equivalent to an ability to read with
understanding. A German, though he or she knows the
letters of the Latin alphabet, will not even be able to utter
(let alone understand) an English text unless told about
certain phonetic rules. Things become even more compli-
cated when we think of the ancient habit of writing without
spacing between words (‘scriptio continua’) and the first step
of the grammatical exercise, the ‘lectio’, which consisted of
more than simply recognizing letter-forms.12 Augustine’s
11
Trans. E. Hill, in Saint Augustine: Teaching Christianity (New York,
1996), 104.
12
On the methodology of reading see M. Irvine, The Making of Textual
Culture: ‘Grammatica’ and Literary Theory (Cambridge, 1994), 68–74
(accent, vocalizing, punctuating, oral delivery); also the essay by Chin in
this volume.
 Augustine’s Hermeneutics 211
comparison makes sense, however, when we consider the
ancient notion of letters: ‘letters’ or ‘scripture’ meant the
conservation of words, texts, or thoughts already understood
and thereafter fixed, as becomes clear from Plato’s Phaedrus,
Aristotle’s De Interpretatione (1. 16a: letters as symbols
of the spoken words), Quintilian,13 and the grammatical
tradition down to Priscian. In this tradition, reading was
connected with an already existing body of (disciplinary)
knowledge. To pursue the analogy: Augustine’s hermeneu-
tics teaches a discipline or technique that enables people
to decipher or retrieve what they already know, or what is
already known. This principle corresponds to the famous
‘hermeneutical circle’ that both assumes a close relation-
ship between the authority of a text and the exegetical
principles applied to it and calls for a fusion of the horizons
(Horizontverschmelzung) of the text and its reader.14 As
Gadamer puts it, hermeneutics is not just a set of rules,
but the way a reader interacts with the text by being part
of it.15
A second point arises in connection with a remark of
Augustine’s in his Tractates on the Gospel according to St
John (24. 2). There he distinguishes between a picture
(‘pictura’) and letters (‘litterae’). While it is sufficient to look
at the beauty of a picture and be delighted by it, the aesthetic

13
Inst. 1. 7. 3. See also Irvine, Making of Textual Culture, 97–104, on
letters and writing.
14
See BA 11(2).438–49; Eden, Hermeneutics and the Rhetorical
Tradition, 58. Heidegger in particular theorized about the hermeneutical
circle (Thiselton, Two Horizons, 194–7), and was answered by Gadamer
with his notion of the fusion of horizons (ibid. 304–8). See also Gadamer,
Hermeneutik i. 270–95; idem, Hermeneutik, ii: Wahrheit und Methode, 2nd
edn. (Tübingen, 1993), 57–65. For Horizontverschmelzung in particular
see Gadamer, Hermeneutik i. 311 f., 380 f., 401; ii. 14, 55, 109, 351, 436,
475.
15
Gadamer, Hermeneutik i. 273 f.; 396: ‘Lesendes Verstehen ist nicht
ein Wiederholen von etwas Vergangenem, sondern Teilhabe an einem
gegenwärtigen Sinn’ (‘The understanding of something written is not
a reproduction of something that is past, but the sharing of a present
meaning’ (Engl. trans. 354)); 398 f.; Hermeneutik ii. 21. Cf. G. Ripanti,
Agostino teoretico dell’interpretazione (Brescia, 1980), 73–86, on the exist-
ential, theological, and philosophical presuppositions that influence all
understanding.
212 Karla Pollmann
appearance of a letter is of only minor importance. What
matters is that letters admonish us (‘commoneris’) not only
to read but also to understand the content they convey, which
exceeds their material appearance. Applying this insight
to the passage highlighted in the proem to De Doctrina
Christiana, we can infer that Augustine’s treatise is to be
understood not as a self-sufficient literary artefact but as a
work that somehow points to a field or fields of knowledge
beyond itself.
Augustine’s De Doctrina Christiana thus proposes a uni-
versally valid ‘discipline of Scripture’ and at the same time
affirms its own strictly relative or indicative function. The
tension between these two aims permeates the whole work,
as we shall now attempt to show in detail.

2. h e r m e n e ut i c s a s a d i s c i p l i n e : f r a m e wo r k
Book 1 sets the normative horizon16 of Augustine’s hermen-
eutics: love (‘caritas’) towards God and one’s neighbour is
the framework and aim of every interpretation of the Bible.17
According to Augustine, ‘caritas’ is the common boundary
between God and humanity, and the one that unites all human
beings among themselves; it is the boundary between history
and eternity, between dynamic desire and final tranquil
enjoyment. The universal dimension of ‘caritas’ is justified
by its goal, the eternal God. Whereas love for temporal
things fades away once the desired object has been obtained,
love for eternal things remains even after possession of them
(Doct. Chr. 1. 38. 42). We shall return to this eschatological
aspect of Augustine’s hermeneutics.
This key Augustinian notion of ‘caritas’, however, con-
tains a hidden weakness that relativizes it as a hermeneutical
concept. It is basically an ethical criterion, which is only
truly fulfilled in practical application and proved by appro-

16
On this term see Thiselton, Two Horizons, 149–54.
17
Pollmann, Doctrina Christiana, 121–47; I. Sluiter, ‘Metatexts and
the Principle of Charity’, in P. Schmitter and M. J. van der Wal (eds.),
Metahistoriography: Theoretical and Methodological Aspects in the His-
toriography of Linguistics (Münster, 1998), 11–27.
 Augustine’s Hermeneutics 213
priate (i.e., selfless) human conduct. Augustine modifies the
pagan concept of love as a desiring and therefore motivating
power, through a combination of Platonic /ρω and Stoic
(ρµ (‘impulse’),18 which belongs to the ethical part of
philosophy. In Book 1 he emphasizes that compliance with
the commandments of the Bible, as the expressed will of
God, has to be the final aim of every truly successful biblical
exegesis. This ethical finality is re-emphasized in Book 4,
where Augustine asserts that a speaker’s (i.e., a preacher’s)
life-style may be more persuasive than any oration, and thus
it comes to frame the whole work. A similar thought had
already been formulated by Origen,19 who distinguishes
between purely theoretical disciplines, like geometry, whose
final aim is the understanding of their own contents, and
disciplines with a practical aim, like medicine, in which
knowledge is acquired in order then to be applied in another
sphere.20 To the latter category belongs ‘knowledge and
service of the Word’ (‘notitia ministeriumque sermonis’),
the final aim of which is works of mercy (‘opera’). Theo-
logically, this view could be supported by the biblical injunc-
tion in Ephesians 3: 19, ‘to recognize the love of Christ
which surpasses all knowledge’ (‘supereminentem scientiae
caritatem Christi’). This text is quoted in a passage of Book
2 of De Doctrina Christiana where Augustine emphasizes
the conditions for proper Christian behaviour (‘omnis actio
christiana’), without mentioning intellectual activity.21 Simi-
larly, 1 Corinthians 8: 1—‘knowledge puffs up, but charity
edifies’ (‘scientia inflat, caritas aedificat’)—forms almost a
mantra of De Doctrina Christiana.
Besides the ethical category of ‘caritas’, there is a dogmatic
restriction on the hermeneutics proposed in De Doctrina
Christiana. Augustine stipulates that a proper interpretation

18
Augustine, Div. Qu. 35. 2: ‘amor’ = ‘appetitus’.
19
Origen, In Lucam 1 (GCS 35.8–9).
20
Similarly, Cicero, De Officiis 1. 42. 150 f., distinguishes between
‘artes liberales’, ‘artes sordidae’, and a third, intermediate class of ‘artes,
quibus . . . non mediocris utilitas quaeritur, ut medicina, ut architectura,
ut doctrina rerum honestarum’.
21
Doct. Chr. 2. 41. 62: ‘bene operari in Christo et ei perseveranter
inhaerere, sperare caelestia, sacramenta non profanare’.
214 Karla Pollmann
of Scripture should conform to the rule of faith,22 which he
briefly presents in Book 1 by offering theological explana-
tions of the Christian doctrine of God and his church,
according to the scheme of the Apostolic Creed. The clear
implication is that this Creed was universally accepted by the
Church—admittedly more a postulate than an actual reality
in Augustine’s time (as indeed today).
The universal potential of Augustinian biblical her-
meneutics can thus be seen as limited, to some degree, by the
ethical and dogmatic conditions imposed upon it.

3. h e r m e n e ut i c s a s a d i s c i p l i n e : st r u c t u r e
In the opening sections of Books 1 and 2 Augustine gives a
clear systematic division23 of the contents of De Doctrina
Christiana, explaining the intended arrangement of the
material in the four books of his treatise (see Fig. 9.1).24
At the beginning of Book 1 Augustine first states the
theme of De Doctrina Christiana: namely, the treatment of
Holy Scripture (‘tractatio scripturarum’). Books 1–3 will
deal with ‘the mode of finding out what has to be under-
stood’ (‘modus inveniendi, quae intelligenda sunt’, 1. 1. 1),

22
See M. Fiedrowicz, Prinzipien der Schriftauslegung in der alten Kirche
(Bern, 1998), 151 n. 2; B. Studer, Schola christiana (Paderborn, 1998),
215 f.; and, for Augustine’s successful negotiation between doctrinal
‘auctoritas’ and free ‘ratio’, Forman, Augustine and the Making of a
Christian Literature, 100–28.
23
i.e., a dihaeresis or ‘partitio’ into the general (‘genus’) and the par-
ticular (‘species’). Such a systematic division is characteristic of the pagan
textbook (‘ars’ or, in Greek, technē), which is the generic literary model for
Augustine’s Doct. Chr.: Pollmann, Doctrina Christiana, 89–104.
24
This indicates that Augustine had the general plan of De Doctrina
Christiana in mind when he started to write in 396 and kept to it when he
finished the work in 426/7. The sketch in Fig. 9.1 is a slightly modified
version of the analysis offered in Pollmann, Doctrina Christiana, 90. For
recent summaries see M. Simonetti, Biblical Interpretation in the Early
Church: An Historical Introduction (Edinburgh, 1994; original Italian
edn., Rome, 1981), 107 f., and F. Young, Biblical Exegesis and the
Formation of Christian Culture (Cambridge, 1997), 270–7. In Doct. Chr. 2,
Augustine offers some additional subdivisions of the ‘doctrinae
gentilium’. For detailed discussion, see Klingshirn’s essay in this volume.
 Augustine’s Hermeneutics 215

Figure 9.1 Contents of De Doctrina Christiana

and Book 4 with ‘the mode of making known what has been
understood’ (‘modus proferendi, quae intellecta sunt’, ibid.).
At first glance it may seem strange that Augustine’s her-
meneutics is divided between a mode of finding, that is,
216 Karla Pollmann
of understanding the sense of a text, and a mode of perform-
ance, that is, communicating that sense to others. Modern
hermeneutics—and before Augustine, that of Tyconius—
confines itself to the former.25 It has been suggested that
Augustine’s bipartition corresponds to the rhetorical div-
ision between ‘inventio’, the discovery of ideas and subject-
matter for a speech, and ‘elocutio’, their stylistic elaboration,
but this view is open to serious objections. First, the system
of rhetoric is customarily divided into five parts (‘inventio’,
‘dispositio’, ‘elocutio’, ‘memoria’, and ‘actio’). Secondly,
Augustine states explicitly at 4. 1. 2 that he is not composing
a rhetorical textbook (‘rhetorica praecepta’; cf. ‘ars rhe-
torica’, 4. 2. 3). Thirdly, as can be seen from Figure 9.1, the
traditional subject-matter of rhetorical ‘inventio’ and
‘elocutio’ does not convincingly fit the actual contents of De
Doctrina Christiana. As an alternative explanation, we might
suppose that Augustine was influenced by the originally
Stoic distinction of an internal word (logos endiathetos) and
an external word (logos prophorikos), which is taken over by
religious thinkers from Philo onwards; indeed, Augustine
briefly refers to it here (1. 13. 12), as well as in other works of
his. However, the distinction in that case is one of specula-
tive theology, used to account for the nature of Jesus Christ
as the internal word of God in the external shape of a histor-
ical human being. It is never taken for a linguistic model in
De Doctrina Christiana.
The most convincing parallel is a fragment of Theophras-
tus as quoted in Ammonius’ commentary on Aristotle’s De
Interpretatione: ‘The relation of speech (λ"γο) is twofold,
first in regard to the audience, to which speech signifies
something [i.e., communication: Doct. Chr. 4], and secondly
in regard to the things about which the speaker intends to

25
I have so far been unable to find a comparable dichotomy in modern
hermeneutics, though Croatto, Biblical Hermeneutics, 82 f., emphasizes
the ‘sequential rotation, in which the word generates the text and the text
generates the word’. It is noteworthy that modern hermeneutical discus-
sion, when dealing with Augustine at all, generally concentrates on his
Confessions or quotes him at second-hand. This is true, e.g., for P. Ricoeur,
M. Foucault, J. Derrida, and, to a lesser degree, H.-G. Gadamer.
 Augustine’s Hermeneutics 217
persuade the audience [i.e., understanding: Doct. Chr. 1–3].’26
Augustine may be thought to have applied the twofold rela-
tion or quality of speech as found in Theophrastus to his
hermeneutics, making it the basis for his major subdivision
of De Doctrina Christiana between Books 1–3 and Book 4.
This would imply (a) that he sees a structural analogy
between λ"γο and hermeneutics,27 and (b) that he regards
the ‘tractatio scripturarum’ as a fundamental, if not the
fundamental, activity of Christian understanding and
communication.28
The presence of such a theoretical framework would
also help to explain the structure of the Confessions,
which Augustine composed after breaking off work on De
Doctrina Christiana. It is often suggested that there is not
much coherence between the first ten books that deal with
Augustine’s life up to his baptism and the final three
books that contain a model exegesis of the beginning of

26
διττ# γὰρ ο'ση τη̂ του̂ λ"γου σχ&σεω . . ., τη̂ τε πρ τοὺ
ακροωµ&νου, ο= κα σηµανει τι, κα τ# πρ τὰ πράγµατα, >π.ρ ?ν ( λ&γων
προτθεται πεσαι τοὺ ακροωµ&νον. See further Pollmann, Doctrina
Christiana, 170–3. Granted, the ‘ars grammatica’ was similarly divided
into a ‘scientia interpretandi’ (the science of interpreting) and a ‘ratio recte
scribendi et loquendi’ (principles for writing and speaking correctly):
Irvine, Making of Textual Culture, 6. But Augustine in Doct. Chr. is
not interested in correct speech but in persuasion, even at the price of
sacrificing grammatical correctness (‘latinitas’) if necessary: K. Vössing,
Schule und Bildung im Nordafrika der Römischen Kaiserzeit (Brussels,
1997), 233 n. 901. Hence rhetorical principles are more important for his
hermeneutics than grammatical ones; see also n. 41 below.
27
Already Philo had linked hermeneutical questions with an effort to
penetrate the problem of language: K. Otte, Das Sprachverständnis bei
Philo von Alexandrien: Sprache als Mittel der Hermeneutik (Tübingen,
1968); I. Christiansen, Die Technik der allegorischen Auslegungswissenschaft
bei Philo von Alexandrien (Tübingen, 1968). Gadamer, Hermeneutik i.
387–409, also sees a close link between language and hermeneutics.
28
This echoes the Neoplatonic idea that a person after the vision of the
One will return and, if possible, bring word of the soul’s heavenly inter-
course to others (Plotinus, Enneads 6. 7. 35; 6. 9. 11); see also Augustine,
C. Faust. 22. 54 (persons living a contemplative life are aflame with the
love of generating, for they desire to teach what they know); Gregory the
Great, Homiliae in Ezechielem 1. 5. 13; 2. 2. 4; and the essay by Conybeare
in this volume.
218 Karla Pollmann
Genesis. If Book 10 (on memory) and Book 11 (on time)
29

provide a generalizing link between the life of Augustine and

the general experience of life in time, then the exegesis given
by him mainly in Books 12 and 13 may be regarded as a
particular form of the only true and possible continuation of
a person’s life after his or her conversion to Christianity.
Given the anthropological model of De Doctrina Christiana,
it makes sense that the one, all-embracing aspect of a
Christian life-and-speech worth mentioning at this stage
would be that of an activity of biblical exegesis directed
towards others, this being the ‘end’ or ultimate meaning of
any Christian’s life after conversion, according to Augustine.
A successful exegesis will always, finally, mean a morally
good life-style.30
Theophrastus’ ‘pragmatic dimension of the sign’, as it
may be called, fits exactly with the immediate needs of
Augustine in De Doctrina Christiana, especially with regard
to his final aim of persuading people and goading them to
ethical action. In the same spirit, Augustine explains to
his audience in a sermon (date of delivery unknown) on
Christian ‘disciplina’ (= ‘learning’31) that Christian ‘dis-
ciplina’ means the proper love of God and one’s neighbour,

