Ammianus Marcellinus and the Representation of Historical Reality 0801435269, 9780801435263

Citation preview

T h e T o w n sen d Lectures The Department o f Classics at Cornell University is fortunate to have at its disposal the Prescott W. Townsend Fund— established by Mr. Townsends widow, Daphne Townsend, in 1982. Since 1985, income from the fund has been used to support the annual visit o f a distinguished scholar in the field o f classics. Each visiting scholar delivers a series o f lectures, which, revised for book publication, are published by Cornell University Press in Cornell Stud­ ies in Classical Philology. During the semester o f their residence, Townsend lecturers effectively be­ come members o f the Cornell Department o f Classics and teach a course to Cornell students as well as deliver the lectures. The Townsend Lectures bring to Cornell University, and to Cornell Uni­ versity Press, scholars o f international reputation who are in the forefront o f current classical research and whose work represents the kind o f close reading o f texts that has become associated with current literary discourse, or reflects broad interdisciplinary concerns, or both.

C O R N E L L STU D IES IN CLA SSIC A L P H ILO LO G Y E D IT E D B Y

FR E D E R IC K M. AH L * K E V IN C L IN T O N JO H N E. C O L E M A N * JU D IT H R. G IN SB U R G G. M. KIRKW O O D * DAVID M A N K IN G O R D O N M. M ESSIN G * A L A N J. N U SSBA U M H A YD EN PE LLIC C IA * PIETRO P U C C I JE F F R E Y S. R Ü ST E N * D AN U TA SH A N Z E R VO LU M E LVI Ammianus Marcellinus and the Representation of Historical Reality by Timothy D. Barnes

A L S O IN T H E T O W N S E N D L E C T U R E S

Art ifccs o f Eternity: Horace's Fourth Book o f Odes by M ichael C . J. Putnam

Socrates, Ironist and Moral Philosopher by G rego ry Vlastos

Culture and National Identity in Republican Rome by E rich S. Gruen

Horace and the Dialectic o f Freedom: Readings in

Epistles i ”

by W . R. Johnson

Anim al Minds and Human Morals: The Origins o f the Western Debate by Richard Sorabji

Platonic Ethics, O ld and N ew by Julia Annas

Ammianus Marcellinus and the Representation o f Historical Reality Timothy D. Barnes

Cornell University Press Ithaca and London

Copyright © 1998 by Cornell University All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, Sage House, 5 1 2 East State Street, Ithaca, N e w York 14850. First published 1998 by Cornell University Press Printed in the United States o f America Cornell University Press strives to use environmentally responsible suppliers and materials to the fullest extent possible in the publishing o f its books. Such materials include vegetable-based, lo w -V O C inks and acid-free papers that are recycled, totally chlorine-free, or partly composed o f non wood fibers. Library o f Congress Cataloging-in-Publication Data Barnes, Tim othy David, Ammianus Marcellinus and the representation o f historical reality / Tim othy D . Barnes, p.

cm. —

(Cornell studies in classical philology ; v. 56. Th e Townsend lectures)

Includes bibliographical references and index. ISBN 0 -8 0 14 -35 2 6 -9

i. 2.

(cloth : alk. paper)

Ammianus Marcellinus.

Rërum gestarum libri.

Rom e— History— Empire, 2 8 4 - 4 7 6 — Historiography. 3.

Historiography— Rome.

I. Title.

II. Series.

DG316.B37 1998 937'.7,202 — dc2i 98-19791 Cloth printing

1

0

9

8

7

6

5

4

3

2 1

CONTENTS

Preface Editions, Translations, and Commentaries Abbreviations I II III IV V VI VII VIII IX X XI X II X III X IV XV

The Impartial Historian Reality and Its Representation Symmetry and Structure Narrative and Excursus Dating, Emphasis, and Omission Origin and Social Status The Greek Template Christian Language and Anti-Christian Polemic Things Seen and Things Read Enemies, Animals, and Stereotypes Empresses and Eunuchs Tyranny and Incompetence The N ew Achilles Past, Present, and Future Tacitus, Ammianus, and Macaulay

vii ix xi I

11 20 32 43 54 65 79 95 107 120 129 143 166 187

A P P E N D IC E S

1 2 3 4

The Text o f Ammianus The Structure o f Livy s Ab Urbe Condita Ammianus’ References to the Lost Books Consular Dates in the Res Gestae [v]

201 209 213 218

C ontents

5 6 7 8 9 10

Formal Excursus Ammianus’ Use o f Accentual Clausulae Corrupt and Mistaken Dates Missing Praefecti Urbis Maximinus and the Trials at Rome under Valentinian The Movements o f Valens

222 225 231 237 241 247

Bibliography Indexes Names o f Persons and Places Modern Scholars Passages o f Ammianus Discussed

255

[vi]

279 287 289

PREFACE

The present book is a twice rewritten and greatly expanded version o f the seven Townsend Lectures that I composed and delivered at Cornell University in the Fall Term o f 1994. I am most grateful to Prescott and Diana Townsend for their generosity in leaving the bulk o f their estate to support an endan­ gered academic discipline, to the Cornell Department o f Classics for inviting me to spend a term with them, and to all those who made my time in Ithaca so very pleasurable, especially Carol Kaske and Danuta Shanzer. M y serious interest in Ammianus Marcellinus goes back to 1986, when I made the surprising discovery that, virtually without exception, those who had written about the historian since Otto Seeck in 1906 had misunderstood both the chronological principles on which Ammianus arranges his narrative and the literary structure o f the Res Gestae. M y interest was further kindled as I came to realize how much the Roman Empire o f Ammianus differed from the mid-fourth-century world that my researches into the career o f Athana­ sius were revealing to me— a discrepancy that clamored for explanation. Since I started to think about Ammianus, both graduate and undergraduate students in Toronto have helped me constantly by compelling me to clarify my ideas in class; and I am extremely grateful to Paul Burton, Angela Kalinowski, Ron Pathen, and Sarah Pothecary for allowing me to use essays on Am­ mianus that they wrote for me and to Rodney Ast, Gordon Nixon, Michael Redies, and Aara Suksi for practical assistance. I owe more than I can easily express in words to my colleagues in the Toronto Department o f Classics not only for letting me pester them with questions arising from a text in which hardly any o f them has a real interest, but also, more generally, for collectively creating a lively academic environment in which it is a delight to teach and [vii]

P reface

to conduct research. I know that the others will excuse me i f I single out for mention Alexander Jones, who gave me lucid guidance on astronomical mat­ ters. I am also grateful to Rob Prichard and Paul Gooch for teaching me how autocratic and bureaucratic structures o f power normally react to incompe­ tence and corruption. It was only after I had formulated most o f my conclusions about Ammia­ nus’ treatment o f Christians and Christianity that I became aware o f the work o f David Woods, whose acute researches into the imperial bodyguard un­ der Julian have produced results convergent with mine. I am most grateful to Dr. Woods for his generosity in allowing me to read several o f his articles be­ fore publication: they have rescued me from uncritical acceptance o f received views on a number o f important points. I must also record with gratitude the generous financial support given me by the Social Sciences and Humanities Research Council o f Canada. The preparation o f the final text has led me to reflect on some important lessons that I learned before I became a pupil o f Ronald Syme in 1964. The first serious book o f history that I can remember reading (in 1956) was the first edition o f Alan Bullock’s remarkable Hitler, A Study in Tyranny (London, 1952): the passage o f time has powerfully confirmed the sureness and accu­ racy o f Bullock’s historical judgments and established his book as a model o f how a historian can write objectively about men and events for which he feels the deepest repugnance. Bullock’s Hitler and Stalin: Parallel Lives (London, 1991) imparts the same lesson— and reading it has reinforced my conviction that Ammianus failed in his obligation as a historian to strive to transcend per­ sonal bias. When I was an undergraduate at Oxford, I studied Plautus and attended Eduard Fraenkel’s seminars. From Fraenkel I learned both to be attentive to the Greek elements in any Latin writer and to respect the classical scholarship o f Wilhelmine Germany. Hence, when I began to read Ammianus closely and discovered that Ulrich von Wilamowitz-Moellendorff, Friedrich Leo, Eduard Norden, and Fraenkel himself all believed that Ammianus thought in Greek rather than Latin, I took the idea seriously— and soon saw that it provides the interpretative key to a difficult author. T im o t h y B a r n e s

Toronto 31 August 1997

[viii]

EDITIONS, TRANSLATIONS, AND COMMENTARIES

References to Ammianus Marcellinus are normally given as bare numbers in parenthesis in the main text (e.g., 14 .1.1). The text o f Ammianus used or quoted is that o f the Teubner edition by W. Seyfarth (Leipzig, 1978), unless a deviation is explicitly signaled. The following editions, translations, and com­ mentaries are cited by the name o f the editor(s), translator(s), and/or com­ mentator^) alone when it is clear which passage is under discussion and which edition or commentary is intended. 1. H e n ri de Valois (Valesius). A m m iani Marcellini Rerum Gestarum qui de X X X I supersunt

libri X V I I I . Paris, 1 6 3 6 . (N o tes are reprinted in the vario ru m edition o f j . A . W a g ­ n er and K . G . A Erfiirdt. L e ip zig, 18 0 8 ). 2. C . U . C la rk . A m m ia n i M arcellini Rerum Gestarum libri qui supersunt. B erlin , 1 9 1 0 , 1 9 1 5 . 3. T h e L o e b edition b y j . C . R o lfe. Am m ianus Marcellinus. L o n d o n , 1 9 3 5 , 19 4 0 , 19 3 9 . 4. T h e B u d é edition , A m m ien M arcellin: Histoire , o f w h ic h the fo llo w in g volum es have been available to m e: E . G a lletier and J . Fontain e. T o m e 1: Livres X I V —X V I . Paris, 19 6 8 . G . Sabbah. T o m e 2: Livres X V I I —X I X . Paris, 19 7 0 . J.

Fontain e. T o m e 3 : Livres X X —X X I I . Paris, 19 9 6 .

J . Fo n tain e. T o m e 4: Livres X I I I - X X V . Paris, 1 9 7 7 .

M .-A . M arié. Tome 5 : Livres X X V I - X X V I I I . Paris, 1984 . 5. P. de Jo n g e . Sprachlicher und historischer Kommentar z u Am m ianus Marcellinus X I V 1 - 7 . G ro n in g e n , 1 9 3 5 . -------- . Sprachlicher und historischer Kommentar z u Am m ianus Marcellinus X IV , 2. Hälfte (c. 7 - 1 1 ) . G ro n in g e n , 1 9 3 9 . -------- . Philological and Historical Com m entary on Am m ianus Marcellinus X V , G ro n in g e n , 19 4 8 .

[ix]

1-3 .

E d i t i o n s , T r a n s l a t i o n s , and C o m m e n t a r i e s -------- . Philological and Historical Com m entary on Am m ianus Marcellinus X V , 6 —13 . G ro n in g e n , 1 9 5 3 . -------- . Philological and Historical Com m entary on Am m ianus Marcellinus X V I . G r o n in ­ gen , 1 9 7 2 . -------- . Philological and Historical Comm entary on Am m ianus Marcellinus X V I I . G r o n in ­ gen , 1 9 7 7 . -------- . Philological and Historical Comm entary on A m m ianus Marcellinus X V I I I . G r o n in ­ gen , 19 80 . -------- . Philological and Historical Comm entary on Am m ianus Marcellinus X I X . G r o n in ­ gen , 19 8 2 . J . den B o eft, D . den H en gst, and H . C . Teitler. Philological and Historical Com m entary

on Am m ianus Marcellinus X X . G ro n in g e n , 19 8 7 . -------- . Philological and Historical Comm entary on Am m ianus Marcellinus X X I . G r o n in ­ gen , 1 9 9 1 . J. den B o eft, J. W . D rijvers, D . den H en gst, and H . C . Teitler. Philological and H istori­

cal Commentary on Am m ianus Marcellinus X X I I . G ro n in g e n , 1 9 9 5 . (T h e three m ost recent volum es are cited co llectively as “ the D u tc h co m m en tato rs.” ) 6. W . Se y fa rth ’s edition w ith a G e rm a n translation and b r ie f co m m e n ta ry : A m m ianus

Marcellinus: Römische Geschichte. Schriften und Quellen der alten Welt. B e rlin : 1, 19 6 8 ; 2, 19 6 8 ; 3 , 19 7 0 ; 4, 1 9 7 1 . 7 . J . Szidat. Historischer Kommentar z u Am m ianus Marcellinus, Buch X X —X X I . Teil I: D ie

Erhebung Ju lia n s; Teil II: D ie Verhandlungsphase; Teil I I I: D ie Konfrontation. Historia Einzelschriften 3 1 , 38 , 89. W iesb ad en , 1 9 7 7 , 1 9 8 1 , 19 9 6 . 8. T h e P en gu in translation b y W . H a m ilto n and A . W a lla ce -H a d rill. A m m ianus M ar­

cellinus: The Later Rom an Em pire (a . d . 3 5 4 -3 7 8 ) . H a rm o n d sw o rth , 19 8 6 .

The English translations o f Ammianus are my own. For passages o f narra­ tive I have freely adapted the version by Hamilton and Wallace-Hadrill, mod­ ifying it wherever I thought that I could recapture Ammianus’ tone or precise meaning more faithfully. For passages omitted by Hamilton and WallaceHadrill, I have consulted Rolfe s translation in preparing my own.

[x]

ABBREVIATIONS

The names o f ancient authors other than Ammianus are normally given in full: for the abbreviations used for the titles o f their works and for the standard col­ lections o f inscriptions and papyri, the following works o f reference should be consulted. H o m b lo w e r, S ., and A . Sp a w fo rth , eds. Tltc O xford Classical D ictionary 3 . O x fo r d , 19 9 6 , x x ix -liv . H . G . L id d ell and R . Sco tt. A G reek—English Lexicon9, rev. H . S . Jo n e s. O x fo rd , 19 4 0 , x v i-x lii. W . H . L a m p e. A Patristic G reek Lexicon. O x fo rd , 1 9 6 1 , ix - x li i i .

G.

The full titles o f periodicals and serials that are cited in the notes by conven­ tional abbreviations are given in the bibliography, but the following books (plus a recently reedited chronicle) are always cited with abbreviated titles. A lfö ld i, Conflict ( 19 5 2 ) A . A lfö ld i, A Conflict o f Ideas in the Late Roman Em pire. The Clash between the Senate

and Valentinian /, trans. H . M attin g ly. O x fo rd , 1 9 5 2 . Athanasius ( 19 9 3 ) T . D . B arn es, Athanasius and Constantius. Theology and Politics in the Constantinian

Em pire. C a m b r id g e , M ass., 19 9 3 . B itter, Kampfschilderungen (19 7 6 ) N.

B itter, Kampfschilderungen bei Am m ianus Marcellinus. Diss. E rla n g e n -N ü rn b e rg ,

1 9 7 5 ; p ub l. B o n n 19 7 6 . B lo ck le y, A m m ianus ( 1 9 7 5 ) R.

(P.) C . B lo ck le y , A m m ianus Marcellinus. A Study o f H is Historiography and Political

Thought. C o lle c tio n Lato m u s 1 4 1 . Brussels, 1 9 7 5 .

[xi]

A bbreviations

Blomgren, Quaestiones (1937) S. Blomgren, D e Sermone A m m iani M arcellini quaestiones

variae.

Uppsala Universitets

Ârsskrift 1937: 6 . Uppsala, 1937. Chastagnol, Fastes (1962) A. Chastagnol, L es Fastes de la Préfecture de Rom e au Bas-Em pire. Études Prosopographiques 2. Paris, 1962. Cognitio Gestorum (1992) J. den Boeft, D. den Hengst, and H. C. Teitler, eds., Cognitio Gestorum. T h e H isto­ riographie A rt o f Am m ianus Marcellinus. Koninklijke Nederlandse Akademie van Wetenschappen: Verhandelingen, Afd. Letterkunde, Nieuwe Reeks 148. Amsterdam, 1992. (1981) T. D. Barnes, Constantine

Constantine

and Eusebius.

Cambridge, Mass., 1981.

(1987) R. A. Bagnall, Alan Cameron, S. Schein, and K. Worp. Consuls o f the Later Rom an Em pire. Atlanta, 1987. Demandt, Z eitkritik (1965) A. Demandt, Zeitkritik und Geschichtsbild im Werk Am m ians. Diss. Marburg, 1963; publ. Bonn, 1965. Consuls

Descr. cons.

R. W. Burgess, ed.,

T he Chronicle o f H ydatius and the Consularia Constantinopoli-

tana. Two Contemporary Accounts o f the F in a l Years o f the Rom an Em pire.

Oxford, 1993,

215-45. Burgess’ text supersedes the standard edition by Mommsen, T. Chronica M inora. M onumenta Germ aniae Historica 1, Auctores Antiquissimi 9. Berlin, 1892, 205—47. References are given to the relevant consular year, and where there is more than one entry for the year, to the number assigned to it by Mommsen and Burgess (e.g., 378-3). Drexler, Am m ianstudien (1974) H. Drexler, Am m ianstudien. Spudasmata: Studien zur Klassischen Philologie und ihren Grenzgebieten, ed. H. Hommel and E. Zinn. Hildesheim / New York, 1974Elliott, Am m ianus (1983)

T. G. Elliott,

Am m ianus Marcellinus and Fourth C entury H istory.

Sarasota/Toronto,

1983Ensslin, Am m ianus (1923) W. Ensslin, Z u r Geschichtschreibung Klio 16. Leipzig, 1923. H istory and Historians (1984)

und Weltanschauung des A m m ianus Marcellinus.

History and Historians in Late A n tiqu ity ,

B. Croke and A. M. Emmett, eds. Sydney,

1983[xii]

A bbreviations Jo n e s,

LRE

A . H . M . Jo n e s,

The Later Roman Empire, 284-602.

O x fo rd , 19 6 4 .

In the A m e ric a n reprint in tw o volum es, pages 1 0 7 1 - 1 5 1 8 correspond to 3 . 1 - 4 4 8 in this edition .

Julian Apostata ( 19 7 8 ) R . K le in , ed ., Julian Apostata. Wege der Forschung 509. D arm stadt, M a tth e w s, Ammianus (19 8 9 ) J . F. M a tth e w s, The Roman Empire of Ammianus. L o n d o n , 19 8 9 .

19 7 8 .

M a tth e w s, Aristocracies ( 1 9 7 5 ) J . F. M a tth e w s,

Western Aristocracies and Imperial Court, a .d . 364-425.

O x fo rd , 1 9 7 5 :

p h o to gra p h ic reprint w ith postscript, 19 9 0 .

New Empire ( 19 8 2 ) The New Empire ofDiocletian and Constantine. C a m b rid g e , M ass., 19 8 2 . Städte und Steuern (19 8 6 ) E . P ack, Städte und Steuern in der Politik Julians. Untersuchungen zu den Quellen eines Kaiserbildes. C o lle c tio n Lato m u s 19 4 . Brussels, 19 8 6 . PLRE i A . H . M . Jo n e s, J . R . M artin d ale, a n d j. M o rris. The Prosopography of the Later Ro­ man Empire. 1: a .d . 260-395. C a m b rid g e , 1 9 7 1 . PLRE 2 T . D . B arnes,

P ack,

J. R . M artin d ale,

The Prosopography of the Later Roman Empire. 2: A.D. 395-527.

C a m b rid g e , 19 8 0 .

Reading the Past (19 9 0 ) G.

C la rk e , ed., w ith B . C r o k e , A . E m m e tt N o b b s, and R . M o rd e y .

Reading the Past

in Late Antiquity. Rushcutters B ay, 19 9 0 . R ik e , Apex Omnium ( 19 8 7 ) R . L . R ik e , Apex Omnium. Religion in the "Res Gestae’’ ofAmmianus. The Transforma­ tion of the Classical Heritage 1 5 . B erk eley, 19 8 7 . R o sen , Ammianus ( 19 8 2 ) K . R o sen , Ammianus Marcellinus. Erträge der Forschung 1 8 3 . D arm stadt, 19 8 2 . R o sen , Studien ( 19 7 0 ) K . R o sen , Studien zur Darstellungskunst und Glaubwürdigkeit des Ammianus Marcelli­ nus. B o n n , 19 7 0 . T h is w o rk is a reprint, w ith a slightly m o d ified title, o f R osen’s H e id e lb e rg dissertation o f 19 6 6 , p revio u sly revised and published in M an n h e im in 19 6 8 .

Méthode ( 19 7 8 ) G . Sabbah , La Méthode d’Ammien Marcellin. Recherches sur la construction du discours his­ torique dans les “Res Gestae.’’ Paris, 19 7 8 . S e e c k , Regesten ( 1 9 1 9 ) O . S e e c k , Regesten der Kaiser und Päpste für die Jahre 311 bis 476 n. Chr. Vorarbeit zu einer Prosopographie der christlichen Kaiserzeit. Stuttgart, 1 9 1 9 . Sab bah ,

[xiii]

A bbreviations Sy m e , Am m ianus (19 6 8 ) S y m e , R . Am m ianus and the Historia Augusta. O x fo r d , 19 6 8 .

Thompson, Am m ianus (1947) E. A. Thompson, The Historical Work o f Am m ianus Marcellinus. Cambridge, 1947. Vanderspoel, Themistius (1995) J. Vanderspoel, Themistius and the Imperial Court. Oratory, C ivic D uty, and Paideiafrom Constantins to Theodosius. Ann Arbor, 1995. Viansino, Lexicon G . V iansin o, A m m iani M arcellini rerum gestarum Lexicon. H ild e s h e im / Z ü r ic h / N e w Y o rk , 19 8 5 (tw o volum es w ith separate pagination) W an k e, Gothenkriege (19 9 0 ) U . W an k e, D ie Gotenkriege des Valens. Studien z u Topographie und Chronologie im un -

teren Donauraum von 366 bis 3 7 8 n. Chr. E u ro p äisch e H o ch sch u lsch rifte n , R eih e III: G esch ich te und ihre H ilfsw issenschaften 4 1 2 . B e r n / F r a n k f u r t / N e w Y o rk , 19 9 0 .

[xiv]

A m m ianu s M arcellinus and the R epresen tation o f H istorical R eality

A tru ly great historian w o u ld reclaim those m aterials w h ic h the novelist has appropriated. — M acaulay

[I] THE IMPARTIAL HISTORIAN

At the close o f his history, Ammianus Marcellinus described himself as “ a sol­ dier and a Greek” (31.16.9). He was bom about 330 into the local aristocracy o f one o f the cities o f Roman Syria or Phoenicia, and his father was probably a career soldier who rose to a position o f some importance in the reign o f the emperor Constantius, who ruled the East from 337 to 361 (Chapter VI). Am ­ mianus entered the Roman army as an officer in an élite corps around 350 and first appears in his narrative as extant in the year 354 (14.9.1, 11.5). It is not known how long he served beyond 359, when he disappears from his narra­ tive after escaping from Amida when the Persian king took it by storm and returning safely to Antioch (19 .8 .5-12 ). Ammianus reappears in his narrative in 363, when his use o f the first-person plural indicates that he joined Julian’s expedition into Persia at Circesium (23.5.7, cf. 6.30) and returned to Antioch with the defeated Roman army after its failure (25.10.1: Antiochiam venimus). After 363, however, Ammianus dis­ appears from his narrative completely, except for isolated first-person state­ ments that reveal that he was residing in or near Antioch in 372, no longer a soldier (29.1.24, 2.4), and a remark that implies that he was still there in 378 (31 . 1.2). Subsequently, Ammianus traveled in Greece, where he saw a ship carried almost two miles inland near Mothone in Laconia by the tsunami o f 21 July 365 (26.10.19), saw at least part o f the coasts o f Thrace and the Black Sea (22.8.1, 27.4.2), and probably traversed the Balkans, seeing the bones o f Ro­ mans and Goths killed in fighting near Marcianople in the autumn o f 377 (31.7.16 ).’ He had also toured Egypt (17.4.6, 2 2.15.1) before he came to 1 Th e precise location seems to be uncertain: Wanke, Gotenkriege (i99 °). 1 5 7 - 6 0 .

[I]

T he I m p a r t i a l H i s t o r i a n

Rome, probably shortly after 380. There, it is plausibly inferred from the bit­ terness with which he refers to the event (14.6.19), he had the misfortune to be expelled as a foreigner during a food shortage, probably in 384 on the or­ ders o f Symmachus as praefectus urbi.2 Ammianus was a Greek, his native language was Greek, and he thought in Greek (Chapter VII). Yet he wrote a history on an enormous scale in Latin, which covered a period o f almost three centuries, from the accession o f the em­ peror Nerva in 96 to the disastrous Battle o f Adrianople on 9 August 378 and its immediate aftermath, probably in a total o f thirty-six books (Chapter III). Only the second half o f the Res Gestae has survived. The extant part begins in the middle o f the historian’s account o f the activities o f the two emperors dur­ ing the campaigning season o f 353 (14 .1.1), and it occupies six hundred pages in modern critical editions. It is by far our fullest, most precise and most reli­ able narrative source for military campaigns and political events at the imperial court in the fourth century.3 Hence Ammianus has inevitably provided the basis for all modern narrative accounts o f the period from 353 to 378. He has also pervasively influenced all modem interpretations o f the period, including those by innovative historians who refuse to give narrative sources their tra­ ditional privileged position.4 Ammianus’ history is thus fundamental to modern understanding and inter­ pretation o f the fourth century. All the more necessary, therefore, to investi­ gate its strengths and weaknesses, its biases, its literary quality— in short, to ask how Ammianus depicts a period that is usually seen through his eyes. Am ­ mianus has traditionally been regarded as belonging to the select canon o f great historians who have penned reliable and impartial histories o f their own times. Edward Gibbon included in his “ vindication” o f the last two chapters o f the first volume o f his History o f the Decline and Fall of the Roman Empire, which analyzed the development o f Christianity down to the early fourth cen­ 2 Alan Cameron, J R S 54 (1964), 28. Th e expulsion o f peregrini from Rome is explicitly at­ tested in 384: Symmachus, Ep. 2.7; Ambrose, De Officiis 3 . 7 . 4 7 - 5 1 , cf. J.-R . Palanque, R E A 33 (19 31), 3 4 9 - 5 2 , 355, who points out that a similar expulsion is not documented during any o f the other known food shortages o f the late fourth century. Ammianus complains that, while the practitioners o f the liberal arts, though very few in number, were all expelled, the fans and pre­ tended fans o f mimae were allowed to remain, together with 3,000 dancing girls, their accom­ panying choruses, and their trainers (14.6.19). 3 For a full appreciation o f the scope and virtues o f Ammianus’ narrative, see Matthews, A m ­

mianus (1989), 3 3 - 2 2 8 . 4 See, recently, Averil Cameron, The Later Roman Empire, a . d . 284—430 (Cambridge, Mass., 1993), esp. 1 9 - 2 1 , 7 3 - 7 4 , 8 5 - 8 9 , 1 3 3 - 3 7 . M uch greater independence o f Ammianus is shown by Peter Brown, whose classic World o f Late Antiquity (London, 19 71) seems to mention him only twice and very briefly ( 11 5 , 120).

[2]

T he I m p a r t i a l H i s t o r i a n

tury in a manner that offended many o f his Christian contemporaries, an as­ sessment o f the “ character and credit” o f the ecclesiastical historian Eusebius o f Caesarea, whom he disparaged and compared unfavorably with Ammianus, and three early modem historians whom he greatly admired: S in ce the o rig in o f T h e o lo g ic a l Faction s, som e historians, A m m ian u s M arcellinus, F ra -P a o lo , T h u an u s, H u m e , and perhaps a fe w others, have deserved the singu­ lar praise o f h o ld in g the balance w ith a steady and equal hand. In dependen t and u n co n n ecte d , th ey con tem p lated w ith the sam e indifferen ce, the o p inions and interests o f the co n ten d in g parties; or, i f they w e re seriously attached to a par­ ticular system , th ey w e re arm ed w ith a firm and m oderate tem per, w h ic h en ­ abled th em to suppress their affections, and to sacrifice their resentm ents. In this small, b u t venerable S y n o d o f historians, E u seb iu s can n o t claim a seat.5

The three “ independent and unconnected” historians whom Gibbon names with Ammianus reflect his own interests and predilections— and one o f them conspicuously lacks the qualities for which Gibbon commended him. Gibbon had praised de Thou as “ a moderate philosopher” in his second commonplace book in the early i770s.6Jacques Auguste de Thou (1553 - 16 17 ) was a moderate Catholic who supported political accommodation with protes­ tants and was one o f the drafters o f the Edict o f Nantes in 1598 granting tol­ eration to the Huguenots. His History of His Time, which covers the period from 1543 to the early seventeenth century and was frequently reprinted in French and English after its original publication in Latin in Paris between 1604 and 1620, was placed on the index o f forbidden books by the Spanish Inquisition.7 David Hume ( 17 11- 17 7 6 ) was a friend whom Gibbon admired deeply both as a philosopher and a historian.8 His History of England from the Invasion of Julius Caesar to the Revolution in 1688, first published in three separate pairs o f volumes between 1745 and 1762, commended itself to Gibbon by both its s E. Gibbon, A Vindication o f Some Passages in the Fifteenth and Sixteenth Chapters o f the History o f the Decline and Fall o f the Roman Empire (London, 1779), 1 1 o - 1 1 ; The English Essays of Edward Gibbon, ed. P. A . Craddock (Oxford, 1972), 299. 6 English Essays (1972), 203. 7 The multifarious editions o f the History are catalogued by S. Kinser, The Works o f JacquesAuguste de Thou (The Hague, 1996), 6 - 7 8 . 8 Decline and Fall, 7.308, n. 101 (B) = 3.10 5 7 , n. 89 (W) (quoted at n. 14). See also 1.2 5 5, n. 90 (B) = 1 . 2 5 1, n. 86 (W ); English Essays, 338 (I, 2: “ M r. Hume told me,” etc.). References to Gibbon’s Decline and Fall are given by volume and page numbers in both the 19 2 6 —1929 reprint o f J. B . B u ry’s “ revised library edition” in seven volumes, which is often styled the second edi­ tion (London, 1 9 0 9 -1 9 1 4 ) (B), and the critical edition in three volumes by David Womersley (London, 1994) (W ).

[3]

T he I m p a r t i a l H i s t o r i a n

Tory standpoint and its distinguished style, which enabled it to remain a stan­ dard work for more than a century,9 and Gibbon alluded to Hume as a pre­ eminent Scotsman in terms that tempted a modern historian o f Rome to in­ dulge in a rare moment o f self-revelation.101 Pietro Sarpi (15 5 2 -16 2 3 ) is a very different character. He was a servite friar who enjoyed the protection o f the city o f Venice in political battles against con­ temporary popes: he wrote about the Reformation and Counter-Reformation o f the sixteenth century with passion, not detachment, and no one now would seriously claim any degree o f objectivity for his lstoria del concilio tridentino, which was first published in London in 16 19 under the anagrammatical pseu­ donym Pietro Soave Polano and “ represent[s] the Council o f Trent as being solely a conspiracy against reform o f the church.” 11 Gibbon read Sarpi in his youth: his early commonplace book o f 1755 commends him as “ one o f the most learned men o f his time.” 12 In the first volume o f Decline and Fall, Gib­ bon adduced a rescript o f Diocletian (without giving the relevant reference to the Codex Justinianus) “ on the respectable authority o f Fra-Paolo” and praised him most warmly as “ in learning and moderation . . . not inferior to Grotius.” 13 And in the penultimate chapter o f his sixth volume he delivered a final and highly commendatory verdict: Sarpi was a “ worthy successor” o f the “ noblest historians,” the Florentines Guicciardini and Machiavelli, and this trio, together with Davila, “ were justly esteemed the first historians o f mod­ ern languages, till, in the present age, Scotland arose to dispute the prize with Italy herself.” 14 Literary taste and intellectual fashions have changed greatly since Gibbon’s day. Although his Historia delle guerre civili in Francia, published in Venice in 1630, went through some two hundred printings in several European languages during the seventeenth and eighteenth centuries, Enrico Caterino Davila (1576 —1631) is today completely forgotten. Hume’s philosophical works are still read with profit, but not his history o f England; de Thou is 9 See, for example, the appreciative notice by Leslie Stephen, Dictionary o f National Biogra­ phy 10 (London, 19 2 1 [1892]), 2 1 5 - 2 6 . 10 Syme, Ammianus (1968), 22. n. 3, quoting Gibbon, Decline and Fall, 3.4 7 (B) = 1.10 0 1 (W ): “ Such reflections tend to enlarge the circle o f our ideas, and to encourage the pleasing hope that N e w Zealand may produce, in some future age, the Hume o f the Southern Hemisphere.” 11 F. L. Cross and E. A . Livingstone, Oxford Dictionary o f the Christian Church2 (Oxford, 1974), 1 2 3 6 - 3 7 , cf. H. Jedin, Der Quellenapparat der Konzilgeschichte Pallavicinos: Das Papstum und die

Wiederlegung Sarpis im Lichte neuerschlossener Archivalien (Rome, 1940); Geschichte des Konzils von Trient, 2 (Freiburg, 1957), 4 4 1 - 4 4 . 12 English Essays (1972), 2 3 - 2 4 . 13 Decline and Fall, 2.52 , n. 139 (B) = 1.492, n. 13 7 (W ); 2 .14 8 , n. 18 7 (B) = 1.580, n. 186 (W ): the first quotation is taken from Gibbon, Vindication (1779), 1 2 - 1 3 = English Essays (1972), 238. 14 Decline and Fall, 7.308, n. 10 1 (B) = 3 .10 5 7 , n. 89 (W).

[4]

T he I m p a r t i a l H i s t o r i a n

generally ignored, except by historians with a professional interest in the six­ teenth century;15 and Sarpi’s voluminous tomes remain unopened even by those who proclaim his supreme importance as an intellectual figure.16 Gibbon’s library contained two copies o f Hume’s history and the historical works o f Sarpi and de Thou in both French and their original language o f publication (Italian and Latin, respectively).1718He had naturally read Ammi­ anus long before he penned his Vindication, but at the time o f its composition he was already at work on the next installment o f his history, whose second volume draws heavily on the Res Gestae.™ Gibbon pays Ammianus a series o f fulsome tributes that, although justly both famous and familiar, will bear yet another repetition. For they show how little the prevailing estimate o f Am ­ mianus changed during the next two centuries. When Gibbon considered the ecclesiastical politics o f the reign o f Constantius, he gave Ammianus a privi­ leged position as an unbiased witness: The sentiments of a judicious stranger who has impartially considered the progress of civil or ecclesiastical discord are always entitled to our notice; and a short passage of Ammianus, who served in the armies, and studied the char­ acter, of Constantius, is perhaps of more value than many pages of theological invectives.19 When he came to Julian, Gibbon appealed to “ the unexceptionable testi­ mony o f Ammianus Marcellinus” and commended the intrinsic value o f his account o f the reign both in general and in particular: The philosophic soldier, who loved the virtues without adopting the prejudices ofhis master, has recorded, in his judicious and candid history of his own times, the extraordinary obstacles which interrupted the restoration of the Temple of Jerusalem.20 15 See, W . M cC uaig, Carlo Sigonto: The Changing World o f the Late Renaissance (Princeton, 1989), 7 0 —72; B. B . Diefendorf, Beneath the Cross: Catholics and Huguenots in Sixteenth-Century

Paris (N ew York, 19 91), 4, 8 1, 83, 97, 1 6 9 - 7 1 . 16 G . Steiner, No Passion Spent: Essays, 1978-1996 (London/Boston, 1996), 4: “ I have, a dozen times, slunk by Sarpi’s leviathan history o f the Council o f Trent (one o f the pivotal works in the development o f western religious-political argument). ” J. H. Plumb, The Death ofthe Past (Boston, 1970), 126, makes the less momentous claim that “ Sarpi was a far greater historian than Livy.” 17 G . Keynes, The Library o f Edward Gibbon2, St. Paul’s Bibliographies N o. 2 (Dorchester, 1980), 156, 245, 267. 18 O n Gibbon’s use o f and attitude toward Ammianus, see D. Womersley, The Transformation o f “ The Decline and Fall o f the Roman Empire” (Cambridge, 1988), 1 6 9 - 8 1 . 19 Decline and Fall, 2.380 (B) = 1.793 (W). 20 Decline and Fall, 2.485 (B) = 1 .8 9 0 - 9 1 (W ).

[5]

T he I m p a r t i a l H i s t o r i a n

And when he reached the accession o f Theodosius, he bade Ammianus an emphatic and moving farewell: It is not w ith o u t the m o st sincere regret that I m ust n o w take leave o f an ac­ curate and faithful gu ide, w h o has co m p o sed the h isto ry o f his o w n tim es w ith ­ o u t in d u lgin g the p rejudices and passions w h ic h usually affect the m in d o f a co n tem p o ra ry.21

Gibbon was not blind to the subjective and personal elements in Ammi­ anus: he noted “ the sarcasm o f an impartial historian” and argued that in the two angry digressions on the city o f Rome “ the judicious reader” would “ per­ haps detect the latent prejudices and personal resentments which soured the temper o f Ammianus himself.” 22 Gibbon sensed that there was an apparent contradiction between the impartiality o f historical judgment that he prized so highly and the often emotional style o f the historian, but he made a rigid distinction between the content o f Ammianus’ work and his manner o f pre­ sentation: “ The sincerity o f Ammianus would not suffer him to misrepresent facts or characters, but his love o f ambitious ornaments frequently betrayed him into an unnatural vehemence o f expression.” 23 Leading Roman historians and Latin scholars in Wilhelmine Germany wrote in virtually identical terms. Otto Seeck declared Ammianus to be cool and unbiased in his judgments, un­ moved by the religious conflicts o f his time, a pagan whose beließ were purely theoretical, not a living creed.24 Hermann Peter contrasted the “ rhetorical whitewashing” o f Ammianus’ account o f Julian’s victory at Strasbourg with “ the calm and moderation which constitute the main feature o f his being.” 25 Gibbon’s high estimate o f Ammianus as an impartial historian has contin­ ued to be shared and repeated by most who have written about both the his­ torian and the Roman Empire in the fourth century until very recently: Hugo Jones, for example, saluted him as “ a great historian, a man o f penetrating in­ telligence and o f remarkable fairness.” 26 One o f the most lucid and precise statements o f this traditional view can be found in M. L. W. Laistner’s Sather Lectures o f 1947, whose intelligent chapter on Ammianus remains worth reading and pondering. Laistner picked out “ conspicuous fairmindedness” as 21 Decline and Fall, 3 .12 8 (B) = 1.10 7 3 (W). 22 Decline and Fall, 2.262 (B) = 1.686 (W ); 3 . 3 1 1 - 1 8 (B) = 2 . 1 7 5 - 8 1 (W ), where Gibbon blends and paraphrases Ammianus 14.6 and 28.4. 23 Decline and Fall, 2.264, n. 18 (B) = 1.687, n. 16 (W). 24 O. Seeck, R E 1 (1894), 18 5 1. 25 H. Peter, Wahrheit und Kunst: Geschichtschreibung und Plagiat im klassischen Altertum (Leipzig/ Berlin, 19 11 ), 403. 26 Jones, L R E 116 .

[6]

T he I m p a r t i a l H i s t o r i a n

Ammianus’ salient characteristic and pronounced that, apart from a general and undisguised dislike o f Germans, he displays an obvious bias on only two matters: One of these is the description of Roman society, which is a kind of satire in­ spired by a long-established literary tradition and perhaps by the rigors of his ear­ lier life; the other is a bitter outburst near the end of his work against the whole tribe of lawyers. Was this invective . . . inspired merely by certain general no­ tions aroused by observing frequent malpractices in the administration of jus­ tice? It may be so, but it is tempting to imagine that behind his vitriolic bitter­ ness lay some personal experience of having been bested by a smart attorney. With his sensitivity to Latin, Laistner recognized the tendentious nature o f Ammianus’ excursus on Rome and lawyers (14.6.2-26; 28.4.6—35; 30.4.3 — 22), and Ammianus himself confesses his personal animosity against lawyers when he refers to the indignity that he had suffered at their hands in the East (30.4.4: super eius indignitate . . . quam in illis partibus agens expertus sum). Nevertheless, Laistner accepted the prevailing scholarly estimate o f the qual­ ity o f Ammianus’ narrative without enquiring whether personal animus, con­ cealed in apparently judicious language rather than openly avowed, could be a pervasive feature o f his work.27 Gibbon’s favorable assessment o f Ammianus has recently received a full and able restatement in John Matthews’ large book o f 1989, which conveys its main message in its very title. The Roman Empire of Ammianus argues both that the historian depicted the society in which he lived fairly and accurately and that the Roman Empire o f the second half o f the fourth century really was as Ammianus depicts it: It will be obvious that he is a wonderfully eloquent witness of almost every as­ pect of the life and society of his times. In breadth of interest, wealth of circum­ stantial detail and power of observation, he rivals any other Greek or Roman his­ torian known to us from any period, and outclasses most. As contemporary historians, only Thucydides and possibly Polybius have any prior claim to our admiration, and Ammianus’ world is so much vaster, its political structures more forbidding, and its cultural complexity far greater than theirs: all seen with the observant eye of an individual fascinated by all forms of human conduct, a still living challenge to the modern historian of his age.28 27 M . L. W . Laistner, The Greater Roman Historians (Berkeley, 1947), 1 4 1 - 6 1 , 1 8 1 - 8 3 : the passage quoted occurs on p. 158. 28 Matthews, Ammianus (1989), 228.

[7]

T he I m p a r t i a l H i s t o r i a n

Moreover, a subsequent essay by Matthews defends the accuracy o f Ammi­ anus’ excursus on lawyers, which Laistner diagnosed as the product o f per­ sonal resentment: it argues that the historian described “ the real world’’ in which he lived, that “ his accounts o f it are full o f precisely observed detail, which often occur when the rhetoric is most intense,” and that the rhetori­ cal aspects o f his narrative method should not be overrated in relation to the circumstantial.29 Thus, although Matthews concedes the force o f Ammianus’ rhetoric, he consistently emphasizes the “ precisely observed detail” over the possibility that such details may be subservient to rhetoric and prejudice. The traditional estimate o f Ammianus that has held sway from Gibbon to Matthews confronts the obvious problem that this supposedly impartial and dispassionate historian writes with unusual violence and ferocity. How can a method o f expression that so often distorts and contorts the phrase, the sen­ tence, even the paragraph, be reconciled with Ammianus’ postulated serenity o f historical judgment? Appeal has naturally been made to modem artistic and literary analogies. Jacques Fontaine depicted Ammianus as a “ romantic histo­ rian” comparable to Chateaubriand, while simultaneously arguing that the subjective and emotional elements in his text reflect the world in which he lived, and hence do not impair the lucidity, equanimity or detachment o f his historical judgments.30 But that is to treat the contradiction as if it were unim­ portant and to avoid rather than to solve a real problem. When Sir Ronald Syme came to Ammianus at the age o f sixty as a result o f his newly discovered interest in the Historia Augusta, he immediately and in­ stinctively recognized the problem: A t the outset an historian proclaim s that he w ill tell the truth. W h a t else w as he to say? T h e profession belongs to standard co n ven tio n . . . . W ith A m m ian u s the th in g is n o t a m ere co n ven tio n . R ath e r an obsession and a passion. It is insistent and pervasive. T ru th and h o n esty w ill n o t be fou n d in the c o u rt and the councils o f the C aesars. . . . A m m ian u s was a truthful m an. H is portrayal o fju lia n verges o n p a n e g y ric, but con veys pertinent criticism and perm its a balanced estim ate. E lse w h e re h e is d o m ­ inated b y a n u m b er o f prejudices. H e has n o th in g g o o d to report o f the lo w e r classes, for exam p le. In general, em o tio n ( o f the m o re h o n o u rab le sort), the im -

30

J. F. Matthews, Cognitio Gestorum (1992), 57. J. Fontaine, B A G B (1969), 4 1 7 - 3 5 , esp. 4 18 : “ ce vieil officier, lucide, équanime, moralisa­

teur, un peu distant.” Fontaine rejects the application o f the favorite term “ baroque” to A m ­ mianus, and he has recently tried to relate his style to a “ Theodosian aesthetic” (Cognitio Gesto­

rum [1992], 2 7 - 3 7 ) .

[8]

T he I m p a r t i a l H i s t o r i a n p o rtan ce o f the th em e, and the style he has elected im pel h im to exaggeratio n . T h a t has happen ed to o th er historians. T h e H is to ry (as extant) depends largely on his o w n exp e rie n ce and m editation : the m em oirs o f an old soldier d evelo ped , adorn ed, and rein forced b y o th e r infor­ m ation. . . . H is o w n testim on y is firm , valid and n o t to be discounted.-1 1

Syme’s analysis, although explicitly reasserting the traditional estimate o f Am ­ mianus as an impartial historian, also consciously subverts it. For Syme sees distortion not merely in a handful o f passages and in Ammianus’ confessed predilection for Julian, but as a pervasive and characteristic feature o f the Res Gestae: although Syme’s Ammianus was at heart “ a truthful man,” he is nev­ ertheless ‘‘dominated by a number o f prejudices,” and he habitually exagger­ ates under the influence o f emotion, theme, and style. Some years after Syme wrote, the traditional assessment o f Ammianus was subverted in another, even more troublesome, fashion. Tony Woodman ad­ duced the third o f the brief paragraphs quoted from Syme to redeem the rep­ utation o f Velleius Paterculus, whose history, dedicated to M. Vinicius during his consulate in a .d . 30, has usually been regarded as sycophantic and men­ dacious. Specifically, Woodman claimed that Syme’s verdict on Ammianus “ could with equal justice be applied to the Tiberian portion o f Velleius’ nar­ rative.” 3132 Syme immediately protested against this revaluation o f Velleius, for the obligation to produce panegyric does not excuse mindless repetition o f the official version o f events when it is patently false— and Velleius’ language reflects the propaganda and rhetoric o f the government o f Tiberius while Se­ janus was his chief minister.33 As concerns Ammianus, the parallel is valid in at least one particular. Ammianus’ loyalty to and partiality for his old com­ mander Ursicinus inevitably recalls that o f Velleius for his old commander-in­ chief Tiberius.34 Hence Woodman’s misguided attempt to rescue the reputa­ tion o f Velleius as a historian indirectly impugns that o f Ammianus. The 31 Syme, Ammianus (1968), 94. 32 A . J. Woodman, Velleius Paterculus: The Tiberian Narrative (2 .9 4 -13 1) (Cambridge, 1977),

55-56. 33 R. Syme, Roman Papers, 3 (Oxford, 1984), 1 0 9 0 - 1 1 0 4 . Syme adduced Velleius’ statement that the war against Tacfarinas in Africa, which lasted seven years, “ auspiciis consiliisque eius [sc. Tiberius] brevi sepultum est” (2 .1 29.4) as the clearest example o f “ mendacity in Velleius.” Th e phrase bellum/a sepultum/a, which Velleius also uses four times in his account o f Augustus (2.7 5.1, 8 2.1, 89.3, 90.1) is now known to reflect contemporary official phraseology: the senatus consulturn de Cn. Pisone patre proclaims that Piso tried to arouse a civil war by invading Syria after the death o f Germanicus, although “ all the evils o f civil war had been buried by the numen o f Augustus and the virtues o f Tiberius” (lines 4 5 - 4 9 , esp. 47: omnibus civilis belli sepultis malis). 34 Ammianus’ partiality for Ursicinus was demonstrated by Thompson, Ammianus (1947), esp. 4 2 - 5 5 — without noting the parallel.

[9]

T he I m p a r t i a l H i s t o r i a n

comparison o f the two writers poses an awkward and unavoidable question: can the traditional estimate o f Ammianus withstand scrutiny in the light o f modern techniques o f both historical research and literary criticism? It has in fact begun to crumble. H alf a century ago, a classic o f modem literary criti­ cism offered a fundamental reassessment o f Ammianus, which most historians o f Late Antiquity have been strangely reluctant to apply consistently to the evaluation o f his narrative.35 35

The present work deliberately refrains from any general assessment o f modern scholarly

research into the Res Gestae: For guidance readers should consult the full and helpful bibliogra­ phies in Seyfarth’s preface to his Teubner edition, i.x x v -x lv ii; Rosen, Ammianus (1982), 183 — 2 2 1 (more than 400 items); Matthews, Ammianus (1989), 5 5 4 —7 1 (with an introductory discus­ sion o f earlier bibliographical surveys). O n the date o f composition, a compelling case that A m ­ mianus completed the Res Gestae no later than 390 or 39 1 has been made by C . R T. Naudé, A J A H ç (1984, pubi. 1990), 7 0 —94; Matthews, Ammianus (1989), 1 7 - 2 7 . [IO ]

[IO]

[Il] REALITY AND ITS REPRESENTATION

Erich Auerbach published his classic study Mimesis in 1946, and it was su­ perbly translated into English by Willard Trask in 1953.' Unfortunately, Trask rendered Auerbach’s subtitle in a way that has unintentionally misled many readers and some critics. What is rendered into English as The Representation of Reality in Western Literature is subtly, but significantly, different from Dargestellte Wirklichkeit in der abendländischen Literatur. Whereas the original German means “ reality (as) represented in western literature” with emphasis on the first noun, the English version replaces the noun qualified by a participial adjective by two nouns and thus shifts the emphasis from “ reality” to its “ representation.” It is mistaken, therefore, to criticize Auerbach as if his book were exclusively a work o f literary criticism.12 On the contrary, as Rene Wellek noted in an im­ portant and influential review, Auerbach always moves beyond the analysis o f style “ to reflections on the attitude o f a writer toward reality and his tech­ nique o f reproducing it” and thence to “ reflections about periods and cul­ tures, social conditions and assumptions,” so that his book “ can be viewed as 1 E . Auerbach, Mimesis: Dargestellte Wirklichkeit in der abendländischen Literatur (Bern, 1946); English translation by W . R. Trask, Mimesis: The Representation o f Reality in Western Literature (Princeton, 19 53). In an epilogue, Auerbach defined his subject as “ die Interpretation des W irk­ lichen durch literarische Darstellung oder ‘ Nachahmung’ ” (494). Trask translates this as “ the in­ terpretation o f reality through literary representation or ‘ imitation’ ” (534), thereby concealing Auerbach’s subtle shift from the abstract “ Wirklichkeit” o f the book’s title to the concrete “ das W irkliche” in the epilogue. 2 O n the historical or historiographical aspects o f Mimesis, see esp. W . W . Holdheim, C L IO 10 (19 8 1), 14 3 - 5 4 ; T. Bahti, After Strange Texts: The Role o f Theory in the Study o f Literature, ed. G . S. Ja y and D . L. M iller (Univ. Ala., 1985), 1 2 7 —45.

[II]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

a short history o f the human condition.” 3 For it is a study o f the development o f Western culture from its twin roots in Homer and the Old Testament to the twentieth-century novel. As is natural with a work o f genius and profound originality, Mimesis was seriously misrepresented by reviewers and commentators. Ernst Curtius, for example, appealed to Otto Regenbogen for the proposition that one o f Auer­ bach’s two main theses was that only modern realism broke through the an­ cient doctrine o f the three styles,4 even though Auerbach had expressely as­ serted that “ the rule o f differentiated styles cannot possibly apply” to the account o f Peters denial o f Christ in the Gospel according to Mark, since the depiction is “ entirely realistic” and the mingling o f styles “ was rooted from the beginning in the character ofJewish-Christian literature.” 5 When Auer­ bach penned a dignified reply to Curtius, he not only corrected his critic’s misapprehension about his view o f stylistic doctrines, but also explained the nature o f his book: his method (he asserted) was not sociological; what he had offered was an interpretative essay written by a particular individual in a par­ ticular situation at a particular date.6 Ammianus’ history, too, was written by a particular individual in a partic­ ular situation at a particular date, but it also claims to describe the world in which he lived. From Ammianus, Auerbach selected for discussion an episode in Book XV, which he thereby made famous.7 In his review o f the English translation o f Mimesis, Wellek commended Auerbach for selecting passages “ known to only a few specialists” and posed a question that was obviously not intended to be rhetorical when he asked “ Who has read the gruesome story o f the arrest o f Peter Valvomeres in Ammianus Marcellinus?” 8 Ammianus’ ac­ count o f the arrest o f Peter Valvomeres (15.7.4 —5) provides both the title and the starting point o f the third chapter o f Auerbach’s book. In his discussion o f this episode, as throughout Mimesis, Auerbach combines literary criticism with general verdicts about the temper o f the age: T h e inciden t is so treated that it p ro du ces a stro n gly sensory im p ressio n — to such an exten t in fact that m an y readers w ill feel it unpleasantly realistic. A m ­ m ianus has orien ted it entirely towards gestures: the c o m p a ct c ro w d set against the im p osin g p refect as he dom ineers o ver them . T h is elem en t o f the sensory and J R. W ellek, Kenyon Review 16 (19 5 4 ), 299; cf. C . Landauer, German Studies Review 1 1 (19 88), 84: “ Auerbach is constructing a model o f W estern culture w h ich is all-inclusive.”

4 E. R. Curtius, Romanische Forschungen 64 (19 5 2 ), 57. 5 Auerbach, Mimesis ( 1 9 5 3 ) , 4 1.

ft E. Auerbach, Romanische Forschungen 65 (19 5 4 ), 5 - 1 5 , 17: “ ein B u ch , das ein bestim m ter M ensch , in einer bestimmten Lage, zu A n fan g der 19 40 er Jahre geschrieben hat.”

7 Auerbach, Mimesis (1953), 50 —60. K Wellek, Kenyon Review

16 ( 1954). 299-

[I2]

R e a l i t y a n d I ts R e p r e s e n t a t i o n th e g e s tu ra l is p r e p a r e d f o r f r o m th e fir s t — t h r o u g h th e c h o ic e o f w o r d s a n d s im ­ iles . . . — a n d r e a c h e s its c lim a x in th e s c e n e at th e S e p t e m z o d iu m w h e n L e o n t iu s , s ittin g in h is c a r r ia g e w it h fla s h in g e y e s c o n fr o n ts th e “ s n a k ily ” h is s in g m o b lik e an a n im a l ta m e r, u n m o v e d as th e y r a p id ly v a n ish . A r io t , a s o lita r y m a n t r y in g to q u e ll it b y th e p o w e r o f h is e y e s, th e n s te p p in g i n — s o m e h a rsh w o r d s , a r in g ­ le a d e r s m u s c u la r b o d y ra is e d h ig h , fin a lly a flo g g in g . T h e n all is q u ie t , a n d , b y w a y o f c o n c lu s io n , w e g e t a ra p e a n d th e s u b s e q u e n t ca p ita l p u n is h m e n t.

For Auerbach, the world o f Ammianus is a grim world, and the historian him­ self a profound pessimist: E v e r y w h e r e h u m a n e m o t io n a n d r a t io n a lity y ie ld to th e m a g ic a lly a n d s o m b e r ly s e n s o ry , to th e g r a p h ic a n d th e g e stu ra l. . . .

A m m ia n u s ’ w o r ld is s o m b e r : it is fu ll o f s u p e r s titio n , b lo o d fre n z y , e x h a u s t io n , fe a r o f d e a th , a n d g r im a n d m a g ic a lly r ig id g e stu re s ; a n d to c o u n te r b a la n c e all th is th e re is n o t h in g b u t d ie e q u a lly s o m b e r a n d p a th e tic d e t e r m in a t io n to a c ­ c o m p lis h an e v e r m o r e d iffic u lt, e v e r m o r e d e s p e ra te task: to p r o t e c t th e E m p ir e , t h re a te n e d fr o m w it h o u t a n d c r u m b lin g w it h in . . . .

W it h g lit t e r in g w o r d s a n d p o m p o u s ly d is to r te d c o n s t r u c t io n s la n g u a g e b e g in s to d e p ic t th e d is to r te d , g o r y , a n d sp e c tra l re a lity o f th e a g e . . . .

G r o t e s q u e a n d s a d istic , sp e c tra l a n d s u p e rs titio u s , lu s t in g f o r p o w e r y e t c o n ­ s ta n tly t r y in g to c o n c e a l th e c h a tt e r in g o f th e ir t e e t h — so d o w e se e th e m e n o f A m m ia n u s* r u lin g class a n d th e ir w o r ld . . . .

J u d g e d b y classical sta n d a rd s, th e s ty le , b o th in d ic t io n a n d s y n ta x , is o v e r r e fm e d a n d e x a g g e r a t e d ly s e n s o r y ; its e ffe c ts are p o w e r fu l, b u t d is to r te d . Its e ffe c ts are as d is to r te d as th e re a lity it re p re se n ts. A m m ia n u s ’ w o r ld is v e r y o fte n a c a r ic a tu r e o f th e n o r m a l h u m a n e n v ir o n m e n t in w h ic h w e liv e ; v e r y o fte n it is lik e a b a d d re a m . . . . S t r ik in g o n ly in th e s e n s o ry , re s ig n e d a n d as it w e r e p a r a ly z e d d e s p ite its s tu b b o r n r h e to r ic a l p a ssio n , h is m a n n e r o f w r i t i n g h is t o r y n o w h e r e d isp la y s a n y t h in g r e d e e m in g , n o w h e r e a n y th in g th at p o in ts to a b e tt e r fu tu r e , n o w h e r e a fig u r e o r an a c t a b o u t w h ic h stirs th e r e fr e s h in g a tm o s p h e r e o f a g re a te r f r e e ­ d o m , a g r e a t e r h u m a n ity .9 9

Quoted from Auerbach, Mimesis (1953), 53, 5 3 - 5 4 , 55, 56, 57, 5 9 - 6 0 . Auerbach provided

his own translation, which “ attempts to preserve the strangely baroque style o f the original’’ ( 5 1 - 5 2 ) . The original reads as follows: dum has exitiorum communium clades suscitat turbo feralis,/ urbem aeternam Leontius regens/ multa spectati iudicis documenta praebebat,/ in audiendo celerior,/ in discep­ tando iustissimus,/ natura benevolus/ licet auctoritatis causa servandae/ acer quibusdam

[13]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

This analysis contains an inherent contradiction. Does Ammianus’ style faith­ fully reflect the world that he describes? O r does it turn it into a “ bad dream” that caricatures and distorts it? Auerbach appears to assert both these mutually incompatible propositions. What is “ the normal human environment in which we live” ? Is Auerbach’s “ we” here the human race in general or modem west­ ern men and women? It is necessary to distinguish between the literary analy­ ses in Mimesis and its historical interpretations. John Matthews has already drawn this distinction in a paper with the pro­ grammatic title “ Peter Valvomeres, Re-arrested.” 10 Matthews accuses Auer­ bach o f being evasive and scores easy points against some o f the historical assumptions that underlie his literary observations, particularly against Auer­ bach’s claim that Ammianus fails adequately to indicate the social and histor­ ical context o f the riot that the prefect Leontius suppressed. Yet, like most who have recently written about Ammianus, Matthews accepts the central contention o f Auerbach’s analysis— that “ the prime quality o f the passage is the pictorial imagery.” Nor does Matthews reject Auerbach’s diagnosis o f the relation between Ammianus’ text and the historical reality that it depicts. For Matthews argues that Ammianus’ emphasis on pictorial effect and gesture cor­ responds to central aspects o f public life in Late Antiquity: hence, where Auer­ bach emphasizes the elements o f unreality and distortion, Matthews contends videbatur et inclinatior ad damnandum. / prima igitur causa seditionis in eum concitan­ dae vilissima fuit et levis. / Philoromum enim aurigam rapi praeceptum/ secuta plebs om­ nis velut defensura proprium pignus/ terribili impetu praefectum incessebat ut timidum,/ sed ille stabilis et erectus / immissis apparitoribus/ correptos aliquos vexatosque tormen­ tis/ nec strepente ullo nec obsistente/ insulari poena multavit. / diebusque paucis secutis/ cum itidem plebs excita calore, quo consuevit,/ vini causando inopiam/ ad Septemzodium convenisset, celebrem locum ,/ ubi operis ambitiosi N ym phaeum / Marcus condidit im­ perator. / illuc de industria pergens praefectus/ ab omni toga apparitioneque rogabatur enixius,/ ne in multitudinem se arrogantem immitteret et minacem / ex commotione pristina saevientem. / difficilis ad pavorem recta tetendit/ adeo, ut eum obsequen . . . de­ sereret/ licet in periculum festinantem abruptum. / insidens itaque vehiculo/ cum speciosa fiducia contuebatur acribus oculis/ tumultuantium undique cuneorum / veluti serpen­ tium vultus/ perpessusque multa dici probrosa / agnitum quendam inter alios eminentem/ vasti corporis rutilique capilli interrogavit,/ an ipse esset Petrus Valvomeres, ut audierat, cognomento. / eumque, cum esse sonu respondisset obiurgatorio,/ ut seditiosorum ante­ signanum olim sibi compertum / reclamantibus multis / post terga manibus vinctis suspendi praecepit. / quo viso sublimi / tribuliumque adiumentum nequiquam implorante/ vulgus omne paulo ante confertum/ per varia urbis membra diffusum/ ita evanuit, ut turbarum acerrimus concitor/ tamquam in iudiciali secreto/ exaratis lateribus/ ad Picenum eiceretur,/ ubi postea ausus eripere virginis non obscurae pudorem/ Patruini consularis senten­ tia/ supplicio est capitali addictus. ( 1 5 .7 .2 - 5 ) I have marked all the clausulae with a hasta [/]: m y clausulation sometimes diverges from that o f Seyfarth.) 10 Matthews, Homo Viator: Classical EssaysforJohn Bramble (Bristol/Oak Park, 1987), 2 7 7 - 8 4 .

[14]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

that in the fourth century both reality and social relations “ conform to a the­ atrical mode o f expression.” 11 In this attempt to establish that “ reality and social relations” conformed to “ a theatrical mode o f expression” in the fourth century, however, Matthews blurs an important distinction that Auerbach had emphatically asserted. Auerbach acknowledged that Ammianus resembled Tacitus in certain ways, but he saw a fundamental difference between both the historical reality and its representation in the two historians: “ a comparison with Tacitus (he wrote) serves to show how much stronger the magical and the sensory has become at the expense o f the objectively rational.” 112 For Matthews, in contrast, the similarities outweigh the differences. In his hands, Ammianus’ representation o f reality becomes dramatic in the sense o f depict­ ing historical action in the manner o f a tragedy, as Tacitus does. For Matthews, the arrest o f Peter Valvomeres, like other episodes in Ammianus, “ is presented almost like a scene from a play, the contrasting emotions and postures o f the sides preparing for the dialogue between the central characters.” Matthews sees dramatic action in Ammianus, not merely dramatic tableaux: “ the urban prefect Leontius approaches the rioting crowd almost as if in a crowd scene from a Shakespearean play, addressing the mob leader man to man, in a con­ frontation o f startling intimacy.” 13 The Shakespearean analogy is deeply misleading. It confuses two distinct senses o f the word theatrical— and it ignores the profound differences between what could be seen on the dramatic stage in Tacitus’ day and in the fourth century.14 I f Tacitus and Ammianus are both “ theatrical,” they are so in two very different ways (Chapter XV). Ammianus focuses not on the development o f plot or the internal attitudes, emotions, and motives o f the historical actors, as Tacitus does, but on the visual aspects o f the historical drama, and he presents the drama itself not as a developing plot, but as a series o f dramatic tableaux, that are discontinuous even when they succeed one another closely.15 Like contemporary poets, Ammianus practices what Michael Roberts has called “ the jeweled style.” 16 The analysis that Roberts gives o f Ammianus’ fa11 Matthews, Homo Viator (1987), 279, cf. Ammianus (1989), 4 6 0 - 6 1 . Th e visual character o f Ammianus’ portrayal o f persons and events was illustrated in a justly famous paper by R. M acMullen, Art Bulletin 46 (1964), 4 3 5 - 5 5 . 12 Auerbach, Mimesis (1953), 53. 13 Matthews, Homo Viator ( 1987), 279, 280. 14 For a brief description o f these changes, see Roman Theater and Society, ed. W. J. Slater (Ann Arbor, 1996), 1 6 1 - 8 0 . 15 J. Fontaine, L e Transformazioni della Cultura nella tarda Antichità: Atti del Convegno tenuto a Catania, Universita degli Studi, 2 7 sett.—2 Ott. 1982 (Rome, 1985), 7 9 5 - 8 0 8 . H e attributes to A m ­ mianus an “ optique nécessairement théâtrale, voire, si l’on ose ce néologisme, amphithéâtrale” (803). 16 M . Roberts, The Jeweled Style: Poetry and Poetics in Late Antiquity (Ithaca, 1989), esp. 132 .

[15]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

mous description o f Constantius’ triumphal entry into Rome in 357 (16 .10 .4 — 1 2) can be applied to many scenes in his work, including the arrest o f Peter Valvomeres: “ the effect is o f a series o f brilliantly eye-catching but discrete vi­ sual impressions, which in part by their very brilliance deter the viewer from attempting to piece together the individual scenes into a coherently ordered whole.” 17 I f this analysis o f Ammianus’ method o f presenting reality and his­ torical events is correct, it raises the question o f how far this stylization has distorted his depiction o f historical reality. For such stylization must operate at the unconscious and semiconscious levels as well as the conscious. The his­ torian’s representation o f reality cannot be, as Gibbon asserted, merely a mat­ ter o f surface rhetoric. Nor may it simply be assumed that his distorted style and presentation faithfully reflect an unpleasant and distorted historical real­ ity. Auerbach made that assumption, and it could be correct, but such an as­ sumption needs to be proved before being made the basis o f either a histori­ cal or a literary interpretation. Do the findings o f historical research confirm Ammianus’ essential veracity or call it into doubt? Much recent work on both Ammianus and the fourth cen­ tury either delivers or implies a negative verdict on one whom Amaldo Momigliano characterized as “ the lonely historian,” 18 a designation that sub­ tly emphasizes the disjunction between the historian and the world he depicts. Edward Thompson’s slim volume o f 1947 marked a watershed in historians’ approaches to Ammianus. Although Thompson reiterated that “ the general accuracy o f Ammianus’ monumental work even in matters o f the minutest de­ tail cannot seriously be called in question,” 19 he demonstrated how unfair in general, and how misleading and even inaccurate in detail, is Ammianus’ ac­ count o f the actions and policies o f the emperor Constantius. Thompson’s specific arguments were too often a priori and too often based on unexamined assumptions, yet he established two central facts about the Res Gestae beyond all reasonable doubt. The first was that Ammianus’ account o f Constantius’ court and military policies is colored by his admiration for his commanding officer and friend Ursicinus.20 Thompson showed that there is good reason to believe that Ammianus has exaggerated the merits o f Ursicinus and that sym­ pathy for the general after he was dismissed as magister militum led the histo­ rian to his hostile view o f the emperor, which was then strongly reinforced 17 M . Roberts, Philologus 13 2 (1988), 183. 111

A . Momigliano, Annali della Satola Superiore Normale di Pisa, Classe di lettere e filosofia,

Ser. 3.4 (1974). I 3 9 3 - I 4 0 7 19 Thompson, Ammianus (1947), 40. 20 Thompson, Ammianus (1947), 4 0 - 5 5 .

[ 16]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

by his admiration for Julian, whose propaganda against Constantius he some­ times repeats. Second, as a Marxist and a Communist, Thompson was sensitive to Am ­ mianus’ class bias. Thompson recognized a textbook bourgeois when he read one. He did not invent the idea that the historian was a curialis by legal origin and status: Wilhelm Ensslin had already argued in 1923 that curial status helps to explain Ammianus attitudes and Weltanschauung.21 But Thompson detected the attitudes o f a curialis in so many passages that it seemed impossible any longer to doubt that Ammianus belonged to the “ upper middle class,” the curial class that was both “ oppressed and oppressive,” and that his status as a curialis provided a basis for understanding both his career and much in his history.22 The task o f reassessing the quality o f Ammianus as a historian has subse­ quently been prosecuted with vigor by several scholars with very different ap­ proaches.23 Yet not all such enquiries have applied the relentless logic that such a réévaluation requires. In the 1970s, although Roger Blockley produced both a monograph on the historiography and political thought o f Ammianus and a series o f articles in which he consistently called the historians factual accuracy into question, he too often appealed to Ammianus’ political and moral as­ sumptions or invoked the copying o f rhetorical models as omnibus explana­ tions for postulated distortion rather than employing all the available evidence to prove specific inaccuracies, unfair bias, or the suppression o f relevant facts.24 Moreover, despite the implications o f his arguments about individual episodes, Blockley paradoxically reaffirmed the traditional general estimate o f Ammi­ anus as a sincere and honest historian, “ faithful to the truth.” 25 21 Ensslin, Ammianus (1923), esp. 4 - 6 . 22 Thompson, Ammianus (1947), 2, 15, n. 6, 68, 8 1 - 8 5 , 1 2 8 - 2 9 , followed by R. Pack, C P 48 (1953), 8 0 - 8 5 ; G. A. Crump, Ammianus Marcellinus as a Military Historian (Historia Einzel­

schriften 2 7 ,19 7 5 ) , 5 - 1 3 ; N . J. E. Austin, Ammianus on Warfare: An Investigation into Ammianus' M il­ itary Knowledge (Collection Latomus 165, 1979), 1 2 - 1 3 ; T. D. Barnes, Reading the Past (1990), 62. 23 Esp. C . P. T. Naudé, Ammianus Marcellinus in de Lig van die Antieke Geskiedskrywing (Diss. Leiden, 1956); Acta Classica 1 (1958), 9 2 - 1 0 5 (battles and sieges); Demandt, Zeitkritik (1965); Rosen, Studien (1970); Bitter, Kampfschilderungen (1976): for critical comments on some o f Rosens main arguments, see N . J. E. Austin, Historia 22 (1973), 3 3 1 - 3 5 ; G. Calboli, Bollettino di Stndi Latini 4 (1974), 6 7 —103. J. Szidat, Cognitio Gestorum (1992), 1 0 7 - 1 6 , has recently argued that exaggeration does not lead Ammianus to distort historical reality in any important way. 24 R. P. C . Blockley, A JP 93 (1972), 4 3 7 —50: Ammianus was “ strongly influenced" by J u ­ lian’s propaganda and hence “ must be used with caution” ; Latomus 31 (1972), 4 3 3 - 6 8 : Am m i­ anus’ account o f Gallus is “ suspect on grounds both literary and historical” ; Ammianus (1975): Phoenix 31 (1977), 2 1 8 —3 1: the account o f the Battle o f Strasbourg is shaped so as to produce “ a microcosm o f Ammianus’ overall attitude toward Roman-German relations” ; Studies in Latin Literature and Roman History 2 (Collection Latomus 168, 1980), 4 6 7 -8 6 : Ammianus had an “ almost paranoid dislike” o f Constantius. 25 Blockley, Ammianus (1975), 100, 10 1, 136.

[17]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

Some recent assessments o f Ammianus have been markedly more critical. In 1979, Chantal Vogler published an analysis o f the administration o f Con­ stantius in which she was perforce compelled to evaluate the main sources for his reign: she stigmatized Ammianus’ account o f Constantius’ dealings with his Caesars as a mass o f subjective impressions requiring correction because the historian was carried away by his sympathies for Julian, and she observed that “ historical truth does not always win.” 26 In a monograph on Ammianus published in 1983, Thomas Elliott correctly identified some cases o f gross dis­ tortion, although he formulated his case too brusquely and sometimes too carelessly to convince others that he was right.27 In particular, the thesis ar­ gued by Elliott (following Salvatore D ’Elia)28 that Ammianus is a pagan apol­ ogist who treats Christianity unfairly was answered by David Hunt in an ar­ ticle that was hailed as a definitive refutation29— prematurely, since its central arguments are demonstrably fallacious.30 Most recently, John Drinkwater has called the historian’s bluff over his account o f the alleged rebellion o f Silvanus, which he helped to suppress (15 .5 .15 —34).31 Although Ammianus’ account o f this “ reluctant usurper” has always been accepted largely at face value,32 there are strong grounds for scep­ ticism. Admittedly, it very quickly came to be accepted that Silvanus pro­ claimed himself Augustus at Cologne (15.5.16) and reigned for twenty-eight days.33 But the alleged usurpation leaves no trace whatever in the Roman im­ perial coinage o f nearby Trier.34 Hence Silvanus never minted coins in his name. That fact implies that in reality he never laid claim to the imperial power: 26 C . Vogler, Constantins II et Vadministration impériale (Strasbourg, 1979), 44. 27 Elliott, Ammianus (1983), cf. J. M . Alonso-Nûnez, J R S 76 (1986), 328: “ Elliott has exag­ gerated the bias introduced by Ammianus in his narrative, to reach the surprising conclusion that he was not an impartial historian. . . . His verdict that Ammianus was a pagan apologist and thus an anti-Christian historian representing the pagan reaction is completely distorted.“ That does not do justice to Elliott’s recognition o f how bias can be disguised (Ammianus [1983], 1 2 — 13, with a telling modern analogy). 28 Elliott, Ammianus (1983), 2 0 5 - 2 1 , cf. S. D ’Elia, Studi romani 10 (1972), 3 7 2 - 9 0 , partly an­ ticipated by A. Selem, Rivista di Cultura Classica e Medioevale 6 (1964), 2 2 4 - 6 1 . 29 E. D. Hunt, C Q , N .S. 35 (1985), 18 6 -2 0 0 : heartily commended by Matthews, Ammianus (1989)» 4 3 5 - 5 1 , 5 4 6 , n. 22, 547, n. 3 2 - 3 3 . 30 See Reading the Past (1990), 7 5 - 8 2 ; C P 88 (1993), 6 7 - 7 0 ; Chapter IX . 31 J. F. Drinkwater, Studies in Latin Literature and Roman History 7 ( Collection Latomus 227, 1994), 5 6 9 - 7 6 . 32 Syme, Ammianus (1968), 5 , 1 1 . Unlike many others, Syme correctly allowed for the lacuna in 15.5.30. 33 Julian, Orat. 1, 48bc; Pan. Lat. 3 ( 1 1 ) .1 3 .3 ; Eutropius, Brev. 10 .13 ; Jerome, Chronicle 2 39 d Helm; Epitome 42.10. 34 R IC 9 . 1 6 5 - 6 7 . The contrast to the revolt o f Poemenius against Magnentius in 353 is striking (ib. 1 6 4 - 5 ) . It need hardly be added that C IL 10.6945 = IL S 748 (Aversa) cannot be a milestone o f Silvanus.