29
For the structural problems of this change of subject in the Confes-
sions see J. J. O’Donnell, Augustine: Confessions, iii (Oxford, 1992), 250–2;
N. Fischer and C. Mayer (eds.), Die Confessiones des Augustinus von Hippo
(Freiburg, 1998), 19–59, emphasizing the formal unity of the Confessions,
considered as that of a Christian protreptikos in the form of a dialogue
with God; and J. Holzhausen, ‘Augustin als Biograph und Exeget: Zur
literarischen Einheit der Confessiones’, Gymnasium 107 (2000), 519–36,
whose hypothesis that God has to be made to speak through exegesis in a
biography (p. 536) is consistent with the argument advanced here.
30
In the Confessions, Augustine does not so much emphasize the neces-
sity of a morally good life-style as the goal of all truly successful exegesis
as point to the eschatological end of a Christian life, the seventh day of
eternal rest (Conf. 13. 35. 50–38. 53); cf. Fischer and Mayer (eds.), Die
Confessiones des Augustinus von Hippo, 603–52. For the eschatological
dimension of ‘caritas’ in Doct. Chr., see further below.
31
Augustine, Disc. Chr. 1, echoes the etymology that would derive
‘disciplina’ from discere, which was popular in antiquity (e.g., Varro,
De Lingua Latina 6. 62) but has now been discarded: G. Jüssen and
G. Schrimpf, ‘Disciplina, doctrina’, in Historisches Wörterbuch der
Philosophie, i (Basel and Stuttgart, 1971), cols. 256–61, at 256.
 Augustine’s Hermeneutics 219
which is more important than material wealth. Moreover, he
makes it clear towards the end of Book 4 that he has not just a
clerical, but a wider, potentially universal readership in mind
as addressees for his hermeneutics, speaking of
a teacher of the truths by which we are delivered from eternal evils
and conducted to eternal good things, wherever these are being
presented, whether to the people, or privately to one person or
several, whether to friends or enemies, whether in unbroken dis-
course or in conversation, whether in treatises or in books, whether
in letters either lengthy or brief. (4. 18. 37)
All Christians can read and try to understand Scripture; it
will then be their task to communicate the results in sundry
ways.32
It is noteworthy that in Book 4 Augustine, despite his gen-
erally universal scope, gives two African examples to make
his point. First (at 4. 10. 24), he says that the Latin word ‘os’
is ambivalent, meaning ‘bone’ when the o is short and
‘mouth’ when it is long. He recommends for the sake of
clarity that the dedicated teacher, when speaking to the
unlearned, not shrink from saying ‘ossum’ (vulgar Latin for
‘bone’) rather than ‘os’, so as to avoid misunderstanding on
the part of those unable to distinguish between short and
long vowels. Interestingly, Augustine had already discussed
the ambivalence of ‘os’ thirty years earlier, before breaking
off work on De Doctrina Christiana (3. 3. 7). Then too he had
pleaded for the common barbarism ‘ossum’ as a way of con-
veying the correct meaning without ambiguity, but without
the specific reference to Africa. The new specificity of the
later reference perhaps reflects the writer’s own greatly
increased experience of the conditions of preaching in North
Africa over the intervening decades. (Note also 4. 24. 53,
where, as an example of a speech in the grand style, he cites a

32
See in more detail Pollmann, Doctrina Christiana, 69–75;
G. G. Stroumsa, ‘Milk and Meat: Augustine and the End of Ancient
Esotericism’, in A. and J. Assmann (eds.), Schleier und Schwelle, i
(Munich, 1997), 251–62; I. Sluiter, ‘Communication, Eloquence and
Entertainment in Augustine’s De Doctrina Christiana’, in J. den Boeft and
M. L. van Ploo-van de Lisdonk (eds.), The Impact of Scripture in Early
Christianity (Leiden, 1999), 245–57, at 250–9.
220 Karla Pollmann
sermon of his own given in Caesarea in Mauritania around
418.33)
Broadly speaking, in De Doctrina Christiana Augustine
combines several fairly heterogeneous fields, including ethics
(‘caritas’) and dogmatics (the Creed and Trinitarian
thought) in Book 1, semiotics and grammar in Books 2 and 3,
rhetoric in Book 4, and the pagan liberal arts at the end of
Book 2, in order to show the propaedeutic usefulness of the
encyclopaedic knowledge of his time for an understanding
of the Bible. In this he goes further than Tyconius, who had
combined theological and grammatical-rhetorical categories
in his Liber Regularum.34 It is therefore not surprising to find
scholars criticizing Augustine’s hermeneutics for its lack of
coherence. The strictly systematic structure of De Doctrina
Christiana makes it clear, however, that Augustine aimed at
such coherence. The failure of modern readers to recognize
that degree of systematicity probably results from the fact
that, as far as we know, Augustine’s organization of his
material does not follow any traditional system (τ&χνη/‘ars’).
For instance, he adopts neither the full-blown Stoic doctrine
of signs nor any encompassing grammatical or rhetorical
scheme. Instead, he takes up particular elements of such sys-
tems and transforms or enlarges them for his own hermen-
eutical purpose, which is to indicate how one should handle
traditional disciplines, which methods can be used to obtain
an understanding of the Bible and to expound it to others,
and what the limits and final aim of this undertaking are.
Thus conceived, Augustine’s hermeneutics becomes a
kind of ‘meta-method’ or ‘meta-discipline’ embracing all
other disciplines by indicating their instrumental service for
understanding the Bible, and by denying them a specifically
Christian usefulness in their own right. This is an ambitious

33
See for this K. Pollmann, ‘African and Universal Elements in the
Hermeneutics of Tyconius and Augustine’, in P.-Y. Fux, J.-M. Roessli,
and O. Wermelinger (eds.), Augustinus Afer—Saint Augustin: Africanité et
universalité, Actes du colloque international Alger-Annaba, 1–7 April 2001
(Fribourg, 2003), ii. 353–62.
34
Cf. Pollmann, Doctrina Christiana, 55–6; and Eden, Hermeneutics and
the Rhetorical Tradition, 61, who distinguishes between broadly legal and
broadly stylistic rules.
 Augustine’s Hermeneutics 221
correction of De Ordine, written in 386, where dialectic had
been called the ‘discipline of disciplines’ (2. 13. 38) and the
‘artes’ were commended as a means of gradually ascending
to the highest truth (2. 2. 5, 35–44; 2. 18. 47). In De Doctrina
Christiana, Augustine explicitly revokes such claims.35

4 . h e r m e n e ut i c s a s a d i s c i p l i n e : to o l s
The combination of these heterogeneous fields in De
Doctrina Christiana is made possible by what Augustine uses
as his chief hermeneutical tool: the sign.
All things, including methods and disciplines, can have the
function of a sign. At 2. 1. 1 Augustine gives a definition of
the sign: ‘a sign is a thing (res), which, besides the outward
appearance it presents to the senses, causes something else
to come out of it into one’s knowledge’ (‘signum est enim
res praeter speciem, quam ingerit sensibus, aliud aliquid ex
se faciens in cogitationem venire’). This definition is not
exactly Stoic, because the Stoics made a systematic distinc-
tion between the sensible material σηµανον (‘signifier’) and
the incorporeal, intelligible σηµαιν"µενον (‘signified’),
which can also be called λεκτ"ν (‘sayable’). The latter plays
no part in the system of De Doctrina Christiana, where
Augustine is interested only in the relation between ‘res’ and
‘signum’. As Ammonius testifies (De Interpretatione 17.
2–8), Aristotle thought that in a linguistic context ‘it is not
necessary to conceive of anything else additional . . . which
the Stoics postulated and decided to name a “sayable”’ .36 In

35
Note esp. Doct. Chr. 2. 37. 55: ‘tantum absit error, quo videntur sibi
homines ipsam beatae vitae veritatem didicisse cum ista didicerint’. This
sceptical attitude of the older Augustine towards scientific knowledge (for
which cf. also Ep. 101. 2; Retr. 1. 3. 4; similarly Civ. 10. 29: ‘ad Deum per
virtutem intelligentiae pervenire paucis esse concessum’) had few follow-
ers among later thinkers: H. M. Klinkenberg, ‘Artes liberales/artes
mechanicae’, in Historisches Wörterbuch der Philosophie, i. 532; C. Horn,
Augustinus (Munich, 1995), 58–61; Studer, Schola christiana, 182, 226–9
(on dialectic as an acceptable tool for Christians).
36
A. A. Long and D. N. Sedley, The Hellenistic Philosophers, i (Cam-
bridge, 1987), 198 (λεκτ"ν is the Augustinian ‘dicibile’ in De Dialectica);
W. and M. Kneale, The Development of Logic (Oxford, 1962), 140, prefer
to translate it ‘what is meant’; cf. also Heßbrüggen-Walter in this volume,
esp. 186, 190–1, 194–202.
222 Karla Pollmann
De Dialectica, written in 387, Augustine mentions the
Stoic λεκτ"ν (‘dicibile’), which plays no role in De Doctrina
Christiana roughly ten years later. Again we can see that
Augustine knew the criterion in question and omitted it
deliberately. It was the rhetorical tradition, however, that
knew the twofold definition of ‘sign’ as something sensible
and intelligible. Cicero, for example, says that ‘a sign is
something apprehended by one of the senses and indicating
something that seems to follow logically as a result of it’.37
This formula correlates with a holistic definition of human
beings as consisting of body and soul.38
As may be gathered from the work of a writer like Sextus
Empiricus, there was considerable debate about the nature
of signs in antiquity. Augustine’s De Dialectica shows that he
was well aware of this discussion. Hence we can conclude that
in De Doctrina Christiana he consciously adopted elements
of peripatetic and rhetorical theories of language and sign.
The rhetorical focus was enlarged by Augustine himself,
who dilates the system by introducing ‘signa ignota’ and
‘ambigua’, and ‘propria’, and ‘translata’ (unknown, ambigu-
ous, proper, and transferred signs), which are all rhetorical
categories.39 This proceeding bears out the general observa-
tion that Augustine in his hermeneutics laid great emphasis
on the rhetorical or communicative aspect of persuasion,40
37
Cicero, De Inventione 1. 39. 48: ‘signum est quod sub sensum aliquem
cadit et quiddam significat quod ex ipso profectum videtur’. Cf. the
Aristotelian logical tradition: Aristotle, Analytica Priora 70a. 7–9;
Rhetorica ad Alexandrum 12. 1430b, where only the intellectual, not the
sensual, aspect is mentioned.
38
Mentioned, e.g., at Doct. Chr. 1. 26. 27; C. Faust. 22. 27; Beata Vita 2.
7; Ep. 3; Serm. 150. 5, following the Aristotelian tradition, and not
the Platonic, as he does in other places; cf. M. Simonetti, Sant’Agostino:
L’istruzione cristiana (Verona, 1994), 399 n. 6.
39
See Pollmann, Doctrina Christiana, 180–1.
40
Cf. Eden, Hermeneutics and the Rhetorical Tradition, 42, directed
against statements by Irvine—e.g., Making of Textual Culture, 170: ‘The
model for interpreting texts in De doctrina christiana is grammatica, not
rhetorica’; 178 (Doct. Chr. called a Christian ‘ars grammatica’); but cf. 183
(‘grammatico-rhetorical’ elements). We should note that Augustine at
Doct. Chr. 3. 29. 40 states explicitly that he does not intend to write an ‘ars
grammatica’; but neither does he want to write an ‘ars rhetorica’ (4. 1. 2).
Chin’s essay in this volume renews the claim for the ‘grammaticality’ of
Doct. Chr., but on different grounds.
 Augustine’s Hermeneutics 223
meaning in this case the actualization of the biblical message
and its application to the needs and interests of the indi-
vidual reader or hearer. Augustine implicitly assumes that
the Bible is the universal book that can replace all other
reading. To sketch the principles of its interpretation is thus
already to adumbrate a universal hermeneutics.
We may ask why Augustine decided to introduce the ques-
tion of the sign into his hermeneutics at all,41 and here we can
look again at the text already mentioned from his Tractates
on the Gospel according to St John, where he distinguishes
between the nature of a picture and that of a letter. Horace in
Ars Poetica 361 says that poetry is like a picture (‘ut pictura
poiesis’), reflecting the Aristotelian mimetic theory of litera-
ture. Augustine does not want the Bible (or his own work,
least of all De Doctrina Christiana) to be mimetic, but to be
semiotic—in other words, full of signs admonishing us, like
letters, to decipher and understand at a deeper level.42 One
should not neglect the material surface of those signs (i.e.,
the ‘literal’ sense of the Bible), but the literal sense is not
necessarily an end in itself. Augustine thus manages to
integrate both the historical-critical and the allegorical
approaches to the biblical text, on the basis of a system
that is specifically Christian in character. The sign as the
principle tool of Augustine’s hermeneutics allows for its
universal dimension, since (a) all things can be signs anyway,
and (b) signs in virtue of their relativity or relatedness are
not meant to stand by themselves but need a wider context in
order to be intelligible—a context that implies a limitation of
their objectivity.43
As Kathy Eden rightly points out,44 Augustine develops a
hermeneutics of the middle way, avoiding the excesses either
of a purely literal reading, associated by him with the Jews

41
He states already at Doct. Chr. 1. 2. 2 that ‘every discipline deals with
things or with signs, but things are learned by signs’ (‘omnis doctrina vel
rerum est vel signorum, sed res per signa discuntur’).
42
For a similar distinction between reference and mimesis see
R. Lundin, C. Walhout, and A. C. Thiselton, The Promise of Hermeneutics
(Grand Rapids, Mich., 1999), 50–7, following Gadamer.
43
See, e.g., Doct. Chr. 2. 2. 3. Gadamer, Hermeneutik i. 424 f. (on
Augustine, Trin. 15. 10–15), ii. 174–83.
44
Eden, Hermeneutics and the Rhetorical Tradition, 61–2.
224 Karla Pollmann
in their carnal servitude (Doct. Chr. 3. 9. 13), or of an exag-
geratedly allegorical reading, associated by him with the
pagans (3. 7. 11), in order to promote the spiritual liberty
(‘christiana libertas’) of an alternative interpretation theory
(3. 8. 12; 3. 9. 13).

5 . h e r m e n e ut i c s a s a d i s c i p l i n e :
t h e o lo g i c a l s e tt i n g
Putting the fundamental claims of De Doctrina Christiana
into theological perspective, Augustine emphasizes repeat-
edly that it is God who in fact guarantees the success of both
understanding and communication. Here of course he
touches upon his doctrine of grace, and so embeds the uni-
versal claim of his hermeneutics in the even more universal
and divine frame which is simultaneously its content, aim,
and support.45
At the end of Book 1 (39. 43), Augustine quotes 1 Corin-
thians 13: 8: ‘As for prophecies (prophetiae), they shall be
done away with, as for tongues (linguae), they shall cease, as
for knowledge (scientia), it shall be done away with.’ For
him, this eschatological goal is already fulfilled in the desert
hermits who live without books, solely according to the prin-
ciples of hope, faith, and charity. They do not even need the
Bible except for the purpose of instructing others.46 The
same is true of De Doctrina Christiana: it too serves to
instruct others, but will ultimately pass away.
The end of the work is therefore threefold:
Pragmatic: The goal of the treatise will be attained once
people have understood how to interpret the Bible and can
do it for themselves: in this respect, De Doctrina Christiana is
like a knowledge of ‘letters’.
Ethical: The aim of the treatise, and of the understanding
of the Bible which it seeks to inculcate, will be fulfilled once
people begin to live according to the double commandment
of love for God and for one’s neighbour: the hermeneutics of
De Doctrina Christiana implies an ethics.
45
On the ‘hermeneutical circle’, as thus inscribed and theoretically
justified by Augustine, see above, n. 14.
46
See BA 11(2).480–3.
 Augustine’s Hermeneutics 225
Eschatological: Love alone (‘caritas’), the goal of all exe-
gesis, will endure beyond the end of time, leaving even faith
and hope behind (1. 39. 43): De Doctrina Christiana is an
instrument in the salvation-historical process.
The eschatological criterion has important implications
for Augustine’s developing view of the kind(s) of disciplinary
knowledge fit to be espoused by Christians. Already in De
Genesi contra Manichaeos, written about seven years before
he became a bishop, in 388/90, Augustine linked human
knowledge expressed in human words to the post-lapsarian
state of humanity and foresaw the eschatological destruction
of this kind of knowledge. He is commenting on Genesis 2: 5:
And therefore humanity, already toiling [because of the Fall]
received from the clouds rain that was needful on earth, that is,
teaching with human words. Hence we were given to understand
that human beings toiling on earth—that is, established in the dry-
ness of their sins—need to receive divine teaching through human
words, like rain from clouds. But such knowledge will perish.
(1 Cor. 13: 8)47
Aside from this theological devaluation of human learning,
Augustine also has a pragmatic objection to disciplines
which may be hard and thorny to learn and which can in
some cases be more quickly and easily acquired through
observation and imitation of others who practise them (Doct.
Chr. 2. 37. 55).
The merely temporary status of ‘doctrina’ is confirmed by
the scheme of the seven-step ascent to God outlined at the
beginning of Book 2 (7. 9–11):48
Step 1: fear of God (cf. Matt. 5: 3)

47
Gen. adv. Man. 2. 5. 6: ‘et ideo iam laborans in terra necessariam
habet pluviam de nubibus, id est doctrinam de humanis verbis. . . . ut
intellegeremus laboranti homini in terra, id est in peccatorum ariditate
constituto, necessariam esse de humanis verbis divinam doctrinam
tamquam de nubibus pluviam. Talis autem scientia destruetur.’
48
For a wider context and other instances, see M. Parmentier, ‘The
Gifts of the Spirit in Early Christianity’, in den Boeft and van Ploo-van de
Lisdonk (eds.), Impact of Scripture, 58–78. M. Nussbaum, ‘Augustine and
Dante on the Ascent of Love’, in G. B. Matthews (ed.), The Augustinian
Tradition (Berkeley and Los Angeles, 1999), 61–90, does not discuss this
passage.
226 Karla Pollmann
Step 2: modest piety, even when we do not understand
Scripture (Matt. 5: 4)
Step 3: knowledge (‘scientia’ = ‘doctrina christiana’), leading
to mourning (Matt. 5: 5)49
Step 4: courage, hunger for justice (Matt. 5: 6)
Step 5: counsel with mercy, love of neighbour (Matt 5: 7)
Step 6: dying to this world, purging of the eyes to see God,
putting neither one’s neighbour nor oneself before truth
(Matt. 5: 8)50
Step 7: wisdom (‘sapientia’) in peace and tranquillity (Matt
5: 9)51
49
See the more explicit statements in Serm. 347. 3: ‘merebuntur
scientiae gradum, ut noverint non solum mala praeteritorum peccatorum
suorum, de quibus in primo gradu poenitentiae dolore fleverunt, sed etiam
in quo malo sint huius mortalitatis et peregrinationis a Domino, etiam
cum felicitas saecularis arridet’ (‘They will earn the level of knowledge, so
that they do not only understand the evils of their previous sins, about
which they wept on the first level of their penitence, but also understand
how bad their situation is in their mortality and pilgrimage away from the
Lord, even if worldly happiness smiles on them’) (PL 39.1525); Serm.
Dom. Mont. 1. 3. 10, 1. 4. 11.
50
i.e., the step of ‘understanding’ (Hill, Saint Augustine: Teaching
Christianity, 163 n. 11) which, oddly enough, Augustine does not actually
name here. But cf. Serm. Dom. Mont. 1. 4. 11: ‘intellectus congruit mundis
corde tamquam purgato oculo, quo cerni possit quod corporeus oculus
non vidit nec auris audivit nec in cor hominis ascendit, de quibus hic
dicitur: “beati mundicordes” [Matt. 5: 8]’ (‘Understanding coincides with
those pure in heart, as if their eye had been cleansed, through which one
can see what the corporeal eye did not see and the ear did not hear and
what did not enter the human heart, about which it is said here: “Blessed
are the pure in heart”’ ).
51
For a clearer correlation, see Serm. Dom. Mont. 1. 3. 10: ‘postrema est
septima ipsa sapientia, id est contemplatio veritatis, pacificans totum
hominem et suscipiens similitudinem dei, quae ita dicitur: beati pacifici,
quoniam ipsi filii dei vocabuntur [Matt. 5: 9]’ (‘Finally there is the
seventh level, which is wisdom itself, that is the contemplation truth,
which pacifies the human being as a whole and receives the likeness of
God, of which is said: “Blessed are the peacemakers, for they shall be
called the children of God”’ ). ‘Pacificus’ may be understood as either
‘pacifying’ or ‘peaceful’: see A. Souter, A Glossary of Later Latin to 600
AD (Cambridge, 1949, and reprints), s.v. Here Augustine takes the latter
sense for granted, allowing himself a Platonic-Stoic interpretation of the
highest state of the human soul as one of tranquillity; C. van Lierde, ‘The
Teaching of St Augustine on the Gifts of the Holy Spirit from the Text
of Isaiah 11: 2–3’, in F. Van Fleteren et al. (eds.), Augustine: Mystic and
Mystagogue (New York, 1994), 5–110, at 55–9.
 Augustine’s Hermeneutics 227
Augustine here lists the gifts of the Holy Spirit attributed to
the messianic king in Isaiah 11: 2–3,52 but works backwards
from the last, the fear of the Lord, to the first, the spirit
of wisdom. He reverses the Old Testament order because
‘the fear of the Lord is the beginning of wisdom’.53 He seems
to be the first to interweave the Beatitudes of the New
Testament with the gifts of the Holy Spirit from the Old
Testament.54 The Beatitudes represent programmatically
the new moral order and the new reality that Jesus came to
proclaim. As the founder of this new kingdom, Jesus already
represents the messianic king prophesied in Isaiah; by
linking that prophecy with the Beatitudes, Augustine
makes it valid for every Christian taught by Jesus, who came
to fulfil the Old Testament (Matt 5: 17). The messianic per-
spective of Isaiah is seen as something that can be fulfilled
by or in a Christian individual, who thereby becomes part of
its eschatological realization.

52
LXX: ‘et requiescet super eum spiritus Domini: spiritus sapientiae et
intellectus, spiritus consilii et fortitudinis, spiritus scientiae et pietatis,
et replebit eum spiritus timoris Domini’ (‘And the spirit of the Lord
shall rest upon him: the spirit of wisdom and understanding, the spirit
of counsel and courage, the spirit of knowledge and piety, and the spirit of
the fear of the Lord will fill him completely’).
53
Psalm 110 (111): 10; Sirach 1: 16: ‘initium sapientiae timor Domini’
(‘The fear of the Lord is the beginning of wisdom’). The order of the
saying was first changed by Hilary of Poitiers, In Psalmos 118: 5. 16
(around 365); see also Ambrose, In Psalmos 118: 5. 39. Cf. Augustine’s
explanation in Serm. 347. 2; see also Serm. Dom. Mont. 1. 4. 11.
54
The Beatitudes had already been described as steps towards the
ultimate goal of perfection by Gregory of Nyssa in his Eight Speeches on
the Beatitudes (delivered 387), which had been adapted by Ambrose in his
commentary on Luke (written between 388 and 392). The gifts of the
Holy Spirit had been seen as guides towards God and Christian perfection
by Irenaeus of Lyon, Adversus Haereses 3. 24. 1; Demonstratio 9: see A.
Mutzenbecher, in CCSL 35, pp. xiii–xvi. Augustine seems to be the first to
combine the Beatitudes of Matt. 5: 3–10 (economized from eight to seven)
with the seven gifts of the Holy Spirit of Isa. 11: 2–3. There is nothing like
this in the commentaries on Isaiah by Eusebius and Jerome. (Origen’s
commentary on Isaiah is lost apart from some fragments.) Augustine can
be seen advancing this view in his Serm. 347. 2 ff. (PL 39.1524–6, date of
delivery unknown) and his Serm. Dom. Mont. 1 (written between 392 and
396), where, as already noted, we find all the elements of Doct. Chr. 2 in
greater detail. See also En. Ps. 11: 7; Ep. 171A.
228 Karla Pollmann
Characteristically, Augustine not only changes the order
of the gifts of the Holy Spirit but also lines them up as a
progressive sequence in which every step must follow the one
before, whereas the Old Testament passage provides merely
a cumulative enumeration.55 Moreover, the messianic-pro-
phetic context has now changed. In Isaiah, the messianic
king is endowed with God’s spirit (11: 2), which can be
taken to imply that every individual needs the help of the
Holy Spirit for a successful ascent.56 But whereas the
passage in Isaiah depicts a future vision that will transform
the whole of society and establish a new and just order,
Augustine concentrates on the spiritual progress of the
individual.57 Here Neoplatonic influence is visible, though
the language and ideas are otherwise wholly biblical. The
idea of progress and ascent is Neoplatonic, as is that of the
final vision in a state of calm (compare Plotinus, Enneads 6.
9. 11) and of fellow human beings left behind.58 For
Augustine, however, these effects can only be worked by

55
See also Augustine, Serm. 347. 2 f. (PL 39.1524 f.). There is no
progression noted in Eusebius’ or Jerome’s commentaries on Isaiah or
in Jerome’s on Matthew: CCSL 35, p. ix. But in other contexts such
moral progression had already been spelled out: e.g., Origen, In Numeros
Homiliae 27 passim (esp. GCS 30.263, 272: ‘ordinem profectuum’, and
276); Ambrose, In Lucam 5. 50–2 (CCSL 14.152–3), and esp. 60: ‘vide
igitur ordinem . . . nisi pauper fueris, mitis esse non poteris’ (CCSL
14.155), based on Gregory of Nyssa, Beat. Or.
56
e.g., Doct. Chr. prol. 8; 4. 15. 32; B. Kursawe, Docere–delectare–
movere: die officia oratoris bei Augustinus in Rhetorik und Gnadenlehre
(Paderborn, 2000), 39–41, at 46 f. Already Gregory of Nyssa saw the
ability of a human to perform the Commandments implicit in the
Beatitudes as a gift of God; Beat. Or. 7: %µο δοκε κα τ /ργον %φ’ @ τν
τοσου̂τον µισθν %παγγ&λλεται Aτερον δ!ρον ε0ναι (PG 44.1281A).
57
According to Augustine, this can only rarely be achieved in this life,
and then only temporarily; R. Teske, ‘St Augustine and the Vision of
God’, in Van Fleteren et al. (eds.), Augustine: Mystic and Mystagogue,
287–99, at 299. A similar opinion is expressed by Plotinus, Enn. 6. 9. 10.
58
This ‘anti-social’ element is made explicit only in the model of ascent
as described in Doct. Chr., and is not made sufficiently clear in the other-
wise excellent commentary in BA 11(2).472. Cf. Plotinus, Enn. 6. 7. 35
(someone having the vision of the One forgets all other objects of
contemplation).
 Augustine’s Hermeneutics 229
divine grace; there is thus a genuinely Christian quality to
his presentation.59
What is most important for our purpose, though often
ignored, is that Augustine allocates the work of De Doctrina
Christiana itself to the third step, near the beginning of a
person’s progress towards the understanding and vision of
God and the grasping of wisdom. There is a substantial
gap between ‘scientia’ (also ‘doctrina’ or ‘disciplina’)60 as a
purely rational faculty of knowledge, the understanding of a
clearly defined area or subject-matter with an ethical aim,
and ‘sapientia’, the contemplation of eternal truths.61 This
emphasis is again Neoplatonic,62 and corrects Augustine’s
earlier, more optimistic ideas about the capacity of dialectic