[18]

R e a l i t y a n d I ts R e p r e s e n t a t i o n

paradoxically, once that is granted, most o f Ammianus’ account o f the sup­ pression o f Silvanus becomes more (not less) plausible.35 The time is thus ripe for a systematic investigation o f the structure, nature, and quality o f Ammianus’ Res Gestae, which combines literary and historical approaches. In the chapters that follow, the formal structure o f the Res Gestae and its arrangement o f material are laid out (Chapters III-V ). Next, the histo­ rian’s origin, social status, culture, and attitudes are investigated (Chapters V I— IX). Third, Ammianus’ presentation o f both the main historical actors and several supporting characters in the period for which the history is extant are assessed (Chapters X -X III). Then it is asked how Ammianus interpreted the whole sweep o f Roman history and what he expected o f the future (Chap­ ter X IV ). In conclusion, Ammianus is compared briefly with two other great historians: Tacitus, whom he had read and whom he imitated in various ways, and Macaulay, who uncannily resembles him as a historian (Chapter XV). The aim is to provide a companion to Ammianus on the model o f Sir Charles Firth’s classic companion to Macaulay’s History of England, which concen­ trated on the modem historian’s fundamental historical beliefs and his man­ ner o f depicting persons and events.36 35 Drinkwater, Studies in Latin Literature, 7 (1994), 575 - 7 6, suggesting that Ursicinus had Silvanus killed “ out o f self-interest.” 36 C . Firth, A Commentary on Macaulay's History of England, ed. G. Davies (London, 1938). Macaulay s unfair denigration o f the Duke o f Marlborough and William Penn and his white­ washing o f his hero William 111 had been exposed in a series o f essays by John Paget published between 1858 and i860 and later collected together in book form as The New “Examen” (Edin­ burgh/London, 1861). Although the great historians reputation all but obliterated Pagets crit­ icisms at the time, as Winston Churchill complained in the preface to his reissue o f his book ([Halifax, 1934], ix —xv), their basic validity is now taken for granted by all who write about Macaulay: Firth, Commentary (1938), 2 6 3 - 7 6 ; H. R. Trevor-Roper, Tlte Romantic Movement and the Study of History (London, 1969), 8 - 9 ; J. Clive, Not by Fact Alone (N ew York, 1989), 72.

[Ill] SYMMETRY AND STRUCTURE

The twentieth century, or at least English-speaking culture in the twentieth century, has lost the feeling for symmetry and formal structure that came in­ stinctively to earlier ages. Formal structure, that is, the division o f long works into books among the ancients and their division into volumes in the modem age, was important for all historians in the classical tradition until the nine­ teenth century. The subsequent loss o f this sense o f the architecture o f a his­ tory that also sets out to be a literary masterpiece can be charted through a comparison o f the original publication o f Edward Gibbon’s History of the De­ cline and Fall of the Roman Empire in the late eighteenth century and the edi­ tion by John Bagnall Bury, which has been the standard edition o f Gibbon’s work used by scholars during the twentieth century.1 Gibbon published the first volume o f his Decline and Fall in 1776, the sec­ ond and third together in 178 1, and the last three o f the six volumes together in 1788.12 Each installment has a carefully balanced structure and each volume a careful and deliberate internal arrangement, though without any perceptible concern for formal symmetry in the number o f chapters. The first volume had a total o f sixteen chapters. The central block o f eleven chapters is a political 1 First published in 18 9 6 -19 0 0 . For the normal mode o f referring to Decline and Fall in the present work, see Chapter I, n. 8. 2 For the pervasive differences between the three installments o f 1776, 17 8 1, and 1788 on practically every level, see D. Womersley, The Transformation of uThc Decline and Fall of the Ro­

man Empire ” (Cambridge, 1988). Womersley’s subsequent critical edition (London, 1994) pre­ serves the architecture o f Gibbon’s masterpiece. For, although each o f its three volumes includes two o f Gibbon’s original six, a facsimile o f the title page o f the first edition o f each o f Gibbon’s original volumes and Gibbon’s prefaces to his second and fourth volumes are printed in their orig­ inal positions, immediately preceding Chapters I, X V II, X X V I I, X X X I X , X LV III, and LVIII.

[2 0]

Sy m m e t r y and St r u c t u r e

and military narrative o f the history o f the Roman Empire from 180 to 324; these are framed by three introductory chapters that survey the stable empire o f the Antonines and two concluding chapters that describe and analyze the rise o f Christianity down to the “ Great Persecution” o f 3 0 3 - 3 13 . The second and third volumes had ten and twelve chapters, respectively. They took the main narrative to 476 and concluded with an epilogue discussing the causes o f the “ Fall o f the Roman Empire in the West,” which Gibbon had composed as a separate brief essay before the publication o f the first volume.3 The final three volumes, with a total o f thirty-three chapters, formed a clear composi­ tional unit too, whose function was not merely to take the story down to 1453, but also to explain the origin o f the world in which Gibbon lived.4 . Each volume o f the second installment (though not perhaps o f the third) exhibits a deliberate and careful arrangement o f the historical material. The second volume covers the five and a half decades between Constantine’s defeat o f Licinius in 324 and the Battle o f Adrianople, concluding with an epilogue on the settlement o f the Goths. Hence it includes the whole o f the period covered by the extant books o f Ammianus. Gibbon follows his principal source for the years 3 5 3 -3 7 8 in giving generous space and great prominence to the emperor Julian, but he balances this by assigning an emphatic and important place to Athanasius as bishop o f Alexandria from 328 to 373-5 The volume has ten chapters in all. It begins with one chapter on the foundation o f Constan­ tinople and Constantine’s organization o f the “ new empire” and two chap­ ters that narrate the political and military history o f the Roman Empire from 324 to Julian’s proclamation as Augustus in 360. The central place is filled by two chapters on the Christian church in the fourth century and three on the brief reign ofjulian. Athanasius, whom Ammianus had deliberately treated as a marginal figure (15.7.7—10),6 occupies a position close to the very middle o f Gibbon’s second volume, where he is presented as a champion o f religious freedom who defied the tyranny o f Constantius, and as a man who possessed 3 The precise date o f composition (17 7 2 or 1773) is disputed: see P. R. Ghosh, J R S 73 (1983), 1 —23; P. B. Craddock, Edward Gibbon, Luminous Historian, 1772—1794 (Baltimore/London, 1989), 8 - 1 1 ; P. R. Ghosh, J R S 81 (19 91), 1 3 2 - 5 6 . 4 P. B. Craddock, Roman Images, ed. A . Patterson (Baltimore, 1984), 6 3 - 8 2 . Chapter LV, for example, observes at the outset that “ the disciples o f Mahomet still hold the civil and religious sceptre o f the Oriental world” and concludes with the words “ the light o f knowledge which arose on the western world” (6.129, 166 [B] = 3.440, 470 [W ]). 5 O n Gibbon’s treatment o f Athanasius, his dependence on Tillemont, and his disingenuous answer to the charge o f plagiarism, which has too often been accepted uncritically, see Corolla Torontonensis. Studies in Honour o f Ronald Morton Smith, ed. E. I. Robbins and S. Sandahl (Toronto,

1994). 14-256 Ammianus here writes as if he had not mentioned Athanasius before, despite his impor­ tance in imperial politics in the 340s and in 350: see Athanasius (1993), 1 6 6 - 6 8 .

[2l]

Sy m m e t r y and St r u c t u r e

“ a superiority o f character and abilities which would have qualified him, far better than the degenerate sons o f Constantine, for the government o f a geat monarchy.” 7 The final pair o f chapters reflect the final books o f Ammianus: one takes the main narrative down to 375, and the last frames an account o f the campaigns o f 376—378 between a digression on the Huns and the settle­ ment o f the Goths after the Battle o f Adrianople. All this (and much more) is very clear about the structure o f Gibbon’s His­ tory of the Decline and Fall of the Roman Empire, which recent studies have done much to illuminate in detail.8 But it entirely escaped Bury, who believed that, because Gibbon was a great historian who (unlike David Hume) was still read, he could and should be corrected and brought up to date by judicious addi­ tions to his footnotes and appendices supplied by a scientific historian, so that twentieth-century readers could not only enjoy Gibbons masterly prose but also imbibe accurate and up-to-date information about the period that the De­ cline and Fall covered.9B u ry’s edition contains seven volumes, not six, and it ig­ nores every single one o f Gibbon’s carefully chosen divisions between volumes. Two o f B ury’s deviations from Gibbon’s original structure are particularly unfortunate. Bury put the pair o f intimately connected chapters in which Gibbon analyzed the rise o f Christianity at the beginning o f the second vol­ ume instead o f at the end o f the first as the author did. These famous chap­ ters adopted a tone o f barely concealed hostility toward the Christians o f the first three centuries, and they were greeted with an avalanche o f angry cleri­ cal ripostes, to which Gibbon replied in 1779 with A Vindication of Some Pas­ sages in the Fifteenth and Sixteenth Chapters of the History of the Decline and Fall of the Roman Empire.101 Because o f this controversy, the subsequent five vol­ umes o f the Decline and Fall, published in 1783 and 1788, adopt a significantly different tone toward Christianity and Christians: Gibbon mocks gently with an air o f amused detachment where he had earlier mounted a frontal attack employing outright ridicule.11 Worse still, Bury put Gibbon’s “ General Ob­ servations on the Fall o f the Roman Empire in the West,” which he printed with separately numbered footnotes as an epilogue to both his second instal7 Decline and Fall, 2.384 (B) = 1.796 (W). 8 Esp. L. Braudy, Narrative Form in History and Fiction (Princeton, 1970), 2 2 5 - 6 8 ; Craddock, Gibbon (1989), 1 0 1 - 3 , 14 9 ~ 5 9 , 1 8 4 - 9 3 , 2 1 3 - 1 6 , 2 4 1 - 4 4 . 9 See the “ Introduction by the Editor” in B u ry’s first edition o f Decline and Fall (1 [London, 1896], xxxi-lxviii). 10 O n Gibbons immediate reaction to criticism o f these two chapters, which he later misrep­ resented, see the illuminating essay by D. Womersley, Edward Gibbon and Empire, ed. R. M cK itterick and R. Quinault (Cambridge, 1997), 1 9 0 - 2 1 6 . 11 M . C . Noonkester, H T R 83 (1990), 3 9 9 - 4 1 4 .

[2 2 ]

Sy m m e t r y and St r u c t u r e

ment and the third volume o f his Decline and Fall, in a completely unemphatic position in the middle o f his fourth volume.12 The ancients were extremely sensitive to formal, balanced, symmetrical struc­ ture in literary works on a large scale. Dictionaries and grammars in Late An­ tiquity, it was observed long ago, all comprised exactly twenty books, no more and no less, “ in accordance with the fashion which regulated Latin lexico­ graphical works.” 13 Creative and more original writers in less academic gen­ res either observed the conventions o f symmetry or used them to convey or insinuate an implicit message. Quintilian wrote his Institutio Oratoria in twelve books. He could have chosen ten or possibly fifteen instead, but hardly eleven or thirteen. When Apuleius composed his Metamorphoses in eleven books, he was deliberating flouting, or rather exploiting, convention in order to set the eleventh and last book outside the narrative o f Lucius’ adventures in Greece: he could trust his readers to draw the desired inference that it marked an as­ cent to a higher realm o f reality and describes a genuine conversion thgt stands outside the story o f the “ golden ass.” 14 Augustine knew the convention and tried to observe it. His City of God has a total o f twenty-two books, but it falls into two parts, o f ten and twelve books respectively, which could in their turn be divided into two groups o f five books and three o f four.15 A century ago Curt Wachsmuth identified two patterns widely employed by Greek and Roman authors who wrote long histories divided into many books: some historians arranged their work in multiples o f five or ten books (pentads and decades), and the total number o f books remained a multiple o f ten even when the complexity or proportions o f the material destroyed the internal symmetry o f the work. Others composed in hexads, or groups o f six books. Examples o f the first pattern are easy to find: the history o f early Rome by Dionysius o f Halicarnassus, the Elder Pliny’s Bella Germaniae and Josephus’ Jewish Antiquities, each in twenty books; Polybius and Diodorus Siculus in forty books; and the eighty books o f Cassius D io’s Roman History, much o f 12 Decline and Fall, 4 . 1 7 2 - 8 1 (B) = 2 .5 0 8 - 1 6 (W ). 13 W . M . Lindsay, Nonius Marcellus’ Dictionary o f Republican Latin (Oxford, 1902), 1 - 2 , ad­ ducing also Aulus Gellius and Isidore o f Seville. Lindsay held the fact all the more significant be­ cause Nonius’ materials “ did not lend themselves readily to this division, so that the books are awkwardly uneven in size.” u J. Tatum, Apuleius and " The Golden A ss" (Ithaca, 1979), 2 1 - 3 7 , 8 0 - 9 1 . 15 Augustine, Ep. ad Firmum = Ep. 1A .1 Divjak. For groups o f five books o f larger histori­ cal works (e.g., Livy) transmitted together, C . Wachsmuth, Rb. Mus., N . F. 46 (1891), 3 2 9 - 3 1 . Th e thirteen books o f the Confessions are widely believed to result from Augustine’s insertion o f the tenth book into an original twelve (C P 89 [1994], 294).

[23]

Sy m m e t r y and St r u c t u r e

which falls naturally into sections embracing five or ten books.16 And it may be argued that Livy, who clearly divided the extant part o f his history Ab Urbe Condita into pentads and decades, continued this strict arrangement to the very end o f his work, composing a total o f one hundred and forty books rather than one hundred and forty-two, as the manuscripts o f the Periochae im­ ply (Appendix 2). The prime example o f composition in hexads, often themselves divided into balancing triads, is Tacitus.17 His Annals originally contained eighteen books, his Histories twelve. In the fourth century they circulated as a com­ bined work in thirty books.'8 No reader o f Tacitus can miss the clear division between Annals I—III and IV -V I: Book IV opens with a character sketch o f Sejanus and a survey o f the Roman Empire, both deliberately deferred from Tacitus’ account o f the year 14 in the first book, and Book VI closes with the death o f Tiberius and an analysis o f his character.19 There is a similar division in the six books devoted to the reign o f Nero: the second half o f Book X V narrates the suppression o f the Pisonian conspiracy o f 65, and X V I opens with a symbolic foreshadowing o f Nero’s impending downfall.2021As for the Histo­ ries, Books I—III narrate the civil wars o f 69 and conclude with the death o f Vitellius, whereas the fourth book opens with the consolidation o f the Fla­ vian regime in Rome and a senatorial debate in which Helvidius Priscus con­ fronts Eprius Marcellus, through which Tacitus sets the tone o f his treatment o f Vespasian by contrasting the philosopher who criticized him with a delator who was one o f his closest advisers until he perished in 79, presumably toward the end o f Book V I.2' Hexadic composition can be detected even in the forty-four books o f Pom­ peius Trogus’ Historiae Philippicae, which Wachsmuth analyzed as seven groups 16 Pliny, Ep. 3.5.3/4; C . Wachsmuth, Einleitung in das Studium der alten Geschichte (Leipzig, 1895), 8 3 - 8 9 (Diodorus), 440 (Josephus), 639 (Dionysius), 5 9 7 -6 0 0 (Dio), 6 4 4 - 4 8 (Poly­ bius). For pentadic and decadic structure in Dio, which is obscured by textual losses, see A . von Gutschmid, Kleine Schriften, 5 (Leipzig, 1894), 5 6 1 - 6 2 ; F. Millar, A Study o f Cassius Dio (O x­ ford, 1964), 39. 17 E. Wölfflin, Hermes 21 (1886), 1 5 7 - 5 9 . 18 Jerome, Comm, in Zach. 3 .1 4 .1 / 2 (PL 2 5 .1 5 2 2 = C C L 76A .878): Cornelius quoque Tac­ itus, qui post Augustum usque ad mortem Domitiani vitas Caesarum triginta voluminibus exar­ avit. For p roof that the Annals had eighteen books and the Histories twelve, see R. Syme, Tacitus (Oxford, 1958), 2 1 1 - 1 5 , 2 6 3 - 6 6 , 6 8 6 - 6 7 . For subsequent discussion, G . W ille, Der Aufbau der Werke des Tacitus (Amsterdam, 1983), 2 2 2 - 3 4 1 , 3 5 8 - 6 4 7 . 19 Tacitus, Ann. 4 . 1 - 2 , 5 - 6 , 6 . 5 7 , cf. R. H. Martin and A . Woodman, Tacitus: Annals , Book IV (Cambridge, 1989), 1 2 - 1 4 . 20 Tacitus, Ann. 1 6 . 1 - 3 (Didos gold); cf. A . J. Woodman, Tacitus and the Tacitean Tradition (Princeton, 1993), 1 2 7 - 2 8 . 21 O n the structure o f the Historiaey see esp. Syme, Tacitus, 2 1 1 - 1 6 .

[24]

S y m m e t r y and St r u c t u r e

o f six books with a virtual postscript o f two.22 What then o f Ammianus’ thirty-one books? The number is anomalous. It cannot be justified by appeal to the thirty-seven books o f the Elder Pliny’s Natural History because that to­ tal is reached by prefixing an index and lists o f authorities used, which is num­ bered as Book I, to thirty-six books o f the Natural History proper. Wachsmuth accordingly accepted the theory endorsed (though not invented) by Alfred von Gutschmid that a book has been lost between X X X and X X X I, so that the original number o f books was the aesthetically acceptable total o f thirtytwo.23 This theory is not merely implausible, but impossible, since Ammianus does not in fact completely omit western events after 22 November 375: he omits the political history o f the West after the proclamation o f Valentinian II as Augustus (described in 30 .10 .4 -6 ) so that he can avoid writing about the execution o f the general Theodosius in Carthage during the winter o f 375 — 376, an event too embarrassing for a historian to describe during the reign o f his son.24 Moreover, Ammianus follows Tacitus, who may be his conscious model, in adopting a triadic and hexadic structure. The extant eighteen books o f the Res Gestae have the following obvious hexadic structure: XIV—XIX

Gallus and Julian as Caesars (353—359)

X X —X X V

J u lia n as A u g u s tu s ( 3 6 0 - 3 6 3 )

X X V I-X X X I

T h e re ig n s o f V a le n tin ia n in th e W e st ( 3 6 4 - 3 7 5 ) a n d V alen s in th e E a s t ( 3 6 4 - 3 7 8 )

Furthermore, the first two o f these hexads divide neatly into pairs o f triads with identifiable main themes: XIV-XVI

T h e fall o f G a llu s , th e a p p o in tm e n t o f J u lia n as C a e s a r , an d h is in itia l su c c e sse s in G a u l

XVII-XIX

J u l i a n s c o n t in u in g su c c e ss in G a u l c o n tra s te d w it h C o n s t a n ­ tiu s ’ fa ilu r e in M e s o p o t a m ia

X X -XX II

T h e p r o c la m a t io n o f J u lia n as A u g u s tu s, h is c o n flic t w it h C o n s t a n tiu s , a n d h is ru le as s o le e m p e r o r u n til th e e n d o f 3 6 2

XXIII-XXV

J u lia n ’s P e rs ia n e x p e d it io n a n d its im m e d ia te c o n s e q u e n c e s

22

Wachsmuth, Einleitung, 1 1 3 , commended by R. Syme, Roman Papers, 6 (Oxford, 19 91), 370. 23 Wachsmuth, Einleitung, 6 8 2 —83, cf. A . von Gutschmid, Kleine Schriften 5 (Leipzig, 1894), 5 7 -2 —73- Th e theory that a whole book has been lost between Books X X X and X X X I goes back to Chifflet and Valesius in the seventeenth century.

24 O n this obscure episode, see esp. A . Demandt, Historia 18 (1969), 5 9 8 - 6 3 6 ; R. M . Errington, Klio 78 (1996), 4 4 3 - 4 7 -

Sy m m e t r y and St r u c t u r e

It may be deduced by extrapolation, therefore, that the Res Gestae had a hexadic structure throughout: accordingly, it may be proposed that the full text contained the aesthetically acceptable total o f thirty-six books. That hypoth­ esis may seem too adventurous, too speculative, yet it solves the otherwise in­ tractable problem o f the scope and content o f the lost books. The Res Gestae began with the reign o f Nerva, who was proclaimed emperor after the assassination o f Domitian on 18 September 96, and they end with the aftermath o f the Battle o f Adrianople, which was fought on 9 August 378 (31.16.9, 12.10). Book XIV, the first to survive, opens with a reference back to Constantius’ invasion o f Gaul in the summer o f 353 and a transition to east­ ern affairs during the same summer (14 .1.1), and Book X X X I, which concludes the work, concentrates on the events leading up to the disastrous battle in which the emperor Valens perished. Neither pair o f facts is in itself either con­ troversial or problematical. Taken together, however, they imply a series o f questions that are both troublesome and unavoidable. The first six extant books (X IV -X IX ) cover a period o f six years (353-359), the next six (X X -X X V ) a mere four, from January 360 to February 364. At what point precisely did Ammianus begin to describe events in such amplitude? Where in his narrative did he make the transition from dependence on written sources to contempo­ rary history? And, above all, how was he able to cover two and a half centuries in thirteen books? The vast majority o f modern scholars who have studied Ammianus have found no insuperable difficulty in believing that his account o f the years 9 6 -353 occupied only thirteen books. Too often, however, the problem has been mitigated or removed by making the tacit and mistaken as­ sumption that Ammianus’ narrative expanded to the amplitude o f the first ex­ tant books exactly where the surviving portion commences.25 Thus a standard history o f Latin literature declared that we can reconcile ourselves to the loss o f the first thirteen books “ because the more valuable part, in which Ammi­ anus describes the events o f his own time, has been spared destruction.” 26 Sim­ ilarly, G. B. Pighi suggested that Ammianus covered the century between the assassinations o f Domitian and Commodus in four books, the next ninety yean to the accession o f Diocletian in another four, and the period from 20 N o­ vember 284 to the summer o f 353 in five,27 a reconstruction that entails a very cursory treatment o f the reigns o f Diocletian and Constantine. 25 H. Peter, Wahrheit und Kunst: Geschichtschreibung und Plagiat im klassischen Altertum (Leipzig/Berlin, 19 11 ) , 401: “ der das Werk des Tacitus als Reichsgeschichte . . . fortgesetzt hat, zuerst kurz, von dem vierzehnten an (dem J. 353), dem ersten erhaltenen, schon ausführlich.” 26 M . Schanz, Geschichte der römischen Litteratum 4.1 2 (Munich, 1914)» 95. 27 G . B. Pighi, Ammiani Marcellini Rerum Gestarum capita selecta (Neuchâtel/Paris, 1948), ix (with minor errors o f chronology).

[26]

Sy m m e t r y and S t r u c t u r e

There have so far been two sharp challenges to this uncomfortable ortho­ doxy. In 1880 Hugo Michael developed the hypothesis, stated baldly at the end o f his dissertation o f 1874,20 that Ammianus either wrote two separate works separately numbered or composed a history in two parts: he argued that the first work or part, o f unknown size and scale, comprised a history o f the Roman Empire from Nerva to Constantine, whereas the other was a history o f his own day from the death o f Constantine, o f which the first thirteen books are lost. Michael employed two principal arguments, which he based on extant cross-references to the lost portions o f the work: first, Ammianus’ lost account o f the period 3 3 7 -3 5 3 was as full as the extant account o f the years 353 onward, so that the lost Books I—X III covered only the years 337— 353 and cannot have begun at any point earlier than the death o f Constantine; second, Ammianus’ references in extant excursus to earlier lost excursus im­ ply that he wrote two separate works because he avoided repeating digressions on the same subject within the second, partially extant work, but readily re­ peated the substance o f digressions that had occurred in the first.28 29 Michael’s theory soon encountered trenchant criticism from Ludwig Jeep, who was generally held to have disproved it.30 In the 1960s, however, Henry Rowell reiterated Michael’s conclusions and argued again that the extant Books X IV —X X X I are the last eighteen books o f a contemporary history commenc­ ing in 337 “ when Constantine died and a new era began with the accession o f his sons.” 31 Although Alan Cameron declared himself “ wholly convinced” by Rowell’s reformulation o f Michael’s theory,32 Ronald Syme and Frank Gilliam, who both considered it carefully, were more sceptical and rejected it. Yet neither was completely categorical in his denial. Syme appeared to accept the possibility o f “ two works, perhaps, not one,” but argued that “ two sepa­ rate works, separately numbered, do not have to be accepted,” and avowed a preference for envisaging a single work in two parts, Books I - X covering the years 9 6 -3 3 7 with Book X I beginning in 337 and constituting the first book 28 H . Michael, De Ammiani MarceUini studiis Ciceronianis (Diss. Breslau, 1874), 50, Thesis IV. 29 H . Michael, Die verlorenen Bücher des Ammianus Marcellinus (Prog. Breslau, 1880): follow­ ing the fashion o f his time, Michael also believed that a book had been lost between X X X and X X X I (17, adducing the unfulfilled promises in 2 8 .1.5 7 ; 22.8.35). 30 L. Jeep, Rit. Mus., N . F. 43 (1888), 6 0 - 7 2 , accepted by M . Petschenig, Bursians Jahres­ berichte 7 2 (1892, pub. 1893), 3 - 4 . 31 H. T. Rowell, Ammianus Marcellinus. Soldier-Historian of the Late Roman Empire (Cincinnati, 1964), 1 6 - 2 1 . Th e same scholar later argued, unconvincingly, that Ammianus refers to the city o f Rome in 14.6.2 in a fashion that would be impossible if the same work had included accounts o f the imperial visits o f Diocletian in 303 or o f Constantine in 3 1 2 - 3 1 3 , 3 1 5 , and 326 (H. T. R o w ­ ell, Mélanges d'archéologie, d ’épigraphie et d'histoire offerts àJérôme Carcopino [Paris, 1966], 8 4 5 -4 8 ) . 32 Alan Cameron, Claudian. Poetry and Propaganda at the Court o f Honorius (Oxford, 1970), 359, n. 2.

[27]

Sy m m e t r y and St r u c t u r e

o f Ammianus’ history o f his own time.33 Gilliam commended Syme’s conclu­ sions as reasonable and therefore approximately correct, yet explicitly con­ ceded that the results o f his survey o f Ammianus’ references to the period 1 1 7 —285 “ do not exclude a longer account in a separate work.” 34 Hence it was claimed, as recently as 1984, that “ the ghost o f two works cannot yet be laid to rest.” 35 But five years later John Matthews reformulated the orthodox view and asserted that the Res Gestae was not really a history o f the Roman Empire from 96 to 378, as Ammianus appears to state, but “ in essentials a his­ tory o f the present age, its nature defined by the methods outlined in the pref­ ace to Book X V ,” to which the early books served as “ little more than a sub­ stantial introduction.” 36 Neither the orthodox view, even with this modification, nor the hypothe­ ses o f Michael and Rowell remove a very real difficulty. It is not possible to postulate two separate works, since Ammianus refers back to his account o f the sack o f Seleucia by the generals o f Lucius Verus in the winter o f 16 5 —166 with the words “ as we have related before” (23.6.24: ut ante rettulimus). On the other hand, it would be an extraordinary coincidence if Ammianus’ his­ tory suddenly expanded to its present scale precisely at the point where the sur­ viving books happen to begin— surely too extraordinary to be believed except on strong and compelling evidence. Hence there appears to be a dilemma: Ammianus cannot have compressed his history o f the Roman Empire from 96 to 353 into a mere thirteen books, yet the books that survive are the last eigh­ teen books o f a single historical work that began with 96 and ended in 378. Because the dilemma cannot be escaped in either o f the two ways so far at­ tempted, a new solution o f the problem must be sought. One is available in the hypothesis that the transmitted book numbers are erroneous and that each book originally bore a number five greater than its transmitted number. In other words, let it be postulated that Ammianus wrote a total o f thirty-six books, not thirty-one, arranged in six hexads as follows: 1 - 6 (lost) 7 —1 2 (lost)

F r o m N e r v a ( 9 6 - 9 8 ) to D io c le t ia n ( 2 8 4 - 3 0 5 ) C o n s t a n t in e ( 3 0 6 - 3 3 7 )

13 -18 (lost) Constantius’ rise to sole rule over the Roman Empire (337-353) 1 9 - 2 4 (= X I V - X I X ) Gallus and Ju lia n as C aesars (353-359) 2 5 -3 0 (= X X -X X V )

J u lia n as A u g u s tu s ( 3 6 0 - 3 6 3 )

3 1-3 6 (= X X V I-X X X I)

T h e re ig n s o f V a le n tin ia n a n d V a le n s ( 3 6 4 - 3 7 8 )

33 Syme, Ammianus (1968), 8 - 9 . Similar opinions are repeated by Blockley, Ammianus (1975), 1 2 —15; Sabbah, Méthode (1978), 2. 34 J. F. Gilliam, Bonner Historia-Augusta-Colloquium 1970 ( 1972), 1 2 5 —26. 35 A. Emmett, History and Historians (1984), 49. 4 2 36 Matthews, Ammianus (1989), 27, 30.

[2 8 ]

S y m m e t r y and S t r u c t u r e

This hypothesis cuts the Gordian knot o f the traditional problem by chang­ ing its terms. It postulates that what survives is precisely the second half o f Ammianus’ work, and it thus provides enough space in the early books to ac­ commodate long and frequent excursus on the scale implied by Ammianus’ cross-references (Appendix 3). Can anything more be said in its favor? The preface to Book X V confirms that the narrative reached the amplitude o f the extant books before 353. After concluding Book X IV with the execu­ tion o f the Caesar Gallus and an obituary, Ammianus opens the following book with a personal or programmatic statement: u t c u m q u e p o t u i v e r it a t e m s c r u t a r i,/ e a , q u a e v id e r e lic u it p e r a e ta te m / v e l p e r ­ p le x e in t e r r o g a n d o v e rsa to s in m e d io s c ir e ,/ n a r r a v im u s o r d in e c a s u u m e x p o s it o d iv e r s o r u m ;/ re sid u a , q u a e s e c u tu ru s a p e r ie t t e x t u s ,/ p r o v ir iu m c a p tu lim a tiu s a b s o lv e m u s / n ih il o b tr e c ta to r e s lo n g i, u t p u ta n t, o p e r is fo r m id a n te s . / tu n c e n im la u d a n d a est b r e v it a s ,/ c u m m o ra s r u m p e n s in te m p e s t iv a s / n ih il s u b tra h it c o g ­ n it io n i g e s t o r u m . ( 1 5 . 1 . 1 ) U s in g e v e r y e ffo r t to in v e s tig a t e th e tru th , I h a v e set o u t, in th e o r d e r o f t h e ir o c c u r r e n c e , e v e n ts w h ic h m y a g e a llo w e d m e to se e m y s e lf o r to k n o w b y t h o r o u g h q u e s t io n in g o f th o s e w h o t o o k p a rt in th e m . T h e re st o f m y task, w h ic h th e s e c t io n to f o l lo w w ill re v e a l, 1 sh a ll d is c h a r g e to th e b e s t o f m y a b il­ it y in a m o r e p o lis h e d s ty le , w it h n o fe a r o f th o s e w h o d e n ig r a te a w o r k w h ic h th e y t h in k t o o lo n g . F o r b r e v it y is o n l y la u d a b le w h e n it p re v e n ts t e d iu m a n d ir ­ r e le v a n c e w it h o u t d im in is h in g k n o w le d g e o f th e c o u r s e o f e v e n ts .

This brief paragraph should not be misunderstood. Ammianus here announces no change in historical method whatever, nor does he state that that his nar­ rative is about to become more detailed than it had been for the immediately preceding period.37 On the contrary, he continued to use autopsy and the questioning o f eyewitnesses for the period after 354 as before: it follows that he must have marked the change from using written sources to contemporary history at some earlier point, most probably when he came to the death o f Constantine.38 Ammianus does not promise a more expansive narrative from the beginning o f Book XV, and in fact the scale o f treatment does not expand 37 As has often been assumed, as by Peter, Wahrheit und Kunst ( 19 1 1 ) , 4 10 (“ mit erklärter Vollständigkeit” ); Blockley, Ammianus, (19 75), 12 (who contrasts the narrative o fB o o k X V on­ ward with “ a previous, less detailed style, o f which X I V was the last book” ); Rosen, Ammianus (1982),.76 (“ grössere Ausführlichkeit” ); C . W . Fornara, Cabinet o f the Muses, ed. M . Griffith and D . J. Mastronarde (Atlanta, 1990), 16 4 (“ more expansively than hitherto” ). 38 Fornara, Cabinet o f the Muses (1990), 169, argues that the prefaces to Books X V and X X V I provide “ independent and substantial support” to this natural assumption.

[29]

Sy m m e t r y and St r u c t u r e

significantly until the account ofjulian’s Persian expedition in Books X X I I I X X V : Books X V - X X I I cover approximately one calendar year each (from autumn 354 to the end o f 362). What Ammianus promises at the start o fB o o k X V is neither greater detail nor even probably greater accuracy,39 but greater literary polish.40 His promise should be interpreted in the light o f his later statement that his account ofjulian as emperor, though factually accurate and carefully verified, will be a virtual panegyric (16 .1.3: ad laudativam paene ma­ teriam pertinebit). The paragraph that introduces Book X V foreshadows the opening o fB o o k X V I: it offers an anticipatory justification or apology for the full treatment ofjulian, whose imminent entry into the narrative as Caesar it heralds. The transmitted book numbers constitute an obvious obstacle to the hypoth­ esis that Ammianus Marcellinus originally wrote a history in thirty-six books. The obstacle, however, is not insuperable. The grammarian Priscian in the early sixth century is the sole writer o f Late Antiquity to show any acquain­ tance with Ammianus’ work, with the possible exception o f the author o f the Historia Augusta.41 Priscian quotes Ammianus once to illustrate a linguistic phenomenon that is common in Ammianus and his quotation comes from the very beginning o f Book XIV.42 Since the majority o f Priscian’s quotations come from the first book o f any work that he cites, this fact encourages the suspicion that Priscian did not know the preceding books, from which it may be deduced with a high degree o f probability that they had already been lost within little more than a century after their publication.43 I f that is so, then the numbering o f the extant books soon became vulnerable to accidental al­ teration. Because the manuscripts o f Ammianus and Priscian concur, the trans­ mitted book numbers must go back at least to the Carolingian period. Is it far­ fetched to suggest that the process o f alteration began either with Priscian 39 For this interpretation o f 1 5 .1 .1 , see Rowell, Ammianus (1964), 18 (“ with still greater ac­ curacy” ); I. Lana, La storigrafia latina del IVsecolo d.C . (Turin, 1990), 64 (“ con maggiore accuratezza"). It must be conceded, however, that limatius appears to have one o f these senses in I5.I3-240 The Oxford Latin Dictionary gives a single meaning for the adjective limatus: “ (of writings, etc.) Carefully composed, polished; (also, o f writers and orators)” (1030). That Ammianus uses the word in exactly this sense is confirmed by the passages quoted at T L L 7 .2 .1 4 2 2 .1 6 —48, 14 2 3 .4 9 —52. Despite his earlier misleading gloss, Fornara, Cabinet o f the Muses (1990), 16 5, trans­ lates correctly as “ in a more polished style.” 41 Argued by Syme, Ammianus (1968), 2 5 - 7 1 . 42 Priscian, Inst. Gramm. 9 .51 = H. Keil, Grammatici Latini 2 (Leipzig, 18 55), 4 8 6 .2 7 -4 8 7 .3 : nam quartae (sc. conjugationis) in “ tum” faciunt supinum -, ut “ indulsi indulsum” vel “ indul­ tum,” unde Marcellinus rerum gestaram X I I 1I: tamquam licentia crudelitati indulta (14 .1.4 ). 43 L. Jeep, Philologus 67 (1908), 2 1.

[30]

Sy m m e t r y and St r u c t u r e

himself or with an early copyist who corrupted “ xviiii” into “ xiiii” ? Priscian was widely read and authoritative.44 Moreover, where the book numbers given for his quotations can be checked, as for example in the case o f Livy, they are not invariably correct:45 indeed, a survey designed to estimate the value o f Priscian’s testimony about book numbers for the structure o f Sallust’s lost Historiae concluded that his book numbers were erroneous one time in ten.46 It is not implausible, therefore, to suppose that some conscientious scribe or reader “ corrected” the book numbers in a manuscript o f the rarely read Ammianus Marcellinus to make them conform to what he had read in Pris­ cian. And if a parallel case o f erroneous numbers be demanded, appeal can be made to Seneca’s Natural Questions, where the content o f its preface establishes that the traditional Book III must originally have stood at the beginning o f the work.47 The preceding argument can be summed up succinctly. The transmitted book numbers cannot be reconciled with what Ammianus himself says about the scale and compass o f his history, and they fail to correspond to the ob­ servable symmetrical structure o f the extant books. Hence it is reasonable to conclude that they may be erroneous and that the extant Books X IV —X X X I were originally numbered X I X - X X X V I , so that the first eighteen books o f the Res Gestae have been lost and the second eighteen books have survived. 44 Nearly eight hundred surviving manuscripts were listed by M . Passalacqua, / codici di Prisciano (Rome, 1978), including 128 o f the ninth and tenth centuries. G . Ballaira, Per il catalogo dei codici di Prisciano (Turin, 1982), added another one hundred and thirty. 45 L.Jeep , Philologus 68 (1909), 19 - 2 0 . 46 G. Perl, Philologus i n (1967), 2 8 6 - 8 8 . 47 H. M . Hine, A n Edition with Commentary o f Seneca, Natural Questions, Book II (N ew York, 1981), 2 —23; L. Annaei Senecae Naturalium Quaestionium lihri (Leipzig, 1996), xxii—xxv.