59
See van Lierde, ‘Teaching of St Augustine on the Gifts of the Holy
Spirit’, 18–24.
60
H.-I. Marrou, Saint Augustin et la fin de la culture antique, 4th edn.
(Paris, 1958), 554–8, at 562, and, e.g., Doct. Chr. 2. 38. 57 for ‘doctus’ as
opposed to ‘sapiens’. Kursawe, Docere, 144 n. 643, claims that ‘scientia’
usually denotes ‘personal knowledge’, as opposed to ‘objective discipline’,
which is certainly not always the case in Augustine.
61
e.g., Augustine, Serm. 347. 2: ‘sapientia, lumen scilicet mentis
indeficiens’ (PL 39.1524); Serm. Dom. Mont. 1. 4. 11: ‘sapientia, id est
contemplatio veritatis’. Marrou, Saint Augustin et la fin de la culture
antique, 564–9.
62
In Neoplatonic thought, the intellect must be eternally out of its
mind in order to be one with the divine mind. For this paradoxical
self-transcendence of the intellect, see, e.g., Plotinus, Enn. 5. 3. 13; 6. 9.
11; C. Butler, Western Mysticism, 2nd edn. (London, 1960), 338 ff.;
J. J. O’Meara, ‘The Neoplatonism of Saint Augustine’, in D. O’ Meara
(ed.), Neoplatonism and Christian Thought (Albany, NY, 1982), 34–41, at
40. Analogously, in Augustine, seeing eternal truth means leaving
behind the faculties of language and thought: Ver. Rel. 72: ‘transcende et
te ipsum. Sed memento, cum te transcendis, ratiocinantem animam te
transcendere’; Conf. 9. 10. 24: ‘et venimus in mentes nostras et tran-
scendimus eas’; 9. 10. 25: ‘et ipsa sibi anima sileat et transeat se non se
cogitando’; G. Madec, ‘Ascensio, ascensus’, AL i. 465–75, at 469, 473;
Studer, Schola christiana, 277–80; more generally Horn, Augustinus,
61–87. P. Cary, Augustine’s Invention of the Inner Self (Cambridge,
2000), offers a different interpretation, emphasizing that Augustine, like
Plotinus, sees the ultimate vision of God as a perfection of the activity of
the mind, as opposed to Ps.-Dionysius, where the realm of the mind is
indeed left behind in the ultimate vision.
230 Karla Pollmann
and philosophy to enable true understanding and wisdom
(e.g., De Ordine 2. 18. 47–8).63
This relative positioning of the hermeneutic agenda of De
Doctrina Christiana is consistent with the specific quality of
its tool of choice, the sign. As already noted, signs are uni-
versally present, but their very nature is to hint at a different,
higher reality, a reality that they do not and cannot represent
fully in themselves. Besides this cognitive restriction, it
should be borne in mind that Augustine says repeatedly that
all successful biblical interpretation must result in ethically
good behaviour: love towards God and one’s neighbour.
Such behaviour can even be a substitute for right inter-
pretation, whether one is trying to make sense of the Bible
for oneself (1. 39. 43) or attempting to convey its meaning to
63
We must therefore disagree with Marrou, Saint Augustin et la fin de la
culture antique, 564, who, commenting on this passage in Doct. Chr., asks
rhetorically, ‘la scientia, don de l’Esprit, n’est-elle pas la grâce qui nous
permet de retirer un fruit spirituel de l’étude de l’Ecriture et aussi ce fruit
lui-même?’, as if ‘scientia’ could here be identical with ‘sapientia’. Note
that for ‘scientia’ as a ‘theological term’, Marrou quotes only Serm. 347. 2 f.;
Serm. Dom. Mont. 1. 4. 11–12; and Doct. Chr. 2. 7. 10—all places where
Augustine treats Isa. 11: 2 f., where the Latin version includes the term
‘scientia’. It is likely that, here as elsewhere (see Burton’s essay above),
Augustine was influenced by the biblical terminology; however, the con-
text makes it clear that ‘scientia’ is used as a synonym for ‘disciplina’ (on
which see refs. provided by Marrou, 562 no. 2, with Klinkenberg, ‘Artes
liberales’, col. 532). Likewise, not strictly correct is the observation in BA
11(2) ad loc., where it is claimed that Step 3, ‘scientia’, has a privileged
place in the sevenfold ascent to God. This is true, of course, inasmuch as
that is what Augustine wants to talk about in the treatise at hand (as stated
at 2. 8. 12: ‘sed nos ad tertium illum gradum considerationem referamus,
de quo disserere quod dominus suggesserit atque tractare instituimus’
(‘But let us now turn our attention to that third level, on which I proposed
to discuss and consider whatever ideas the Lord may have provided’)),
but it is not true from his overall theoretical point of view. All steps
are indispensable, since all are conditions for a successful ascent, but each
will be superseded in its turn, with the exception of the last. This is so
despite the fact that, naturally, for Augustine the ‘scientia divinarum
scripturarum’ (‘knowledge of the Divine Scriptures’) stands much above
the ‘scientia gentium’ (‘knowledge of the Gentiles’) (Doct. Chr. 2. 42. 63),
following 2 Tim. 4: 3, where the Christian ‘sana doctrina’ (‘wholesome
knowledge’) (singular!) is contrasted with the plurality of false ‘doctrinae’
of heterodox groups and others. See also Doct. Chr. 1. 39. 43, where
he quotes 1 Cor. 13: 8 about the eventual perishing of all prophecies,
languages, and knowledge (‘scientia’).
 Augustine’s Hermeneutics 231
others (4. 29. 61). The intellectual effort of interpretation is
thus subordinated to a wholly ethical perspective, which is
in turn superseded by the final step in the process of ascent,
the possession of wisdom and the vision of God. This vision
or fruition of God, though an act that fellow human beings
will be able to enjoy mutually, will not contain any ethical or
altruistic component.64

6. c o n c lu s i o n : a u n i v e r s a l d i s c i p l i n e
and its limits
The preceding analysis should have made clear not only how
in De Doctrina Christiana Augustine has endeavoured to
provide a systematic, scientifically based theory of biblical
interpretation from a Christian point of view, but also how
he affirms this as the only justifiable intellectual occupation
for a Christian. Other, ‘worldly’ disciplines are useful only in
so far as they help with the understanding of the Bible.
Augustine’s hermeneutical claim is that biblical inter-
pretation is the one true (Christian) discipline, comprising
all others and giving them a perspective. In its theoretical
comprehensiveness, this hermeneutics is a universal dis-
cipline. Not surprisingly, it is difficult to master. (Some of
its requirements, such as familiarity with Greek and Hebrew
texts of the Bible, are known to have exceeded even
Augustine’s own abilities.) We must also conclude that, for
Augustine, the phrase ‘universal discipline’ would be an
oxymoron. Indeed, in his setting of pragmatic, ethical,
dogmatic, theological, and eschatological restrictions on
the ‘meta-discipline’ of the Scriptures, he seems at times
remarkably prescient of modern debates on the limits of
scientific or disciplinary knowledge.
64
e.g., Doct. Chr. 1. 22. 20; 23. 22; 29. 30; cf. Civ. 22. 30.
 bibliography

Amsler, M., Etymology and Grammatical Discourse in Late

Antiquity and the Early Middle Ages (Amsterdam, 1989).
Arnold, D. W. H., and Bright, P. (eds.), De Doctrina Christiana:
A Classic of Western Culture (Notre Dame, Ind., and London,
1995).
Athanassiadi, P., ‘Dreams, Theurgy and Freelance Divination:
The Testimony of Iamblichus’, JRS 83 (1993), 115–30.
Babcock, W. S., ‘Caritas and Signification in De Doctrina Chris-
tiana 1–3’, in Arnold and Bright (eds.), De Doctrina Christiana:
A Classic of Western Culture, 145–63.
Banchich, T. M., ‘Eunapius in Athens’, Phoenix 50 (1996),
304–11.
Barnes, M. R., ‘The Arians of Book V, and the Genre of De
Trinitate’, JTS 44 (1993), 185–95.
Barnes, T. D., ‘Himerius and the Fourth Century’, Classical
Philology 82 (1987), 206–25.
Barthes, R., Critical Essays, trans. R. Howard (Evanston, Ill.,
1972).
—— A Lover’s Discourse (New York, 1978).
—— et al., Exégèse et herméneutique (Paris, 1971).
Bellissima, G., ‘Sant’ Agostino grammatico’, in Augustinus
Magister: Congrès international augustinien, Paris, 21–4 septembre
1954, 3 vols. (Paris, 1955), i. 35–42.
Bernardi, J., ‘Un Regard sur la vie étudiante à Athènes au milieu
du IVe siècle après Jésus-Christ’, Revue des Études Grecques 103
(1990), 79–94.
Bidez, J., Vie de Porphyre, le philosophe néo-platonicien (Hil-
desheim and New York, 1980).
Bö mer, F., Der lateinische Neuplatonismus und Neupythagoreismus
und Claudianus Mamertus in Sprache und Philosophie (Leipzig,
1936).
Børtnes, J., ‘Eros Transformed: Same-Sex Love and Divine
Desire’, in Hägg and Rousseau (eds.), Greek Biography and
Panegyric in Late Antiquity, 180–93.
Bouché -Leclercq, A., Histoire de la divination dans l’Antiquité, i
(Paris, 1879).
Bouhot, J.-P., ‘Augustin prédicateur d’après le De Doctrina
 Bibliography 233
Christiana’, in G. Madec (ed.), Augustin prédicateur (395–411):
Actes du colloque international de Chantilly (5–7 septembre 1996)
(Paris, 1998), 49–61.
Brachtendorf, J., ‘The Decline of Dialectic in Augustine’s
Early Dialogues’, StPatr 37 (2001), 25–30.
Bray, G., Biblical Interpretation Past and Present (Downers
Grove, Ill., 1996).
Bright, P. (ed.), Augustine and the Bible (Notre Dame, Ind.,
1999).
Brown, P. [R. L.], Augustine of Hippo: A Biography (London,
1967; 2nd edn., 2000).
— The World of Late Antiquity: From Marcus Aurelius to
Muhammad (London, 1971).
— Power and Persuasion in Late Antiquity: Towards a Christian
Empire (Madison, 1992).
—‘ A New Augustine’, New York Review of Books, 24 June 1999.
—‘ Introducing Robert Markus’, in ‘The Study of Augustine,
1950–2000: Evolving Disciplinary Contexts’, AugSt 32 (2001),
177–206, at 181–7.
Bruning, B., ‘De l’astrologie à la grâce’, in Collectanea Augustini-
ana: Mélanges T. J. Van Bavel, ii (Leuven, 1990), 575–643.
Bucher, T. G., ‘Zur formalen Logik bei Augustinus’, Freiburger
Zeitschrift für Philosophie und Theologie 29 (1982), 3–45.
Buchwald, F., ‘Musae Varronis’, Museum Helveticum 23 (1966),
215–17.
Burkert, W., ‘Cicero als Platoniker und Skeptiker’, Gymnasium
72 (1972), 175–200.
Burton, P., The Old Latin Gospels: A Study of their Texts and
Language (Oxford, 2000).
Butler, C., Western Mysticism, 2nd edn. (London, 1960).
Cameron, A., ‘Paganism and Literature in Late Fourth Century
Rome’, in M. Fuhrmann (ed.), Christianisme et formes littéraires
de l’Antiquité tardive en Occident (Geneva, 1977), 1–30.
Cameron, M., ‘The Christological Substructure of Augustine’s
Figurative Exegesis’, in Bright (ed.), Augustine and the Bible,
74–103.
Carson, A., Eros the Bittersweet (Princeton, 1986).
Cary, P., Augustine’s Invention of the Inner Self (Cambridge,
2000).
Cavallera, F., ‘Notes chronologiques et hagiographiques sur
quelques sermons de saint Augustin’, Bulletin de Littérature
Ecclésiastique 31 (1930), 21–30.
Chadwick, H., Saint Augustine: Confessions (Oxford, 1991).
Clark, E. A., Reading Renunciation: Asceticism and Scripture in
Early Christianity (Princeton, 1999).
234 Bibliography
Clark, G., ‘Translate into Greek: Porphyry of Tyre and the New
Barbarians’, in R. Miles (ed.), Constructing Identities in Late
Antiquity (London, 1999), 112–32.
Coleman, R. G. G., ‘The Formation of Specialized Vocabularies
in Philosophy, Grammar, and Rhetoric: Winners and Losers’, in
M. Lavency and D. Longrée (eds.), Actes du Ve Colloque de
Linguistique latine (Louvain, 1989), 77–89.
Collart, J., ‘Saint Augustin grammairien dans le De magistro’,
REAug 17 (1971), 279–92.
Courcelle, P., Les Lettres grecques en Occident de Macrobe à
Cassiodore, 2nd edn. (Paris, 1948) = Late Latin Writers and
Their Greek Sources, trans. H. E. Wedeck (Cambridge, Mass.,
1969).
— Les Confessions de saint Augustin dans la tradition littéraire:
Antécédents et posterité (Paris, 1963).
Cribiore, R., Gymnastics of the Mind: Greek Education in
Hellenistic and Roman Egypt (Princeton, 2001).
Cristante, L., Martiani Capellae De Nuptiis Philologiae et
Mercurii Liber IX (Padua, 1987).
Croatto, J. S., Biblical Hermeneutics: Toward a Theory of Read-
ing as Production of Meaning (Maryknoll, NY, 1987; original
Spanish edn., Buenos Aires, 1984).
Dahlmann, H., ‘Marcus Terentius Varro’, in RE suppl. vol. vi
(1935), 1172–277.
—‘ Silent diutius Musae Varronis quam solebant’, in J. Collart
(ed.), Varron: Grammaire antique et stylistique latine (Paris,
1978), 85–9.
D’Alessandro, P., ‘Agostino, Claudiano Mamerto, Cassiodoro e i
Disciplinarum libri di Varrone’, in MOUSA: Scritti in onore de
Giuseppe Morelli (Bologna, 1997), 357–70.
Danié lou, J., ‘Eunome l’Arien et l’exégèse néo-platonicienne du
Cratyle’, Revue des Études Grecques 69 (1956), 412–32.
Dawson, D., Allegorical Readers and Cultural Revision in Ancient
Alexandria (Berkeley and Los Angeles, 1992).
de Ghellinck, J., ‘Un Aspect de l’opposition entre hellénisme
et christianisme: L’attitude vis-à-vis de la dialectique dans les
débats trinitaires’, in Patristique et Moyen Age, iii (Brussels and
Paris, 1948), 247–310.
deGraff, T. B., ‘Plato in Cicero’, Classical Philology 35 (1940),
143–55.
Della Corte, F., Varrone: Il terzo gran lume romano (Genoa,
1954).
den Boeft, J., ‘Nullius Disciplinae Expers: Virgil’s Authority
in (Late) Antiquity’, in L. V. Rutgers et al. (eds.), The Use of
Sacred Books in the Ancient World (Leuven, 1998), 175–86.
 Bibliography 235
—— ‘Divinatio’, AL ii. 518–19.
— and van Ploo-van de Lisdonk, M. L. (eds.), The Impact
of Scripture in Early Christianity (Leiden, 1999).
de Rijk, L. M., ‘ “Enkyklios Paideia”: A Study of its Original
Meaning’, Vivarium 3 (1965), 24–93.
Doignon, J., ‘Problèmes textuels et modèles littéraires dans le
livre I du De ordine de saint Augustin’, REAug 24 (1978), 71–86.
Dolbeau, F., ‘Nouveaux Sermons d’Augustin pour la conversion
des païens et des Donatistes (III)’, REAug 38 (1992), 50–79.
Dulaey, M., ‘L’apprentissage de l’exégèse biblique par Augus-
tin’, REAug 48 (2002), 267–95; 49 (2003), 43–84.
Dü rig, W., ‘Disciplina: Eine Studie zum Bedeutungsumfang des
Wortes in der Sprache der Liturgie und der Väter’, Sacris
Erudiri 4 (1952), 245–79.
Dyroff, A., ‘Über Form und Begriffsgehalt der augustinischen
Schrift De Ordine’, in M. Grabmann and J. Mausbach (eds.),
Aurelius Augustinus: Festschrift der Görresgesellschaft zum 1500.
Todestage des Heiligen Augustinus (Cologne, 1930), 15–62.
Eden, K., ‘The Rhetorical Tradition and Augustinian Her-
meneutics in De Doctrina Christiana’, Rhetorica 8 (1990),
45–63.
— Hermeneutics and the Rhetorical Tradition: Chapters in the
Ancient Legacy and its Humanist Reception (New Haven and
London, 1997).
Eickelman, D. F., ‘The Art of Memory: Islamic Education and
its Social Reproduction’, Comparative Studies in Society and
History 20 (1978), 485–516.
Elsner, J., ‘The Itinerarium Burdigalense: Politics and Salvation
in the Geography of Constantine’s Empire’, JRS 90 (2000),
181–95.
Fedwick, P. J., ‘A Chronology of the Life and Works of Basil of
Caesarea’, in P. J. Fedwick (ed.), Basil of Caesarea: Christian,
Humanist, Ascetic (Toronto, 1981), i. 3–19.
Ferrari, L. C., ‘Augustine and Astrology’, Laval Théologique et
Philosophique 33 (1977), 241–51.
Fiedrowicz, M., Prinzipien der Schriftauslegung in der alten
Kirche (Bern, 1998).
Fischer, N., and Mayer, C. (eds.), Die Confessiones des Augusti-
nus von Hippo (Freiburg, 1998).
Forman, R. J., Augustine and the Making of a Christian Literature
(Lewiston, NY, 1995).
Frank, G., The Memory of the Eyes: Pilgrims to Living Saints in
Late Antiquity (Berkeley and Los Angeles, 2000).
Frantz, A., The Athenian Agora 24: Late Antiquity A.D. 267–700
(Princeton, 1988).
236 Bibliography
Frede, M., Die stoische Logik (Göttingen, 1974).
Froehlich, K. (trans. and ed.), Biblical Interpretation in the Early
Church (Philadelphia, 1984).
Fussl, M., ‘Disciplinae liberales (I–III)’, AL ii. 472–81.
Gadamer, H.-G., Hermeneutik, i: Wahrheit und Methode, 6th edn.
(Tübingen, 1990).
Gallay, P., La Vie de saint Grégoire de Nazianze (Lyon, 1944).
Gasparro, G. S., ‘Addas-Adimantus unus ex discipulis Man-
ichaei: For the History of Manichaeism in the West’, in R. E.
Emmerick, (ed.), Studia Manichaica, iv: Internationaler Kongreß
zum Manichäismus, Berlin, 14.–18. Juli 1997, Berichte und
Abhandlungen, Berlin-Brandenburgische Akademie der Wis-
senschaften, Sonderband 4 (Berlin, 2000), 546–59.
Gnilka, C., Chrēsis: Die Methode der Kirchenväter im Umgang mit
der Antiken Kultur, i: Der Begriff des ‘rechten Gebrauchs’ (Basel,
1984).
Goulet, R., ‘Prohérésius le païen et quelques remarques sur la
chronologie d’Eunape de Sardes’, Antiquité Tardive 8 (2000),
209–22.
Gorman, M., ‘The Diffusion of the Manuscripts of Saint
Augustine’s De doctrina christiana in the Early Middle Ages’,
Revue Bénédictine 95 (1985), 11–24; repr. in his The Manuscript
Traditions of the Works of St Augustine (Florence, 2001),
265–78.
Green, R. P. H., Augustine: De Doctrina Christiana (Oxford,
1995).
— Saint Augustine: On Christian Teaching (Oxford, 1997).
Hadot, I., Arts libéraux et philosophie dans la pensée antique (Paris,
1948).
Hadot, P., Marius Victorinus: Recherches sur sa vie et ses œuvres
(Paris, 1971).
Hagendahl, H., Augustine and the Latin Classics, 2 vols.
(Göteborg, 1967).
Hä gg, T., and Rousseau, P. (eds.), Greek Biography and Panegyric
in Late Antiquity (Berkeley and Los Angeles, 2000).
Halporn, J. W., and Vessey, M. (eds.), Cassiodorus: ‘Institutions
of Divine and Secular Learning’ and ‘On the Soul’ (Liverpool,
2004).
Harrison, C., Augustine: Christian Truth and Fractured
Humanity (Oxford, 2000).
—‘ The Rhetoric of Scripture and Preaching’, in R. Dodaro and
G. Lawless (eds.), Augustine and his Critics (London, 2000),
214–30.
Hedrick, C. W., jun., History and Silence: Purge and Rehabilita-
tion of Memory in Late Antiquity (Austin, Tex., 2000).
 Bibliography 237
Hill, E., Saint Augustine: Teaching Christianity (New York,
1996).
Hoffmann, M., Der Dialog bei den christlichen Schriftstellern der
ersten vier Jahrhunderte (Berlin, 1966).
Holte, R., ‘Monica, “the Philosopher”’ , Augustinus 39 (1994),
293–316.
Holzhausen, J., ‘Augustin as Biograph und Exeget: Zur liter-
arischen Einheit der Confessiones’, Gymnasium 107 (2000),
519–36.
Hombert, P.-M., Nouvelles Recherches de chronologie augustinienne
(Paris, 2000).
Horn, C., Augustinus (Munich, 1995).
Hü bner, W., Die Begriffe ‘Astrologie’ und ‘Astronomie’ in der
Antike, Akademie der Wissenschaften und der Literatur, Mainz,
Abhandlungen der Geistes- und Sozialwissenschaftlichen
Klasse, 1989, no. 7 (Stuttgart, 1990).
—‘ Die “artes liberales” im zweiten Buch von De Ordine’, in
Charisteria Augustiniana Iosepho Oroz Reta dicata (Madrid,
1994), 317–44.
—‘ Disciplinarum libri’, AL ii. 485–7.
Hü lser, K., Die Fragmente zur Dialektik der Stoiker, iv (Stuttgart
and Bad Cannstatt, 1988).
Inglebert, H., Interpretatio Christiana: Les Mutations des savoirs
(cosmographie, géographie, ethnographie, histoire) dans l’Antiquité
chrétienne 30–630 après J.-C. (Paris, 2001).
Irvine, M., The Making of Textual Culture: ‘Grammatica’ and
Literary Theory, 350–1100 (Cambridge, 1994).
Janka, M., Ovid ‘Ars Amatoria’, Buch 2: Kommentar (Heidelberg,
1997).
Jü ssen, J., and Schrimpf, G., ‘Disciplina, doctrina’, in Hi-
storisches Wörterbuch der Philosophie, i (Basel and Stuttgart,
1971), cols. 256–61.
Junod, É., ‘Basile de Césarée et Grégoire de Nazianze sont-ils les
compilateurs de la Philocalie d’Origène? Reexamen de la lettre
115 de Grégoire’, in Mémorial Jean Gribomont (Rome, 1988),
349–60.
Kaster, R. A., ‘Notes on “Primary” and “Secondary” Schools in
Late Antiquity’, Transactions of the American Philological
Association 113 (1983), 323–46.
— Guardians of Language: The Grammarian and Society in Late
Antiquity (Berkeley and Los Angeles, 1988).
Kevane, E., ‘Augustine’s De Doctrina Christiana: A Treatise on
Christian Education’, RechAug 4 (1966), 97–133.
Klingshirn, W. E., ‘The Figure of Albericius the Diviner in
Augustine’s Contra Academicos’, StPatr 38 (2001), 219–23.
238 Bibliography
Klinkenberg, H. M., ‘Artes liberales/artes mechanicae’, in
Historisches Wörterbuch der Philosophie, i (Basel and Stuttgart,
1971), cols. 531–5.
Kneale, W., and Kneale, M., The Development of Logic (Oxford,
1962).
Konstan, D. ‘How to Praise a Friend: St Gregory of Nazianzus’
Funeral Oration for Basil the Great’, in Hägg and Rousseau
(eds.), Greek Biography and Panegyric in Late Antiquity, 160–79.