[IV] NARRATIVE AND EXCURSUS

In a history with a formally symmetrical structure o f triads and hexads, there can be no formal or proportional symmetry in the historical narrative itself without the danger o f distortion. Any historian must allow himself the free­ dom to treat some matters briefly, others at greater length— and all readers feel that Ammianus has devoted an excessive amount o f space to Shapur’s campaign o f 359 in northern Mesopotamia and, within that campaign, to the siege o f Amida, which lasted seventy-three days and from which the historian himself barely escaped with his life (19 .1-9 ). On the other hand, Ammianus’ conscious choice to expand his narrative in 363 so that Julian’s Persian expe­ dition occupies almost three books (X X III-X X V ), although only seven months intervened between the departure o f the Roman army from Antioch on 5 March 363 and its return to Antioch in October (23.2.6—25.10 .1), whereas the preceding nine books had covered on average just over one year each, from eastern events in 353 (14 .1—4) to events in Antioch in the autumn o f 362 (2 2 .13 -14 ), reflects the historical importance o f the failed expedition, and Ammianus’ emphasis has an obvious justification. But precisely how has Ammianus constructed his narrative? What are the blocks o f material from which he has put it together? And how has he arranged them? Following the fashion o f the nineteenth century and himself setting an ex­ ample for many scholars in the twentieth, Otto Seeck invoked Quellen­ forschung: in Books X IV -X X V , he contended, Ammianus derived the organi­ zation o f his material from a mechanical combination o f two main written sources; one o f these sources was purely annalistic and arranged events by the consuls o f each calendar year, whereas the other used the chronological order­ ing principle adopted by Thucydides, who dated events by successive summers [32]

N a r r a t i v e and E x c u r s u s

and winters.1 But there is no good reason to deny an author who has devoted such care to the overall architecture o f his history the ability to choose and organize his own material.12 Since Seeck, only two attempts have been made to analyze the overall struc­ ture o f Ammianus’ narrative. Unfortunately, both proceed from mistaken premises, so that, although each established something important, neither pro­ duced a simple, elegant, and systematic analysis. G. B. Pighi saw that Ammi­ anus put his narrative together as series o f narrative sections corresponding to different centers o f activity in succession, but he went astray by accepting the false proposition that Book X IV begins with the winter o f 3 5 3 - 3 5 4 (instead o f eastern affairs during the summer o f 353), and he vainly attempted to see symmetrical composition within each book.3 Christa Samberger saw the sub­ jective nature o f much o f Ammianus’ presentation, the biographical empha­ sis o f much o f his narrative, and the division o f his material into blocks o f narrative dealing separately with the activities o f the emperor (or emperors), events in the city o f Rome, and eastern or western affairs not directly con­ nected with the imperial court, but she attempted to fit Ammianus’ narrative into the straitjacket o f a primary annalistic structure,4 which it simply does not possess (Chapter V).

1 O . Seeck, R E 1 (1894), 18 4 8 -4 9 ; Hermes 41 (1906), 4 9 3 - 9 4 , 5 2 7 - 3 9 . 2 Reading the Past (1990), 6 8 - 7 0 . 3 G . B. Pighi, I discorsi nelle storie di Ammiano Marcellino (Milan, 1936), 6 1 - 8 1 , esp. 81: “ abbiamo primo dimostrato ehe il libro Ammianeo è costituito da serie di sezioni narrative, corrispondenti a centri d’attività, press’a poco contemporanei oppure successivi secondo le vicende della storia delTImpero . . . , e ehe le sezioni sono disposte simmetricamente e collegate spesso per mezzo di formule speciali.” Pighi analysed Book X I V as being composed o f the following five narrative or thematic blocks: I, East (Gallus), autumn 3 5 3 -Septem ber 354; II, West (Constantius), 10 October 3 5 3 -February/M arch 354; III, Rome (first prefecture o f Orfitus), 8 December 3 5 3 - 3 5 6 ; IV, East (Gallus), summer 354; V, West (Constantius), A pril/M a y -N o vem b er 354. Besides allowing no structural function to the two excursus explicitly marked as such (14.4, 6 . 2 27), this analysis paradoxically treats the last journey, the execution, and the obituary o f Gallus all as belonging to “ la narrazione dei fatti di Costanzo” (67). Pighi’s similar analyses o f other books produce equally implausible corollaries (6 8 -7 4 ). Pighi subsequently reproduced his analysis in summary form in his Ammiani Marcellini rerum gestarum capita selecta (Neuchâtel, 1948), x x iv —xxx. 4 C . Samberger, Klio 51 (1969), 3 4 9 -4 8 2 . She speaks continually o f “ der annalistische Aufbau” o f Books X I V - X X V (352) and “ die gewohnte annalistische M ethode” (378), assert­ ing that “ die kleinste kompositionelle Einheit ist der Jahresbericht” (400). She falsely assumes that Book X I V begins with Gallus* activities during the winter o f 3 5 3 - 5 4 (465, 472), while ac­ cusing Ammianus o f having no concern for accurate chronology (405: “ der Historiker, ungeachtet der genauen Chronologie’*).

[3 3 ]

N a rrati ve and E x c u r s u s

Ammianus in fact organized his material in compositional blocks using the following basic types o f unit for constructing his narrative: 1 . T h e a c tiv itie s o f an e m p e r o r d u r in g th e c a m p a ig n in g se a so n 2 . E v e n ts at o r c o n n e c t e d w it h h is c o u r t d u r in g th e s u c c e e d in g w in t e r 3 . E v e n ts at R o m e u n d e r su c c e ssiv e p re fe c ts o f th e c ity 4. P r o v in c ia l e v e n ts o f a p a r tic u la r s u m m e r o r w in t e r 5. F o r m a l e x c u rs u s e x p lic it ly d e sig n a te d as s u c h b y a fo r m u la o f tra n s itio n e i­ th e r at th e b e g in n in g o r e n d o r at b o th b e g in n in g a n d e n d ( A p p e n d i x 5)

The complexity o f the narrative, therefore, varies considerably according to whether or not Ammianus must coordinate (or even relate to each other) the activities o f more than one emperor in separate theaters o f action and on how strictly he observes the principle o f describing the actions o f eastern and west­ ern emperors in each successive summer and winter. For almost all o f the period covered by Books X IV —X X I there were two emperors resident and active in different places. In these books, therefore, six separate and constant categories o f material can be identified: (A ) T h e a c tiv itie s o f C o n s t a n tiu s a n d p o litic a l a n d m ilit a r y e v e n ts at h is c o u r t (B ) T h e a c tiv itie s o f G a llu s ( X I V ) a n d J u lia n ( X V I - X X I ) a n d p o litic a l a n d m ilit a r y e v e n ts at th e ir c o u rts (C ) P re fe c ts o f th e c it y o f R o m e ; 5 (D ) E a s te rn e v e n ts n a rra te d s e p a ra te ly f r o m im p e r ia l a c tiv itie s (E) W e s te rn e v e n ts n a rra te d s e p a ra te ly f r o m im p e r ia l a c tiv itie s (F )

F o r m a l e x c u rs u s

The material in Books X IV - X X I can be assigned to each o f these categories as follows: Book X IV I

C r im e s o f G a llu s (B )

2 - 4 .1

E v e n t s o f s u m m e r 3 5 3 (D )

4 .2 - 7

T h e S a r a c e n i (F)

5 .1- 5

C o n s t a n tiu s w in te r s at A r le s , 3 5 3 - 3 5 4 (A )

5 .6 - 9

P a u l “ th e C h a in ” (E )

6 .1

O r fit u s , 8 D e c e m b e r 3 5 3 to s u m m e r 3 5 5 6 (C )

5 The certain or probable dates at which each prefect entered and left office are, except where otherwise stated, taken from Chastagnol, Fastes (1962), 1 3 9 —93. 6 For proof that Orfitus was replaced by Leontius in 355, not 356, Phoenix 46 (1992), 2 5 7 59; C P 88 (1993), 6 5 - 6 6 .

[34]

N a r r a t i v e and E x c u r s u s 6 .2 - 2 7

R o m e (F)

7 + 9

T h e g o v e r n m e n t o f G a llu s , 3 5 3 - 3 5 4 (B )

8 Provinces of Oriens (F) 10

C a m p a ig n o f C o n s t a n tiu s to th e u p p e r R h in e in 3 5 4 (A )

11.1-

26

G a llu s ’ re c a ll, a rre st, a n d e x e c u t io n (B )

1 1 .2 7 - 3 4

O b it u a r y o f G a llu s (B )

Book X V 1-2

A c c u s a tio n s at c o u r t (A )

3 .1-

6

3 * 7 —1 1

P u n is h m e n t o f G a llu s ’ s u p p o rte rs (D ) T h e a ffa ir o f A fr ic a n u s (E)

4 .1 + 7 - 1 3 4 .2 -

6

5 .1-

34

C a m p a ig n o f C o n s t a n tiu s , 3 5 5 (A )

T h e B o d e n s e e (F) S ilv a n u s (E )

5-

35 ~ 3 6

6

R e p r e s s io n b y P a u l (E )

7

L e o n t iu s , s u m m e r 3 5 5 to w in t e r 3 5 6 - 3 5 7 (C )

8

J u lia n p r o c la im e d C a e s a r (6 N o v e m b e r 3 5 5 ) a n d sen t to G a u l (A )

9 —1 2

R e a c t io n o f C o n s t a n tiu s to th e s u p p re s s io n o f S ilv a n u s (A )

G a u l (F )

13

T h e c o r r u p t io n o f M u s o n ia n u s a n d P r o s p e r 7 (D )

Book X V I I-

5

J u lia n in 3 5 6 (B )

6 -

8

A c c u s a tio n s at c o u r t (A )

9

P e rs ia n raid s a n d t h e ir s e q u e l (D )

10

C o n s t a n tiu s visits R o m e (28 A p r i l - 2 9 M a y 3 5 7 ) , th e n d e p a rts in h aste f o r I lly r ic u m (A )

II-

1 2 .6 6

1 2 . 6 7 —7 0

C a m p a ig n o f j u l i a n in 3 5 7 (B ) C o n s t a n t iu s ’ r e a c tio n to J u lia n ’s v ic t o r y (A )

Book XVII 1-3

C a m p a ig n o f 3 5 7 c o n t in u e d (B )

4 .1

O r fit u s f o r th e s e c o n d tim e , s p r in g 3 5 7 to s p r in g 3 5 9 (C )

4 .2 -

7

23

O b e lis k s in R o m e (F)

5

E x c h a n g e o f le tte rs b e tw e e n S h a p u r a n d C o n s ta n tiu s (A )

6

Iuthungi invade Raetia (E) Seyfarth’s clausulae in 15 .13 .3 are potentially misleading. Since Prosper was the deputy o f

Ursicinus in the East while the latter was absent in Gaul ( 14 .11 .5 , cf. 17 .5 .1 5 , 14 .1), the sentence should be articulated as follows: hunc Prosper adaequitabat/ pro magistro equitum agente etiamtum in Galliis/ militem regens.

[35]

N a r r at i ve and E x c u r s u s 7 .1-

8

E a r th q u a k e o f 2 4 A u g u s t 3 5 8 8 (D )

7 .9 - 14

E a r th q u a k e s (F)

8 -10

J u lia n s c a m p a ig n o f 3 5 8 (B )

11.1-

4

1 1 .5

R e a c t io n at th e c o u r t o f C o n s t a n tiu s (A )

A r t e m iu s vicarius o f R o m e in p la c e o f th e d e c e a s e d B a s s u s 9 (C )

12 -13 14

C o n s t a n t iu s ’ c a m p a ig n o f 3 5 8 (A )

N e g o tia tio n s w ith P ersia (D )

Book XVIII 1

A d m in is t r a tio n o f J u lia n (B )

2

H is c a m p a ig n o f 3 5 9 (B )

3

E x e c u t io n o f B a r b a tio (A )

4 -10

S a p o r in v a d e s R o m a n t e r r it o r y (A )

Book X IX 1-9

T h e s ie g e a n d c a p tu re o f A m id a (D )

10

T e rtu llu s, ?36 o to a u tu m n 3 6 1 10 1 (C )

11

C a m p a ig n a g a in st S a rm a ta e L im ig a n te s (A )

12 .1-

18

12 .19 - 2 0 13

T ria ls fo r tre a so n (D ) P r o d ig y at D a p h n e (D )

Isa u ria n raid s (D )

Book X X 1

Ju lia n sen d s L u p ic in u s to B r it a in (e a rly 36 0 ) (B )

2

R e c a ll o f U rs ic in u s (A )

3. i

E c lip s e o f th e s u n 11 (D )

3 .2 -

12

4 -5

Ju lia n p r o c la im e d A u g u s tu s, s p r in g 3 6 0 (B)

6 -7

S h a p u r attack s S in g a ra , s u m m e r 3 5 9 (D )

8 .1

E c lip s e s (F)

C o n s t a n tiu s w in te r s in C o n s t a n t in o p le , th e n g o e s east (A )

8 .2 -

22

9 .1-

5

9 .6 —8

Ju lia n w r ite s to C o n s t a n tiu s (B ) C o n s t a n tiu s sen d s L e o n a s to G a u l, s p r in g 3 6 0 (A )

Ju lia n re c e iv e s L e o n a s a n d re p lie s (B )

9 .9

S u s p ic io n o f L u p ic in u s (E )

10

J u lia n ’s c a m p a ig n o f 3 6 0 (B )

11

C o n s t a n tiu s tak es th e fie ld a g a in st th e P e rsia n s (A )

11.2 6 - 3 0

R a in b o w s (F)

8 Ammianus’ exact date is confirmed by Descr. cons. 358.2. 9 Orfitus was still prefect on 25 March 359 ( C 77 i 14 .6 .1), and Bassus died on 25 August 359 (ILS 1286). 10 Th e date at which Tertullus became prefect is not certain (Appendix 7). 11 On the problems raised by this notice, see Chapter IX.

[36]

N a rrati ve and E x c u r s us

Book X X I 1-5 Julian 1.7 - 1 4

p re p a re s fo r w a r (B )

D iv in a t io n (E )

6

A c t iv it ie s o f C o n s t a n tiu s d u r in g th e w in t e r o f 3 6 0 - 3 6 1 (A )

7

C o n s t a n tiu s a d v a n c e s in to M e s o p o t a m ia (A )

8 -10 11

J u lia n a d v a n c e s in to I lly r ic u m (B )

—1 2 .2 0

12 .2 1- 2 5

R e s is ta n c e o f t w o le g io n s at A q u ile ia (E ) A c t io n s o f J u lia n at N a issu s (B )

12 .2 4 e

M a x im u s , a u tu m n 3 6 1 - s p r i n g 3 6 3 12 (C )

13 -15

C o n s t a n tiu s p re p a re s f o r w a r a g a in st Ju lia n , b u t d ie s (3 N o v e m ­ b e r 3 6 1 ) s h o r t ly a fte r le a v in g A n t io c h (A )

16

O b it u a r y o f C o n s t a n tiu s (A )

Between the deaths o f Constantius on 3 November 361, whose obituary and burial conclude Book X X I, and ofjovian on 1 6 —17 January 364, with which Book X X V closes, there was only a single Roman emperor. For this period, therefore, Ammianus’ narrative structure is correspondingly simpler, and he continues this simpler structure with a single main narrative into Book X X V I, which describes the proclamation o f Valentinian and Valens as emperors and their activities until they separated and divided the empire in 364. Hence the narrative structure o f Books X X I I - X X V I may be analyzed as follows: (A ) M a in p o litic a l a n d m ilit a r y n a rra tiv e (B ) P r e fe c ts o f R o m e (w it h c e rta in o r p r o b a b le d ates o f te n u re n o t e d as b e fo re ) (C ) O t h e r e v e n ts (D ) E x c u r s u s , m e t h o d o lo g ic a l a n d r e fle c tiv e as w e ll as fo r m a l

The material in Books X X I —X X V I can be assigned to each o f these four cat­ egories as follows: Book XXII 1 —2

J u lia n w a its u n til h e h ears o f C o n s t a n t iu s ’ d e a th , th e n p ro c e e d s to C o n s t a n t in o p le (A )

3

N e w a p p o in tm e n ts to h ig h o ffic e (A )

4

R e o r g a n iz a t io n o f th e p a la c e (A )

12 Maximus is attested as prefect on 28 January 362 (IC U R , N .S. 4 .117 5 8 = I L C V 3904), cf. C IL 6 .3 14 0 1; Symmachus, Rel. 34.5 [both undated]): his successor was L. Turcius Apronianus Asterius, w hom Julian appointed in Antioch in January 363 (23.1.4,; 2 6 .3 .1 - 6 ) . In defense o f the transmitted date o f C T h 14.4.3 (9 December 363), which Seeck, Regesten (1919), 84, 2 1 1 , followed by Chastagnol, Fastes (1962), 1 5 6 - 1 5 7 , emended to 9 December 362, see Cognitio

Gestorum (1992), 5 —6.

[37]

N a r r at i ve and E x c u r s u s 5

R e e s ta b lis h m e n t o f p a g a n ism (A )

6

Ju lia n an d th e E g y p t ia n litig a n ts (A )

7

O t h e r a c tiv itie s o f Ju lia n in C o n s t a n t in o p le (A )

8

T h e co asts o f T h r a c e a n d P o n tu s (D )

9

Ju lia n travels acro ss A sia M i n o r to A n t io c h (A )

10

Ju lia n in A n t io c h (A )

1 1 .i- 3

T h r e e e x e c u t io n s (C ) ’

1 1 .4 - 11 12

T h e d e a th o f b is h o p G e o r g e in A l e x a n d r i a 13 (C )

J u lia n s r e lig io s ity (A )

1 3 .1-

4

13 .4

T h e b u r n in g o f th e te m p le o f A p o llo a n d its s e q u e l (A )

E a r th q u a k e at N ic o m e d ia a n d N ic a e a (C )

14 .1-

3

J u lia n s d iffic u ltie s w it h th e A n t io c h e n e s (A )

4 -5

H e asce n d s M o u n t C a s iu s (A )

14 .6

R e p o r t o f an A p is b u ll (C )

14 .7 - 8

T h e A p is b u ll (D )

15 -16

E g y p t (D )

Book XXIII J u lia n s fo u r t h c o n s u la te : e v e n ts, a p p o in tm e n ts , a n d o m e n s o f J a n u a r y

I

363 2 -3

(A )

T h e start o f th e P e rsia n e x p e d it io n (A )

4

M ilit a r y e n g in e s (D )

5

Ju lia n at C e r c u s iu m a n d Z a it h a (A )

6 .1-

84

P e rsia (D )

6 .8 5 -8 8

P earls (D )

Book X X IV 1-6

J u lia n s a d v a n c e in to P e rsia as fa r as C t e s ip h o n (A )

7 .1-

2 + la c u n a

< la c u n a >

T h e R o m a n s b e fo r e C t e s ip h o n (A )

T h e a c tio n s o f A rs a c e s , S e b a stia n u s, a n d P r o c o p iu s ( re fe re n c e b a c k at 7 .8 ) (A )

8

T h e start o f th e re tre a t (A )

Book X X V 1- 3 .14 3 .15 - 2 3

13

T h e R o m a n re tre a t a n d th e fatal w o u n d in g o f J u lia n (A ) T h e d e a th o f J u lia n (A )

George was killed on 24 December 361 (Historia acephala 2 .10 Martin), long before Julian

reached Antioch on 18 or 19 July 362: Ammianus 2 2 .9 .15, cf. F. Cum ont, Syria 8 (1927), 3 3 9 “ 4 1. On Ammianus’ error, see Appendix 7.

[3 8 ]

N a r r at i ve and E x c u r s u s 4

O b it u a r y o f J u lia n (A )

5

T h e e le c t io n o f J o v i a n as e m p e r o r (A )

6

T h e R o m a n re tre a t c o n tin u e s (A )

7

S h a p u r o ffe rs p e a c e a n d his h arsh te rm s are a c c e p te d (A )

8

T h e R o m a n a r m y m a rc h e s to N is ib is (A )

9 .1-

6

T h e s u r r e n d e r o f N is ib is (A )

9 .7 -

11

C o m p a r is o n to e a rlie r R o m a n d efe a ts (D )

12 -13

P r o c o p iu s e sc o rts th e b o d y o f J u lia n to T arsu s, th e n d isap p ears (A )

10 .11 0 .3

2

T h e R o m a n a r m y re a c h e s A n t io c h (A )

C o m e t s (D )

10 .4 - 1 3

J o v i a n p r o c e e d s fr o m A n t io c h to B it h y n ia , w h e r e h e d ies (A )

i o. 1 4 - 1 7

O b it u a r y o f J o v ia n (A )

Book X X V I 1.1-

2

T h e p r o b le m s o f w r it in g re c e n t h is t o r y (D )

1.2 -

2 .1 1

1.8 -

14

3 4 .14 .4

T h e e le c t io n o f V a le n tin ia n as e m p e r o r (A ) L e a p y e a rs (D )

A p r o n ia n u s , s p r in g 3 6 3 to e a r ly 3 6 4 14 (B ) 3

V a le n tin ia n p ro c la im s V a le n s his c o lle a g u e (A )

Illn ess o f th e e m p e ro rs (A )

4 .5 -

6

R o m e ’s fr o n tie r s m e n a c e d b y e n e m ie s (C )

5 .1-

5

T h e e m p e ro rs g o to S ir m iu m a n d d iv id e th e e m p ir e , th e n d e p a rt f o r M ila n a n d C o n s t a n t in o p le (A )

5 .6 -

8

T h e y e a r 3 6 5 b r in g s t r o u b le : th e A la m a n n i cro ss th e fr o n t ie r an d P r o c o p iu s re b e ls (A )

5 .9 -

14

5 .15

T h e s tr u c tu r e o f th e n a rra tiv e (D )

6 - 10 .14

V a le n tin ia n a n d th e n o r t h e r n fro n tie rs (A )

T h e re b e llio n o f P r o c o p iu s , its su p p re ssio n b y V alen s an d th e s u b se q u e n t p u r g e (A )

10 .15 - 19

T h e tsu n a m i o f 2 1 J u l y 3 6 5 15 (C )

It is significant that Ammianus felt himself impelled to digress to explain the necessity o f narrating separately contemporaneous events in East and West as soon as he began to describe the activities o f Valentinian and Valens after they set up separate administrations in East and West (26.5.15). In Books X X V I IX X X this separation is scrupulously observed, but without treating the activ­ 14 Apronianus was appointed by Julian in Antioch in January (23.1.4), but Ammianus implies that he had reached Rome by 19 March (23.3.3). 15 For the widespread damage caused by this tsunami and the significance read into it by con­ temporaries, see B. Bousquet and F. Jacques, M E F R (A) 96 (1984), 4 2 3 - 6 1 ; C . Lepelley, ibid., 4 6 3 -9 0 ; M . Henry, Phoenix 39 (1985), 3 6 - 6 1 ; C . Lepelley, Kokalos 3 6 - 3 7 (19 9 0 -9 i)> 3 5 9 - 7 2 .

[39)

N a r r a t i v e and E x c u r s u s

ities o f each o f the two emperors summer by summer and winter by winter as compositional units, as in Books X IV —X X I. Hence the structure o f Books X X V I I - X X X can be analyzed as follows: (A ) E a s te rn p o litic a l an d m ilit a r y e v e n ts (B ) W e ste rn p o litic a l an d m ilit a r y e v e n t s 16 (C ) P re fe c ts o f th e c it y o f R o m e

.

(D ) F o r m a l e x c u rs u s a n d a u th o r ia l c o m m e n ts

The material in Books X X V I I - X X X can be assigned to each o f these four categories as follows: Book XXVII 1-2

T r o u b le w it h th e A la m a n n i (e a rly 36 6 ) (B )

3 .1-

2

3 .3 - 4

P o r t e n t at P is to r ia (B ) S y m m a c h u s , A p r il 3 6 4 - M a r c h 3 6 5 (C )

3 .5 -

10

L a m p a d iu s , A p r i l - S e p t e m b e r 3 6 5 (C )

3 .11- 13

V iv e n tiu s , O c t o b e r 3 6 5 - M a y 3 6 7 (C )

3 .1 4 - 1 5

B is h o p s (D )

4. i

V a le n s p re p a re s fo r w a r a g a in st th e G o t h s (A )

4 .2 -

14

T h r a c e (D )

5

V a le n s ’ G o t h ic w a r, 3 6 7 - 3 6 9 (A )

6

Illn ess o f V a le n tin ia n an d p r o c la m a t io n o f G r a tia n as A u g u s tu s , 3 6 7 (B )

7

C o n d e m n a t io n s o f v a r io u s p e rso n s b y V a le n tin ia n (B )

8

P r o b le m s in B r it a in (B )

9 .1-

5

R o m a n u s in A f r ic a (B )

9 .6 -

7

Isa u ria n raid s (A )

9 .8 - 10

P ra e te x ta tu s, A u g u s t 3 6 7 - S e p t e m b e r 3 6 8 (C )

10

V a le n tin ia n c a m p a ig n s o n th e R h in e (B )

11

P r o b u s as p r a e to r ia n p r e fe c t o f I l l y r i c u m 17 (B )

12

C o n f lic t w it h P ersia o v e r A r m e n ia (A )

Book XXVIII 1 . 1, 5 - 5 6 1.2 -

4

T ria ls at R o m e , 3 6 9 - 3 7 0 to 3 7 4 18 (B ) P h r y n ic h u s (D )

Ul On the movements o f Valentinian and Valens, see, respectively, Seeck, Regesten (1919), 2 1 6 - 4 6 ; Appendix 10. 17 Petronius Probus was prefect o f Illyricum, Italy, and Africa from 368 to 375: Ammianus 2 7 .1 .1 ; 3 0 . 5 . 4 - 1 1 ; Rufmus, H E 1 1 .1 2 , cf. P L R E 1 .7 3 7 - 3 9 . IS On the chronology o f this chapter, see Appendix 9.

[40]

N a r r a t i v e and E x c u r s u s 1.5 7

T h e deaths o f M ax im in u s, Sim pliciu s, and D o r y p h o r ia n u s 19 (D)

2 .1 -

10

Valentinian on the R h in e (B)

2 .1 1 -1 4 3

T h e M arato cu p ren i in Syria (A)

T h eo d o siu s in B ritain (B)

41-2

O ly b riu s, Ja n u a ry 3 6 9 - A u g u s t 3 7 0 (C )

4 .3 -

5

A m p eliu s, Ja n u a ry 3 7 1 - J u l y 3 7 2 (C )

4 . 6 —3 5

T h e R o m an aristocracy (D)

5 .1-

7

R aid b y Saxon es, 3 7 0 (B)

5 .8 -

15

T h e R h in e frontier (B)

6

T ro u bles in Tripolitania, 3 6 4 o n w ard (B)

Book X X IX 1 —2

Treason trials in the East, 3 7 2 o n w ard (A)

3

T h e cru elty o f Valentinian (B)

4

Valentinian cam paigns across the R h in e (B)

< la c u n a >

B appo

< la c u n a >

E u p r a x iu s

5

(attested 2 2 A u g u st 3 7 2 ) (C ) (attested 14 F e b r u a r y 3 7 4 ) 20 (C )

Th eodosiu s* cam paign in M auretania (B)

6 .1 —16

T h e Q u a d i invade Illyricu m (B)

6 .17 -1 9

C lau d iu s, 2 1 M a y - 19 J u l y 3 7 4 (C )

Book X X X i

T h e k illing o f K in g Pap o f A rm en ia, p ro b ab ly 3 7 4 21 (A)

2 .1 -8

C o n flic t w ith Sap o r over A rm e n ia and Iberia (A)

2 .9 3

12

T h e death o f R em igiu s (B)

Valentinian cam paigns on the R h in e, 3 7 4 (B)

4 .1-

2

T h e adm instration o f Valens (A)

4 .3 -

22

5 -6

Valentinian leaves T rie r in the spring and goes to Pannonia w h e re he

Law yers (D)

dies, 1 7 N o v . 3 7 5 (B) 7 -9 10

O b itu a ry o f V alentinian (B) Valentinian II p ro claim ed A ugustus, 2 2 N o v . 3 7 5 22 (B )

Am m ianus here promises to describe their deaths, w hich occurred in the early months o f 3 7 6 (Appendix 9), but in the event breaks o ff his continuous narrative o f western affairs with the proclamation o f Valentinian II as Augustus on 2 2 Novem ber 3 7 5 ( 3 0 .1 0 .4 - 6 ) . 20 O n the large lacuna in B o o k X X I X , see Appendix 8.

21 P L R E 1.6 6 5 -6 6 . 22 Am mianus* dates o f 17 and 22 N o vem ber are confirm ed by Descr. cons. 3 7 5 .2 , 3; Socrates,

H E 4 .3 1.6 -7 .

[41]

N a r r a t i v e and E x c u r s u s

After the proclamation o f the infant Valentinian as Augustus (30 .10 .4 -6 ), Ammianus does not attempt to give a full or connected account o f western events. Hence there are no notices o f the prefects o f the city o f Rome who held office after Claudius (29.6.17-19). Book X X X I focuses on the events leading to the disaster o f Adrianople, and its structure may be analyzed as comprising (A) a single main narrative with (B) two formal excursus and an epilogue: 1

P o rte n ts p r e d ic t in g th e d e a th o f V a le n s (A )

2

T h e H u n s a n d A la n s (B )

3 - 5 .9

U n d e r p re ssu re , th e G o t h s r e q u e s t a n d a re g r a n te d p e r m is s io n to settle so u th o f th e D a n u b e , w h e r e e x p lo it a t io n b y R o m a n o ffic ia ls le a d s to in s u r r e c tio n (A)

5 .1 0 - 1 7 6 -9 io

R o m a n r e c o v e r y fr o m p ast d isasters (B )

T h e G o t h s an d R o m a n tro o p s in 3 7 6 a n d 3 7 7 (A ) D e s p ite tr o u b le w it h th e L e n tie n se s, G r a t ia n p re p a re s to assist V a le n s a g a in st th e G o t h s (A )

11.1-5 1 1.6

V alen s leaves A n t io c h a n d travels to T h r a c e (A ) G r a tia n re a c h e s C a s tr a M a r tis (A )

12-13 14

T h e c a m p a ig n o f A d r ia n o p le (A )

O b it u a r y o f V a le n s (A )

15-16 .7

T h e im m e d ia te a fte rm a th o f th e b a ttle (A )

16 .8

K illin g o f G o t h s t h r o u g h o u t th e E a s t (A )

16 .9

A u t h o r ’s fa r e w e ll to h is re a d e rs (B )

[42]

[V] DATING, EMPHASIS, AND OMISSION

As an annalistic historian, Tacitus could manipulate both his annalistic frame­ work and the interplay between it and the division into books in order to em­ phasize an event or an episode o f his choosing, which he wished to invest with particular significance. ' Unlike Livy, moreover, Tacitus did not always begin his account o f each year with the entry o f the consuls into office and events in Rome at the start o f the year, nor did he record even events at Rome in strict chronological order.12 His account o f the year 15, for example, begins with an anticipatory notice that a triumph was decreed to Germanicus for victories won in a spring campaign against the Chatti and a summer campaign against the Cherusci: Tacitus then proceeds to describe the two campaigns, concluding with the award o f insignia triumphalia to A. Caecina, L. Apronius, and C . Silius, the lieutenants o f Germanicus, for their services during these campaigns (which the Senate must have decreed at the same time as it decreed the triumph recorded earlier), before turning to affairs in Rome and Tiberius’ refusal to allow oaths to be sworn to accept his acta, presumably on 1 January.3 Similarly, Tacitus’ account o f the year 17 begins with a formal notice that Ger­ manicus celebrated a triumph over the Cherusci, Chatti, Angrivarii, and the 1 J. Ginsburg, Tradition and Theme in the “Annals” o f Tacitus (N ew York, 19 81), 9 6 - 1 0 0 ; R. H. Martin and A. Woodman, Tacitus: Annals Book I V (Cambridge, 1989), 1 5 - 1 9 . 2 See the analysis o f material in the first six books o f the Annals in Ginsburg, Tradition (19 81), 5 7 -7 9 , 12 8 -4 3. 3 Tacitus, Ann. 1 . 5 5 - 7 2 . 1 , cf. D. Timpe, Der Triumph des Germanicus: Untersuchungen zu den Feldzügen derJahre 1 4 - 1 6 n. Chr. in Germanien (Bonn, 1968), 4 3 - 5 8 ; Ginsburg, Tradition (19 81), 2 1, 6 7 - 7 2 . Ginsburg reconstructs the actual sequence o f the events recorded by Tacitus for 15 as follows: Ann. 1 . 7 2 . 2 - 8 1 (January to March), 5 5 . 1 b—7 1 (spring and summer), 5 5 . 1 s, 7 2 .1 (autumn).

[43]

D a t i n g , E m p h a s i s , and O mission

other tribes up to the Elbe on 25 May.4 This technique is sometimes relevant to evaluating Tacitus’ accuracy. He begins the year 61 with the stark assertion that “ in the consulate o f Caesennius Paetus and Petronius Turpilianus a grave disaster was sustained in Britain,” then describes the events leading up to the rebellion o f Boudicca, the rebellion itself, and the appointment o f a new gov­ ernor, presumably during the winter o f 6 1 —62, before turning to the events o f 61 in Rome.5 Many modern students o f Roman Britain have shown a per­ verse resolve to reject Tacitus’ explicit and emphatic date for the rebellion in favor o f 60 simply because he notes that Petronius Turpilianus had relinquished his consulate before he was sent to Britain to replace Suetonius Paulinus.6 For the most part in the Annals, Tacitus made the beginning and end o f each book coincide with the beginning and end o f his acount o f a consular year. Sometimes he exploited the coincidence to great effect. The year 57 and Book X III o f the Annals end with the withering o f the ficus ruminalis and its subsequent revival: Tacitus did not need to tell his readers explicitly that the withering o f the tree portended an early death for Nero, its recovery the ad­ vent o f the Flavian dynasty.7 But Tacitus can also make the break between books and the start o f a new year diverge to telling effect. After the death o f Germanicus in 19 and Piso’s rash attempt to recover control o f the province o f Syria, the second book o f the Annals closes with a series o f annalistic no­ tices that lead up to an obituary o f Arminius.8 Book III then opens with the winter journey o f Agrippina from Antioch to Rome carrying the ashes o f her dead husband: the consuls o f 20 appear as both participants in the action and as marking the transition from 19 to 20 when they greet her after she has landed in Italy.9 Ammianus was not an annalistic historian. For him the primary chronologi­ cal units o f his main narrative were the actions o f an emperor during cam4 Tacitus, A n n . 2 .4 1 .2 . T h e date is confirm ed b y the Fasti Ostienses and the Fasti A m iterni (Insc. Ital. 1 3 . 1 . 1 8 5 , 3 .1 8 7 ) . s Tacitus, A n n . 1 4 . 2 9 - 3 9 (Britain), 4 0 - 4 7 (events in Rom e). 6 Tacitus, A n n . 1 4 .3 9 .3 . In favor o f accepting Tacitus’ explicit statement that “ Caesennio Paeto et Petronio Turpiliano consulibus gravis clades in Britannia accepta” (2 9 .1), see K . K . C a r roll, Britannia 10 (19 79 ), 1 9 7 - 2 0 2 . 7 Tacitus, A n n . 1 3 .5 8 ; C . Segal, Ramus 2 ( 19 7 3 ) , 1 0 7 - 2 6 ; H . Y . M c C u llo c h , P hoenix 34 (1980), 2 3 7 - 4 2 . s Tacitus, A n n . 2 . 8 4 - 8 8 , cf. Ginsburg, Tradition ( 19 8 1), 3 6 —38. T h ere is, how ever, no co m ­ pelling reason to reject Tacitus’ date o f 19 for the death o f Arm inius in favor o f 2 1 , as is often done: D . T im p e , A rm in iu s-Stu dien (Heidelberg, 19 70 ), 2 4 - 2 5 ; F. R. D . G oodyear, T he A n n als o f Tacitus, Books 1 - 6 , 2 (Cam bridge, 19 8 1 ) , 4 47. y Tacitus, A n n . 3 .2 .3 , cf. Ginsburg, Tradition ( 1 9 8 1 ), 19 , 58, 106, n. 3, w h o draws attention to the very first words o f B o o k III: “ nihil intermissa navigatione hiberni maris Agrippina etc.”

(3 . 1. 0 -

[44]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

paigning seasons and the intervening winters (Chapter IV). Hence he could not use the interplay o f annalistic and formal structure as Tacitus had. In Am ­ mianus the consuls appear for the most part almost incidentally. Moreover, his continuous and consecutive narrative o f imperial activities season by season ends with the rebellion o f Procopius. The transition between Books X X V I and X X V II is managed as previous transitions had been. Book X X V I con­ cludes its account o f eastern events in 366, then explicitly regresses to register the tsunami o f 21 July 365, whose date is given by day, month, and consuls (26.10.15). Book X X V II opens with a reference to the kalends o f January (2 7 .1.1) , which are later specified as those o f the year 367 (2.1). Thereafter, however, Ammianus ceases to register the consuls with any consistency: those o f 368 are nowhere named, those o f 369 only in passing to date a Saxon raid (28.5.1) , and the Res Gestae, at least as extant, name the consuls o f only two years after 369 (viz., 374 and 377) (Appendix 4). Nonetheless, Ammianus was not unaware that consular dates could be used for emphasis— and he so uses them in relation to his hero Julian.10 For two o f the ten years between 354 and 363, Ammianus notes the transi­ tion from one consular year to the next with incidental notices that have no literary function in ordering his narrative. Thus the consuls o f 354 and 357 are noted when Constantius and Julian depart from winter quarters in the spring o f these yean (14 .10 .1; 16 .11.1) . For the other eight yean, Ammianus exploits the consular formula in a variety o f ways, but almost always to en­ hance his favorable depiction o f Julian. The entry into office o f Arbitio and Lollianus as the consuls o f 355 is not signaled at all when the narrative moves from 354 into 355 with Constantius’ departure from winter quarten into Raetia on campaign (15.4.1). Instead, Ammianus holds back his annalistic notice until he can state that Julian was proclaimed Caesar on 6 November in the consulate o f the year when Arbitio and Lollianus were consuls (15.8.17). On 1 January 356 the newly proclaimed Caesar assumed the consular fasces in Vienne, which provides Ammianus with the opportunity to add a brief panegyric o f the new ruler (16.1). For 358, for the first time in the extant books, Ammianus uses the tradi­ tional consular dating formula with two names in the ablative case: it stands in the middle o f a book, where it marks a sudden, stark transition from Ju ­ lian’s success in Gaul to the impending danger on the eastern frontier (16.5.1: Datiano et Cereali consulibus). The entry into office o f the brothers Eusebius and Hypatius on 1 January 359 opens Book X V III and introduces a vignette o f Julian as an excellent civil administrator and judge (18.1). Similarly, Book X X 10 Cognitio Gestorum (19 9 2), 1 - 8 .