Kü hnert, F., Allgemeinbildung und Fachbildung in der Antike
(Berlin, 1961).
Kursawe, B., Docere – Delectare – Movere: Die officia oratoris bei
Augustinus in Rhetorik und Gnadenlehre (Paderborn, 2000).
la Bonnardiè re, A.-M. (ed.), Saint Augustin et la Bible (Paris,
1986).
Lacan, J., Écrits, trans. A. Sheridan (New York, 1977).
Laks, A., and Most, G. (eds.), Studies on the Derveni Papyrus
(Oxford, 1997).
Lamirande, E., ‘Quand Monique, la mère d’Augustin, prend
la parole’, in A. Zumkeller (ed.), Signum Pietatis: Festgabe für
Cornelius Petrus Mayer OSA (Würzburg, 1989), 3–19.
Langslow, D. R., Medical Latin in the Roman Empire (Oxford,
2000).
Law, V., ‘St Augustine’s “De grammatica”: Lost or Found?’,
RechAug 19 (1984), 155–83.
Lenaz, L., Martiani Capellae De Nuptiis Philologiae et Mercurii
Liber Secundus (Padua, 1975).
Lepelley, C., Les Cités de l’Afrique romaine au Bas-Empire,
2 vols. (Paris, 1979–81).
Lerer, S., Boethius and the Dialogue: Literary Method in ‘The
Consolation of Philosophy’ (Princeton, 1985).
Lettieri, G., L’altro Agostino: Ermeneutica e retorica della grazia
dalla crisi alla metamorfosi del De Doctrina Christiana (Brescia,
2001).
Leyerle, B., ‘Landscape as Cartography in Early Christian
Pilgrimage Narratives’, Journal of the American Academy of
Religion 64 (1996), 119–43.
Lienhard, J., ‘Augustine on Dialectic: Defender and Defensive’,
StPatr 33(1997), 162–6.
Lim, R., Public Disputation, Power and Social Order in Late
Antiquity (Berkeley and Los Angeles, 1995).
Linderski, J., ‘The Augural Law’, ANRW 16.3 (1986), 2146–312.
Long, A. A., ‘Cicero’s Plato and Aristotle’, in J. G. F. Powell (ed.),
Cicero the Philosopher (Oxford, 1995), 37–61.
— and Sedley, D. N., The Hellenistic Philosophers, i (Cam-
bridge, 1987).
 Bibliography 239
Lü tcke, K.-H., Auctoritas bei Augustin, mit einer Einleitung zur
römischen Vorgeschichte des Begriffs (Stuttgart, 1968).
Lundin, R., Walhout, C., and Thiselton, A. C., The Promise
of Hermeneutics (Grand Rapids, Mich., 1999).
Lutz, C. E. (ed.), Remigii Autissiodorensis Commentum in Mar-
tianum Capellam, Libri III–IX (Leiden, 1965).
Lyons, J., Introduction to Theoretical Linguistics (Cambridge,
1968).
MacCormack, S., The Shadows of Poetry: Vergil in the Mind of
Augustine (Berkeley and Los Angeles, 1998).
McGuckin, J. M., Saint Gregory of Nazianzus: An Intellectual
Biography (Crestwood, NY, 2001).
McLynn, N. B., ‘A Self-Made Holy Man: The Case of Gregory
Nazianzen’, JECS 6 (1998), 463–83.
—‘ Gregory of Nazianzen’s Basil: The Literary Construction of
a Christian Friendship’, StPatr 37 (2001), 173–93.
McWilliam, J., ‘The Cassiciacum Autobiography’, StPatr 18/4
(1990), 14–43.
Markus, R. [A.], The End of Ancient Christianity (Cambridge,
1991).
— Signs and Meanings: World and Text in Ancient Christianity
(Liverpool, 1996).
Marrou, H.-I., ‘ “Doctrina” et “disciplina” dans la langue des
Pères de l’Église’, Archivum Latinitatis Medii Aevi (Bulletin du
Cange) 9 (1934), 5–25.
— Saint Augustin et la fin de la culture antique (Paris, 1938;
reissued with a ‘Retractatio’, 1949; 4th edn., 1958).
— Histoire de l’éducation dans l’antiquité (Paris, 1948; Engl.
trans., London, 1956).
— Décadence romaine ou antiquité tardive? IIIe–VIe siècle (Paris,
1977).
Mauch, O., Der lateinische Begriff ‘disciplina’: Eine Wortunter-
suchung (Freiburg, Switzerland, 1941).
Mayer, C. P., ‘Res per signa: Der Grundgedanke des Prologs in
Augustins Schrift De doctrina christiana und das Problem seiner
Datierung’, REAug 20 (1974), 100–12.
Miller, P. Cox, ‘ “Differential Networks”: Relics and Other
Fragments in Late Antiquity’, JECS 6 (1998), 113–38.
—‘ Strategies of Representation in Collective Biography: Con-
structing the Subject as Holy’, in Hägg and Rousseau (eds.),
Greek Biography and Panegyric in Late Antiquity, 209–54.
Moreau, M., Bochet, I., and Madec, G. (eds. and trans.), Saint
Augustin: La doctrine chrétienne, BA 11(2) (Paris, 1997).
Morgan, T., Literate Education in the Hellenistic and Roman
Worlds (Cambridge, 1998).
240 Bibliography
Neuschä fer, B., Origenes als Philologe, 2 vols. (Basel,
1987).
Norris, F. W., ‘Your Honor, My Reputation: St Gregory of
Nazianzus’ Funeral Oration for Basil the Great’, in Hägg and
Rousseau (eds.), Greek Biography and Panegyric in Late
Antiquity, 140–59.
Nussbaum, M., ‘Augustine and Dante on the Ascent of Love’, in
G. B. Matthews (ed.), The Augustinian Tradition (Berkeley and
Los Angeles, 1999), 61–90.
Nutton, V., ‘Ammianus and Alexandria’, Clio Medica 7 (1972),
165–72.
O’Connell, R. J., St Augustine’s Early Theory of Man (Cam-
bridge, Mass., 1968).
O’Daly, G., Augustine’s ‘City of God’: A Reader’s Guide (Oxford,
1999).
O’Donnell, J. J., ‘Paganus’, Classical Folia 31 (1977), 163–9.
—‘ The Demise of Paganism’, Traditio 35 (1979), 45–88.
— Augustine: Confessions, 3 vols. (Oxford, 1992).
O’Meara, J. J., ‘The Neoplatonism of Saint Augustine’, in
D. O’Meara (ed.), Neoplatonism and Christian Thought (Albany,
NY, 1982), 34–41.
Opelt, I., ‘Materialien zur Nachwirkung von Augustins Schrift
De doctrina christiana’, Jahrbuch für Antike und Christentum 17
(1974), 64–73.
Otte, K., Das Sprachverständnis bei Philo von Alexandrien:
Sprache als Mittel der Hermeneutik (Tübingen, 1968).
Pacioni, V., L’Unità teoretica del ‘de ordine’ di S. Agostino (Rome,
1996).
Parmentier, M., ‘The Gifts of the Spirit in Early Christianity’,
in den Boeft and van Ploo-van de Lisdonk (eds.), The Impact of
Scripture in Early Christianity, 58–78.
Penella, R. J., Greek Philosophers and Sophists in the Fourth
Century A.D. (Leeds, 1990).
Pé pin, J., Augustin et la dialectique (Villanova, Pa., 1976).
Petit, P., Les Étudiants de Libanius (Paris, 1955).
Pizzani, U., ‘L’Enciclopedia agostiniana e i suoi problemi’, in
Atti del Congresso internazionale su S. Agostino nel XVI cente-
nario della conversione, Roma 15–20 settembre 1986, Studia
Ephemeridis ‘Augustinianum’ 24 (Rome, 1987), 331–61.
Pollmann, K., ‘La Genesi dell’ermeneutica nell’Africa del secolo
IV’, in Cristianesimo et Specificità regionali nel Mediterraneo
Latino (sec. IV–VI): XXII Incontro di Studiosi dell’antichità
cristiana, Studia Ephemeridis ‘Augustinianum’ 46 (Rome,
1994), 137–45.
— Doctrina Christiana: Untersuchungen zu den Anfängen der
 Bibliography 241
christlichen Hermeneutik unter besonderer Berücksichtigung von
Augustinus, De doctrina christiana (Fribourg, 1996).
—‘ African and Universal Elements in the Hermeneutics
of Tyconius and Augustine’, in P.-Y. Fux, J.-M. Roessli, and
O. Wermelinger (eds.), Augustinus Afer—Saint Augustin:
Africanité et universalité, Actes du colloque international
Alger-Annaba, 1–7 April 2001, 2 vols. (Fribourg, 2003), ii.
353–62.
—‘ Doctrina christiana (De -)’, AL ii. 551–75.
Press, G. A., ‘The Subject and Structure of Augustine’s De Doct-
rina Christiana’, AugSt 11 (1980), 99–124.
—‘ The Content and Argument of Augustine’s De Doctrina
Christiana’, Augustiniana 31 (1981), 165–82.
—‘ Doctrina in Augustine’s De Doctrina Christiana’, Philosophy
and Rhetoric 17 (1984), 98–120.
Reynolds, L. D., and Wilson, N. G., Scribes and Scholars: A
Guide to the Transmission of Greek and Latin Literature (Oxford,
1968; 3rd edn., 1991).
Reynolds, S., Medieval Reading: Grammar, Rhetoric and the
Classical Text (Cambridge, 1996).
Ripanti, G., Agostino teoretico dell’interpretazione (Brescia, 1980).
Rist, J. M., Augustine: Ancient Thought Baptized (Cambridge,
1994).
Ritschl, F., ‘De M. Varronis disciplinarum libris commentarius’,
in Opuscula Philologica, iii (Leipzig, 1877), 352–402.
Rollinson, P., Classical Theories of Allegory and Christian
Culture (Pittsburgh, 1980).
Rousseau, P., Basil of Caesarea (Berkeley and Los Angeles,
1994).
Scanlon, M., ‘Augustine and Theology as Rhetoric’, AugSt 25
(1994), 37–50.
Schä ublin, C., Untersuchungen zu Methode und Herkunft der
Antiochenischen Exegese (Cologne, 1974).
—‘ De Doctrina Christiana: A Classic of Western Culture?’, in
Arnold and Bright (eds.), De Doctrina Christiana, 46–67.
Schindler, A., Wort und Analogie in Augustins Trinitätslehre
(Tübingen, 1965).
Schissel von Fleschenberg, O., Marinos von Neapolis und die
neuplatonischen Tugendgrade (Athens, 1928).
Schlapbach, K., Augustin: Contra Academicos (vel De Academi-
cis) Buch 1, Patristische Texte und Studien 58 (Berlin and New
York, 2003).
Schmaus, M., Die psychologische Trinitätslehre des Hl. Augustinus
(Münster, 1927).
Seeliger, H., ‘Aberglaube, Wissenschaft und die Rolle der his-
242 Bibliography
torica narratio in Augustins Doctrina Christiana’, Wissenschaft
und Weisheit 43 (1980), 148–55.
Shanzer, D., A Philosophical and Literary Commentary on
Martianus Capella’s De Nuptiis Philologiae et Mercurii Liber 1
(Berkeley and Los Angeles, 1986).
—‘ Licentius’s Verse Epistle to Augustine’, REAug 37 (1991),
110–43.
Sieben, H.-J., ‘Die “Res” der Bibel: Eine Analyse von Augusti-
nus, De doctrina christiana I–III’, REAug 21 (1975), 72–90.
Simon, M., ‘Zur Abhängigkeit spätrömischer Enzyklopädien
der Artes liberales von Varro’s Disciplinarum libri’, Philologus
110 (1966), 88–101.
Simonetti, M., Biblical Interpretation in the Early Church: An
Historical Introduction (Edinburgh, 1994; original Italian edn.,
Rome, 1981).
— Sant’ Agostino: L’istruzione cristiana (Verona, 1994).
Sluiter, I., ‘Metatexts and the Principle of Charity’, in
P. Schmitter and M. J. van der Wal (eds.), Metahistoriography:
Theoretical and Methodological Aspects in the History of
Linguistics (Münster, 1998), 11–27.
—‘ Communication, Eloquence and Entertainment in Augus-
tine’s De Doctrina Christiana’, in den Boeft and van Ploo-van de
Lisdonk (eds.), The Impact of Scripture in Early Christianity,
245–57.
Solignac, A., ‘Réminiscences plotiniennes et porphyriennes dans
le début du de ordine de saint Augustin’, Archives de Philosophie
20 (1957), 446–65.
Steenson, J. N., ‘The Date of the Philocalia’, in R. P. C. Hanson
and H. Crouzel (eds.), Origeniana Tertia (Rome, 1988), 245–52.
Steinhauser, K. B., ‘Codex Leningradensis Q.v.I.3: Some
Unresolved Problems’, in Arnold and Bright (eds.), De Doctrina
Christiana, 33–43.
Stock, B., Augustine the Reader: Meditation, Self-Knowledge, and
the Ethics of Interpretation (Cambridge, Mass., 1996).
Strauss, G., Schriftgebrauch, Schriftauslegung und Schriftbeweis
bei Augustin (Tübingen, 1959).
Stroumsa, G. G., ‘Milk and Meat: Augustine and the End of
Ancient Esotericism’, in A. and J. Assmann (eds.), Schleier und
Schwelle, i (Munich, 1977), 251–62.
Studer, B., Schola christiana (Paderborn, 1998).
Sundkler, B., and Stead, C., A History of the Church in Africa
(Cambridge, 2000).
Teske, R., ‘St Augustine and the Vision of God’, in Van Fleteren
et al. (eds.), Augustine: Mystic and Mystagogue, 287–99.
Testard, M., Saint Augustin et Cicéron (Paris, 1958).
 Bibliography 243
Theiler, W., ‘Porphyrios und Augustin’, in Forschungen zum
Neuplatonismus (Berlin, 1966), 160–251.
Thiselton, A. C., The Two Horizons: New Testament Herme-
neutics and Philosophical Description with Special Reference to
Bultmann, Gadamer, and Wittgenstein (Grand Rapids, Mich.,
1980).
Van Fleteren, F., ‘Augustine, Neoplatonism, and the Liberal
Arts’, in Arnold and Bright (eds.), De Doctrina Christiana,
14–24.
— and Schnaubelt, J. C. (eds.), Augustine: Biblical Exegete
(New York, 2001).
— et al. (eds.), Augustine: Mystic and Mystagogue (New York,
1994).
van Lierde, C., ‘The Teaching of St Augustine on the Gifts of
the Holy Spirit from the Text of Isaiah 11: 2–3’, in Van Fleteren
et al. (eds.), Augustine: Mystic and Mystagogue, 5–110.
Verbraken, P.-P., Études critiques sur les sermons authentiques
de saint Augustin, Instrumenta Patristica 12 (The Hague,
1976).
Verheijen, L. M. J., ‘Le De Doctrina Christiana de Saint Augus-
tin: Un manuel d’herméneutique et d’expression chrétienne
avec, en II.19.29–42.63, une charte fondamentale pour une
culture chrétienne’, Augustiniana 24 (1974), 10–20.
Vessey, M., ‘Conference and Confession: Literary Pragmatics in
Augustine’s “Apologia contra Hieronymum”’ , JECS 1 (1993),
175–213.
—‘ The Demise of the Christian Writer and the Remaking
of “Late Antiquity”: From H.-I. Marrou’s Saint Augustine
(1938) to Peter Brown’s Holy Man (1983)’, JECS 6 (1998),
377–411.
— Pollmann, K., and Fitzgerald, A. D. (eds.), History,
Apocalypse, and the Secular Imagination: New Essays on Augus-
tine’s ‘City of God’ (Bowling Green, Oh., 1999).
Veyne, P. (ed.), Histoire de la vie Privée (Paris, 1985).
Voss, B. R., Der Dialog in der früchristlichen Literatur, Studia et
Testimonia Antiqua 9 (Munich, 1970).
Vö ssing, K., Schule und Bildung im Nordafrika der Römischen
Kaiserzeit (Brussels, 1997).
Wickham, L. R., ‘The Syntagmation of Aetius the Anomean’,
JTS 19 (1968), 532–69.
Winkler, K., ‘La Théorie augustinienne de la mémoire à son
point de départ’, in Augustinus Magister: Congrès international
augustinien, Paris, 21–4 septembre 1954, 3 vols. (Paris, 1955),
i. 511–19.
Wright, D. F., ‘Augustine: His Exegesis and Hermeneutics’, in
244 Bibliography
M. Saebø (ed.), Hebrew Bible/Old Testament: The History of its
Interpretation, i, pt. 1 (Göttingen, 1996), 701–30.
Young, F. Biblical Exegesis and the Formation of Christian Culture
(Cambridge, 1997).
Zlatar, Z., The Epic Cycle: Allegoresis and the Western Epic
Tradition from Homer to Tasso (Sydney, 1993).
 index locorum