[45]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

opens by noting the consuls o f 360 in its second sentence as the narrative turns from events in Illyricum to the proclamation ofjulian as Augustus (20.1.2). In 361, by contrast, the entry into office o f the consuls Taurus and Florentius (21.6.5) does not mark the progression o f the narrative from one year to the next: that has already occurred when Julian, after celebrating his quinquenna­ lia (21 . 1 . 4), attends a Christian service on the festival “ which Christians cel­ ebrate in the month o f January and call Epiphany” (21.2.5). The consuls o f 362, however, are noted no fewer than three times: in comment on Julian’s complaint that Constantine conferred consulates on “ barbarians” (21.10.8), whenjulian, then still at Naissus, designated Mamertinus and Nevitta (21.12.5), and in describing Julian’s behavior on 1 January 362 in Constantinople (22.7.1). But Ammianus’ most elaborate use o f a consular notice occurs at the very start o f the triad o f books devoted to Julian’s Persian expedition. On i January 363 Julian assumed the consular fasces in Antioch. The em­ peror was consul for the fourth time, with Flavius Sallustius, the praetorian prefect o f Gaul as his colleague (23.1.1). B y 363 the Kalends o f January had become an important festival throughout the Greek East, especially when the day inaugurated an imperial consulate.11 Ammianus uses the opportunity to enhance his presentation ofjulian as a hero possessing civilitas, which was one o f the traditional virtues o f a pagan Roman emperor, to demythologize an episode o f which Christian writers made much, and to show how the Persian expedition was from the start doomed to defeat as contrary to the expressed will o f the gods. After a sentence marking the transition from one calendar year to the next, Ammianus names the consuls o f 363. An authorial comment follows. It was a novelty for an emperor to take as his consular colleague a privatus, that is, a man not related to him, for no one could recall such a pair o f consuls since Diocletian and Aristobulus. The precedent that Ammianus alleges is doubly peculiar. First, there had been a similar consular pair three years after Diocle­ tian and Aristobulus: the consuls o f 288 were Maximian, the Augustus o f the West, and Pompeius Januarianus.1112 Second, although the consular lists that survive from Late Antiquity do indeed have Diocletiano et Aristobulo for the 11 M . Meslin, La Fête des keilendes de janvier dans l ’empire romain. Etude d ’un rituel de Nouvel A n ( Collection Latomus 1 1 5 , 1970), 5 1 - 7 0 ; J. R. Rea, Proceedings o f the X V I I I International Congress o f Papyrology, Athens 2 5 - 3 1 M ay 1986, 2 (Athens, 1986), 2 0 3 - 0 8 ; Oxyrhynclius Papyri 55 (Lon­ don, 1988), 198 (on P. O xy. 3 8 1 2 .5 - 6 ) . It may be inferred that it owed its increased importance in the Christian Roman Empire to the fact that it took over the functions of, and hence pro­ vided an alternative to, the irremediably pagan Saturnalia o f late December: Meslin, Fête des kalendes (1970), 7 9 - 9 3 . 12 Consuls ( 1987), n o - 1 1 .

[46]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

consuls o f 285, Diocletian’s original colleague in that year seems to have been Caesonius Bassus:13 Aristobulus began as the consular colleague o f Carinus and only became Diocletian’s colleague after Diocletian defeated and replaced him as ruler o f the western empire in the late spring o f 285.14 Disregarding the genuine precedent o f 288 in favor o f 285, Ammianus compares Julian to Diocletian, not to Maximian, the grandfather o f Constantius, with whom he contrasts him. Ammianus surely intended the reader o f Book X X III to recall the com­ ment that Book X V I had made on Constantius’ hieratic pose as he entered the city o f Rome in 357: “ during the whole time o f his reign he never took anyone to sit with him in his carriage nor made a private citizen his consular colleague, as consecrated emperors have done” (16.10.12). That comment, in turn, acquires significance when one asks who the principes consecrati were who as emperors had a privatus as consular colleague. The list o f such consular pairs between 288 and 395 (besides the consuls o f 363) is brief: 3 7 1 G r a t ia n II a n d S e x . P e tr o n iu s P r o b u s 3 7 4 G r a t ia n III a n d FI. E q u it iu s 3 7 7 G r a t ia n I V a n d F I. M e r o b a u d e s 3 8 5 A r c a d iu s a n d F I. B a u t o 3 8 7 V a le n tin ia n III a n d E u t r o p iu s 3 8 8 T h e o d o s iu s II a n d M a te r n u s C y n e g iu s 3 9 0 V a le n tin ia n I V a n d FI. N e o t e r iu s 3 9 2 A r c a d iu s a n d F l. R u fm u s 3 9 3 T h e o d o s iu s I I I a n d FI. A b u n d a n t iu s (E ast)

The apparently plural principes consecrati, therefore, are precisely Gratian, al­ though the remark incidentally reflects praise on Theodosius for the consular appointments o f the later 380s. Ammianus follows his notice o f the consuls o f 363 with a brief account o f Julian’s attempt to rebuild the Jewish temple in Jerusalem.15 It falls into three 13 See J R A 9 (1996), 5 37, n. 26, arguing from A E 1978.782. 14 Consuls (1987), 1 0 4 - 5 . O n the date o f Carinus’ death, see now J R A 9 (1996), 5 3 6 —37. 15 This episode has often been discussed: among recent treatments, note F. Blanchetière, J J S 3 1 (1980), 6 1 - 8 1 ; Y. Lewy, Jerusalem Cathedra 3 (1983), 7 0 - 9 6 ; G. Stemberger, Juden und Christen im Heiligen Land. Palästina unter Konstantin und Theodosius (Munich, 1987), 1 6 3 —74; J. W . Drijvers, Cognitio Gestorum (1992), 1 9 - 2 6 . Jewish sources are notoriously silent about Julian’s attempt to rebuild the Temple; D . Levenson, O f Scribes and Scrolls, ed. H . W. Attridge, J. J. Collins, and T. H . Tobin (Lanham, 1990),

[47]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

unequal sections. First, and at greatest length, conies the emperor’s motive. Although he was busy with preparations for the military expedition that was about to begin, Julian did not neglect other duties: he intended to restore in a lavish fashion the once-famous temple in Jerusalem, which had been besieged by Vespasian and stormed by Titus because he wished his reign to be re­ membered for its great building achievements (23.1.2: imperiique sui memo­ riam magnitudine operum gestiens propagare). Second, Ammianus notes the issuing o f the order for rebuilding: Julian entrusted Alypius o f Antioch with the speedy performance o f the task. Third comes the fate o f the undertaking. When (or perhaps although) Alypius was eagerly tackling the task with the ac­ tive assistance o f the governor o f the province, fireballs devastated the site and the project was abandoned, as the elements themselves forbade its completion (23.1.3: hocque modo elemento destinatius repellente cessavit inceptum). Julians attempt to rebuild the temple in Jerusalem is obviously central to any interpretation o f his personality, his reign, or his religious policies as a whole, and Ammianus’ treatment o f the episode is very relevant to determin­ ing his interests and biases. The placing o f the episode in Ammianus’ narra­ tive is careful: it deliberately minimizes its significance by including it among the annalistic notices that introduce the year 363. Ammianus does not and can­ not intend to date the attempted rebuilding in Jerusalem to January 363. Julian entered on his consulate in Antioch, and what Ammianus dates to the very beginning o f the year is Julian’s charge to Alypius. He then avails himself o f the historian’s freedom to look forward from the immediate context to the consequences o f what he has just described before returning to his chrono­ logical point o f departure. If he needed to learn such a procedure from ah ear­ lier historian, Tacitus could be his model. For Tacitus divided what could have been a virtually continuous account o f Roman dealings with Parthia during the reigns o f Tiberius, Claudius, and Nero into narrative sections each o f which usually covers several calendar years, yet he is careful to anchor each section o f eastern narrative to his main annalistic narrative by linking it, sometimes 2 6 1 - 7 9 . Admittedly, a passage in the Palestinian Talmud is transmitted as referring to Julian (Nedarim 3.2, 37d), and the reference has sometimes been taken as authentic, as by W . Bacher, J Q R 10 (1898), 1 6 8 - 7 2 . But an identical passage elsewhere names Diocletian (Shebuot 2.9, 34d): this must be correct, since the emperor in question visited Palestine, as Diocletian did several times (N ew Empire [1982], 5 0 - 5 1 , 5 4 - 5 5 ) . Hence the translation ofN edarim 3.2 b y j. Neusner, Talmud o f the Laud o f Israel 23 (Chicago, 1985), 52, has “ when Diocletian went down there” with no hint that the transmitted reading has been emended. Although deliberate suppression has sometimes been inferred, the natural deduction is that the Palestinian Talmud, which has several mentions o f Ursicinus’ presence in Galilee during the rebellion o f 3 5 1 —35 2, underwent its final redaction in Tiberias between 35 2 and 363: M . A d le r ,J Q R 5 (1893), 626, n. 1; G . Stemberger, In ­ troduction to the Talmud and M idrash2, trans, and ed. M . Bockmuehl (Edinburgh, 1996), 1 7 0 —73.

[48]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

with specific detail, to events in Rome in a particular consular year.16 Ammi­ anus knew and used this standard technique. It is clearest in Book XIV, where the beginning o f the Isaurian raids is set in the narrative o f eastern events dur­ ing the summer o f 353, but Ammianus marks the passage o f time, refers to winter, and follows the episode through to its end in 354-17 Similarly, at the start o f Book X X III, Ammianus puts the order to Alypius in early January, but says nothing about how long the preparations for rebuilding took. It is a misunderstanding o f the historian’s technique to argue that “ Ammianus dates the whole episode, including its failure, to the time when Julian was in Anti­ och.” 18 It is only the order to rebuild that he unambiguously assigns to the beginning o f the year.19 From the abortive attempt to rebuild the temple in Jerusalem, Ammianus turns back to events in Antioch in early January. He marks the transition with the words isdem diebus (23.1.4), which take the reader back to Julian’s assump­ tion o f the consulate on the first day o f the year. The paragraph into which Seyfarth divides the rest o f the chapter comprises two parts that are very un­ equal in length— four official appointments (4), then four omens (5-7) that set the scene and create the mood for Julian’s departure from Antioch on his expedition into Persia. Ammianus marks the sequence o f events carefully and precisely. First, Julian gave offices and honors to senatorial ambassadors who had come to Syria from Rome; then, after he had made these appointments, he was alarmed by an omen that portended his death. Despite frequent modem assertions to the contrary, Ammianus is correct in dating these three senatorial appointments to Janu­ ary 363— Apronianus as prefect o f the city o f Rome (he received instructions from Julian in Antioch on 17 January on how to prevent Christians from prac­ ticing as lawyers in his court in Rome),20 Octavianus as proconsul o f Africa (to take office in April), Venustus as vicarius Hispaniarum, and Aradius Rufmus as comes Orientis in place o f the emperor’s uncle Julianus, who had just died.21 16

Note esp. Tacitus’ treatment o f the eastern campaigns o f Corbulo in A n n . 1 3 . 5 - 9 , 1 3 . 3 4 . 2 -

4 1; 1 4 . 2 3 - 3 6 ; 1 5 . 1 - 1 7 , 1 5 . 2 4 - 3 1 , cf. K. Gilmartin, Historia 22 (1973), 5 8 3 - 6 2 6 . 17 H S C P 9 2 (1989). 4 1 8 - 1 9 . 18 Bowersock , Ju lia n (1978), 12 1. [t> J . Vogt, K aiser Ju lia n und das Ju d en tu m . Studien zum IVeltanscliauungskampf der Spätantike (M orgenland 30, 1939)1 46. The text should probably be articulated as follows: negotiumque ma­

turandum Alypio dederat/ Antiochensi qui olim Britannias curaverat pro praefectis. For A m ­ mianus’ use o f the pluperfect with a perfect or aorist sense, see Jonge on 14 .7 .12 and the Dutch commentators on 2 0 .3.1, 4.4. Dederat is translated correctly by Seyfarth and Hamilton, incor­ rectly by Rolfe and Fontaine. 20 Constitutio de postulando , published by B. Bischoff and D. Nörr, A b h . M ünchen , Phil.-hist. Kl., N.F. 58 (1963)1 7 . 21 O n the date ofjulianus’ death, see Cognitio Gestorum (1992), 5 - 6 .

[49]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

After these four appointments had been made, Julian was terrified by an omen that he experienced directly. Felix, the comes sacrarum largitionum (an office in which he had probably served since the death o f Constantius), had died suddenly just before the emperors uncle Julianus, and when the Anti­ ochene crowd saw the traditional phrase “ Felix Iulianus Augustus” in an an­ nouncement posted publicly, they read it aloud with a malicious twist: by adding the enclitic word for “ and” after “ Augustus,” they transformed the meaning o f the phrase from the loyal “ Happy < b e > Julian the Augustus” to the sinister and ill-wishing “ Felix, Julianus— and the Augustus.” In relating the death o f Felix, Ammianus has been deliberately and significantly selective: he declines to mention that the Christians o f Antioch, who formed the larger part o f the vulgus to whom he attributed the ill-omened witticism, saw the death o f Felix as a divine punishment.2- For Felix was a convert from Chris­ tianity to paganism who accompanied Julianus and Helpidius, comes rerum pri­ vatarum and himself another convert to paganism, when they entered the great church in Antioch: on this occasion, Felix is reported to have commented sar­ castically on the luxurious sacred vessels with which Constantine and Con­ stantius had equipped the church.23 Previously, Ammianus continues, on the kalends o f January, as Julian as­ cended the steps o f “ the temple o f the Genius,” that is, the shrine o f the god­ dess Tyche, patron deity o f the city o f Antioch,24 an elderly priest collapsed and died. The spectators proclaimed that this portended the death o f the el­ der o f the consuls o f the year, that is, o f Sallustius. Ammianus remarks that this interpretation was mistaken, either out o f ignorance or because o f a desire to flatter: as it turned out, the senior consul o f the year did soon die, since the emperor was senior to any privatus by virtue o f his rank, regardless o f the age o f each man. Next, as preparations were begun for the departure o f the expedition, news arrived o f an earthquake in Constantinople. In this case, there were experts to proclaim to Julian that the earthquake was an unfavorable omen that pre­ dicted the victory o f the side whose territory was being invaded by foreign arms: they therefore urged the emperor to desist from an enterprise that had been shown to be untimely.25 22

P L R E 1.332 , quotes the relevant passages. Th e earliest is Gregory o f Nazianzus, O ral. 5.2:

although Gregory was writing after Julian’s death, there can be little doubt that his interpreta­ tion, which reproduces the familiar pattern o f the death o f a persecutor, was advanced as soon as Felix fell ill. 24 Libanius, O ral. 14.36; Theodoretus, H E 3 .12 .4 . For Helpidius’ career, see P L R E 1.4 15 , Helpidius 6. 24 G . Downey, A H istory o f Antioch in Syria (Princeton, 19 6 1), 7 3 - 7 5 , 384. 25 O n earthquakes as omens in the Greco-Rom an world, see the rich collection o f material as­ sembled by A. S. Pease, M . Tulli Ciceronis de D ivinatione libri duo i (Urbana, 1920), 109, 2 2 7 —28.

[50]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

Finally, at about the same time, presumably about the same time as news o f the earthquake arrived, Julian received a letter from Rome reporting that the Sibylline books had been consulted, as he had ordered, and that they expressly forebade the emperor to advance beyond his frontiers that year. Each o f these four omens is o f a familiar type— the utterance that unex­ pectedly predicts the future, the sudden death, an earthquake, a divine mes­ sage through an acknowledged medium o f communication between gods and men. Each is connected with the beginning o f the year, the two in Antioch explicitly, those at Constantinople and Rome implicitly (23.1.7: eo anno), and all four omens convey a clear apologetic message that is reiterated in Ammi­ anus’ reports o f discouraging omens during the march to Ctesiphon: the gods gave Julian clear and unambiguous warnings o f the disaster that would over­ take his Persian expedition.2627 This elaborate notice ofhappenings in Antioch in January 363 omits or glosses over certain episodes that illustrate antipathy between Julian and the largely Christian populace o f Antioch and the tensions that his anti-Christian poli­ cies produced even in the imperial bodyguard. Ammianus does not conceal the hostility that existed between Julian and the Antiochenes. The emperor’s arrival was ill-omened: he reached the city during the annual festival o f Ado­ nis, so that he was greeted with wailing and lamentation instead o f rejoicing (22.9.15). And when he left, he appointed one Alexander o f Heliopolis as governor o f Syria: Alexander was cruel and bad-tempered, not suitable for such a post at all (Julian conceded), but the greedy and abusive Antiochenes deserved him (23.2.3). The most striking and unusual manifestation o f the hostility was the emperor’s composition and publication o f one o f the strangest literary products o f antiquity, which combines the literary elements o f a cynic diatribe and a speech in praise o f a city— hence its double title Antiochikos or Misopogon (On the City of Antioch or The Beard-hater).21 In it, Julian poses as a cynic philosopher delivering an invective against himself while he lavishes on the city and its inhabitants heavily sarcastic praises that make no attempt to conceal his bitterness and resentment.28 The occasion o f the ridicule that in­ furiated Julian was the celebration o f the kalends ofjanuary and the work was posted up ouside the imperial palace for all to read during that month.29 A111-

26 O n these later omens, see Chapters X III, X IV . 27 A . Marcone, REA ug 30 (1984), 2 2 6 - 3 9 . J. M . Alonso-N unez, Ancient Society 10 (1979), 3 1 1 - 2 4 , discounts the religious tensions evident in the work and asserts that “ among the in­ habitants o f Antioch there was no religious interest, neither pagan nor Christian” (324). 28 Bowersock, Julian (1978), 1 0 3 - 4 ; J. Long, Ancient World 24 (1993), 1 5 - 2 3 . 29 Malalas 3 2 8 . 3 - 4 Bonn, cf. M . Gleason, J R S 76 (1986), 1 0 6 - 1 9 .

[51]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

mianus notices the Misopogott, but he displaces it from its original context and transfers it to the preceding autumn (22.14.2). Modern techniques o f research reveal significant episodes from Julian’s stay in Antioch, which Ammianus plays down or leaves out altogether. The future emperor Valens, it may be argued, was tribune o f the Ioviani Cornuti in the imperial bodyguard until he was discharged after he punched a pagan priest who sprinkled him with holy water as he entered the temple o f Tyche dur­ ing the ceremonies o f 1 January.30 This conclusion is admittedly speculative, since the ecclesiastical historians Theodoretus and Sozomenus tell the story o f Valentinian, not Valens.31 But it is plausible and persuasive, despite the fact that when Ammianus refers to the same episode, he asserts that the priest fell without being struck by anyone (23.1.6: nullo pulsante). Again, it has con­ vincingly been inferred from hagiographical evidence that the trial and exe­ cution in Antioch o f Bonosus and Maximilianus, the tribunes o f the Cornuti Seniores Ioviani and Cornuti Seniores Herculiani, apparently on 21 August or 20 September 362, clarifies an otherwise totally unexplained allusion in Am­ mianus.32 The standard bearer o f the Ioviani (Ammianus records) deserted to the Persians immediately after Jovian was proclaimed emperor because o f a quarrel with him before his accession: ' T h e standard-bearer o f the Jo v ia n i, w h o m Varronianus had o n ce co m m a n d e d , had been o n bad term s w ith the n e w em p ero r w h ile he was still a private citizen because he had insulted and denigrated his father. Fe a rin g d an ger fro m an en ­ e m y w h o had n o w risen above the co m m o n level, he deserted to the Persians and, w h e n he was given an o p p o rtu n ity to tell w h a t he k new , he in fo rm ed S h a pur, w h o was already close at hand, that the m an he feared was dead and that an u n ruly th ro n g o f cam p -fo llo w ers had p roclaim ed the o bscure, lethargic and ef­ fem inate Jo v ia n , w h o was still a m ere protector, as p se u d o -e m p e ro r.33 (2 5 .5 .8 )

Ammianus fails to explain the nature or origin o f the quarrel between Jovian and the standard bearer o f the Ioviani: an explanation must be sought in the fact that the latter (whose name is unknown) replaced Bonosus when he was executed, perhaps over objections voiced by the future emperor.34 •w D. Woods, Studies in Latin Literature and Roman History, ed. C . Deroux, 9 (Brussels, 1998), 463 —86. Valentinian had been discharged in 3 5 7 ( 1 6 . 1 1 . 1 —7) and lived in retirement at Sirmium until the late summer o f 363 ( 2 5 .10 .6 -9 ): D. Woods, Ancient Society 26 (1995), 2 7 3 - 7 7 . ■ " Theodoretus, H E 3 . 1 6 . 1 —5; Sozomenus, H E 6 . 6 4 - 6 . D. Woods, Hagiographica 2 (1995), 2 5 —55. •w It is hard to find an exact translation that adequately conveys the contempt expressed by Ammianus’ Latin: turbine concitato calonum ad umbram imperii Iovianum adhuc protectorem ascitum, inertem quendam et mollem. •’ 4 Woods, Hagiographica 2 (1995), 4 5 - 5 1 ; Journal o f Roman Military Equipment Studies 6 (1995), 6 1-6 8 .

[52]

D

a t i n g

, E

m p h a s i s

,

a n d

O

m is s io n

A still more significant episode omitted by Ammianus is an abortive plot to assassinate Julian. Libanius alludes to it in three o f his Julianic speeches: ten guardsmen intended to kill Julian on the parade ground during exercises, but they got drunk and incautiously revealed their plans.35 Two o f the ten, Ju ventinus and Maximinus, were Christians offended by the constant smell o f pagan sacrifice: according to John Chrysostom, they complained at a banquet that God was being dishonored and his holy laws trampled upon.36 Later Christian writers and calendars identify the pair as Scutarii Gentiles and sup­ ply the date o f their martyrdom as 29 January.37 The other eight alleged plot­ ters were released.38 Why has Ammianus left out a matter so important as dis­ loyalty in the palace guard? His motive seems clear. The failed assassins were celebrated as martyrs after their execution, but Ammianus’ account o f Julian gives the impression that there were no Christian martyrs at all during his reign. Ammianus thus faithfully reflects the emperor’s official propaganda that any Christian put to death had been justly condemned for crimes that had nothing to do with his religion.39 Christian sources name several dozen mar­ tyrs under Julian.40 Ammianus notes the death o f precisely one o f them, and he goes out o f his way to deny that he was in any sense a martyr. He presents Artemius, who had been dux o f Egypt under Constantius, and who was later venerated as a saint and martyr in Constantinople,41 as a notorious malefactor who was accused by the people o f Alexandria and executed “ for a mass o f horrible crimes” (22.11.2). 35 36 37 38 39

Libanius, Orat. 15.4 3; 16 .19 ; 18.199. John Chrysostom, In sanctos martyres hwentinum et M axim inum ( P G 50.574). Theodoret, H E 3 . 1 5 . 4 - 9 ; P. Peeters, A n a l. Boll. 4 2 (1924), 7 7 - 8 2 . Libanius, E p . 1120 . Hence the complaint o f Gregory o f Nazianzus that Julian tried to deprive the church o f

martyrs (Orat. 4.58). 40 H. C . Brennecke, Studien zu r Geschichte der Homöer (Beiträge z u r Historischen Theologie 73, 1988), 1 1 4 - 5 7 ; R .J. Penella, A ncient World 2 4 (1993), 3 1 - 4 3 . 41 On Artemius’ posthumous reputation, see S. M . C . Lieu, From Constantine to Ju lia n : P a­ gan and B yzan tin e Views , ed. S. Lieu and D. Montserrat (London, 1996), 2 1 3 - 2 3 .

[53]

[VI] ORIGIN AND SOCIAL STATUS

Who was Ammianus Marcellinus? Where did he come from? What was his social status? There is no direct and explicit evidence beyond the little that Ammianus has chosen to vouchsafe about himself in the extant books o f his history. He was a soldier and a Greek (31.16.9). Both terms require further definition before their precise import becomes clear (Chapters VII, VIII). The extant books disclose Ammianus’ approximate age and the outlines o f a ca­ reer. He was an adulescens in 357 (16.10.21): when his extant narrative begins, he was serving in the Roman army as a protector domesticus (14.9.1, 18 .8 .11), at­ tached by imperial command to the general Ursicinus, whom he acompanied as a staff officer on various postings and missions between 354 and 3 59.1 In 363 Ammianus took part in Julian’s Persian expedition and returned to Anti­ och with the defeated Roman army (2 3.5.7 -25.10 .1). Thereafter, he resided in the East for at least fifteen years until he migrated to Rome, where he com­ pleted his history about 390.12 In order to discover Ammianus’ origin and social status, therefore, it is nec­ essary to use both indirect indications in his text and any explicit evidence ex­ ternal to the Res Gestae that may be relevant. A letter o f Libanius has tradi­ tionally been used not only to date the composition o f the work, but also to establish Ammianus’ precise local origin. During the second half o f the year 392, the aged Antiochene sophist dispatched a letter to one Marcellinus, a fellow-Antiochene who was in Rome and engaged in literary activity there. Since so much depends on its interpretation, the letter must be quoted in full:

1 Conveniently summarized in P L R E 1.5 4 7 - 4 8 , Marcellinus 13 ; for a fuller exposition, see L. Dilleman, Syria 38 (1961), 9 1 —98; Matthews, Ammianus (1989), 34 —66, 81 —83. 2 Matthews, Ammianus (1989), 1 3 - 3 2 .

[54]

O

r ig in

a n d

S

o c i a l

S

t a t u s

L ib a n iu s to M a r c e llin u s

( i) I e n v y y o u fo r p o sse ssin g R o m e an d R o m e fo r p o sse ssin g y o u . F o r y o u p o s ­ sess a t h in g to w h ic h th e re is n o t h in g s im ila r o n e a rth , an d R o m e o n e n o t in fe ­ r io r to h e r o w n c itiz e n s , w h o d e s c e n d fr o m d e m i- g o d s . (2) It w o u ld , th e re fo re , h a v e b e e n a g re a t t h in g f o r y o u ju s t to liv e in su c h a c it y in sile n c e an d to liste n to sp e e c h e s d e liv e r e d b y o th e rs: R o m e p r o d u c e s m a n y o ra to rs w h o fo llo w in th e fo o ts te p s o f th e ir fo re fa th e rs . B u t in fa c t, as o n e c a n le a rn fr o m th o se w h o a rriv e fr o m t h e re , y o u y o u r s e lf (I g a th e r) h a ve g iv e n s o m e p u b lic re c ita tio n s an d are g o in g to g iv e m o r e , s in c e y o u r w o r k has b e e n d iv id e d in to m a n y s e c tio n s an d th e p ra is e b e s to w e d o n w h a t has a lre a d y a p p e a re d in v ite s a n o th e r in sta llm e n t. . (3) I h e a r th at R o m e h e r s e lf is c r o w n in g y o u r la b o r an d has r e n d e re d a v e r d ic t th at y o u h a v e s h o w n y o u r s e lf s u p e r io r to s o m e an d e q u a l to th e rest. T h is a d o rn s w it h h o n o r n o t m e r e ly th e a u th o r, b u t us t o o to w h o m th e a u th o r b e lo n g s. (4) P le a s e d o n o t sto p c o m p o s in g su c h th in g s an d ta k in g th e m fr o m th e s tu d y to lit e r a r y sa lo n s, an d d o n o t w e a r y o f b e in g a d m ire d , b u t b e c o m e still m o r e fa ­ m o u s y o u r s e lf a n d g iv e us fa m e . F o r it is a m a rk o f th e t r u ly c e le b ra te d c itiz e n to a d o rn h is n a tiv e c it y w it h h is a c h ie v e m e n ts. (5)

M a y y o u c o n t in u e in y o u r p re s e n t h a p p y state! A s fo r m e , u n less o n e o f

th e g o d s aid s m e in m y state o f m o u r n in g , it w ill b e im p o ssib le fo r m e to b e a r it. F o r h e w h o w as m y o n ly so n , n o t a b a d f e llo w a n d b o r n o f a g o o d m o th e r , e v e n i f sh e w a s n o t o f fr e e b ir th , is g o n e a n d b u r ie d a fte r d y in g fr o m g r i e f w h ic h w as t h e re s u lt o f an o u tr a g e . W h o th e y w e r e w h o in su lte d h im , y o u m u st d is c o v e r f r o m o th e rs : e v e n t h o u g h I h a v e s u ffe re d , I m u s t re fra in fr o m n a m in g th e m . (6)

W h ile th e e v il w a s still s e e th in g , C a llio p iu s w a s sn a tc h e d fr o m a m o n g his

b o o k s a n d h is la b o rs. W o u n d fo llo w s w o u n d , an d th e e d u c a tio n o f th e y o u n g d e ­ g e n e ra te s . T h is y o u ca n le a rn t o o fr o m th o se w h o h a v e d iv id e d u p h is estate. F o r m e , w h a t h a p p e n e d b e fo r e h is d e a th , h is d e a th it s e lf an d w h a t has h a p p e n e d a f­ te r h a v e o c c a s io n e d la m e n ta tio n a n d tears, m o s t o f w h ic h flo w c o p io u s ly o v e r w h a t I w r i t e .3 3

Libanius, E p . 1063: my translation o fR . Foersters Teubner text, adapted slightly from C P

88 (1993), 58. There are three other recent English translations: (1) C . W. Fornara, Historia 41 (1992), 332 , which I saw before publication and to which my own rendering owes much; (2) A. F. Norman, Libanius: Autobiography and Selected Letters 2 (Cambridge, Mass., 1992), 4 ^9 ” 33 (no. 188), whose translation o f the crucial terms syngraphe /eus assumes that the recipient o f the letter is the historian— which is precisely the central point in dispute; (3) J. F. Matthews, C Q , N .S . 44 (1994), 2 5 2 - 5 3 , who correctly translates the words in dispute as “ composition” and “ writer,” respectively. Since Libanius’ epistolary style is difficult and frequently elliptical, I have tried to bring out his train o f thought a little more explicitly than Fornara, Norman, and M atthews— and I have occasionally translated rather freely in order to produce reasonably id­ iomatic English. For a German translation o f Foersters text, see G. Fatouros and T. Krischer, Libanios, Briefe griechisch-deutsch (Munich, 1980), 154 “ 57- A French translation o f sections 1 - 5 is offered by

[55]

O

r ig in

a n d

S

o c i a l

S

t a t u s

Until very recently, it was universally assumed that the recipient o f the letter must be the historian Ammianus Marcellinus.4 ln 1987, however, Charles Fornara challenged this traditional identification in a paper that was eventu­ ally published in 1992— after he had generously allowed Glen Bowersock and the present writer to use his typescript in advance o f publication.5 Fornara ar­ gued that the Marcellinus to whom Libanius wrote was not the historian Ammianus Marcellinus: he drew the corollary that the Antiochene origin de­ duced for the historian from this letter o f Libanius must be discarded; instead, he suggested that the historian came from Thessalonica.6 In a long review o f Matthews’ study o f Ammianus, Bowersock accepted Fomara’s thesis that the correspondent o f Libanius could not be the historian and suggested that he might be the Marcellinus to whom one Magnus (presumably Magnus o f N isibis, who taught rhetoric and medicine in Alexandria) dedicated two sets o f acrostic iambic verses incorporated in the medico-magical treatise C yranidés.7 In a reconsideration o f the problem published in 1993 , 1 concluded that, al­ though Fornara’s demolition o f the traditional view was effective, neither the Macedonian origin that he proposed nor the Alexandrian origin suggested by Bowersock was likely to be correct, since Matthews had made a good case for believing that Ammianus could speak Syriac and hence came from the region o f Syria, if not from Antioch itself.8 In 1994 Matthews restated his case for the traditional view,9 which shows signs o f being reinstated as the communis opinio o f enlightened scholarship.10 Now Matthews’ restatement stands or falls on

P. M . Camus, A m m ieti M arcellin, témoin des courants culturels et religieux à la f i n du IV e siècle (Paris, 1967), 2 7 8 - 7 9 : it is based on the translation o f Reiske’s text b y j. Gimazane, A m m ien M arcellin. S a vie et son oeuvre (Diss. Bordeaux, pub. Toulouse, 1889), 4 0 3 - 4 .

4 Matthews, A m m ian us (1989), 8: “ the identity o f Libanius’ correspondent as Ammianus M ar­ cellinus is inescapable” ; 454: Libanius' letter is “ the one certain external reference to Ammianus.” 5 Matthews, A m m ian us (1989), 4 7 8 —79, n. 1, reports Fornara’s conclusion (from a paper de­ livered in Oxford in February 1987) and argues against it. h Fornara, H is t o r ia li (1992), 3 3 3 —44. 7 G. W . Bowersock, J R S 80 (1990), 2 4 7 - 4 8 , citing the brilliant detection o f the acrostics by M . L. West, C Q , N . S. 42 (1982), 480 —8 1, who also improved the text printed by D . Kaimakis, D ie K yraniden (Meisenheim am Gian, 1976), 5 0 - 5 1 , 9 6 - 9 7 . Magnus o f Nisibis was an ia-

trosophist, and his presence in Alexandria is attested in 364 and 388 (P L R E 1. 534, Magnus 7). West reedited two more acrostic poems in the same text that have the author’s name in the gen­ itive as M agnou: a third was subsequently added by R. Führer, Z P E 58 (1985), 270, and possible additional evidence about Magnus is noted by D. Bain, “ O w ls to A thens. ” Essays on Classical S u b ­ jects Presented to S ir Kenneth D over, ed. E. M . Craik (Oxford, 1990), 296, n. 4. " C P 88 (1993), 5 9 —6 1, cf. Matthews, A m m ianus (1989), 6 5 —70. 9 J. F. Matthews, C Q , N .S. 44 (1994), 2 5 2 —69. Comments on Fornara’s arguments have also been offered by S. Rota, K O I N f iN I A 18 (1994), 1 6 5 - 7 7 . 10 K. Rosen, D er neue P au ly 1 (1996), 596: Ammianus "w ar zweifellos der Empfänger des Briefes.”

[56]

O

r ig in

a n d

S

o c i a l

S

t a t u s

the proposition that the two related nouns in the letter, which are translated above as “ author” and “ work,” must mean specifically “ historian” and “ his­ tory.” However, outside the letter to Marcellinus, Libanius uses the nouns sungrapheus and sungraphe only five and nine times respectively.11 Hence, even if Libanius always used the two words elsewhere in the more restricted sense, the inference that they must have the same sense in the letter to Marcellinus would would have a weak statistical basis. For other authors use the word sun­ grapheus in the wider sense. Libanius’ frequent model Isocrates uses it o f him­ self as a writer o f speeches, Libanius’ contemporary the emperor Julian to mean “ writer o f prose” as opposed to poet, and the ecclesiastical historian Theodoretus even applies it to the Syriac poet Ephrem.1112 Moreover, Libanius himself employs this wider sense: he uses the verb sungraplio and the noun sun­ graphe for the activity and products o f orators, and styles himself sungrapheus as the alleged author o f a panegyric on the usurper Procopius.13 Furthermore, it is pure petitio principii to assert that it is implausible to imagine two Marcellini from Antioch in Rome in 39 2:14 Marcellinus is not a rare name, and the An­ tiochene origin o f the historian is in dispute. But doxography does not de­ cide truth. What in the letter tells for or against the traditional identification o f its addressee? The crucial point is one that would surely have been realized long ago, had the letter been read without preconceptions. Both Libanius’ tone and the contents o f his letter imply that his addressee has recently arrived in Rome and has recently made his literary début there: they do not suit a man o f sixty or so who had lived in Rome for nearly a decade.15 Indeed, the remark that it would have been a great thing for the Antiochene Marcellinus merely to live in Rome in silence and to listen to others would have been intolerably offen­ sive and patronizing if addressed to an established historian, especially one who had recently completed a history on the grand scale. Libanius’ tone pre­ cludes his identification as the historian Ammianus Marcellinus. The letter im­

11 G . Fatouros, T. K rischer, and D . N ajo ck , Concordantiac in Libaniutn 1.2 (H ildesh eim / Z ü r ic h / N e w Y o rk , 19 87). 640; 2.3 (1989), 39 i- Som ew hat surprisingly, M atthew s n ow here refers to the published concordances to Libanius’ speeches and letters: he appeals instead to a search in the electronic Thesaurus Linguae G raecae conducted for him by H .-U . W iem er (C Q , N .S . 44 [19 94]. 2 6 3 - 6 4 , n. 58). 12 Isocrates, Antidosis 35; Julian, Misopogon 338b; Theodoretus, H E 2 .3 0 .U . 13 Libanius, Orat. 58.40; E p . 4 06 .1, 793, cf. 8 26 .1; Orat. 1.16 3 . T h e norm al usage is w ell stated b y F om ara, Historia 4 1 (1992), 334: “ these w o r d s . . . define generally either the process o f prose com position o r its result.” 14 A s does M atth ew s, C Q , N .S . 44 (1994), 267.