1. BIBLICAL Colossians
2: 3: 138
Genesis 2: 8–9: 192
2: 5: 225 2: 8: 157
Job 2 Timothy
1: 6: 114 4: 3: 230 n. 63
Psalms
110 (111): 10: 227 n. 53
147: 158 2. OTHER
147: 5: 156
Proverbs Aetius
1: 2: 123 Syntagmation (ed. Wickham)
Wisdom 537f.: 193 n. 24
11: 21: 156 546: 195 n. 30–1
Sirach Alanus de Insulis (Alain de Lille)
1: 16: 227 n. 53 Anticlaudianus
Isaiah 5. 258: 103 n. 125
1: 18: 153, 161 Ambrose
11: 2f: 230 n. 63 Explanatio Psalmorum xii
11: 2–3: 227, 227 n. 54 36: 16: 191
11: 2: 228 36: 64: 124 n. 39
53: 12: 157 43: 2: 123 n. 37
Matthew De Fide
5: 3–10: 227 n. 54 5. 41: 192
5: 3: 225 5. 42: 193
5: 4–9: 226 De Iacob et Vita Beata
5: 8: 114, 226 n. 50 1. 1. 4: 123 n. 39
5: 9: 226 n. 51 1. 5. 17: 123 n. 39
5: 17: 227 De Ioseph
10: 20: 156 13. 75: 123 n. 39
Mark De Isaac vel Anima
15: 28: 157 1. 2: 123 n. 39
Romans 3. 9: 123 n. 39
10: 17: 103 n. 125 In Lucam
1 Corinthians 5. 50–2: 228 n. 55
1: 24: 140, 158 In Psalmos
1: 30: 158 118: 5. 39: 227 n. 53
4: 20: 193 Ammianus Marcellinus
8: 1: 213 21. 1. 7: 119
13: 8: 224, 225, 230 n. 63 21. 1. 8: 119
15: 199–200 21. 1. 10: 119 n. 24
2 Corinthians 22. 16. 17: 119
3: 6: 146 22. 16. 18: 45 n. 59
Ephesians 23. 6. 33: 119
3: 19: 213 29. 1. 31: 119
246 Index Locorum
Ammonius 3. 18. 37: 151
In Aristotelis de Interpretatione Beata Vita
17. 2–8: 221 2. 7: 222 n. 38
Anonymus de Constitutione Mundi Civ.
PL 90.908C: 88 n. 75 5. 7: 121 n. 32
Anthologia Graeca 6. 2: 123 n. 38
8. 91: 46 n. 61 6. 5: 159
8. 100: 46 n. 60 8. 16: 121 n. 32
Apuleius 10. 9: 138
De Dogmate Platonis 10. 10: 138
1. 3: 118 n. 22 10. 27: 138
Florida 10. 29: 76 n. 19, 221 n. 35
20: 96 n. 107 11. 1: 15
Metamorphoses 18. 52: 148 n. 18
2. 20: 124 22. 5: 152
3. 16: 124 22. 8: 45 n. 59
3. 18: 124 22. 30: 231 n. 64
4. 9: 124 Conf.
4. 18: 124 1. 1. 1: 158
9. 29: 125 1. 8. 13: 145
Aristotle 1. 12. 19: 156
Analytica Priora 1. 13. 22: 145, 162
70a. 7–9: 222 n. 37 1. 13. 23: 145
Categories 1. 18. 29ff: 146 n. 12
105. 7–20: 199 n. 37 1. 18. 29: 136, 145, 146
De Interpretatione 2. 3. 5: 148
1. 16a: 211 3. 1. 1: 26 n. 3, 27
Metaphysics 3. 2. 2–4: 27
A1: 103 n. 125 3. 3. 5–6: 30
Protrepticus 3. 3. 5: 27, 28 n. 10
Fr. 17: 103 n. 125 3. 3. 6: 27, 146 n. 12
Rhetorica ad Alexandrum 3. 4. 7–8: 157
12. 1430b: 222 n. 37 3. 4. 7: 27, 150
Arnobius 3. 4. 8: 29 n. 13, 157, 158
Adversus Nationes 3. 5. 9–6. 10: 27
1. 5: 119 n. 25 3. 5. 9: 39 n. 44
2. 6: 122 n. 35 3. 7. 12: 39 n. 43
2. 12: 122 n. 35 3. 11. 19: 40
2. 71: 87 n. 69 3. 11. 20: 40 n. 47
3. 10: 86 n. 68 4. 1. 1: 40 n. 47, 159 n. 37
3. 38: 89 n. 80, 106 n. 135 4. 2. 2: 145, 149, 159
5. 18: 87 n. 69 4. 2. 3–3. 5: 41
6. 1: 86 4. 2. 3: 115
6. 11: 87 n. 69 4. 3. 5: 116, 136
6. 22: 87 n. 69 4. 3. 6: 42, 136
Augustine 4. 4. 2: 41
C. Acad. 4. 4. 7: 41, 41 n. 48
1. 1. 4: 51 n. 6 4. 4. 9: 162
1. 6. 18: 125 4. 6. 11: 41
1. 6. 20: 125 4. 7: 108 n. 146
1. 6. 21: 125 4. 8. 13: 41
1. 7. 21: 125–6, 127 n. 48 4. 13. 21: 37
2. 1. 1: 138 4. 14. 21: 151, 158
2. 6. 14: 58 n. 15 4. 16. 28: 27
 Index Locorum 247
4. 16. 30: 150, 153 10. 43. 70: 138
4. 17. 30: 71 n. 6 11. 2. 2: 15
4. 19. 30: 155 11. 2. 4: 15
5. 3. 3: 136 11. 9. 11: 155
5. 3. 5: 156, 158 11. 12. 14: 162
5. 4. 7: 156 n. 32 12. 4. 6: 162
5. 6. 10: 40 n. 47, 161 12. 10. 10: 155
5. 6. 11: 136, 146 n. 13, 160 n. 39 12. 26. 36: 150
5. 7. 12–13: 40 12. 31. 42: 146
5. 8. 14: 35, 37 13. 18. 23: 158
5. 9. 17: 154, 159 13. 19. 24: 153, 161
5. 12. 22: 149 n. 19 13. 35. 50–38. 53: 218 n. 30
5. 23. 13: 151 C. Cresc.
6. 4. 6: 146 1. 14. 17: 152
6. 5. 7: 163 1. 19. 23: 154 n. 30
6. 6. 9: 161, 162 C. Faust.
6. 7. 11–12: 35 22. 27: 222 n. 38
6. 7. 11: 42 n. 50 22. 54: 217 n. 28
6. 7. 12: 32 Contra Iulianum
6. 9. 14: 35 3. 2. 7: 152
6. 9. 15: 35 Disc. Chr.
6. 12. 20: 158 1: 218 n. 31
7. 6. 8–10: 136 Div. Qu.
7. 6. 9: 116, 136 35. 2: 213 n. 18
7. 6. 10: 136 Doct. Chr.
8. 2. 3: 36 n. 36 praef. 1: 209
8. 5. 10: 148 praef. 4: 146 n. 14
9. 2. 2: 159 praef. 8: 228 n. 56
9. 2. 3–4: 108 n. 146 praef. 9: 169 n. 8, 209–10
9. 3. 6: 162 1. 1. 1: 149 n. 19, 170, 214
9. 4. 7: 159 1. 2. 2: 170, 172, 223 n. 41
9. 6. 14–7. 15: 109 n. 152 1. 4. 4: 170, 171 n. 12
9. 6. 14: 71 n. 6 1. 5. 5: 170, 171 n. 12
9. 9. 19: 161 1. 10. 10: 170
9. 10. 23: 161 1. 13. 12: 216
9. 10. 24: 137, 229 n. 62 1. 22. 20: 231 n. 64
9. 10. 25: 137, 229 n. 62 1. 23. 22: 231 n. 64
9. 11. 28: 161 1. 26. 27: 222 n. 38
10. 8. 16: 15 1. 29. 30: 231 n. 64
10. 9. 16: 148, 153 1. 35. 39: 171 n. 12
10. 12. 19: 155, 163 n. 42 1. 36. 40–1: 172
10. 16. 24: 155 1. 37. 41–38. 42: 172
10. 20. 29: 163 n. 42 1. 38. 42: 212
10. 21. 30: 149 1. 39. 43: 224, 225, 230, 230 n. 63
10. 27. 38: 171 n. 14 2. 1. 1: 221
10. 32. 49–50: 71 n. 6 2. 2. 3: 223 n. 43
10. 32. 50–38. 53: 218 n. 30 2. 3. 4: 172
10. 33. 49: 109 n. 152 2. 7. 9–11: 225
10. 33. 50: 150 n. 21 2. 7. 9: 20, 98 n. 117
10. 34. 53: 15 2. 7. 10: 230 n. 63
10. 35. 56: 15 2. 8. 12: 230 n. 63
10. 42. 67–43. 70: 15 2. 11. 16: 129
10. 42. 67: 138 2. 13. 20: 129
10. 42. 68: 138 2. 16. 23–24: 129
248
Augustine (cont.)
2. 16. 23: 129
2. 16. 25: 129
2. 16. 26–17. 27: 129
2. 16. 26: 105, 129
2. 16. 27: 104
2. 17. 27: 79 n. 38
2. 18. 28: 130
2. 19–42: 177, 178 n. 37
2. 19. 29: 130, 197
2. 20. 30–24. 37: 130
2. 20. 30: 130, 131, 132, 133
2. 21. 31–22. 34: 131
2. 21. 32: 134
2. 22. 33: 134
2. 23. 35–24. 37: 131
2. 23. 35: 131, 131 n. 56
2. 23. 36: 130, 131
2. 24. 37: 131
2. 25. 38–26. 40: 132
2. 25. 38: 132, 134
2. 25. 39–40: 132
2. 25. 39: 107 n. 142
2. 26. 40: 132
2. 27. 41–30. 47: 132
2. 27. 41: 139
2. 28. 44: 133, 133 n. 58
2. 29. 45: 130, 132, 133, 134
2. 29. 46: 73 n. 9, 133, 134
2. 30. 47: 134
2. 31. 48–38. 56: 132, 135
2. 31. 48–35. 53: 135, 152 n. 25
2. 31. 48ff: 73 n. 9
2. 31. 48: 135, 198 n. 35
2. 31. 49: 73 n. 9, 199
2. 32. 49: 201
2. 32. 50: 135, 197–8
2. 35. 53: 201, 204
2. 36. 54–37. 55: 135
2. 37. 55: 149 n. 19, 221 n. 35, 225
2. 38. 56: 135, 156 n. 33
2. 38. 57: 129, 135, 229 n. 60
2. 39. 58–9: 135
2. 39. 58: 178
2. 39. 59: 175, 178
2. 40. 60: 19, 157 n. 35, 175
2. 40. 61: 177, 179
2. 41. 62: 213 n. 21
2. 42. 63: 129, 135, 179–80, 230 n. 63
3. 1. 1: 169 n. 5, 180
3. 2. 2: 176 n. 31
3. 3. 6: 142 n. 2
3. 3. 7: 219
3. 6. 11: 107 n. 142
Index Locorum
3. 7. 11: 224
3. 8. 12: 224
3. 9. 13: 224
3. 10. 16: 171 n. 12
3. 25. 35–6: 13 n. 28
3. 29. 40: 73 n. 9, 145 n. 11
4. 1. 2: 216
4. 2. 3: 216
4. 3. 5: 73 n. 9
4. 10. 24: 219
4. 10. 25: 142 n. 2
4. 15. 32: 228 n. 56
4. 18. 37: 219
4. 20. 41: 129 n. 50
4. 24. 53: 219–20
4. 29. 61: 231
4. 60: 149 n. 19
En. Ps.
11: 7: 227 n. 54
Ep.
3: 222 n. 38
21: 16 n. 29
27: 62 n. 27
101: 72 n. 7, 109 n. 153
101. 2: 221 n. 35
101. 3: 81 n. 40
171A: 227 n. 54
Gen. adv. Man.
2. 5. 6: 225
Imm. An.
1. 1: 124 n. 43, 139
Lib. Arb.
1. 1. 2: 124 n. 41
Ord.
1. 1. 3: 110 n. 154
1. 2. 5: 51 n. 6
1. 3. 6: 106 n. 134
1. 3. 9: 126
1. 4. 10: 127
1. 5. 14: 126, 127
1. 6. 15: 127
1. 8. 24: 106 n. 134, 124 n. 40
2. 2. 5: 221
2. 2. 6–7: 54 n. 10
2. 2. 35–44: 221
2. 4. 13–5. 14: 72
2. 7. 45: 78 n. 34
2. 8. 25: 59 n. 22, 187
2. 9. 26: 188
2. 9. 27: 128
2. 10. 29: 49 n. 1
2. 11. 30: 77, 188
2. 11. 31: 64 n. 30, 76, 76 n. 19, 77,
188
 Index Locorum 249
2. 11. 32: 98 1. 3. 10: 226 n. 49, n. 51
2. 12. 35–15. 43: 76 1. 4. 11–12: 230 n. 63
2. 12. 35–13. 38: 44 n. 57 1. 4. 11: 226 nn. 49–50, 227 n. 53,
2. 12. 35ff: 73 229 n. 61
2. 12. 35: 6 n. 11, 78–79, 79 n. 35, Sol.
98, 101, 147 n. 15 1.1: 107 n. 144, 110 n. 157
2. 12. 36–15. 43: 72 1. 1. 1: 51 n. 7
2. 12. 36: 77, 77 n. 24, 107 1. 9. 3: 108 n. 148
2. 12. 37: 79 2. 7. 14: 160 n. 38
2. 13. 38: 77, 190, 221 2. 11. 19–20: 108
2. 14. 39ff: 97 n. 110 2. 11. 20–1: 124 n. 42
2. 14. 39–41: 103 n. 130 2. 11. 20: 58 n. 17, 59 n. 18
2. 14. 39: 76 n. 21, 78, 79, 79 n. 38, 2. 25. 1: 81 n. 41
97 n. 113, 103 n. 129, 105 n. 133, 2. 32. 8: 108 n. 148
107 n. 140 2. 35. 1: 108 n. 150
2. 14. 41: 78, 103 n. 125, 105 In Iohannis Evangelium Tractatus
2. 15. 42: 91 n. 84, 107 n. 140, 108 CXXIV
n. 147, 128 24: 2: 211–12
2. 15. 43: 103 n. 128, 107 n. 140 Trin.
2. 16. 45: 108 5. 6. 7: 203, 204 n. 45
2. 18. 47–48: 230 5. 7. 8: 204 n. 46
2. 18. 47: 70 n.5, 187, 221 Util. Cred.
2. 20. 53: 97 n. 110, 107 n. 140 7. 16: 25 n. 1
2. 20. 54: 6 n. 11, 79, 79 n. 35, 79 n. 17. 35: 117 n. 13
38 Ver. Rel.
Quant. An. 72: 229 n. 62
33. 70–6: 98 n. 117 Augustine (?)
33. 72: 72 De Rhetorica
Retr. 11: 207 n. 4
1. 3: 189 Ausonius
1. 3. 2: 105, 128 Griphus Ternarii Numeri
1. 3. 4: 221 n. 35 ep.: 96 n. 107
1. 4. 4: 108 n. 151 30–3: 96 n. 107
1. 5. 2: 139 Basil of Caesarea
1. 5. 6: 70 n. 3, 81 n. 40, 110 n. 156 Ep. 14: 47 n. 64
1. 6: 6 n. 10, 71, 71 n. 6, 72, 83 n. 51, Caesarius of Arles
141 Sermo 99: 181 n. 45
1. 11: 15 n. 29 Callimachus
1. 13. 14: 15 Aetia prol. 11–12: 107 n. 138
1. 22. 1: 113 n. 2 Fr. 398: 107 n. 138
2. 4: 13 n. 28 Cassiodorus
Serm. Institutiones
12. 1: 114 n. 4, 114 n. 5 1. 4. 1–3: 207 n. 3
12. 4: 114 n. 6 1. 7. 3: 211 n. 13
12. 12: 140 1. 8. 1–21: 207 n. 3
150. 5: 222 n. 38 2. 3. 2: 99 n. 120, 99 n. 121
347. 2f: 227 n. 54, 228 n. 55, 230 n. 2. 3. 22: 7 n. 17
63 2. 7. 2: 93 n. 93, 102, 118 n. 20
347. 2: 227 n. 53, 229 n. 61 2. 7. 3, 4: 100 n. 122
347. 3: 226 n. 49 Censorinus
Sermo de Oboedientia 10. 3: 150 n. 21
5: 27 n. 8, 28 n. 10 Charisius
Serm. Dom. Mont. Ars Grammatica
1: 227 n. 54 4. 4: 173 n. 21
250
Cicero
Academica
1. 1–2: 111
1. 2: 85
1. 3: 111 n. 158
1. 3. 9: 87 n. 72
1. 3. 12: 81 n. 44
1. 7. 25: 143 n. 5
2. 7. 21: 64 n. 30
De Amicitia
1. 24. 38: 154 n. 29
Disputationes Tusculanae
1. 5: 155 n. 31
De Divinatione
1. 1. 2: 118
1. 2. 3: 118 n. 14
1. 6. 12: 116 n. 9
1. 19. 36: 118
1. 41. 91: 118 n. 14
1. 50. 112: 134 n. 60
1. 52. 118: 131 n. 54
2. 23. 50: 118 n. 14
2. 42. 88ff: 133 n. 59
2. 42. 88: 118 n. 19
Epistulae ad Atticum
4. 16. 2: 111
13. 12. 3: 111, 111 n. 161
13. 19. 3: 81 n. 44
Epistulae ad Familiares
1. 9. 23: 87 n. 72
6. 6. 7: 118 n. 18
16. 10. 2: 87 n. 72
De Fato
28: 142 n. 2
De Finibus
3. 2. 5: 143 n. 16
5. 1–5: 36 n. 34
5. 29. 87: 156 n. 34
5. 49: 87 n. 72
De Inventione
1. 5: 149 n. 20
1. 6–7: 147 n. 16
1. 39. 48: 222 n. 37
2. 116–54: 207 n. 4
De Natura Deorum
1. 11: 113
2. 1. 3: 163
2. 10: 118 n. 16
De Officiis
1. 37. 134–5: 153 n. 29
1. 42. 150f: 213 n. 20
Orator
32. 113–4: 153 n. 26
62: 87 n. 72
Index Locorum
De Oratore
1. [42.] 187: 188
1. 167: 149 n. 20
1. 245: 147 n. 16
De Re Publica
1. 10. 16: 156 n. 34
3. 2. 3: 77 n. 28
De Senectute
49. 6: 155 n. 31
Topica
54–7: 196 n. 33
Claudianus Mamertus
De Statu Animae
2. 8: 82 n. 46, 94 n. 98, 100
2. 9: 83, 94 n. 98
Ep. ad Sapaudum: 91 n. 84, 100
Clement of Alexandria
Protrepticus
4. 46: 86 n. 65
Digest
1. 3: 207
50. 17: 207
Diogenes Laertius
2. 79: 90 n. 82
9. 12: 84 n. 59
Donatus
Ars Grammatica
3. 4. 5: 173 n. 21
Ennius
Annales
1. 113: 173 n. 21
Eunapius
Vitae Sophistarum
483–5: 34 n. 26
483: 32 n. 19, 35 n. 30
486: 35 n. 30
487: 33 n. 23, 33 n. 25
492: 36 n. 36
493: 35 n. 30, 43 n. 53
502–3: 43 n. 53
502: 46 n. 62
Eunomius
Apologia
12. 8–12: 194
Festus
De Verborum Significatu
s. v. ‘Tages’: 117 n. 11
Firmicus Maternus
De Errore Profanarum Religionum
18. 1: 125
Mathesis
2. 30. 14: 119 n. 25
5, praef. 1: 123 n. 36
5. 1. 16: 120 n. 27
 Index Locorum 251
5. 1. 17: 120 n. 27 17: 32 n. 19
Fulgentius 18: 32 n. 19
Mythologiae 26: 32 n. 19
1. 15: 95 27: 32 n. 19
Gellius, Aulus 44: 32 n. 21
Noctes Atticae 54. 2: 32 n. 20
praef. 5: 85 n. 6 59–60: 32 n. 19
1. 25. 17: 85 65: 34 n. 27
10. 1. 6: 99 n. 120 69. 7–9: 32 n. 20
18. 15. 2: 99 n. 120 69. 8: 32 n. 19
Gregory of Nazianzus Hilary of Poitiers
De Vita Sua In Psalmos 118: 5. 16: 227 n. 53
128–9: 38 n. 41 Horace
211–64: 30–1 Ars Poetica
212–13: 31 n. 16 361: 223
242–4: 40 Odes
Epistulae 1. 28. 1–2: 155 n. 31
1: 47 n. 63 3. 21. 9–10: 154 n. 29
2: 47 n. 63 Irenaeus of Lyon
4–6: 47 n. 64 Adversus Haereses
5: 31 n. 17 3. 24. 1: 227 n. 54
30: 42 n. 51 Demonstratio
34: 44 n. 56 9: 227 n. 54
115: 37 Isidore of Seville
192. 3: 34 n. 28 Origines
233: 36 n. 34 1. 2: 69 n. 1
235: 36 n. 34 2. 23: 99 n. 121
Orationes Jerome
7. 6–8: 45 Comm. in Epistulam Pauli ad
7. 6: 35 n. 31, 38 n. 41, 46 n. 61 Galatas
7. 9: 46 n. 60 2. 11: 36 n. 36
21: 42 n. 51 Chronicon
43. 14–24: 28–9 s. a. 353: 36 n. 36
43. 17: 33, 34 n. 29 Epistulae
43. 20: 33, 36 n. 34 99. 2: 124 n. 39
43. 21: 30, 39, 42 n. 51 100. 10: 124 n. 39
43. 22: 30, 32–3, 33 n. 24 Libanius
43. 23: 44, 44 n. 57 Epistulae
43. 24: 40 278: 36 n. 36
43. 50: 29 n. 13 390. 5: 33 n. 23
43. 70: 29 n. 13 390. 6–7: 33 n. 25
Gregory of Tours 715: 34 n. 27
Decem Libri Historiarum Orationes
10. 18: 91 n. 85 1. 19: 34 n. 26
Gregory the Great 14. 35: 35 n. 32