15 B o w ersö ck , J R S

80 (1990), 2 47; Fom ara, Historia 41 (1992), 3 3 3 - 3 4 ; T . D . Barnes, C P 88

(1993). 58.

[57]

O

r ig in

a n d

S

o c i a l

S

t a t u s

plies that its recipient, who listens to speeches in a city famous for its orators and gives epideixeis, is an orator or sophist rather than a historian. He need not be a rhetor composing and reciting a collection o f speeches,16 but he could be an iatrosophist who has recited part o f a treatise on medical rhetoric.17 There is thus no certain external contemporary evidence for Ammianus’ career or literary activity.18 What then does the text o f the Res Gestae indi­ rectly reveal about the origin and social status o f its author? The latter is the easier to discover. Two passages denouncing Julian’s policy o f forcing men o f curial origin to perform curial duties have long been recognized to reflect the personal anger o f one who was affected by the policy: It w a s h a rsh a n d r e p r e h e n s ib le th a t u n d e r h im o n l y w it h d iffic u lt y d id a n y o n e w h o m t o w n - c o u n c illo r s w is h e d to c o - o p t o b ta in fa ir tr e a tm e n t u n d e r th e la w , e v e n i f h e w a s p r o t e c te d b y s p e c ia l p r iv ile g e s , b y le n g th o f s e r v ic e , o r b y c le a r p r o o f th a t h is o r ig in w a s e ls e w h e r e a lto g e th e r , to s u c h a d e g r e e th a t m a n y in te r­ r o r b o u g h t e x e m p t io n fr o m m o le s ta tio n b y se c re t p a y m e n ts . ( 2 2 .9 .1 2 )

E q u a lly in to le r a b le w a s th e fa c t th at h e a llo w e d s o m e p e rso n s to b e w r o n g f u l ly d ra fte d in to m e m b e r s h ip o f t o w n c o u n c ils w h o w e r e e it h e r f r o m a n o t h e r c it y o r fa r r e m o v e d f r o m lia b ility to s u c h c o n s c r ip tio n t h r o u g h s p e c ia l p r iv ile g e s o r b y t h e ir o r ig in . ( 2 5 .4 .2 1 )

The implicit self-reference is hard to mistake. Yet the precise import o f these passages requires careful consideration. It is mistaken to deduce that Ammi­ anus had been threatened with inclusion in the municipal ordo, “ although he thought that his military service entitled him to exemption.” 19 In both pas­ sages emphasis falls very heavily on the Latin word origo. How many different categories o f person does Ammianus specify whom Julian unjustly compelled or allowed to be compelled to serve as decurions? One category is described in identical terms in both passages: those who had been granted specific ex­ emption from curial service (privilegia). Moreover, those who are originis pe­ nitus alienae in the first passage must surely be identical to the peregrini, that is, the citizens o f another city, in the second. In the first passage the remaining 16 As argued by Fornara, Historia 41 (1992), 3 3 6 - 3 7 . 17 As suggested by Bowersock, J R S 80 (1990), 2 4 7 - 4 8 . 18 O. Seeck, R E 1 (1894), 1846, suggested that the unnamed addressee o f Symmachus, E p . 9. n o , might be Ammianus (“ Symin. ep. IX 1 1 0 könnte wohl an ihn gerichtet sein” ). Th e con­ jecture, promoted to a certainty by many who subsequently wrote about the historian, was dis­ proved by Alan Cam eron , J R S 54 (1964), 1 5 - 1 8 . For a full bibliography, see S. Roda, Comm ento storico al Libro I X d ell’Epistolario di Q. A u relio Simmaco (Pisa, 19 81), 2 4 2 —45.

19 Thompson, A m m ian us (1947), 81: accepted in R eading the Past (1990), 62.

O

r ig in

a n d

S

o c i a l

S

t a t u s

category is defined in terms o f stipendiorum numerus, whereas in the second the third exempt group are origine longe discreti. Hence the correct deduction from the pair o f passages taken together is that Ammianus claimed exemption from curial obligations both on the grounds o f his own service as a soldier and be­ cause his father or grandfather had removed himself from the curial class and the attendant obligations by entering imperial service.20 The inference that Ammianus came from a family that had already risen above curial status is confirmed by the fact that he was a protector domesticus as a young man in his twenties. The standard study o f the scholae palatinac doc­ uments the fact that protectores domestici were normally the sons o f military officers o f high rank.21 Eight are known from the fourth century besides Am­ mianus himself: although the origin o f three o f them is unknown,22 Ammi­ anus records the parentage o f the other five. They all came from very similar backgrounds. Herculanus was the son o f Hermogenes, the magister equitum o f Constantius who was lynched in Constantinople in 342 when he attempted to arrest the bishop Paul (14.10.2). The emperor Jovian was the son o f Var­ ronianus, tribune o f thejoviani and comes domesticorum (25.5.4, 8). The impe­ rial brothers Valentinian and Valens were the sons o f Gratianus, who rose to be comes rei militaris in Britain (30.7.2—3). And Masaucio was the son o f Cre­ tio, who commanded the troops in Africa for Constantius as comes rei militaris per Africam from circa 350 to 361 (26.5.14). Ammianus, it is clear, “ must have had connections.” 23 If all the other known protectores domestici o f his genera­ tion were the sons o f generals, then he too may have been. Whether or not Jean Gimazane was justified in suggesting, more than a hundred years ago, that the historian was the son o f the Marcellinus who was comes Orientis in 3 49,24 his instinct for Roman social norms was correct: to be a protector domes­ ticus at such an early age, Ammianus must surely have belonged to a family o f high status. Moreover, although subsequent writers about Ammianus have shown coolness towards Gimazane s identification o f Ammianus’ father,25 sev­ eral considerations (it may be noted) tell indirectly in its favor. 20 O n such exemptions in the early fourth century, see esp. Jones, L R E 6 9 - 7 0 , 1 3 5 - 3 6 , 7 4 0 - 4 1 ; F. Millar, J R S 73 (1983). 9 1 - 9 6 . 21 R. I. Frank, Scholae Palatinae. T h e Palace G u a rd in the Later Rom an E m pire (Papers and M on o­ graphs o f the Am erican A cadem y in R om e 23, 1969), 73. Frank adduces Ammianus as “ the most fa­ mous example” o f the sons o f decurions who “ formed a second group o f Roman officers” (76 — 79), but none o f the others was a protector domesticus, while two o f Frank’s alleged military officers are civilian palatine officials (P L R E 1.2 77 , Eleusius; 554, Marcianus 8). 22 Romanus and Vincendus (2 2 .11.2 ) and Equitius (26.1.4). 23 R. S. O . Tomlin, T h e Rom an World, ed. J. Wacher, 1 (London / N ew York, 1987), 1 1 5 . 24 Gimazane, A m m ien (1889), 2 4 - 2 7 . 25 A conspicuous exception was L. Dautremer, A m m ien M arcellin. E tu d e d ’histoire littéraire (Diss. Paris, publ. Lille, 1899), 1 1: “ pour qu’Am m ien ait été placé si jeune dans l’état-major d’ un

[59]

OttiGiN

S

a n d

o c i a l

S

t a t u s

Ammianus was Greek, but he did not come from Asia Minor, since he jeers at the customs o f the area (15.7.6: Asiatici mores). He knew Antioch well and admired the city. But that does not make him a native o f the city. For he lived there for a long period, probably almost twenty yean from 363 or thereabouts to shortly after 380. His local knowledge is obvious and profound: for ex­ ample, he alludes to the unexpected appearance o f the Penian king Shapur in the early 250s while the citizens o f Antioch were absorbed in the theater (25.5.3)— a story told elsewhere only in the Antiochene writen Libanius and Malalas (though Eunapius o f Sardis has it in a garbled form).26 But no amount o f local knowledge could prove that the historian was a native o f Antioch. Ammianus knew and admired Antioch because he lived in the city for a num­ ber o f years. Similarly, Ammianus’ praise o f noster Hypatius for his conduct in 372 (29.2.16) reflects friendship, not community o f origin: Hypatius was the brother o f Constantius’ second wife and a native o f Thessalonica, who lived in Antioch while Ammianus resided there and then migrated to Rome when he was appointed praefectus urbi in 379 (Chapter X). Ammianus, it appears, could speak Syriac: not only was he a personal friend ofjovinianus the satrap o f Corduene, who recognized him when he was sent to Corduene on a reconnaisance mission (18.6.20—21), but also he could move around in the Mesopotamian countryside (18.8.7—13, 19 .8 .5 -12 ). That might seem to support an Antiochene origin.27 On a dispassionate assessment, however, fluency in Syriac indicates only that Ammianus came from the gen­ eral area o f Syria and Palestine. Does his text in any way disclose anything more precise? He reveals his origin, it may be suggested, in an idiosyncratic linguistic usage. Ammianus uses the plural Syriae in preference to the singular Syria, and he uses it to associate Phoenicia, the Lebanon and Palestine with Antioch. The satrap Jovinianus was not, as is often stated, “ sent to Syria as a hostage” : 28 he was a hostage in Syriis (18.6,20). Elsewhere, Ammianus has per Syrias (26.3.2), Syriae omnes (2.10.1) and Syriarum provinciae (22.15.2). Pride in his eastern ori­ gin is obvious in Ammianus’ survey o f the provinces o f Oriens in Book X IV : N e x t S y r ia sp re a d s fo r a d is ta n c e o v e r a b e a u tifu l p la in . T h is is fa m e d f o r A n ­ t io c h , a c it y k n o w n to all th e w o r ld , a n d w it h o u t a r iv a l, so r ic h is it in im p o r te d

général en chef en qualité de prolector doesticus, il fallait que son père occupât un poste élevé et disposât d’une influence considérable.” 26 Libanius, Oral. 24.38, 60.2; Malalas 2 9 5 .2 0 -2 9 6 .2 0 Bonn; Eunapius, Vit. phil. 6 .5.2 (465). Eunapius trots out the story as if the episode occurred very shortly before Eustathius’ embassy to Persia in 358 (Ammianus 1 7 .5 .1 5 , 14 .1). 27 Matthews, Ammianus (1989), 44, 5 5 - 5 7 . 28 Matthews, Ammianus (1989), 44.

[6 .0 ]

O

r ig in

a n d

S

o c ia l

S

t a t u s

a n d d o m e s tic c o m m o d it ie s ; lik e w is e fo r L a o d ic e a , A p a m e a , an d also S e le u c ia , m o s t flo u r is h in g c itie s fr o m t h e ir v e r y o r ig in . A f t e r th is c o m e s P h o e n ic ia , ly in g at th e fo o t o f M o u n t L ib a n u s , a r e g io n fu ll o f c h a r m a n d b e a u ty , a d o r n e d w it h m a n y g re a t c itie s; a m o n g th ese in a ttra c tiv e ­ n ess a n d th e r e n o w n o f t h e ir n a m e s T y r e , S id o n an d B e r y t u s are c o n s p ic u o u s , a n d e q u a l to th e se a re E m e s a a n d D a m a s c u s , fo u n d e d in d ays lo n g past. T h e last o f th e S y r ia s is P a la e stin a . . . ( 1 4 . 8 . 8 - 9 )

Ammianus’ praise o f Tyre, Sidon, and Berytus may reflect the pride o f a man from Phoenicia. Neither Berytus nor Sidon had produced any outstanding intellectuals under the Roman Empire, although Berytus was the patria o f Domitius Leontius, who served Constantius as praetorian prefect from 342 to 346 and would thus have been in a position to start a fellow townsman on a career.29 From Tyre, however, had come both Domitius Ulpianus, the fore­ most o f Roman jurists,30 and the philosopher Porphyry, who also knew Ara­ maic or Syriac.31 It may be relevant that Herculanus, on whom Ammianus be­ stows a gratuitously favorable notice, inherited a house at Tyre.32 Herculanus went to Gaul in 354 to report to Constantius about the misdeeds o f Gallus (14.10.2). Ammianus may have known him before they served together: the two young protectores domestici might even have been boyhood friends. There may exist, moreover, an external item o f evidence concerning the historian. One o f a number o f letters from Libanius to his former pupils Apol­ linaris and Gemellus in Tarsus concludes with the following two sentences; P le a s e , m y b o y s , s h o w c o n c e r n f o r y o u r te a c h e r e ith e r b y in v it in g h im to C i l i ­ cia o r s e n d in g s o m e t h in g f r o m th e re . Y o u m a y b e p e rs u a d e d to d e sp ise m o n e y w it h o u t m y fr e q u e n t a d m o n it io n s in th e p ast b y th e m a n w h o b e a rs this le tte r: to j u d g e fr o m h is d ress h e is e n liste d in th e a rm y , b u t in fa c t h e is e n r o lle d a m o n g p h ilo s o p h e r s ; h e has im ita te d S o c r a te s d e s p ite h a v in g g a in fu l e m p lo y m e n t — th e fin e A m m ia n u s .33

The letter appears to have been written in 360.34 At this date, the future his­ torian could have taken it from Antioch, where he was (it may be assumed) 29 O n Leontius’ prefecture, see Z P E 94 (1992), 2 5 1 - 5 2 , 2 5 3 - 5 4 : his origin is inferred from the fact that the ordo o f Berytus, following “ honorific decrees o f the province o f Phoenice ratified by the verdict o f the emperors,” honored him with a statue (IL S 1234).

30 P IR 2 D

169. 31 Porphyry, Vita Plotini 1 7 . 4 - 1 5 .

32 Libanius, Ep. 828 (363). 33 Libanius, Ep. 2 3 3 .4 (my translation). 34 O . Seeck, D ie B riefe des Libattius (Texte und Untersuchungen 30, 1906), 374.

[öl]

O

r ig in

a n d

S

o c i a l

S

t a t u s

stationed after the dismissal o f Ursicinus. For Libanius’ description fits what is known about the culture and formation o f Ammianus Marcellinus.35 Nomenclature is no obstacle to identifying Libanius’ philosophical soldier as the future historian. It is true that Priscian called the historian Marcellinus little more than a century after his death.36 But naming practices in the fourth century were not wholly consistent. Julian’s praetorian prefect, for example, is Saturninius Secundus on three inscriptions, Secundus Salutius on his first mention in Ammianus (22.3.1), Salutius on subsequent mentions in Ammi­ anus, Salutius in the manuscripts o f his treatise On the Gods and the Universe, Salutius, Salustius, or Sallustius in most other literary texts, and Secundus in the law codes.37 Similar alternations o f name can be documented for others in the fourth century 38 and Ammianus may not have cared to give undue prominence to a name that so clearly indicated his eastern origin after he came to Rome. For, although the name Marcellinus is too indistinct to prove anything, Wil­ helm Schulze noted long ago, in another context, both that the cognomen Am­ mianus is very rare in Italy and that all its known bearers are o f Greek origin.39 Subsequent discoveries have strengthened his observation or inference.40 Moreover, where the origin o f possessors o f names beginning in Am(m)i—can be verified, it is usually Semitic, like the woman whose ossuary in Jerusalem states her name and provenance as Ammia from Scythopolis in both Greek and Hebrew.41 It seems that no Ammia, Ammias, Ammion, or Ammianus is 35

See Chapters V II, VIII, X IV : the letter was argued to refer to the historian by M . Biidinger,

A m m ian us Marcellinus und die Eigenart seines Geschichtsim kes (Denkschriften W ien, Phil.-hist. Cl.

44.5, 1895), 9. v. p riscjarii i nst G ram m . 9 .51 (quoted in Chapter HI, n. 42). 37 P L R E 1 . 8 1 4 - 8 1 7 , Saturninius Secundus Salutius 3, cf. J. L. Desnier, R E A 85 (1983), 53 - 6 5 , who argues that the fourth-century author took the name o f the historian Sallust as a pseudonym. 38 For example, Ammianus refers to Q . Flavius Maesius Egnatius Lollianus signo Mavortius as Lollianus when stating the names o f the consuls o f 335 (15 .8 .17 ), but as Mavortius when he appears in his narrative as praetorian prefect (16.8.5). 39 W. Schulze, Z u r Geschichte lateinischer Eigennam en (A bh. Göttingen, Phil.-hist. K l., N .F. 5.5, 1904), 12 1 n. i, citing C I L 9 .120 7 (Aeclanum); I G 14 .13 8 1, 1446 (Rome), to which add C I G 9725 (Rome). 40 There are few men called Ammianus o f any prominence before the fourth century: only an epigrammatist o f the second century and Statilius Ammianus, who is attested as prefect o f Egypt in 2 7 1 - 7 2 : see, respectively, R. Reitzenstein, R E 1 (1894) 1845, Ammianus 1; G. Bastiniani, Z P E 17 (1975), 3 17 , cf. J. R. Rea, O xyrhynchus Papyri 40 (Oxford, 1972), 24, 82 (on P. O x y. 2923). Otherwise, a search in the Thesaurus Linguae Graecae reveals the following approximate distribution o f epigraphically attested Ammiani: (1) pre-Christian: Athens 4, Ephesus 5, Caria 5; (2) Christian: Asia M inor 1 1 , Cyprus 1, Rome 2; (3) Jewish: 2. 41 C I J 13 7 2 = P. Thomsen, Z D P V 44 (19 2 1), 120, no. 204a, cf. H. Wuthnow, D ie semiti­ schen Menschennamen in griechischen Inschriften und Papyri des vorderen O rients (Diss. Tübingen, pub.

Leipzig, 1930), 1 9 - 2 1 .

[62]

O

r ig in

a n d

S

o c i a l

S

t a t u s

to be found in Greek inscriptions before the middle Hellenistic period,42 and o f the ten women with the name Ammia listed in a recent compendium o f ancient Athenians, no fewer than four are stated to have a levantine origin— one each from Antioch, Jerusalem, Samaria, and Sidon.43 Ammianus came from Syria or Phoenicia, his city o f origin being perhaps ei­ ther Tyre or Sidon. He was the son o f an officer in the Roman army who achieved high rank in the early years o f the reign o f Constantius, he had per­ sonal connections in the military élite, and he took the first steps in what ought to have become a distinguished career.44 Three o f Ammianus’ close co­ evals from very similar backgrounds even became emperor. Yet something went wrong with his career and the bright hopes o f his youth were blighted. Why? Unlike Thucydides, who recorded both his failure to relieve Am­ phipolis and the fact that he was exiled for twenty years,45 Ammianus is reti­ cent about himself. Perhaps he was damaged by his close association with U rsicinus, who was officially blamed for the loss o f Amida in 359 (20.2.2-5). Yet Ammianus served as an officer in Julian’s army, which invaded Mesopotamia in 363. Perhaps, therefore, he compromised his career under Julian, as the or­ ator Himerius, who abandoned his official chair o f rhetoric in Athens in the winter o f 36 1—362 and rushed enthusiastically to join the court o f Julian, ap­ pears to have done.46 There are good reasons for holding that Ammianus was brought up a Christian (Chapter VIII). I f he became an apostate in 36 1-36 2 , still more if he helped to enforce Julian’s anti-Christian policies, then his way to further preferment may have been barred under Julian’s Christian successors. B y the time that he wrote his history, Ammianus was a disppointed and em­ bittered man. His use o f the word potentes reflects the attitudes o f one not ac­ customed to enter the portals o f real power.47 The plural potentes occurs four 42 See P. M . Fraser and E. Matthews, A Lexicon o f G reek Personal Nam es, 1. The Aegean Is­ lands, C yprus, Cyrenaica (Oxford, 1987), 3 1 ; 2. Attica (Oxford, 1994), 25; 3A. Peloponnese, Sicily, Western Greece, and Southern Italy (Oxford, 1996), 33; A. B. Tataki, Ancient Beroea: Prosopography and Society (Athens, 1988), 1 0 3 - 5 .

43 J. S. Traill, Persons o f Ancient Athens (Toronto, 1995), 80. 44 N . J. E. Austin, A m m ian us on Warfare: A n Investigation into A m m ian u s' M ilitary Knowledge (Collection Latomus 165, 1979), 16 4 —65, concludes that “ Ammianus' viewpoint for the narra­ tion o f military history is generally that o f a member o f the headquarters staff rather than o f a field officer’ and that “ he was a staff officer attached to various commanders, and not on com­ bat duties.” 45 Thucydides 4 . 1 0 5 - 7 , 5 .2 5 .1, cf. A. W . Gomme, Historical Com m entary on Thucydides (O x­ ford, 1956)» 5 8 4 - 8 8 . 46 C P 82 (1987), 2 2 0 - 2 5 . 47 Contrast his use o f potiores in a wide variety o f favorable senses (Viansino, Lexicon 2 .3 2 6 - 2 7 ) .

[63]

O

r ig in

a n d

S

o c i a l

S

t a t u s

times. One passage is neutral: Valentinian takes a decision without consulting the potentes (27.10.10: nullo potentium in conscientam arcani adhibito). In the other three passages, however, the connotations o f the word are extremely negative. Under Constantius, the potentes in the palace coveted and appropri­ ated the property o f condemned persons (16 .8.11), and Antoninus, who had suffered great losses through the cupidity o f certain persons, ran the risk o f being unjustly ruined because he railed against the potentes (18.5.1). Under Julian, in contrast, the potentium tumor was broken (30.4.1).

[64]

[VII] THE GREEK TEMPLATE

In the final paragraph o f his history, Ammianus describes himself as “ a former soldier and a Greek” (31.16.9: haec ut miles quondam et Graecus. . . pro vir­ ium explicavi mensura). The tone o f the words that Ammianus uses to de­ scribe himself has been a matter o f some debate.1 Are they a proud boast? And is Ammianus claiming that he has both the practical experience o f affairs and the culture needed to be an authoritative historian? O r is he, as has recently been argued, “ offering an apology rather than staking a claim” ? And can his words be construed as “ the apology o f a mere soldier (“ ut,” in a concessive sense) to have ventured on an occupation more characteristic o f civilian, even aristocratic, pursuits than o f the military profession, and o f a native-born Greek to have written in the Latin language” ?12 The new interpretation is not persuasive, even if the word ut could mean “ although” in this context. For, while any author may preface his work with an apology and protest as he proceeds that he has only a middling talent (16 .1.2, 23.4.1: mediocre ingenium), a historians epilogue ought to reflect pride in the achievement o f completing a difficult task, and Ammianus’ self­ characterization precedes a claim that he has, to the best o f his ability, avoided ever misrepresenting the truth (31.16.9: opus veritatem professum/ numquam, ut arbitror, sciens/ silentio corrumpere vel mendacio). Moreover, for Ammianus the term Graecus is more than an epithet assigning a man to a par1 For a careful discussion o f modern interpretations, see G. Calboli, Festschrift fu r Robert M uth (Innsbruck, 1983)» 3 3 “ 5 3 2 J. F. Matthews, H istory and Historians (1983), 3 0 - 3 1 ; A m m ianus (1989), 4 6 1, cf. H. Drexler, Am m ianstudien (1974)1 179 n. 122; R. Browning, Cam bridge H istory o f Classical Literature 2 : Latin Literature , ed. E. J. Kenney (Cambridge, 1982), 749.

t«5]

T

he

G

r eek

T

e m p l a t e

ticular culture: it conveys a strong commendation. The only other singular Graecus in the extant books is the Augustan historian Timagenes, whom Am ­ mianus quotes as an authority on the ethnography o f Gaul. Timagenes is in­ troduced as a Greek by both scholarly expertise and language (15.9.2: et dili­ gentia Graecus et lingua). The order is highly significant: it is his scholarly expertise that stamps Timagenes as a Greek even more than the language in which he wrote. Ammianus should not be measured against contemporary Roman aristo­ crats who dabbled in history in the reign o f Theodosius.3 He intended to evoke a very different standard o f comparison. He set himself in the noble tra­ dition o f Thucydides and Polybius, historians who appealed to their own di­ rect and indirect knowledge o f contemporary events as conferring authority on what they wrote.4 When Ammianus described his method o f writing the history o f his own lifetime in the preface to Book XV, he used words that he surely intended to recall Thucydides: “ Using every effort to investigate the truth, I have set out, in the order o f their occurrence, events which my age allowed me to see myself or to know by thorough questioning o f those who took part in them” (15 .1.1). And the one explicit mention o f Polybius in the extant books is immensely revealing in what it implies about Ammianus him­ self. At the siege o f Pirisabora, the emperor Julian took part in an unsuccess­ ful attempt to force one o f the gates o f the city: “ He himself was unhurt, but his face was suffused with a modest blush. For he had read that Scipio Aemilianus with the historian Polybius, an Arcadian from Megalopolis, and thirty soldiers, had successfully undermined a gate at Carthage by an attack o f this sort.” (24.2.16). Since the young Ammianus served as a soldier in Julian’s Per­ sian expedition, it seems an irresistible inference that he is here comparing his relation to his hero to that o f Polybius to Scipio Aemilianus.5 Whether the implied comparison is deliberate or unconscious, it implies that Ammianus thought o f himself as a historian in the tradition o f Polybius— and hence also o f Thucydides. 3

Such as Nicomachus Flavianus, who was permitted to dedicate his lost A n n ales to the em­

peror Theodosius (IL S 2948: annalium quos consecrari sibi a quaestore et praefecto suo voluit). Flavianus has often been postulated as an important source for both Anunianus and other writ­ ers o f the late fourth century: J. Schlumberger, D ie "Epitom e de C aeso ribu s": Untersuchungen z u r heidnischen Geschichtsschreibung des 4. Jahrhunderts n. Chr. ( Vestigia 18, 1974); F. Paschoud, C in q Études sur Z osim e (Paris, 1975), 6 3 - 1 8 3 (Eunapius); Histoire A uguste 5 .1: Vies d ’A urélien , Tacite

(Paris, 1996), 1 0 —12 , 2 3 0 - 3 1 ; B. Bleckmann, H istoria 44 (1995), 8 3 - 9 9 (Anunianus). * F. W . Walbank, Historical Com m entary on Polybius i (Oxford, 19 57), 6 —16; Speeches in G reek H istorians (Oxford, 1965); Polybius (Berkeley, 1972), 3 2 - 9 6 .

5 Sabbah, M éthode (1978), 9 2 - 1 0 1 , 5 9 3 - 6 0 0 ; T. D. Barnes, R eading the Past (1990), 6 7 - 6 8 .

[66]

T

he

G

r eek

T

e m p l a t e

The giants o f classical scholarship in Wilhelmine Germany saw the essential Greekness o f Ammianus’ history very clearly. In his survey o f artistic prose from Gorgias to the Renaissance, Eduard Norden declared boldly that Am­ mianus “ thinks in Greek.” 6 Norden confessed that he had read little o f the Res Gestae (only a couple o f books), but he instantly recognized the author’s ha­ tred o f Constantius and partiality for Julian, he found his style excessively emotional, and he illustrated its underlying Greek structures o f thought by quoting the Greek equivalent o f individual words such as conducentia (18 .1.1: = ta sumpheronta) and by offering a Greek retroversion o f whole clauses (14.6.8, 10.16).7 The Hellenist Wilamowitz, the Latinist Friedrich Leo, and the Byzantinist Karl Krumbacher accepted that characterization wholeheart­ edly: in a survey o f Greek and Latin language and literature, they presented Ammianus as a Greek or early Byzantine historian, not among the Latin his­ torians o f Rome.8 Eduard Fraenkel agreed: he observed that Ammianus’ use o f the word dies reflected rules learned in the schoolroom rather than the us­ age o f a native speaker o f Latin.9 At a more mundane level, a series o f dissertations by young scholars and Pro­ grammschriften by teachers in German gymnasia found a large number o f Greek usages and turns o f phrase when they examined Ammianus’ diction and syn­ tax. Georg Hassenstein presented his dissertation on the syntax o f Ammianus at the University o f Königsberg in 1877: he started from the assumption that as a Greek Ammianus could not have avoided Graecisms, which he proceeded to find especially in his free use o f adjectives and participles as nouns, in ap­ parent peculiarities in his use o f the pronoun is and the third-person reflexive 6 E. Norden, A n tik e Kunstprosa2 (Leipzig, 1909), 648: “ zwar ist dieses Gräcisieren kein beabsichtiges, sondern die natürliche Folge der Unfähigkeit des Schriftstellers, sich in korrektem Latein auszudrücken: er denkt griechisch.” Norden expressly distinguished Ammianus’ uninten­ tional use o f Greek locutions from the learned and deliberate Graecisms employed by Latin lit­ erary artists from Ennius to Apuleius and Tertullian, which are documented by E. Löfstedt, S y n tactica. Studien und Beiträge z u r historischen S y n ta x des Lateins , 2 (Lund, 19 33), 4 0 6 - 3 1 . Löfstedt,

it may be observed, applied Norden’s criteria to the jurist Gaius, arguing that Graecisms such as the use o f ut followed by an indicative or by an accusative and infinitive to express purpose point to an eastern origin ( 4 3 1 - 3 2 ) . 7 Norden, K unstprosa2 (1909), 6 4 6 —50. Compare the recent comment on 1 5 .1 .1

by

C . W . Fornara, C abinet o f the M uses, ed. M . Griffith and D. J. Mastronarde (Atlanta, 1990), 166: “ the sentence is structurally Greek.” * U. Wilamowitz-Moellendorff, D ie griechische und lateinische Literatur und Sprache2 (Berlin and Leipzig, 1907), 2 0 1, comparing Ammianus to Eunapius, Olympiodorus and Priscus; K. Krum­ bacher, ibid. 266; F. Leo, ibid. 389: “ der griechischen, nicht der längst verlassenen römischen Bahn des Geschichtschreibung folgend.” 9 E . Fraenkel, Glotta 8 ( 19 17 ) , 55: “ D ieser G riech e steht eben nicht inm itten des lebendigen Flusses der Sprache, er hat sein Latein in der Schu le gelernt.”

[67]

T

h e

G

r eek

T

e m p l a t e

(including the genitive sui instead o f the adjective sum), and the frequency o f the indicative mood in reported speech.101 Others soon offered confirmation. H. Ehrismann could only explain oddities in Ammianus’ use o f moods and tenses by the hypothesis that he was a Greek who learned Latin late.11 Anton Reiter, a schoolteacher at Amberg, agreed with Hassenstein that Ammianus’ use o f the indicative in oratio obliqua was an authentic Graecism.12*Hassenstein had also suggested that Ammianus’ free use o f the preposition per might be influenced by the Greek dia: 13 the dissertation on Ammianus’ use o f preposi­ tions by Gustav Reinhardt documented the case at length.14 And a systematic study o f Ammianus’ vocabulary, syntax and style by F. Liesenberg, which he seems never to have completed, found extensive Greek influence on Ammi­ anus’ vocabulary: he noted locutions such as exmmc (21.10.2 = apo nun) and suasionis opifex (30.4.3 = pathous demiourgos) and strengthened Reinhardt’s in­ terpretation o f per as the equivalent o f dia.15 Much scholarly writing on Ammianus in the second half o f the twentieth century, however, even in Germany, has ignored or discounted the views o f Norden, Wilamowitz, and Leo, choosing to emphasize the Roman side o f Am ­ mianus to the detriment o f the Greek, and hence preventing a proper under­ standing o f the historian. Articles with programmatic titles such as “ Ammi­ anus Marcellinus as a Roman Historian” and “ Ammianus Marcellinus as a Late Antique Roman Historian” have proclaimed the historian to be fundamen­ tally Roman rather than G reek;16 the concluding chapter o f John Matthews’ large study (which seems nowhere to refer to Norden) is entitled “ The Ro­ man and the Greek,” not “ The Greek and the Roman” ; 17 and it has subse­ quently been argued that, although Ammianus may have been educated pri­ marily in Greek as a boy, his knowledge o f Greek literature was superficial and 10 G. Hassenstein, D e syntaxi A m m ian i MarceUini (Diss. Leipzig, pub. Königsberg, 1877), 3, 8- 9.

3 0 -3 1, 37- 38. 11 H . Ehrismann, D e temporum et modorum usu A m m ianeo (Diss. Straßburg, 1886), 6. 12 A . Reiter, D e A m m ia n i M arcellini usu orationis obliquae (Prog. Am berg, 1887), 6 8 - 6 9 . ,J Hassenstein, D e syn taxi (1877), 32. 14 G . Reinhardt, D e praepositionum usu apud A m m ian um (Diss. Halle, pub. Cöthen, 1888),

8 -4 5 . 15 F. Liesenberg, D ie Sprache des A m m ian us M arcellinus (Prog. Blankenburg, 18 8 8 —1890), 3 .1, 2 .1 5 , 1 . 13, 3 - 1 1 —1516 Respectively, H. Tränkle, A n tik e und A b en d la n d 11 (1962), 2 1 —33; K. Bringmann, A n tik e und A ben d lan d 19 (19 73), 4 4 —60.

17 Matthews, A m m ian us (1989), 4 5 2 - 7 2 . His earlier essay on Ammianus’ historical evolution held a better balance between the Greek and the Roman sides o f Ammianus, concluding that Greek was “ the dominant literary influence on his manner o f writing” (H istory and Historians [19 83], 3 0 - 4 1 . esp. 3 9 )-

[68]

T

he

G

r eek

T

e m p l a t e

derivative, whereas he had acquired a “ deep familiarity with the Latin liter­ ary tradition.” 18 Yet the best Latin philologists have never doubted that Ammianus some­ times lapses into the Graecisms characteristic o f one whose first language was Greek,19 and evidence has continued to accumulate to confirm Norden’s view that he thinks in Greek. In 1962 a brief article by Wolfang Seyfarth heavily reinforced the point: discussing the phrase exacerbantia tela in Ammianus’ story o f the third-century Persian capture o f Antioch (25.3.2), which earlier editors had emended in various ways, he defended the transmitted reading as “ ein echter Gräzismus” and observed that Graecisms were a characteristic feature o f Ammianus’ Latinity.20 The parallel o f Claudian reinforces the point: as Alan Cameron trenchantly observed, neither Claudian nor Ammianus can be un­ derstood except “ in terms o f a Greek, not a Latin, literary tradition.” 21 The interpreter o f Ammianus needs cultural sensitivity as well as the oldfashioned scholarly virtues. It is significant that the German scholar Joachim Classen stated the correct balance between “ Greek and Roman in Ammianus Marcellinus’ History” in an article composed in English while he was teaching m Nigeria22— and that the present writer lives in an officially bilingual coun­ try whose prime minister at the time o f writing is a Francophone with an im­ perfect command o f English. Moreover, the two studies that have decisively defined the Greekness o f Ammianus were produced by scholars working in a social and intellectual milieu that made them alert to cultural diversity. Hermann Schickinger was a teacher at the state gymnasium in the small Moravian town ofNikolsburg, on the main road from Vienna to Brünn (Brno): the town is now called Mikulov and lies just north o f the Czech-Austrian border. In the late nineteenth century, its population was one quarter Jewish, an unusually high proportion for the region.2-1 In the Programm for the acad­ 18 C . W . Fornara, H istoria 41 (1992), 4 2 0 - 3 8 . 19 E. Löfstedt, Vermischte Studien z u r lateinischen Sprachkunde und S y n ta x (Lund, 1936), 200, 208: Blomgren, Quaestiones (1937), 57, n. 1. 20 W. Seyfarth, K lio 40 (1962), 6 3 —64. 21 Alan Cameron, Renaissances before the Renaissance, ed. W . Treadgold (Stanford, 1984), 4 6 - 47.

22 C . J. Classen, M useum Africum 1 (1972), 3 9 - 4 8 . 2-' V. de St. Martin, Dictionnaire de geographic universelle 4 (Paris, 1890), 150: “ 5650 habitants, 7640 avec la communauté israélite” (from the census o f 1881). Jew s thus comprised about 26 per­ cent o f the population ofN ikolsburg (1,990 out o f 7,640), even though overall they constituted only about 2 percent o f the total population o f Moravia and less than 5 percent o f the total pop­ ulation o f Austria-Hungary: see the figures and estimates set out in A. Nossig (ed .). Jü dische Sta ­ tistik (Berlin, 1903), 4 3 2 - 5 2 ; J. Thon, D ie Ju d e n in Oesterreich (Berlin-Halensee, 1908), 5 —9. Nineteenth-century gazetteers note the salient characteristics ofN ikolsburg as its castle, its syn-

[69]

T

he

G

r eek

T

e m p l a t e

emic year 18 9 6 -18 9 7 Schickinger published a study o f “ The Graecisms in Ammianus Marcellinus” in which he set out the grammatical, syntactical, and lexical usages in the historian that are Greek.*24 Although not all o f the Grae­ cisms which he detected are real proofs that Ammianus’ mother tongue was Greek,25 Schickinger had a keen eye. He noted several apparent transpositions o f Greek words or expressions into Latin, such as the phrase “ mother o f cities” (mater urbium) for the metropolis o f a province (21.10.2; 26.1.3) and the use o f the Latin verb docere as if it were its Greek equivalent (16.8.10: tonstri­ ces docuit filias = “ taught his daughters to be barbers” ).26 Moreover, he ob­ served that Ammianus does not share the normal fondness o f Latin historians for the historical infinitive: in Gardthausen’s edition, he found a single solitary example in six hundred pages,27 which had rightly been emended away even before he wrote (29.3.7: mandarat magistris equitum auditoribus princeps,/ ut etc.).28 Ammianus, let it be added, does not avoid the historical infinitive because it had fallen into desuetude: it probably still survived in popular speech;29 Claudian uses it more than forty times;30 and Jerome employs it even in letters and lives o f saints.31 Most striking o f all, the historical infinitive is common in writers contemporary with Ammianus who, like him, were strongly influenced by Sallust: the Christians “ Hegesippus” and Sulpicius Severus no less than the pagans Aurelius Victor and the author o f the Epitome de Caesoribus.32 agogues, and the high proportion ofjew s: N u ovo dizionariogeografico universale 4 .1 (Venice, 18 31), 2 0 1; W. G . Blackie, Im perial G azetteer 2 (Edinburgh/London, 18 55), 4 89 —90; J. C . Smith, Harper's Statistical G azetteer 0 / the World (N ew York, 1855), 1260.

24 H. Schickinger, Programm des Staats-G ym nasium s in N ikolsburg 24 (18 9 6 -9 7 ), 1 3 - 3 0 . Schickinger’s work is unfortunately absent from the bibliographies in R. Moes, L es Hellénismes de l ’époque théodosienne. Recherches sur le vocabulaire d ’origine grecque chez Am rnien, C laudien et l ’ H is­ toire A uguste (Strasbourg, 1980), viii—xi; Matthews, A m m ianus (1989)» 5 5 4 ~ 6 o (“ Works directly relevant to Ammianus” ). 25 Seyfarth, K lio 40 (1962), 63, n. 5. For example, Schickinger, Programm N ikolsburg 24 ( 1 8 9 6 -

1897), 30, himself noted that the adjective semestris in the sense “ lasting half a month” rather than “ o f six months’ duration” already occurs in Apuleius, M et. 11.4 . 26 Schickinger, Programm N ikolsburg 24 ( 18 9 6 -18 9 7 ), 29. 27 Schickinger, Programm N ikolsburg 24 ( 18 9 6 -18 9 7 ), 27. 2K In favor o f the pluperfect mandarat in place o f the infinitive mandare , see O. Günther, Quaestiones A m m ianeae criticae (Diss. Göttingen, 1888), 5 5 - 5 6 , cf. H. Kallenberg, Quaestiones grammaticae A m m ian eae (Diss. Halle, 1868), 42. In 16 .11 .9 and 31.9*3 redire and transire had long

before been emended to rediere and transiere , respectively: on other passages where V offers a his­ toric infinitive, see J. den Boeft, Cognitio Gestorum (1992), 17. 29 J. B. Hofmann, D ie lateinische Umgangssprache 3 (Heidelberg, 19 5 1), 5 0 - 5 1 . 30 J. J. Schlicher, C P 10 (1915)» 6 5 - 6 6 , 73. 31 E.g., E p . 6 0 .1 0 .7 - 8 (Nepotianus* activities as a priest); Vita H ilarionis 10.4, 12.4, 29.8.

32 Schlicher, C P 10 (19 15 ), 7 1 , 74.

[70]

T

he

G

r eek

T

e m p l a t e

Jan den Boeft, a Dutch scholar writing in English, has brought the insights o f modern linguistic theory to bear with telling effect, for the phenomenon o f bilingualism in the modern world has been much studied— indeed, in many countries it is a pressing political problem. Den Boeft shows how, al­ though Ammianus is functionally bilingual in Greek and Latin, Greek is dom­ inant in various ways, the most obvious (and often noted) being his use o f the phrase “ as we call it in Greek” (e.g., 17 .7 .11: quas Graece syringas appellamus; 23.6.20: transire enim diabainein dicimus Graeci).-13 Greek was Ammianus’ first language. There is no way to discover at what age he learned Latin. But, although he may have been functionally bilingual, Ammianus’ Latin remains awkward and always reflects the Greek substratum o f his thought. Computer analysis and modern statistical techniques have shown how greatly the clausulae employed by Ammianus differ from those o f all other an­ cient Latin authors. In 1987 Steven Oberhelman examined the provenance o f Ammianus’ style by comparing the clausulae in the Res Gestae with those in more than a hundred Latin prose texts written between circa 200 and circa 450, and he concluded that Ammianus stands apart from all the other Latin writ­ ers o f the period in that his clausulae do not display any o f the “ metrical ten­ dencies” detectable to some extent in all o f them.3334 Oberhelman identified the basis o f Ammianus’ prose rhythm as “ the Greek accentual system that he knew from his rhetorical training” : his practice was to adhere to the Greek cursus, but with a preference for accentual clausulae that replicate familiar Latin metrical clausulae, since he needed to adapt the Greek system to take ac­ count o f the heavier, more spondaic nature o f the Latin language, where the accent can never fall on the final syllable o f any disyllabic or longer word and where long syllables and hence paroxytone accentuation preponderate.35 Ammianus was a Greek by language, education, and culture long before he ever entertained the idea o f writing history in Latin. It is his Greek cast o f mind that explains one o f the most unusual and distinctive features o f the Res Gestae: the large number, the variety and expansiveness, and the structural

33 J. den Boeft, Cognitio Gestorum (1992), 9 - 1 8 . O n 22.9.7 (apo tou pesein quod cadere nos dicimus), see ibid. 12: it is not an exception, as is often claimed (e.g., Ensslin, Ammianus [19 23], 30; Matthews, Ammianus [1989], 107), but should be translated, in conformity with other pas­ sages, “ from pesein, which is our word for radere.'' 34 S. M . Oberhelman, Q U C C , N .S. 27.3 (1987), 7 9 - 8 7 - This article belongs to a series on prose rhythm in Latin prose from the second century to the sixth: S. M . Oberhelman and R. G . Hall, C P 79 (1984), 1 x 4 - 3 0 ; 80 (1985), 2 1 4 - 2 7 ; C Q , N .S . 3$ (1985). 2 0 1 - 1 4 ; 36 (1986), 2 0 8 -2 4 ; S. M . Oberhelman, C P 83 (1988), 1 3 6 - 4 9 ; C Q , N .S. 38 (1988), 2 2 8 - 4 2 . 35 O n Ammianus’ clausulae, see further Appendix 6.

T

he

G

r eek

T

e m p l a t e

prominence o f formal excursus, which it is misleading (though conventional and perhaps forgivable) to call digressions.36 The excursus in a sense provide the web for the literary and intellectual fabric o f the work. This central fact about the Res Gestae deserves especial emphasis because the recent Penguin translation o f Ammianus omits almost all the excursus. An­ drew Wallace-Hadrill confesses in his preface to Walter Hamilton’s translation that the omission “ leaves a slightly'unbalanced impression o f Ammianus’ writ­ ing” and acknowledges that “ the digression was an integral feature o f Ammi­ anus’ approach to historical writing” ; nevertheless, he justifies the omission on the grounds that the excursus are “ frankly tedious to a modem reader.” 37* But it misrepresents an author, no less than a composer o f music, to improve his work by leaving out passages that offend the taste o f a later age.30 Ammianus explicitly marks thirty-one passages as formal excursus (Appen­ dix 5). They deserve to be considered as a group to the exclusion o f other pas­ sages that he does not so mark explicitly, however interesting the latter may be. Ammianus writes about the mating o f palm trees (24.3.12—13) and ex­ plains how midges limit the lion population in Mesopotamia: they attack the lions’ eyes and drive them mad from the irritation so that they either plunge into the river and drown or else blind themselves by constant scratching (18.7.5). But since Ammianus does not mark these passages as formal excursùs with the traditional formulae o f transition, they should be excluded from consideration o f the function o f the excursus in the Res Gestae as a whole. Serious Greek and Latin historians had always included digressions, most commonly geographical, but also cultural and ethnographical.39 But Ammi­ anus uses them on a scale unparalleled since Herodotus.40 On the other hand, he uses another standard device for varying or enlivening a narrative very sparingly. Whereas Tacitus had used speeches frequently to analyze motives or situations, and had even included private exchanges between an emperor and his minister,41 Ammianus includes few long speeches, all delivered on cer­ emonial occasions.42 The emperor Constantius delivers formal speeches to 36 E.g., A. Emmett, H istory am i Historians (19 S 3 ), 4 2 - 5 3 . 37 A . Wallace-Hadrill, A m m ianus Marcellinus (Harmondsworth, 1986), 25. .w por example, the epilogue o f M ozart’s D on G iovanni, which is now rightly included in performance. 39 P. Pédech, L a M éthode historique de Polybe (Paris, 1964), 5 1 5 - 9 7 . 40 M . Caltabiano, M etodologie della ricerca sulla tarda antichità , ed. A . Garzya (Naples, 1989), 2 8 9 -9 6 . 41 For example, Tiberius and Sejanus (A n n . 4 .3 9 -4 0 ) and Nero and Seneca (A n n . 14.53 — 56), where Tacitus maliciously makes the imperial pupil more philosophical than his tutor. 42 G. B . Pighi, I discorsi nelle Storie d ’A m m ian o Marccllino (Milan, 1936), 2 9 - 3 0 . He also lists twelve briefer utterances as “ discorsi storiografici minori” ( 3 0 - 3 1 ) .

[7 2 ]

T

he

G

r eek

T

e m p l a t e

his army when he grants peace to the Alamanni and Sarmatae Limigantes ( 1 4 .1 0 .n - 1 6 , 17 .13 .2 6 -3 3 ), when he proclaims Julian Caesar (15 .8 .5 -8 , 10, 12 - 14 ) , and when he is about to march against Julian in 361 ( 2 1.13 .10 - 15 ) . Julian addresses his army when he urges delay before the Battle o f Strasbourg in 357 (16 .12 .9 —12), when he is proclaimed Augustus in 360 (20.5.3-7), when he is about to invade Illyricum in 361 (2 1.5.2 -8 ), when he is about to enter Persian territory in 363 (23.5 .16 -2 3) and at Pirisabora (24.3.3 ~ l)- And, as he lies mortally wounded, Julian delivers a philosophical discourse to his friends (2 5.3.15 —20). Valentinian addresses the army both when he is pro­ claimed Augustus (26.2.6-10) and when he confers the imperial purple on his son Gratian (27.6.6—9, 12 —13). Ammianus also includes three letters quoted at length. Shapur writes to Constantius in 358 (17.5.3 —8) and Constantius replies (17.5.10 —15), and Julian writes to Constantius after his proclamation as Augustus (20.8.5—17). Ammianus also departs from conventional practice by entirely avoiding the familiar genre o f paired speeches in which opposing generals exhort their troops and answer each other when about to enter the fray o f battle. Tacitus provides the most spectacularly unreal example o f the genre when he makes Agricola and Calgacus debate the merits o f Roman imperialism: in what is virtually a formal controversia, the highland chieftain denounces Rome with a school rhetoric indistinguishable from that o f the Roman general, who speaks as if he knows exactly what has just been said on the opposing Scottish moun­ tainside.43 Paired speeches before battle constituted a minor literary genre that was not only fictitious, but also recognized as such by the ancients.44 As a sol­ dier, Ammianus knew that such speeches did not occur in real life, where there was only the possibility o f a few exhortatory remarks. Ammianus’ avoidance o f pre-battle speeches and the overall sparsity o f speeches in the Res Gestae contrast strikingly with the abundance o f excursus. The two phenomena must surely be connected. Ammianus has made a con­ scious choice. Why? The conventional answer to this question is to appeal to the model o f Herodotus. Like Herodotus, it is assumed, Ammianus recited his work in order to entertain: since an educated audience o f the late fourth cen­ tury expected excursus and judged them “ as a sort o f virtuoso cadenza,” Am­ mianus was giving his hearers what they wanted— and doing so allowed him both to slow down the tempo o f his narrative to place dramatic emphasis on 43 Tacitus, Agr. 3 0 - 3 4 , cf. R. M . Ogilvie and I. A. Richmond, Cornelii Taciti de Vita Agricolae (Oxford, 1967), 2 5 3 - 5 4 , 265. 44 M . H. Hansen, Historia 42 (1993), 1 6 1 - 8 0 . This central point is not contested by C . T. H. R. Ehrhardt, Historia 44 (1995), 1 2 0 - 2 1 .

f7 3 ]

T

he

G

r eek

T

e m p l a t e

an episode that followed an excursus and to put Roman history “ in perspec­ tive” by illustrating the diversity o f the empire and o f the barbarians that be­ set it.45 This answer may be correct in part, but it is founded on a mistaken assumption and it is incomplete. The only evidence that Ammianus recited his history is the letter o f Libanius to Marcellinus in 392: if Libanius wrote that letter to another Marcellinus (Chapter VI), then there is no evidence that any part o f the Res Gestae was recited to please an audience: on the contrary, Am­ mianus sets himself in the tradition o f Thucydides and Polybius in implicit opposition to that o f Herodotus. Comparison with Herodotus explains nei­ ther why Ammianus includes scientific excursus nor why he so often gives a learned exposition o f explanatory theories for natural phenomena.46 When he expatiates on earthquakes, eclipses, the rainbow or the plague, he is concerned to classify the phenomena and to report what seem to him the plausible ex­ planations set out by traditional Greek wisdom. Earthquakes, he informs his readers, occur in several ways: T h e y a re e it h e r ( 1) brasmatiae w h ic h , a r o u s in g th e e a rth fr o m th e b o t t o m lik e a tid e , p r o p e l e n o r m o u s m asses u p w a r d s — as in A s ia D e lo s a ro se fr o m th e sea, a n d T h e r a , A n a p h e , R h o d e s , w h ic h u se d to b e c a lle d O p h iu s a a n d P e la g ia in fo r m e r c e n t u r ie s , a n d w a s o n c e d r e n c h e d in a g o ld e n s h o w e r ; a n d E le u s is in B o e o t ia a n d ' V u lc a n u s in th e T y r r h e n ia n S e a a n d s e v e ra l m o r e isla n d s; o r (2) clim atiae w h ic h ru sh s id e w a y s a n d o b liq u e ly fla tte n c itie s , b u ild in g s a n d m o u n ta in s ; o r (3) chas­ matiae w h ic h , s u d d e n ly o p e n in g u p ab ysse s w it h t h e ir g r e a t e r m o t io n , s w a llo w p arts o f th e e a rth , as in th e A t la n tic a n isla n d la r g e r th an th e w h o l e o f E u r o p e , in th e C r is a e a n G u l f H e lic e a n d B u r a , a n d in th e C im in ia n p a rt o f Ita ly t h e t o w n o f S a c c u m u m w e r e d r a g g e d in to th e g a p in g m o u t h o f E r e b u s a n d a re h id d e n in e te rn a l d a rk n e ss. A m o n g th e se th re e ty p e s o f e a rth q u a k e s , mycematiae a re h e a rd w it h a t h r e a t e n in g ro a r, w h e n th e e le m e n ts b u rst t h e ir b o n d s a n d le a p u p o f t h e ir o w n a c c o r d o r slip b a c k as th e e a rth su b sid e s. F o r t h e n , o f n e c e ssity , th e c r a s h in g a n d r u m b lin g o f th e e a rth m u s t r e s o u n d lik e th e b e llo w in g o f b u lls. ( 1 7 . 7 . 1 3 / 1 4 )

Similarly, plagues are o f three types: S c ie n tis ts a n d e m in e n t p h y s ic ia n s te ll us th at in fe c tio u s d ise a se s a re b r o u g h t o n b y e x c e s s o f c o ld o r h e a t, o r o f m o is t u r e o r d ry n e s s. T h a t is w h y p e o p le w h o liv e in d a m p o r m a rs h y p la c e s s u ffe r f r o m c o u g h s a n d o p h th a lm ia a n d th e lik e , w h e r e a s th e in h a b ita n ts o f h o t c o u n t r ie s a re d e h y d r a te d b y b u r n in g fe v e r . . . . O th e r s a s -

45 Wallace-Hadrill, A m m ian us (1986), 25, cf. Matthews, A m m ian us (1989), 462. 46 O n the literary function o f the scientific excursus, see D. den Hengst, Cognitio Gestorum

(1992). 39-46-

[74]

T

he

G

reek

T

e m p l a t e

se rt th a t th e a ir, lik e w a te r , ca n b e in fe c te d b y th e s te n c h o f d e a d b o d ie s o r th e lik e a n d g r a v e ly a ffe c t m e n ’s h e a lth , o r at a n y ra te th at m in o r illn esses a re c a u se d b y a su d d e n c h a n g e o f a ir. T h e r e a re also s o m e w h o say th at w h e n th e a tm o ­ s p h e re is t h ic k e n e d b y u n u s u a lly d e n se e x h a la tio n s fr o m th e e a rth it b lo c k s th e b o d y ’s n a tu ra l v e n ts a n d s o m e tim e s c au ses d e a th . . . . T h e first ty p e o f p e s tile n c e , w h ic h ca u se s th e in h a b ita n ts o f e x c e s s iv e ly d r y re g io n s to b e v is ite d b y fr e q u e n t fe v e rs , is c a lle d e n d e m ic . T h e s e c o n d , w h ic h o c c u r s p e r io d ic a lly a n d affects th e s ig h t a n d c a u se s d a n g e r o u s h u m o rs in th e b o d y , is e p id e m ic . T h e th ird is p la g u e , w h ic h a lso lasts o n ly f o r a tim e , b u t b r in g s d e a th w ith lig h t n in g s p e e d . ( 1 9 . 4 . 2 —7)

The genre o f discourse in passages such as these is not historical. Nor is it truly scientific, the product o f deep enquiry or research. It is rather doxographical.47 But the aggregation o f doxographies on a multitude o f varied subjects adds up to something approaching an encyclopedia. Theodor Mommsen pointed out this unifying feature o f Ammianus’ ex­ cursus at the end o f a long and famous paper on the historian’s geographical knowledge.48 Paradoxically, however, although Mommsen’s paper is well known and his conclusions about Ammianus’ geographical sources have often been quoted with approval,49 his general remarks on the function o f the ex­ cursus have had little resonance in subsequent scholarship.5051His conclusions, therefore, deserve to be restated. Behind the separate geographical excursus divided and worked up schematically, Mommsen detected conscious inten­ tion and a deliberate plan, whereby the historian intended to weave into his history a description o f the whole inhabited world. Mommsen waxed elo­ quent on Ammianus’ deficiencies— awful carelessness, empty words conceal­ ing o f lack o f knowledge, a parade o f erudition ineffectually veiling profound ignorance, an idle striving to know all. But he partly excused the dark shad­ ows o f Ammianus’ inadequacy as the inevitable consequence o f the fact that he belonged to an unhappy generation o f squatters in the ruined world o f a great past.sl Mommsen’s harsh and unsympathetic assessment o f both Am­ mianus and his intellectual environment owes more to his own lifelong anti­ clericalism than to a correct appreciation o f the historian’s achievement. More­ over, he seriously underestimated Ammianus’ learning. Ammianus did not crib his scientific material from encyclopedias. Rather, he gave his real, if limited, 47 R. von Scala, Festgabe z u Ehren M a x Büdinger's uon seinen Freunden und Schülern (Innsbruck, 1898), 1 1 7 - 5 0 . 48 T. Mommsen, G es. Sehr. 7 (Berlin, 1909), 4 2 3 - 2 4 . 44 E.g., Syme, A m m ianus (1968), 105. 50 Observe, however, Ensslin, A m m ianus (1923), 16; A. Solari, Rendiconti L in cei 8 4 (19 49 ), 2 1. 51 Mommsen, G es. Sehr. 7 (1909), 423: “ der Fluch jener unseligen, auch aufdem geistigen G e­ biet in der Trümmerwelt einer grösseren Vergangenheit kümmerlich hausenden Generationen.”

[75]

T

h e

G

r eek

T

e m p l a t e

scientific, ethnographic, and philosophical knowledge the form o f entries in an encyclopedia. He informed his readers about the crocodile and the hippopota­ mus, about comets and atoms, and even about utterly unfamiliar peoples in southern Russia— the Byzares, Sapires, Tibareni, Mossynoeci, Macrones, and Philyres (22.8.31: populi nulla nobis adsuetudine cogniti). That is a parade o f knowledge for its own sake, designed more to impress than to instruct, but it is knowledge that Ammianus acquired as a result o f his study o f Greek ethno­ graphical, scientific, technical, philosophical, and historical writers. It has been conventional to belittle Ammianus’ erudition. A century ago, an attempt was made to derive most o f the material in most o f Ammianus’ ex­ cursus from a single doxographical handbook, whose author was none other than Posidonius.52 More recently, it is argued that when Ammianus names Ptolemy as an authority on astronomy (20.3.4), he has “ in all likelihood” not consulted Ptolemy directly, and that, although he may have consulted “ pop­ ular Greek astronomical treatises” such as those by Cleomedes, Theon o f Smyrna, and Geminus, similarities to Cicero’s De Re Publica and Calcidius’ commentary on Plato’s Timaeus prove that his working sources were in fact Latin.53 Hence a systematic enquiry concludes that, out o f the one hundred and six passages where Ammianus names or cites a Greek writer, no more than twenty show direct knowledge o f the author named or cited.54 But it is surely improbable a priori that one who thought in Greek preferred Latin over Greek sources. Let one striking example where Ammianus’ direct ac­ quaintance with a text to which he explicitly refers has been denied be taken as a test case for his learning. Ammianus names the philosopher Plotinus twice, both times in excursus. One mention is a mere gloss: the Alexandrian philosopher Ammonius Saccas is identified as the teacher o f Plotinus (22.16.5). The other reference is more substantial: It is m a d e c le a r b y th e im m o r t a l p o e m s o f H o m e r th a t it w a s n o t th e g o d s in h e a v e n w h o c o n v e r s e d w it h h e r o e s o r s to o d b y a n d h e lp e d t h e m as t h e y fo u g h t , b u t th e fa m ilia r s p irits w h ic h b e lo n g e d to t h e m . B y t h e ir s ig n a l assista n c e , it is sa id , P y t h a g o r a s , S o c r a te s , N u m a P o m p iliu s , th e e ld e r S c ip io , M a r iu s (in th e o p in io n o f so m e ) a n d O c t a v ia n , w h o w a s th e firs t to b e a w a rd e d th e title o f A u - 32

32 R. von Scala,

Festgabe zu Einen M a x Büdinger’s (18 9 8 ), 12 2 : “ eine einheitliche und zw ar

doxographische W eisheitsquelle” ; 150 : “ kann diese doxographische Q uelle schwerlich eine an­ dere als Poseidonios gew esen.” 33 D . den H engst, M n em osyn e4 39 (19 8 6 ), 1 3 7 . 54 I. Lana, Politica, cultura e religione n e ll’Impero romano (secoli F. C o n ea, I. Gualandri and G . Lozza (Naples, 19 9 3 ), 2 3 - 4 0 .

[76]

I V- VI) tra O riente e Occidente, ed.

T

h e

G

r eek

T

e m p l a t e

g u stu s, a tta in e d fa m e , as also H e r m e s T ris m e g is tu s, A p o llo n iu s o f T y a n a an d P lo tin u s , w h o w a s b o ld e n o u g h to w r it e o n th is e s o te r ic s u b je c t a n d to d e m o n ­ stra te b y p r o fo u n d r e a s o n in g f r o m w h a t b e g in n in g s th e se sp irits are jo i n e d to th e so u ls o f m o rta ls a n d , as it w e r e , t a k in g t h e m in th e ir b o s o m s p r o t e c t t h e m as lo n g as it is a llo w e d , a n d in s tr u c t t h e m in h ig h e r tru th s, i f th e y fe e l th at th e y a re p u re a n d h a v e k e p t th e m s e lv e s sp o tless a n d fr e e fr o m a n y sta in o f sin b y a ss o c ia tin g w it h th e b o d y in an im m a c u la te fa s h io n . ( 2 1 . 1 4 . 5 )

Ammianus here summarizes Plotinus’ treatise “ On Our Allotted Guardian Spirit” 55— and the natural assumption should be that he has read the text o f the treatise.56 It may be true, as the Dutch commentators state, that this brief paraphrase is “ totally inadequate,” but that is not a sufficient reason for con­ cluding that Ammianus bases it not on direct knowledge o f Plotinus, but on the account that Porphyry offers in his Life of Plotinus o f how Plotinus came to write it.57 The most superficial acquaintance with modern scholarship ought to show how fallacious the inference from misrepresentation to igno­ rance can be. Moreover, all that Porphyry’s Life says about the content o f the treatise is that Plotinus “ sets out to explain the differences between spiritcompanions.” 58 Hence the truth is probably much more complicated than has been supposed. Ammianus appears to make a perceptible effort to report Plot­ inus’ often obscure train o f thought,59 yet the phrase tamquam gremiis suis sus­ ceptas- tuentur seems to reflect the title o f the treatise, which Porphyry, not Plotinus, gave it.60 Furthermore, the Dutch commentators have observed a no­ table fact about the language that Ammianus employs in this summary o f Plo­ tinus. He uses the noun colluvio and the adjective immaculatus in a metaphori­ cal sense. Both usages are attested elsewhere only in Christian writers: hence Ammianus transmits this brief précis o f a treatise o f Plotinus to his readers through a Christian filter, presumably in his own mind. The evaluation o f this passage involves a fundamental fact about Ammianus. The standard assumption o f late has been that he acquired all his knowledge o f Neoplatonism in Rome in the 380s.61 But why may Ammianus not have Ennead 3.4. Symb. Osl. 22 (19 4 2 ), 63. 57 A s suggested by P. H enry, Plotin et l ’Occident (Louvain, 19 34 ), 1 9 6 —99. H e n ry ’s cautious ss Plotinus,

56 S. Eitrem ,

proposal (“ il se peut do n c” etc.) has been elevated to a certainty b y later writers. 5X Porphyry,

Vita Plotini 1 0 . 3 3 - 3 5 .

s‘' F o r a m odern synopsis o f this difficult treatise, see A . H . Arm strong,

Plotinus 3 (C a m ­

bridge, M ass., 19 6 7 ), 1 4 1 . V ila Platini 1 0 .3 1 , 2 5 .3 , cf. H enry, Platin (19 3 4 ), 19 8; P. H e n ry and Platini Opera (Paris/Brussels, 1 9 5 1 ) , 3 10 . 61 Besides the D u tch com m entators, note esp. J. Szidat, Mus. Hclv. 39 (19 8 2 ), 1 3 2 - 4 5 ; M atthew s, Ammianus (19 89 ), 4 2 9 - 3 0 . fin Porphyry,

H .-R . S ch w y ze r,

[77]

T

he

G

r eek

T

e m p l a t e

taken an interest in philosophy in his youth? Even if the collected edition o f the Enneads that Porphyry completed in 300 remained unknown in the East until the last third o f the fourth century, individual treatises by Plotinus cer­ tainly circulated much earlier.62 Ammianus could have come across the trea­ tise that he summarizes in Phoenicia or elsewhere in the Syrian region in the 340s or 350s, and he could have studied it with a pupil o f Porphyry. In 360, it should be recalled, Libanius referred to an Ammianus in military service who was a philosopher underneath his uniform: he could be the future historian.63 62 J . M . Rist, Basil o f Caesarea: Christian, H um anist, Ascetic, ed. F. J . Fed w ick (Toronto, 1 9 8 1) , 137 -22 0 . 63 Libanius, E p . 2 3 3 .4 .

[78]

[VIII] CHRISTIAN LANGUAGE AND ANTI-CHRISTIAN POLEMIC

Ammianus’ eastern origin and his Greek cast o f mind are very relevant to as­ sessing his religious beliefs and his treatment o f Christianity. When he wrote Graecus, what he heard inside his own mind was the Greek word Hellen, which has a very different semantic range from its Latin equivalent.1 For, al­ though Hellen could indicate merely that someone was culturally Greek,12 by the late fourth century it often had the specific meaning o f “ pagan,” especially when used by anyone at all hostile to Christianity.3 This sense o f the word is first clearly documented in Porphyry’s polemic Against the Christians: proba­ bly writing circa 300, he complained that Origen, though “ a Greek educated in Greek learning,” had deviated into “ barbarian recklessness” by becoming a Christian,4 and represented the Christians as apostates from both Hellenism and Judaism.5 Eusebius o f Caesarea accepted the dichotomy between Hel­ lenism and true religion, but claimed that the traditional religions o f the Ro­ man Empire were a declension from the original pure religion o f the Hebrew patriarchs, which the coming o f Christ had renewed in the Christian church.6 Hellenism in Late Antiquity (Ann A rbor, 1990), 9 - 1 3 . Ancient World 2 4 (19 9 3), 2 5 ” 2 93 N o te Eunapius’ usage in his Lives of Philosophers and Sophists: once certainly to designate in­ 1 G . W . Bow ersock,

2 Alan C am ero n,

habitants o f G reece (6.6.2 [466]) and once possibly (6.4.7 [465]), but five times clearly meaning no n -C h ristian (6.5.3 [465]; 7 - 3 - 1 2 [4 77]; 10. 6. 3 [490], 6.6 [4 9 1], 8.2 [493])-

H E 6 .1 9 .7 = Porphyry, Against the Christians, frag. 39 Harnack. O n the date and B IC S 38 (19 9 4), 5 3 - 6 5 . 5 Eusebius, Praep. Ev. 1 . 2 . 1 - 4 = Porphyry, Against the Christians, frag. 1 Harnack, cf. B IC S 4 Eusebius,

context o f the w o rk , w h ich are controverted, see 38 (1994)» 6 5.

H E 1 . 2 . 2 - 4 . 1 ; Praep. Evan. 7 . 8 - 8 . 1 4 ; Dem. Evan. 1 . 6 . 2 9 - 7 6 , 7 . 1 . 8 0 - 1 5 4 , Les Vues historiques d ’Eusèbe de Cêsarée durant la période prénicéenne (Dakar, 1 9 6 1) ; T . D . Bam es, Constantine ( 1 9 8 1 ) , 1 2 6 - 2 7 , 1 8 4 - 8 5 . 6 Eusebius,

8 . 3 . 6 - 1 5 ; J. Sirinelli,

[79]

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

The emperor Julian simply reversed Eusebius’ values while retaining his basic antithesis and asserted a fundamental incompatibility between Greek culture and Christianity, between Hellenism and barbarism.7 Hence, when Ammi­ anus described himself as miles quondam et Graecus, he was declaring his reli­ gious allegiance in unambiguous terms.8 Since Ammianus’ religious beliefs are central to interpreting his history, they have by no means been neglected by modern students o f the historian. Initially at least, it may seem surprising that scholars from the seventeenth century onward have attributed to Ammianus every shade o f opinion within the religious spectrum— at the one extreme a Christian, a monotheist sym­ pathetic toward Christianity, a man who genuinely respected Christianity,9 at the other a committed and militant pagan striving to rescue something o f Rome’s traditional religions from the discredit attendant onjulian’s catastrophic failure.101 The center o f the spectrum has o f course appealed most to scholars who wish to appear to be judicious: hence Ammianus has most often been presented as a man o f genuine, but broad-minded piety who accepted the doctrines o f no specific religion, an adherent o f a “ neutral monotheism” who was neither a Christian nor a committed believer in the old gods.11 On this majority view, Ammianus was hostile to fanaticism and excess o f any sort: as a moderate and tolerant pagan, he believed strongly in religious toleration and respected Christianity, even if he sometimes criticized the conduct o f indi­ vidual Christians and Christian groups.12 7 See the numerous passages registered in their index b y J. Bidez and F. C u n io n t, Imp. Cae­ saris Flavii luliani Imperatoris Epistulae Leges Poematia Fragmenta varia (P a ris/London, 19 2 2 ), 30 2. * A . Stoian, Latomus 2 5 (19 6 7), 7 3 - 8 1 ; J. H eyen , Latomus 26 (19 6 8), 1 9 1 - 9 6 . 9 For a survey o f m odern opinions, see Rike, Apex Omnium (19 8 7 ), 1 —5. T h e v ie w o f the seventeenth-century scholars Pithou and Chifflet that Am m ianus was a Christian was decisively refuted by G ronovius in the introduction to his edition (Leiden, 16 9 3), reproduced in the var­ iorum edition o f W agner and Erfurdt (Leipzig, 1808), xlii; sym pathy is detected by Dem andt,

Zeitliritik (19 6 5), 82; a “ positive evaluation o f monastic asceticism” b y V. N e ri, Ammiano e il cristianesimo. Religione e politiea nelle "Resgestae" di Ammiano Xlarcellino (Bologna, 19 8 5), 70. 10 Rike, Apex Omnium (19 8 7 ), 8, 6 9 - 1 3 3 . T h is bo o k is criticized harshly by M atthew s, w h o accuses Rike o f “ insecure technique in handling certain aspects o f the text” (Ammianus [19 8 9 ], 5 4 5. n. 10).

11 A s argued b y E . W itte, Ammianus Marcellinus quid indicaverit de rebus divinis (Diss. Jen a, 18 9 1) , 5 8 —5912 So , for example, Sym e,

Ammianus (19 68), 1 3 7 - 3 8 ; J. F. M atthew s, Ancient Writers, 2 (N e w

York, 19 8 2 ), 1 1 18 ; E . D . H unt, C Q , N .S . 3 5 (19 8 5 ), 1 8 6 - 2 0 0 , strongly com m ended by

Ammianus (19 89 ), 546, n. 2 2. For criticism o f this view , Reading the Past (1990), 7 5 C P 88 (19 9 3), 6 7 - 6 9 . H unt has attempted to reassert it in Studia Patristica 2 4 (19 9 3 ), 108 —

M atthew s, 82;

1 3 , w h ere he admits “ satirical exaggeration” in the com parison o f Christians to w ild beasts (2 2.5.2 ). In his book, Jo h n M atthew s is deeply inconsistent: sometimes he adopts a position close to that argued here, as w h e n he speaks o f “ a polem ic o f distortion” o r o f “ the dismantling o f ap­ parently authentic connections” in order to m inim ize the role o f Christians and draws attention to Am m ianus’ use o f standard polem ical techniques

[80]

(Ammianus [19 8 9 ], 4 3 5 , 4 4 1 , 4 5 1) ; yet he de­

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

W hy has there been such a wide spectrum o f views? Why have different scholars discovered such a variety o f divergent Weltanschauungen in the same text? The reason is not that scholars have simply imported their own precon­ ceptions into their interpretation. It is the contrary one that virtually all mod­ ern views have some foundation in Ammianus’ text. Divergence arises be­ cause a simple and straightforward answer is sought, when there is a deep and irremovable inconsistency in what the historian says about Christianity. At the conscious level, Ammianus sets out to marginalize Christianity by deliberately understating the role that Christians and Christianity played in the political history o f the fourth century. The most widely acknowledged omission was noted by Norman Baynes in 19 10 .13 It concerns Armenia. Am­ mianus offers a moderately detailed account o f Roman dealings with Arme­ nia in the reign o f Valens (27.12, 2 9 .1.1- 4 , 30 .1-2 ). The Armenian historian Pawstos Buzand presents the conflict between King Pap and the katholikos Nerses as the central fact o f the politics o f the kingdom in the 370s: the king, who was hostile to Rome, had the pro-Roman katholikos murdered.14 Ammi­ anus never mentions Nerses or his opposition to the king, despite its relevance to the events that he describes. There is a similar, but far more serious, omis­ sion o f ecclesiastical politics in the Roman Empire. Book X V introduces Athanasius as if Ammianus had never mentioned him before (15 .7 .7 -10 ). Yet Athanasius’ career as bishop o f Alexandria from 328 to 373 (with significant interruptions and periods o f exile) shows that ecclesiastical affairs were at the center o f imperial politics in the reign o f Constantius.15 In 350, for example, Constantius’ praetorian prefect was on his way to Egypt to arrest Athanasius when news came o f the death o f Constans: fearful that Egypt might transfer its allegiance to the usurper Magnentius, the emperor reversed his policy and insincerely assured the bishop o f Alexandria o f his undying friendship and support.16 Ammianus also leaves out completely the ecclesiastical affairs that occupied a large amount o f Constantius’ time during the years for which his

tects genuine sym pathy for captured Christian virgins and for marty rs as victim s o f persecution and expressly denies that Am m ianus was “ a polemical w riter in the manner, say, o f Eunapius o f Sardis” (436, 445)13 N . H . Baynes,

EH R 25 (19 10 ), 6 2 5 - 4 3 , reprinted in Byzantine Studies and Other Essays REArm , N .S . 4 (19 6 7 ), 2 9 7 - 3 2 0 ; Blockley, Ammianus ( 19 7 5 ), 6 2 - 7 2 ; Elliott, Ammianus (19 8 3), 1 8 0 - 8 2 . 14 Fo r the relevant passages ( 5 . 1 , 2 1 - 2 4 , 3 0 - 3 1 ) , see N . G . Garsoïan, The Epic Histories attrib­ uted to Pawstos Buzand (Buzandaran Patniut'iwnk) (Cam bridge, Mass., 1989), 1 8 5 - 8 7 , 2 0 2 - 5 ,

(London, 19 5 5 ), 1 8 6 - 2 0 8 . Fo r subsequent discussion, see N . G . Garsoïan,

2 1 0 - 1 3 (translation); 3 0 6 - 6 7 , 3 1 7 - 1 9 , 3 2 3 - 2 4 (com m entary); 3 9 5 - 9 6 (on Nerses); 3 9 7 - 9 8 (Pap). P cawstos 5 .1 records the installation o f Pap circa 37 0 by Terentius and Addaeus: A m m i­ anus names only Terentius ( 2 7 .1 2 .1 0 ) .

15 Athanasius

( 1 9 9 3 )» 1 6 5 - 7 3 .