Homiliae in Ezechielem Licentius
1. 5. 13: 217 n. 28 Carmen ad Augustinum
2. 2. 4: 217 n. 28 1–15: 80
Himerius 2: 81
Orationes Lucian
2: 36 n. 34 Herodotus
9: 32 n. 21 1: 85 n. 60
10: 40 n. 45 Lucretius
14: 32 n. 19 1. 657: 84 n. 59
252 Index Locorum
Macrobius In Lucam
Somnium Scipionis 1: 213 n. 19
1. 15. 12: 120 In Numeros Homiliae
1. 6. 44: 120 27: 228 n. 55
Saturnalia Philocalia
1, praef. 8: 174 n. 26 1–15: 39 n. 43
1. 24: 119 n. 23 6: 39
1. 24. 9–25: 174 n. 25 12–15: 39
Martianus Capella 26: 39 n. 43
De Nuptiis Philologiae et Mercurii Ovid
(ed. Kopp) Tristia
1. 28: 98 n. 114 2. 273: 151 n. 23
1. 36: 121 Palladius
2. 112: 90 n. 83 Historia Lausiaca
2. 118: 90 n. 83 64: 38 n. 38
2. 120: 90 n. 83 Paulinus of Nola
2. 124–5: 122 Ep. 8: 62 n. 27
2. 126: 97 n. 109 Petronius
2. 138: 90 n. 83 Satyricon
2. 220: 95 n. 103 88: 149–150, 158 n. 36
3. 221–2: 95 n. 104 Philostratus
3. 225–7: 174 n. 27 Vita Apollonii
3. 229: 79 n. 37, 101, 147 n. 15 1. 20: 106 n. 137
3. 230: 174 n. 27 Plato
3. 263: 174 n. 27 Phaedrus
3. 326: 93 n. 90, 174 n. 27 244C–E: 116 n. 9
3. 335: 92 n. 86 266b: 77
4. 335: 93 n. 92, 93 n.95, 93 n. 96, 101 Philebus
5. 426: 92 n. 87 18B6–D12: 77
5. 565: 92 n. 88 Republic
6. 578: 93 n. 95, 101 7. 516c: 81 n. 41
6. 639: 93 n. 95, 101 7. 516e: 81 n. 41
6. 662: 93 n. 95, 101 7. 522c: 81 n. 42
7. 729: 93 n. 89 7. 525d: 81 n. 42
8. 817: 93 n. 93, 93 n. 95, 102 7. 526e: 81 n. 42
9. 891: 96 n. 105, 98 n. 116, 102 7. 527b: 81 n. 42
9. 892: 121 7. 529a: 81 n. 42
9. 893: 121, 122 7. 531c: 81 n. 42
9. 894–6: 121 7. 531d: 81 n. 42
9. 898: 122 7. 532a: 81 n. 42
9. 928: 93 n. 95, 102 7. 532e: 81 n. 43
9. 998: 96 n. 106 7. 533d: 76, 81 n. 42
Maximus of Tyre 7. 539e–540a: 98 n. 115
10. 9: 94 n. 99 Sophistes
Nonius Marcellus 242d: 84 n. 59
135. 9–11: 97 n. 112 Pliny the Elder
135. 10: 99 n. 120 Naturalis Historia
551. 13: 97 n. 112 29. 4. 65: 85 n. 64
551. 15: 99 n. 120 35. 76: 143 n. 7
Olympiodorus Plotinus
Fr. 28: 34 n. 29 Enneads
Origen 1. 4. 4: 56 n. 13
In Leviticum Homiliae 1. 6: 78
9. 2: 124 n. 39 3. 7. 5: 56 n. 12
 Index Locorum 253
4. 4. 6: 55 n. 11 10. 75: 119
5. 1. 2: 187 n. 10 Servius Danielis
5. 3. 13: 229 n. 62 In Aen.
6. 7. 35: 217 n. 28, 228 n. 58 1. 2: 119 n. 24
6. 9. 10: 228 n. 57 1. 305: 120 n. 29
6. 9. 11: 217 n. 28, 228, 229 n. 62 1. 422: 119 n. 24
Plutarch 2. 692–3: 119 n. 23
Quaestiones Conviviales 3. 60: 119 n. 23
746J: 90 n. 83 3. 90: 119 n. 23
Porphyry 3. 359: 119 n. 23
Isagoge 4. 453: 119 n. 23
1. 5–6: 77 4. 662: 119 n. 23
Vita Pythagorae 5. 530: 119 n. 23
46–7: 77 n. 23, 78 8. 314: 120 n. 29
Possidius Sidonius Apollinaris
Indiculum operum Augustini Carmina
iv. 29–33: 113 n. 3 2. 90: 94 n. 97
Priscian 14, praef.: 94 nn. 97–8
Partitiones 23. 151: 94 n. 97
1. 9, 2. 44, 3. 67, etc.: 169 n. 4 Epistulae
7. 141: 173 n. 21 2. 9. 5: 84 n. 54, 94 n. 97
Proclus 4. 3: 89 n. 78, 100
In Prim. Euclid. Elem. Libr. Comm. 4. 3. 1: 94 n. 97
24. 21–7: 78 4. 3. 5: 89 n. 79
Pseudo-Plutarch 4. 10: 94 n. 97
De Liberis Educandis 5. 2: 88, 99, 100
7D (10): 90 n. 82 8. 6. 18: 94 n. 97
Quintilian Socrates
Institutio Oratoria Historia Ecclesiastica
1. 4. 1–3: 207 n. 3 4. 26. 6: 31 n. 18
1. 4–5: 158 n. 36 Stobaeus
1. 7. 3: 211 n. 13 3: 90 n. 82
1. 8. 1–21: 207 n. 3 Suetonius
2. 14. 1–4: 147 n. 16 De Grammaticis
2. 16. 2: 149 n. 20 6. 2: 85
9. 2. 31: 153 n. 28 Syncellus, Georgius
Remigius of Auxerre Chron. 382: 38 n. 38
Commentum in Martianum Capellam, Synesius
Libri III–IX (ed. Lutz) Epistulae
298: 121 n. 32 54: 36 n. 35
Rhetorica ad Herennium 136: 36 n. 35
4. 52. 65: 153 n. 28 Tertullian
Sallust Apologeticum
Iugurtha 22. 9: 139 n. 76
30. 4: 151 n. 22 Varro
Sappho De Lingua Latina
Fr. 31: 171, 182 5. 1: 112
Seneca 6. 62: 218 n. 31
Ep. 88. 20: 147 n. 15 Virgil
Servius Aeneid
In Aen. 3. 88–9: 127
1. 733: 119 n. 24 10. 875: 126
4. 166: 119 n. 24 Vita Caesarii Arelatensis
5. 530: 119 n .23 1. 9: 123 n. 38
254 Index Locorum
Vitruvius 7, praef. 14: 84 n. 57, 99 n. 121
De Architectura 9. 2. 1: 118 n. 22
1. 1. 16: 118 n. 22 9. 6. 2: 118 n. 22
 general index
Items relating directly to Augustine are indexed under his name, in
three categories: (1) career and intellectual development, (2) liter-
ary works (in chronological order), (3) statements on disciplinary
knowledge.
Abelard 185
Aetius 191, 193–6
Alexandria, medical schools 45
Albicerius, diviner 115, 125
allegorical interpretation 207, 224
Alypius 32, 35
Ambrose of Milan 19, 110, 123, 146,
151, 160, 191–3, 196, 198, 199, 201
Ammianus Marcellinus 119
Ammonius 216, 221
Anomians, see Arians
Antiochus of Ascalon 77, 81
Apollonius of Tyana 106
Apostolic Creed 214, 220
Apuleius 124
architecture 69, 84, 88 n. 76, 96–9
Arians 186, 191–6, 202–4, 190
Aristippus 90
Ariston of Chios 90
Aristotle, Peripatetic tradition 27,
152, 195, 202, 202, 211, 216, 221,
223
arithmetic, mathematics 103, 156–7
Arnobius of Sicca 86, 89, 122
astrology, astronomy 70, 93, 118
Athens, educational centre 24–45
Auctor ad Herennium 153
auctoritas 162, 188
Aulus Gellius 85
Aurelius Opilius 85
Augustine:
career and intellectual development:
student at Madaura 148; student
at Carthage 25–48; reads Cicero’s
Hortensius 37, 157; Manichee
27, 37, 39–40, 52, 142, 146, 160;
teacher at Thagaste 40–3; at
Milan 48, 49; conversion 49,
141; at Cassiciacum 6, 25–6,
47–8, 49–65; Platonism,
Neoplatonism 52–65, 75–81,
137–9, 142, 145, 148, 150, 171,
186–9, 228–9; develops
incarnational theology 64–5,
138–40; ordained presbyter 15;
ordained bishop 113; composes
hermeneutical treatise 209
literary works (in chronological
order): De Pulchro et Apto 151;
Cassiciacum dialogues 5, 50–1,
61, 124–8, 186; Contra
Academicos 115, 125–6, 127,
151–2; De Ordine 6, 52–65, 71–2,
73–5, 76–81, 97–8, 105–6, 107–8,
126–8, 184–6, 187–90, 221;
Soliloquia 51–2, 107–8, 142, 160;
De Immortalitate Animae 139;
‘disciplinarum libri’ 6, 9, 47–8,
70–2, 141–2; De Dialectica 222;
De Quantitate Animae 71–2; De
Genesi contra Manichaeos 225; De
Musica 71–2, 150 n. 21; De
Magistro 142, 185; De Vera
Religione 15; De Utilitate
Credendi 15; Contra Adimantum
113; De Doctrina Christiana 3, 7,
9–11, 13, 15, 18, 72–3, 75, 79,
104–5, 106, 110, 128–35, 138, 139,
142–3, 167–83, 184–6, 197–202,
204, 208–31; survey of ‘liberal arts’
in Book 2: 5, 9, 128–35, 177–80,
197; Confessiones 13, 15, 18, 109,
115, 117, 135–8, 141–64, 27–8;
Contra Cresconium 152; Ep. 101
ad Memorium 72, 109; De
Trinitate 202–4; In Iohannis
Evangelium Tractatus 211, 223;
De Civitate Dei 17–18, 109, 116,
256 General Index
Augustine (cont.)
138, 152, 209; Contra Iulianum
184; Retractationes 5–6, 71–2, 73,
141, 143, 189, 209; Opus
Imperfectum contra Iulianum 184
statements on disciplinary
knowledge: on agriculture 134;
on arithmetic, mathematics 129,
135, 156–7; on ascent to
Truth/God 6, 71, 81, 97–8,
107–8, 137, 186–91, 221, 225–31;
on astrology, astronomy 108, 128,
131, 133–4, 135, 136; on biblical
scientia 129, 226; see also tractare,
tractatio; on bibliomancy 137; on
canon of liberal arts 72, 76, 98, 141,
see also Augustine: career and
intellectual development:
‘disciplinarum libri’; on canon of
mantic arts 121 n. 32; on love
(amor, caritas) and interpretation
170–2, 212–14, 220, 225; on
dialectic, logic 108, 117, 135,
151–5, 160–2, 184–205, 221; on
divination 113–40; on
doctrina(e) 128–35, 178, 180,
215, 225, 229; on dreams,
oneiromancy 137; on
eschatological criterion of
interpretation 225; on fine arts
132; on geometry 108, 135,
155–6; on grammar 107–8, 117,
145–8, 167–83; on history 133; on
manufacturing arts 134; on
medicine 132, 134; on memory
53–65, 149; on Muses 104–9,
129–30; on music 105, 109, 129,
150; on natural history 133; on
nautical art 134; on performance
arts 134; on pictura versus
litterae 211–12, 223; on
philosophy, sapientia 157–8,
187–9, 226–7, 229–30; on
rhetoric 108, 135, 149–51; on rule
of faith 214; on signs 129, 130,
132, 168–72, 215, 218, 221–4, 230;
on theurgy 138
Barthes, R. 173
Basil of Caesarea 28–31, 32–7, 37–40,
40–47
Boethius 8, 110
Bömer, F. 82
Brown, P. 1–3
Buchwald, W. 85, 86
Caesarea in Palestine, rhetorical schools
of 47
Caesarius, brother of Gregory of
Nazianzus 45–6
Caesarius of Arles 123
Carson, A. 171, 182
Cassiodorus 7, 69, 73, 144
Charisius 173
Chrysippus 83
Cicero, Ciceronianism 37, 50, 78, 81 n.
44, 85–7, 89, 111–12, 113, 116,
117–18, 134, 143, 147, 149, 155,
156, 157, 163, 222
Claudianus Mamertus 8, 82–3, 88–9,
91–94, 97, 99, 100
Clement of Alexandria 86
colloqui, colloquium 161
Cresconius the Donatist 154, 158
Cyprian 161, 177, 179
D’Alessandro, P. 82
Dahlmann H. 86–7
daimones, demons 116, 139
dialectic, logic 151–5, 160–2, 184–205,
221
διαλ&γοµαι, διάλογο 153–4, 160
Diomedes, grammarian 169–70
dimensio figurarum 155
disciplina(e) 58–60, 116, 117–24,
135–8, 143, 218, 229, 230 n.
disputare, disputatio 152–3, 160
disserere 153
divination 113–40
doctrina(e) 128–35, 178, 180,
229–30
Donatus, grammarian 170, 173, 176
Dyroff, A. 6 n. 12, 73, 75
Eden, K. 223
eloquentia 149–50, 159
encyclopaedia(s), encyclopaedism 70,
73, 84
enkyklios paideia 2, 5 n. 9
Ennius 173
Eunapius of Sardis 31–3, 37, 43, 46
Eunomius 191, 193–6
facundia 150–1
Faustus the Manichee 40, 136, 146,
160, 161
Firmicus Maternus 120, 122, 125
Fulgentius of Ruspe 95
 General Index 257
Gadamer, H.-G. 211 Macrobius 8, 120, 174
geometry 155–6, 213 Mani, Manichees, see Augustine: career
grammar 145–8, 167–83, 206–7, and intellectual development:
210–11 Manichee; Faustus the Manichee
Gregory of Nazianzus 25–48 mantic arts or disciplines 121–3
Gregory the Wonderworker 39 Marius Victorinus 148
Markus, R.A. 168
Hagendahl, H. 73 Marrou, H.-I. 2–13, 26, 184
Hadot, I. 6–8, 73–88, 99, 109, 132 Martianus Capella 7, 8, 69, 73, 75,
Heraclitus 84 90–7, 97–8, 120–2, 132, 143,
Hermagoras 207 174
‘hermeneutical circle’ 211 mathematics, see arithmetic
hermeneutics 11, 38, 206–31 Maximus of Tyre 94
Hierius 37, 151 medicine 44–5, 69, 96–9, 213
Himerius 31–4, 37, 40 memory 53–65, 149
Homer 207, 208 metrics 90
Horace 223 Monica 137, 154, 159, 160, 161
Muses 84–8, 88–97, 122
interrogare, interrogatio 142, 162 music 150
Isidore of Seville 8, 69
Nebridius 42, 136, 162
Jerome 11 n. 24 Neoplatonism 6–8, 74–5, 76, 80–2,
Julian, emperor 119, 148 137–8, see also Plato, Platonism
Julian of Eclanum 152 numerare, numerus 156–7
Julianus Pomerius 123
Oomes, F. 85
Kant, I. 205 oratoria 147–8
Kühnert, F. 97 Origen 37–40, 207, 213
Ovid 151
Lactantius 179
Late Antiquity 2 paideusis 44–5
Libanius 33–5 Pépin, J. 184, 198
Licentius 62 n. 27 Peripatetic tradition, see Aristotle
Carmen ad Augustinum 72, 80–1, 97 Petronius 149
liberal arts or disciplines 9 philosophy 52 n. 9, 157–8
contrasted or combined with the Plato, Platonism 77, 78–82, 107, 108,
Bible, biblical culture 3–4, 10–12, 116, 138, 151, 211, 213, see also
110, 143, 162 Neoplatonism; Augustine: career
listed or enumerated 6, 44, 45, 81, and intellectual development:
84, 88–9, 91, 93, 96 n. 107, 100, Platonism, Neoplatonism
101–2, 143, 158 n. 36, see also Pliny the Elder 143
Augustine: statements on Plotinus 55, 56, 78, 228
disciplinary knowledge; trivium, Pompeius, grammarian 176
quadrivium postmodernism 2
medieval canon 5 n. 9, 6, 8–9, 69, Porphyry 7, 76–8
74, 110 Praetextatus, Vettius Agorius 120
personified 88–90, 90–7, 98, 107, Priscian 169 n. 4, 173, 174, 176, 211
110, 174, see also Muses Proclus 78
litterae, litteratura 146–8, 210–12, 223, Prohairesius 31–4
224 Providence 162
loan-words 144 Pythagoras, Pythagoreanism 73, 75,
logic, see dialectic 78, 79 n. 38
λ"γο 216–17
loquacitas 160 quadrivium, see trivium
258 General Index
Quintilian 147, 149, 153, 211 Theiler, W. 82
Theophrastus 216–17, 218
ratio, reason 63–4, 78, 98, 106, 107, tractare, tractatio 168–9, 172 n. 16,
188–9 176, 209, 214
rhetoric 149–51, 207 Trinitarian theology 191–6, 202–4,
Ritschl, F. 5 n. 9, 69, 73–4, 75–88, 190, 220
98–9 trivium, quadrivium 5 n. 9, 98–9, 103
Tyconius 11 n. 24, 208, 209, 210, 216,
sapientia 123, 157–8, 226–7, 229–30 220
scientia 122–3, 129, 226, 228–30
Sallust 150–1 Varro, Marcus Terentius 7, 69–70,
Seneca the Younger 147 130, 147
sermo, sermocinari, sermocinatio 153–5 De Grammatica 79
Servius 119–20, 170, 176 De Lingua Latina 85, 111–12
Sextus Empiricus 222 Disciplinarum Libri 6–7, 69–112,
Sidonius Apollinaris 83–4, 88–9, 91, 118; also known as Musae 84–8
93, 95, 99, 100 Vindicianus 136
Simplicianus of Milan 148 Virgil 120, 174, 176, 207, 224
Simplicius 199 Vitruvius 84, 143
Stoicism 199, 202, 213, 216, 220, 221
Suetonius 85 Zeno 83
Synesius 36