16 Athanasius,

Hist. Ar. 5 1 .4 ; Apol. ad Const. 2 3 , cf. Athanasius (19 9 3), 1 0 4 - 5 . 18 1 ]

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t i a n

P

o l e m ic

account is extant. Between the summer o f 353 and the spring o f 360, the em­ peror Constantius was close at hand to supervise the conduct o f several im­ portant gatherings o f bishops: the Council o f Arles in the winter o f 35 3—354, the Council o f Milan in the summer o f 355, the small gathering o f bishops that produced the so-called “ dated creed” in his presence at Sirmium on 22 May 359, and the Council o f Constantinople in January 360. Moreover, during the summer and autumn o f 359, the emperor was deeply involved in convening, then in coercing the Councils o f Ariminum and Seleucia, and finally in the complicated negotiations that produced the homoean creed on the last day o f December.17 None o f these councils earns the merest mention in Ammianus’ narrative o f these years. Nor does the ecclesiastical background to Julians rebellion against Constantius. Julian was able to establish his independence o f Con­ stantius largely because western Christians rallied to his support as the only way to avoid the imposition in the West o f the official homoean creed which the Council o f Constantinople proclaimed in January 360.18 In addition, like other historians o f Late Antiquity, Ammianus adopts an external stance when he writes about Christianity and the Christian church. This has normally been regarded as a mere literary convention that reveals nothing about Ammianus’ attitude toward Christianity: a similar stance, using very similar language, can be seen in historians down to the sixth century, in­ cluding writers such as Procopius and Agathias, who were undoubtedly Chris­ tians.19 The argument is valid in general, but it has been given a misleading application. In contrast to Ammianus, Procopius makes his Christianity clear in several passages.2021But when Ammianus twice glosses the word synodus with the phrase ut appellant (15.7.7; 2 1.16 .18 ), he implies that one o f the common­ est forms o f political assembly in the fourth century is an exotic rarity. At a deeper level, however, the Res Gestae exhibit a feature that contradicts Ammianus’ external stance toward Christianity. Any satisfactory analysis o f his religious attitudes must recognize and explain the fact that Ammianus often uses Christian language and Christian modes o f thought and expression with­ out any apparent sign o f self-consciousness. This important linguistic phe­ nomenon was detected several decades ago by G. B. Pighi,2' and is now docu­ 17 For these events, see, briefly,

Athanasius (19 9 3), 1 0 9 - 2 0 , 1 3 6 - 5 1 .

18 Athanasius (1993), 14 9 - 5 1, 153 —5419 Alan and Averil Cam eron, C Q , N .S . 14 (19 6 4), 3 1 2 - 2 8 . 20 Averil C am eron,

Historia 1 5 (1966), 4 6 6 - 8 2 ; Procopius and the Sixth Century (Berkeley,

1985), 33 - 46 , 113 - 3321 G . B . Pighi, Studi dedicati alla memoria di Paolo Ubaldi (Milan, 19 3 7 ) , 4 1 - 7 2 ; R A C i (19 50 ), 3 8 6 -9 4 .

[82]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

mented on a massive scale by the Dutch commentators (who have so far reached Book X X II). Ammianus thus uses Christian language while implicitly dis­ paraging the importance o f Christianity. What might be the explanation o f this apparent contradiction? The analogy ofjulian, whose writings exhibit a simi­ lar linguistic phenomenon,2223suggests that Ammianus, too, was an apostate Christian. Ammianus betrays the artificiality o f his external stance when writing about Christianity in his use o f the ordinary words, common to both Greek and Latin, for priest, bishop, and church. He uses the word presbyter twice in the ex­ tant part o f his history. On the first occasion, he mentions, without glossing the word, a “ certain man o f Epirus who was a priest o f the Christian religion” (29.3.4: Epirotem aliquem ritus Christiani presbyterum). In Book X X X I, however, when precisely the same phrase occurs for the second time, the word presbyter is glossed, for emphasis and with polemical intent. On the morning o f the Battle o f Adrianople, the Gothic leader Fritigern sent envoys to the emperor Valens. The embassy is described as comprising “ a priest (as they themselves call them) o f the Christian religion together with other humble fellows” (31.12.8: Christiani ritus presbyter, ut ipsi appellant. . . cum aliis humilibus). A few lines later Ammianus informs his readers that this same Christian priest (idem Christianus) was privy to the plans o f Fritigern and gave Valens a secret and dishonest letter (31.12.9). That looks like guilt by associa­ tion— and the association is deliberate. The extant books use episcopus for “ bishop” seven times,21 but Ammianus never glosses it, though he does also use circumlocutions such as Christianae legis antistes (15.7.6). And he uses ecclesia (“ church” ) without inhibition four times in the extant books as a word whose meaning needs no explanation.24 In his account o f the killing o f Silvanus, however, Ammianus avoids the word ecclesia by using an apparently classical circumlocution: he states that Silvanus fled ad conventiculum ritus Christiani (i 5.5.31). This attempt at classicism fails in a most interesting way: the word conventiculum in the sense “ place o f meeting” is predominantly Christian.25 Ammianus sometimes unconsciously accepts Christian values as well as Christian vocabulary. In his obituary ofjulian, he states at the outset that he will assess the dead emperor in terms o f the four traditional cardinal virtues o f the philosophers (self-control, wisdom, justice, and courage) together with the 22 See G . J . M . Bartelink, Vig. Chr. 1 1 (19 5 7 ), 3 7 - 4 8 ; C h apter X III. 23 Viansino, Lexicon 1.4 7 3 . 24 Viansino, Lexicon 1.4 56 . 25 T L L 4 .8 4 4 - 4 5 .

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

practical gifts o f military skill, dignity, prosperity, and generosity (25.4. i).26 Yet his detailed enumeration o f Julians virtues starts with the statement that the emperor was “ so spectacularly and incorruptly chaste that after the loss o f his wife he never tasted the pleasures o f sex” (25.4.2). Despite the allusions to Plato and Bacchylides that follow, the frame o f reference is Christian: Ammi­ anus claims that not even his closest attendants ever suspected Julian o f har­ boring any libido during the whole o f his adult life (25.4.3).27 A Christian stan­ dard o f morality also lies behind the way in which Ammianus praises Julian for disdaining a beautiful captive girl offered to him after the storming o f Maiozamalcha. As Ammianus observes, Alexander and Scipio provided the model for such a refusal.28 But Julian refused even to look upon the captive, lest his resolve be broken by desire (24.4.27: ne frangeretur cupiditate). To avoid even lusting after a beautiful woman is a Christian more than a traditional Greek or Roman, virtue.29 Such Christian language and Christian assumptions are, however, overlaid with prejudicial anti-Christian vocabulary taken from the discourse o f reli­ gious polemic. Since Ammianus thought in Greek, it was easy and natural for him to transpose into Latin the polemical vocabulary used against the Chris­ tians by contemporary Greek pagans. Denigration o f the type employed by Ammianus can be found in the emperor Julian, the orator Libanius, Eunapius’ Lives of the Philosophers and Sophists,M> the historian Zosimus in that section o f his history where he follows Eunapius, and Damascius’ Life of Isidore in the sixth century.31 26 O n this obituary, see Drexler,

Ammianstudien (19 74 ), 9 4 - 1 3 6 . T h e order in w h ich A m ­

mianus lists the four cardinal virtues is untraditional, but was also adopted by M am ertinus in his

(Pan. Lat. 3 [ 1 1 )). A n innova­ Abli. Mainz, Geistes- und Sozialwiss. Kl. 1968, 5 2 4 n. 1; A . Guida, Un anonimo panegirico per I'imperatore Giuliano (Anon. Paneg. lui. Imp.)

panegyric o f Julian delivered in Constantinople on 1 January 3 6 2 tion o f the reign o f Julian is detected by H . Gärtner,

(Florence, 1990), 7 5 . 1 have argued that the em peror praised in the fragm entary panegyric edited by Guida may be Constantine rather than Julian, and hence that the n e w order, w h ich put selfcontrol first, begins under the first Christian emperor: Akten des 21. Internationalen Papyrologenkongresses in Berlin 2 (Leipzig, 19 9 7), 6 7 —70. 27 C . J. Classen, Museum Africum 1 (19 7 2 ), 4 4 - 4 $ , aptly com paring Jero m e, Commentary on Ephesians 1.2 (PL 26.464). 2X Am m ianus appears to be cop yin g Aulus Gellius, Noctes Atticae 7 . 8 . 1 —3. Scip io ’s motives w ere political, according to Polybius 1 0 .19 .3 ~T> Livy 2 6 .50 , cf. H . H . Scullard, Scipio Africanus: Soldier and Politician2 (London, 19 70 ), 64. v> M atthew 5.2 8 : “ I f a man looks on a w om an w ith a lustful eye, he has already com m itted adultery w ith her in his heart.” O n the anti-Christian polem ic in this w ork, see G . J. M . Bartelink,

Vig. Chr. 2 3 (1969),

2 9 3 - 3 0 3 . Unfortunately, he begins by contrasting Eunapius with “ A m m ien M arcellin, histo­ rien impartial, qui ne dénigre jamais les Chrétiens dans son œ uvre historique” (293). 31

T h e annotation in the excellent edition by C . Zin tzen ,

Damascii Vitae Isidori reliquiae

(Hildesheim, 19 67), identifies m any anti-Christian allusions: although Damascius provides a very full repertoire o f insults, I refrain from adducing him to avoid any suspicion o f anachronism.

[84]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

When Ammianus applies the derogatory term “ corpse” to the relics o f the martyr Babylas, he seems to be echoing a specific passage ofjulian. The Misopogon speaks o f the. “ corpse o f Daphne” and “ the shrine o f the corpse” in a tone and a context that make it clear that Julian is referring to the shrine and the relics o f the martyr Babylas.32 Ammianus refers to the same relics in the same way: after fire destroyed the shrine o f Apollo at Daphne in October 362, Julian decided to remove “ the bodies buried nearby” (22.12.8: circumhumata corpora statuit exinde referri). Except for introducing a generalizing plural, Ammianus repeats Julian’s statement that he had removed “ the corpse o f Daphne” : he had read the Misopogoti (22.14.2) and the phrase stuck in his mind.33 Both Christians and pagans derided the places o f worship o f the other group as tombs. The Christian apologists o f the second and third centuries called temples tombs;34 in the fourth century, pagans reversed the taunt and called Christian churches tombs,35 and the emperor Theodosius denounced the conventicles o f the Manichees as “ tombs o f deadly mysteries.” 36 Both polem­ ical meanings o f the word sepukhrum can be found in Ammianus. He reports that, when bishop George looked at the temple o f the Tyche o f Alexandria, he exclaimed “ How long will this tomb stand?” (22.11.7: quam diu sepulchrum hoc stabit?). Ammianus uses the term in propria persona to associate Christianity with cowardice and military incompetence. Sabinianus, who had replaced the historians commanding officer and friend Ursicinus as magister militum per Orientem with overall responsibility for the defense o f the eastern frontier, was (he alleges) lazy and inefficient.37 He “ was leading a life o f dis­ solute luxury among the cemeteries o f Edessa, secure, I suppose, in the belief 32 Julian,

Misopogoti 3 6 ib c . Libanius’ m onody on the burning o f the temple o f A pollo at

D aphne, com posed imm ediately after the fire on 2 2 O cto b er 36 2 , calls the m artyr’s remains “ some corpse”

(Oral. 60.5).

33 Saints are also ridiculed as nekroi in Julian,

Gal., frags. 4 3 . 2 6 , 4 8 . 7 , 6 1 . 2 , 8 1 . 4 —5 Masaracchia

(where the “ fresh corpses” are contrasted w ith “ the ancient corpse,” i.e., Christ), cf. R . Asmus,

Woch. kl. Phil. 2 4 (19 0 7), 1 5 2 . T h e same taunt occurs at the end o fju lia n ’s attack on the “ un­ CR, N .S . 15 (19 6 5), 4 3, convincingly conjectured “philottekron — co rpse-lo vin g" as an epithet applied to “ the life o f w retched w om en” ( Orat. 9(6], 203C2). T h e designation nekros was applied to Jesus b y Celsus: O rigen, Contra Celsum 3 .4 1 , 4 3; 7 .6 8 , cf. P. de Labriolle, La Réaction païenne. Lit polémique antielirétienne du I n au VIe siècle (Paris, educated cynics,” w here D . A . Russell,

19 34 ). 4 1 5 . 34 G . W . Clarke,

The “ Octavius” of Minucius Felix ( Ancient Christian Writers 39 , 19 74), 2 1 0 —

1 1 . T h e Alexandrian epigrammatist Palladas parodied Christian rhetoric circa 400 by equating the pagan gods w ith dead men: Alan Cam eron, J R S 55 (19 6 5), 2 3 - 2 8 . 35 Julian,

Orat. 7, 2 2 8 c ; Ep. 84a, 429dl Misopogoti 344a, 3 5 7 c , 36 1a ; Gal., frags. 7 9 .1 2 , 8 1 , 82 Vit. pltil. 6 .1 1 .8 , 10 (472).

Masaracchia; Eunapius, .V, Q j

7,

1 6 .5 .7

(8 M a y 3 8 1) : consueta feralium m ysteriorum sepiilchra.

37 Sabinianus has usually been taken to be a civilian w h o did not belong to the officer cadre from w h ich

magistri militum w ere usually drawn: so, most explicitly, A . D em andt, Chiron 10

[85]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t i a n

P

o l e m ic

that the dead were in no position to disturb the peace” (18.7.7). This dismis­ sive passage contains two polemical phrases: the Edessena sepulchra among which Sabinianus wasted his time are the martyrs’ churches outside the city that the pilgrim Egeria saw some twenty-five years later,38 and in Ammianus vita remissior is not merely characteristic o f Christians, it is virtually coded lan­ guage for the religion itself.39 In his letter to the people o f Bostra, Julian defined Christians as “ those who have turned away from the gods to corpses and relics.” 40 Following his lead, Ammianus calls Christians “ atheists” and “ criminals,” a usage that some re­ cent students have failed to recognize in the historian. Ammianus describes a reign o f the emperor Julian in which there are no Christian martyrs, in which no Christian is punished for his religion (Chapter V). Yet he does mention one prominent martyr by name: he is Artemius, whose extant passion, though composed centuries later, draws so heavily on the lost ecclesiastical history o f Philostorgius that it is an important historical source for the middle o f the fourth century.41 Artemius had brought to Constantinople the relics o f Tim ­ othy in 356 and o f Andrew and Luke in 357, and after his death he came to be revered as a martyr in the imperial capital.42 In Egypt, however, Artemius was not remembered with respect or affection. He became dux Aegypti in the late 350s and in this capacity harassed the supporters o f Athanasius, occupied die Serapeum, and used troops against the populace o f Alexandria, thus mak­ ing himself detested by both Christians and pagans.43 Artemius was tried and executed at Antioch in October 362.44

Historia 40 ( 19 9 1) , 4 9 4 - 5 0 0 . H ow ever, the crucial phrase is transmitted b y victus quidem senex: although Seyfarth prints cultus (Heraeus), B en tleys unctus is surely

(1980), 6 1 1 ; J. Szidat, V as

preferable— and carries no implications about Sabinianus* status. Fo r all that Am m ianus says, he

East Roman Foreign Policy. Formation and Conductfrom Diocletian to Anastasius (Leeds, 19 9 2), 180 , n. 54, w h o argues that Jero m e, Vita Malchi 1 1 , indicates that Sabinianus was probably dux Mesopotamiae may have had a long (if undistinguished) military career before 35 9 : see R. R C . Blockley,

in the 330s. 3H

Itinerarium Egeriae 19 .4 , cf. R Maraval, Lieux saints et pèlerinages d’Orient (Paris, 19 85),

35 0 -52. 39 C o m p are

solutions vitae mollities, w h ich seems to refer to the conversion o f Constantine

( 3 1 . 5 . 1 4 , cf. C h ap ter X I V ) . 40 Ju lian,

Ep. 1 1 4 , 4 38c.

Die Shriften des Johannes von Damaskos 5 (Patristische Texte und Studien 29, 19 88), 2 0 2 - 4 5 . 42 Passio Artemii 1 6 - 1 8 ; Zonaras 1 3 . 1 1 . 2 8 , cf. Descr. cons. 35 6 , 3 5 7 .1 (where Artem ius is not named): on his cult, see now S. M . C . Lieu, From Constantine to Julian: Pagan and Byzantine Views, ed. S. Lieu and D . M ontserrat (London, 1996), 2 1 3 - 2 3 . 43 Fo r A rtem ius’ career, see, briefly, PLR E 1 .1 0 2 , Artem ius 2; for a full treatment, J. D u m ­ mer, A f P i i (19 71)» 1 2 1 - 4 4 . 44 T h e place is stated as D aphne by Passio Artemii 67: on the date, see A p p en d ix 7. 41 T h e full text is edited by B . Kotter,

[86]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t i a n

P

o l e m ic

Ammianus heartily approved and presented Artemius as a notorious crim­ inal who was accused by the people o f Alexandria and tried and condemned to death “ for a mass o f outrageous crimes” (22.1 i . i ). The tendentious phrase atrocium criminum moles refers to Artemius’ actions as a subordinate loyally car­ rying out the religious policies o f Constantius. Theodoretus supplies the charge that Ammianus omits: Artemius was beheaded and his property con­ fiscated because he had destroyed idols in Egypt under Constantius.45 Ammi­ anus voices sentiments very similar to those Eunapius later expressed: the Egyptian monks at Canopus “ collected the bones and skulls o f criminals who had been put to death for numerous crimes, men whom the law courts o f the city had condemned to punishment.” 46 When he expresses explicit opinions about Christianity and Christian be­ havior, Ammianus assiduously tries to appear to deliver a balanced verdict. That does not entail, as has so often been maintained, that he had a realistic and even-handed attitude.47 Similar apparently balanced verdicts abound in Gibbon, who often employs formal balance to sharpen the rhetorical force o f his denigration o f Christianity.48 Ammianus, too, had mastered the art o f “ grave and deliberate irony.” The best illustration o f his skill in deploying the technique is his discussion o f the secret interview that the bishop o f Bezabde was alleged to have had with the Persian king Shapur shortly before his city was taken by storm: T h e b is h o p in c u r r e d a s u s p ic io n , w h ic h in m y o p in io n w a s ill- fo u n d e d , th o u g h it o b ta in e d w id e c u r r e n c y , th at in a s e c re t m e e t in g h e h a d in fo r m e d S h a p u r w h ic h p a rts o f th e c it y - w a lls h e s h o u ld a tta c k as w e a k o n th e in s id e a n d th u s v u l ­ n e r a b le . T h is s t o r y w a s s u b s e q u e n tly r e n d e r e d p la u s ib le b y th e fa c t th at a fte r h is v is it th e e n e m y s ie g e - e n g in e s b e g a n d e lib e r a te ly to b a tte r th e p la c e s w h ic h w e r e in s e c u r e a n d c r u m b lin g th r o u g h d e c a y , w it h g r e a t s h o u ts o f t r iu m p h as i f th e m e n w h o d ir e c t e d t h e m w e r e a w a re o f c o n d it io n s in s id e th e c ity . ( 2 0 .7 .9 )

What is the overall effect o f this passage? It is a serious misreading o f Ammi­ anus to argue that Ammianus “ defends the bishop o f Bezabde against unjust suspicions” or that he had no intention to “ incriminate him on a charge o f 45 Th eodoretus, H E 3 .1 8 .2 . 46 Eunapius, Vit. p h il. 6 . 1 1 . 8 (472). 47 M atthew s, A m m ian us (19 89 ), 4 50 , claims that “ his criticism was reserved for cases and cir­ cumstances in w h ich it was undeniably justified.” 48 F o r exam ple, the otiose and malicious parenthesis in the follow ing footnote: “ Apollonius o f Tyana was b o m about the same time as Jesus C h rist. His life (that o f the former) is related in so fabulous a m anner by his disciples, that w e are at a loss to discover w h eth er he was a sage, an impostor, or a fanatic” (D ecline and F a ll 1 .3 0 5 n. 70 (B) = 1. 3 1 5 n. 63 [W ]).

[87]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

treachery.” 49 Writing a generation later in a different part o f the world, the historian could simply have omitted a rumor that he disbelieved or consid­ ered undeserving o f mention. Ammianus chose to include the story because he had learned from Tacitus how to use rumors to suggest disreputable con­ duct or a dishonest motive while taking no authorial responsibility for the du­ bious information thus conveyed.50 The passage is a “ progressive insinuation,” where Ammianus lets the inherent plausibility o f the rumor and its confirma­ tion by the subsequent course o f events outweigh his initial (and insincere) disclaimer.51 Ammianus could neither know nor report what the bishop ac­ tually said to Shapur; yet he states very clearly that after the interview the Per­ sian king attacked the defenses o f Bezabde where they were weak and could be breached. In Ammianus’ opinion, Christians were intrinsically unpatriotic. The key text that best reveals Ammianus’ bias is his analysis o f the ecclesi­ astical policies o f Constantius in the obituary o f the emperor at the end o f Book X X I: C o n f u s in g th e p la in a n d s im p le C h r is t ia n r e lig io n w it h th e s u p e rs titio n o f an o ld w o m a n , a n d m o r e in te re s te d in c o m p lic a t e d e n q u ir ie s th a n in s e ttlin g d isp u te s w is e ly , h e a r o u s e d v e r y m a n y q u a rre ls w h ic h h e fe d w it h v e rb a l a r g u m e n t as th e y sp re a d w id e ly , so th at, w it h h o rd e s o f b is h o p s s c u r r y in g h it h e r a n d th ith e r o n p u b ­ lic m o u n ts to a tte n d w h a t th e y ca ll sy n o d s, h is a tte m p ts to fo r c e e v e r y r ite to c o n ­ fo r m to h is w h im o n ly s u c c e e d e d in h a m s tr in g in g th e p o s t s e r v ic e . ( 2 1 . 1 6 . 1 8 )

This passage used to be paraded as proof o f Ammianus’ perceptiveness.52 In fact, it is a tissue o f absurdities, but absurdities that echo Christian complaints about Constantius’ interference in ecclesiastical affairs. After Constantius was dead, it was convenient for Christians to blame him for the constant quarrels o f bishops and the frequent synods o f his reign. For blaming Constantius dis­ guised the unpalatable fact that between 335 and 351 no fewer that ten coun­ cils o f eastern bishops had either condemned Athanasius, Marcellus o f Ancyra, and other worthy bishops or implicitly impugned the creed o f the “ ecumeni­ cal” Council o f Nicaea in 325 or both.53 The theological controversy that dominated these years had begun while Licinius still ruled the East, and the 49 So, respectively, M atthew s, A ncient Writers 2 (19 8 2 ), 1 1 1 8 ; H unt, C Q , N .S . 35 (19 8 5), 196. 50 O n Tacitus’ use o f rum or and innuendo, see esp. I. S. R yberg, T A P A 7 3 (19 4 2 ), 3 8 3 - 4 0 4 ; R. Sym e, Tacitus (O xford, 19 58 ), 3 1 4 - 1 6 . 51 Sabbah, M éthode (19 78 ), 4 14 . 52 M atthews, A ncient Writers 2 (19 8 2 ), 1 1 1 8 ; H unt, C Q , N .S . 35 (19 8 5 ), 186. In his book, M atthew s is m arkedly m ore cautious and distances him self from Am m ianus’ verdict (Am m ianus [19 8 9 ], 2 6 3, 4 38 , 449). 53 Fo r the details, see Athanasius (19 9 3), 1 9 - 1 2 0 .

[88]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t i a n

P

o l e m ic

dispute over whether Athanasius was the rightful bishop o f Alexandria had begun with his election in 328.54 How could Constantius have started con­ troversies that began when he was an infant and had raged for nearly twenty years before he was able to intervene in any way? So far from attempting to impose his own wishes, as Ammianus alleges, Constantius refrained from intervening personally in the theological debates for more than twenty years. Admittedly, after 353, he attempted to secure western acceptance o f the decisions o f the eastern Council o f Sirmium o f 3 51. But he played no part in formulating the doctrines set out in the synodical let­ ter o f that council. It was only in frustration that he intervened directly in 359 by convening the separate eastern and western Councils o f Seleucia and Ariminum to ratify a creed drawn up in his presence, on the twenty-second anniversary o f the death o f his father. And the result o f these councils and the negotiations that followed them was not doctrinal complexity, but extreme doctrinal simplification: the homoean creed laid down merely that God the Son is like the Father, but carefully omitted to specify in what sense or senses Father and Son are alike.55 Moreover, Valerio Neri has now shown that Am ­ mianus’ complaint that Constantius could not distinguish between an absoluta et simplex religio and an anilis superstitio and his phrase scrutando perplexius ex­ press a verdict which any catholic Christian o f the reign o f Theodosius could wholeheartedly accept.56 The satirical quality o f Ammianus’ assessment o f Constantius comes out clearly in his allegation that throngs o f bishops scurrying to and fro hamstrung the cursus publicus (21.16.18). Most modern discussions have failed to ask two obvious questions that are relevant to evaluating this remark: how many people used the cursus publicus regularly? and how many bishops used it on spe­ cial occasions? The system was enormous and available for anyone on official business:57 it must therefore have been used every year by several thousand governors, officials, soldiers, agentes in rebus, and private citizens with influ­ ence. But most councils o f bishops were small gatherings o f the bishops o f a single province or o f a group o f neighboring provinces. Even in the reign o f Constantius, most councils were small affairs with predominantly local atten­ dance. Despite the high numbers (well into the hundreds) bandied about by ec­ clesiastical historians, ancient and modern, the Councils o f Arles in 3 5 3 -3 5 4 and Milan in 355 were both attended by no more than about thirty bishops.58 54 Constantine ( 19 8 1), 2 0 1 - 7 , 2 3 0 - 4 2 ; Athanasius (19 9 3), 1 4 - 2 5 , 55 Athanasius (19 9 3), 1 3 6 - 5 1 . 56 V. N eri, Cognitio Gestorum (19 9 2), 5 9 - 6 5 . 57 Jones, L R E 8 3 0 - 3 4 . 58 Athanasius (i993)>

1 1 7 , 2 7 5 - 7 7 , n. 4 5 - 4 7 .

[89]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t i a n

P

o l e m i c

Between 337 and 361 there were only two dates at which more than one hundred bishops gathered together in a single city.59 The first was in 343, when ninety western and seventy six eastern bishops assembled in Serdica, al­ though they never all met together as a single council.60 It is significant that the eastern bishops voiced exactly the same complaint that Ammianus re­ peats— that the cursus publicus was “ being worn out and reduced to noth­ ing.” 61 The second occasion was in 359, when four hundred western bishops were compelled to spend several months in Ariminum under the watchful eye o f the praetorian prefect Taurus,62 and one hundred and sixty eastern bishops met at Seleucia in Cilicia under the supervision o f the comites Bassidius Lauricius and Leonas.63 During the summer o f 359 Ammianus traveled post-haste with Ursicinus from Samosata across Asia Minor and into Europe, then in the reverse direction from the River Hebrus back to Mesopotamia (18.4.7, 6.5, 8). A suspicion thus inevitably arises that his complaint about “ crowds o f bish­ ops” reflects the personal irritation o f a man in a hurry who arrived at a man­ sio only to find that bishops had taken all the fresh horses that his party needed to continue their urgent journey.64 I f that suspicion is well founded, then Am ­ mianus’ remark, which has impressed so many for its perceptiveness, repre­ sents a misleading generalization o f a single occurrence, exaggeration based on personal inconvenience— the procedure o f a satirist like Juvenal rather than a historian.65 Christian language, prejudicial vocabulary, explicit polemic, and satire do not exhaust the variety o f material relevant to Ammianus’ attitude toward Chris­ tianity. There is also covert polemic against Christianity, and it is a pervasive feature o f the Res Gestae. The detection o f hidden polemic involves an obvi­ ous danger: the eager exegete is likely to find what he seeks whether it is re­ ally there or not. That Ammianus makes such covert attacks on Christianity can be established only by providing cogent examples. An acute enquiry has discovered covert anti-Christian polemic in a passage where Ammianus de59 Athanasius (19 9 3), 7 1 - 8 1 . (>0 T h e precise numbers on each side are deduced from Athanasius, H ist. A r. 1 5 .3 ; Socrates, H E 2 .2 0 .5 . Lists survive, albeit incom plete, o f the bishops w h o subscribed the eastern and west­ ern synodical letters: C S E L 6 5 . 7 4 - 7 8 (73 eastern bishops); C S E L 6 5 . 1 3 2 - 3 9 ; E O M I A 1 . 5 4 5 59 (61 and 59 western bishops respectively), cf. Athanasius (19 9 3), 260 , n. 1 0 - 1 2 . 61 C S E L 6 5 . 6 4 . 2 4 - 2 5 : cursusque ipse publicus attritus ad nihilum deducitur. 62 Sulpicius Severus, Citron. 2 . 4 1 - 4 4 . 63 Fo r the num ber, see Athanasius, S y n . 1 2 . 1 ; Socrates, H E 2 .3 9 .5 ; on the role played by Lauricius and Leonas, Athanasius (19 9 3), 1 4 6 - 4 8 . 64 R eading the Past (19 9 0 ), 80. 65 O n the personal elements in the poet, see still G . H ighet, Ju v e n a l the Satirist. A S tu d y (O x ­ ford, 19 5 4 ), exaggerated in detail, but instinctively correct in principle.

[90]

C

h r i s t i a n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

scribes the state o f the eastern armies before Julian reformed them in the win­ ter o f 3 6 1—3Ö2.66 The soldiers o f Constantius were soft, corrupt, and addicted to luxury: they slept on soft beds, their drinking goblets weighed more than their swords, and they recited effeminate ditties instead o f martial hurrahs (22.4.6: cum miles cantilenas meditaretur pro iubilo molliores etc.).67 Am ­ mianus alludes (it seems) to the recitation o f prayers by soldiers, a practice instituted by Constantine.68 There is something analogous in Libanius and Zosimus: the latter alleges that the three hundred and sixty men whom Con­ stantius gave Julian when he sent him to Gaul (that is, part o f the scholae palatinae as his bodyguard) knew only how to pray, the former that when Con­ stantius removed the best o f Julian’s bodyguard in 360, he left him soldiers capable o f nothing else than praying.69 All three authors thus equate Christian prayer with military incompetence. The end o f Book X IX repays close examination as an exercise in disguised anti-Christian polemic. What editors print as the last two chapters com­ prises one substantial episode narrated at length (19 .12 .1- 18 ) and two quasiannalistic notices (19 .12 .19 -2 0 , 19.13). The placing o f these three items is significant: Book X IX concludes a hexad, and these notices immediately pre­ cede the proclamation o f Julian as Augustus. Paul “ the Chain,” whom Ammianus had introduced once before (14.5.6), is introduced again as tartareus ille notarius, an expert in the arts o f bloodshed who did not shrink from fraud to ruin the innocent. An opportunity for vil­ lainy arose from a trivial incident. The remote oracle o f the god Besa in U p­ per Egypt was normally consulted by submitting questions in writing, which were retained in the shrine after the god had answered them. Someone sent some o f these papers to Constantius. In a rage, the emperor sent Paul to search out those who had submitted the offensive questions, arrest them, and bring them to trial before Modestus, the conies Orientis (who normally resided in Antioch). Paul carried out his assignment, and the trials were held at Scythopolis in Palestine. Ammianus explains that Scythopolis was chosen because it was both remote and mid-way between Antioch and Alexandria, from where many o f the accused came (19 .12 .1—8).70 By secretior, it has been suggested, 66 D . W oods, J T S , N .S . 49 (19 9 8), 1 4 5 - 4 8 . 67 D espite T L L 3.2 8 5 —86, cantilenae has a clearly derogatory sense here, whereas iubilunt o c­ curs predom inantly in Christian texts ( T L L 7 . 2 . 5 8 8 - 8 9 ) . 68 Eusebius, I^C 4 . 1 9 - 2 0 . 6y Zosim us 3 .3 .2 (quoting a lost letter o f Julian him self); Libanius, O r at. 18 .9 4 , cf. A m m i­ anus 2 0 .4 .3. 70

O n the trials, w h ich Am m ianus describes at some length ( 1 2 . 9 - 1 8 ) , and their political

background, see J. Straub, Heidnische Geschichtsapologetik in der christlichen Spätantike. Unter­ suchungen über Z e it und Tendenz der Historia A ugusta (Bonn, 19 6 3), 6 3 - 7 4 ; H . Funke, J A C 10 (19 6 7 ), 1 5 1 - 6 5 ; R . von H aehling, J A C 2 1 (19 7 8 ), 7 4 - 1 0 1 .

[91]

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

Ammianus “ presumably means that it was distant from centres o f pagan influence.” 71 More relevant in the present context is the demonstrable fact that Ammianus has left out the principal reason for Paul’s journey to Alexan­ dria. Two local sources record that Paul arrived in the city on 23 June 359, published an imperial edict in favor o f George, whom the Council o f Sir­ mium in 351 had appointed bishop o f Alexandria in place o f Athanasius, and used force to cow Athanasius’ supporters.72 The first o f the two brief notices that follow resembles an annalistic notice from Livy in both content and language (19 .12 .19 —20). At Daphne “ that pleasant and splendid suburb o f Antioch” was bom a child with two mouths, two sets o f teeth, a beard, four eyes and two very small ears. Ammianus inter­ prets the birth o f such a deformed child as a warning that the state was being deformed: such portents, he explains, often indicate the future, but “ since they are not expiated publicly as in the days o f the ancients, they pass without anyone hearing or knowing o f them.” 73 This lament comes from the heart: Ammianus believed in the validity o f all the traditional forms o f divination (Chapters V, XIV ). He then proceeds, in a passage unfortunately omitted in the Penguin translation, to record a renewal o f raiding by the Isaurians (19 .13.1). As a result, he continues: V

T o q u ie t e n th e m b y fo r c e o r p e rs u a s io n L a u r ic iu s w a s s e n t as g o v e r n o r w it h th e a d d e d r a n k o f comes, a m a n w it h u n d e r s t a n d in g o f c iv il a ffa irs, w h o c o r r e c t e d m a n y th in g s w it h th reats r a th e r th a n s e v e rity , so e ffe c t iv e ly th at t h o u g h h e h e ld th e p r o v in c e fo r a lo n g tim e n o t h in g h a p p e n e d w h ic h m ig h t b e d e e m e d w o r t h y o f n o t ic e .74 ( 1 9 . 1 3 . 2 )

This, the very last sentence o f Book X IX , is oddly over-emphatic. Together with the comes Leonas, Bassidius Lauricius presided over the Council o f Seleucia in late September 3 59:75 he was a praeses oflsauria who probably received the rank o f comes precisely for the purpose o f supervising the council,76 since 71 J. Binns, Ascetics ami Ambassadors of Christ. The Monasteries of Palestine 3 14 - 6 3 1 (O xford, 1994)» 1 3 4 -

72 Historia acephala 2 .5

M artin;

Festal Index 29.

73 T h e extant books o f L iv y have no double birth o f this sort, but they record three births o f an infant o f indeterminate sex as prodigies requiring expiation ( 2 7 .3 .5 , 2 7 . 1 1 . 4 , 3 1 .1 2 .6 ) .

74 I accept Seyfarths interpretation o f animadversione dignnm as “ bem erkensw ert” : in the Budé edition, G . Sabbah translates “ qui justifiât une répression.” 75 Epiphanius,

Panarion 7 3 .2 5 ; Socrates, H E 2 . 3 9 - 4 0 ; Sozom enus, H E 4 .2 2 .

76 A n inscription records that Bassidius Lauricius restored a fort w h ich had lo n g been o ccu ­

latrones, garrisoned it, and named it Antioch ia, stating his title and rank as v. c., com(es) et praeses (ILS 740). T h e com bination o f comes and praeses is anomalous: R . Scharf, E A 16 (19 9 ),

pied b y

147- 51*

[92]

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r is t ia n

P

o l e m ic

it seems to follow from Ammianus that he was appointed governor in 359 when the council was about to take place. The assertion that nothing worthy o f note happened in his province while he was governor can hardly be an in­ nocent misapprehension. Even Ammianus’ geographical excursus are sometimes subtly polemical. The excursus on the shores o f Thrace and the Black Sea contains an implicit de­ nial o f the importance o f Constantine’s new Christian city on the Bosporus: “ the left bank (sc. o f the Propontis) is overlooked by the port o f Athyras, Selymbria, Constantinople, the ancient Byzantium, an Athenian colony, and the promontory o f Ceras” (22.8.8). Fourth-century Constantinople was much more than “ the ancient Byzantium” with which Ammianus here equates it.77 In reality, it more than matched contemporary Antioch, which he praised as “ a city known to the world, with which no other could compete in abound­ ing with both imported and home-grown supplies” (14.7.8). Moreover, Am­ mianus’ praise o f Antioch and other cities in the Syrian region whe;n he de­ scribes the provinces o f Oriens in Book X IV culminates in a most remarkable and effective covert insult to Christianity: T h e last o f th e S y r ia s is P a le s tin e , e x t e n d in g o v e r a g re a t e x te n t o f t e r r it o r y a n d a b o u n d in g in c u ltiv a te d a n d w e ll- k e p t lan d s; it also has s o m e sp le n d id c itie s , n o n e o f w h ic h y ie ld s to a n y o f th e o th e rs , b u t t h e y riv a l o n e a n o th e r , as it w e r e , b y p lu m b - lin e . T h e s e a re C a e s a r e a , w h ic h H e r o d e s b u ilt in h o n o r o f th e e m ­ p e r o r O c ta v ia n u s , E le u th e r o p o lis , a n d N e a p o lis , a lo n g w it h A s c a lo n an d G a z a , b u ilt in a fo r m e r a g e . In th ese d is tric ts , n o n a v ig a b le r iv e r is a n y w h e r e to b e se e n , b u t in n u m e ro u s p la c e s n a tu ra l w a r m sp rin g s g u sh fo r th , a d a p te d to m a n y m e ­ d ic in a l u se s. B u t th e se re g io n s also m e t w it h a lik e fa te , b e in g fo r m e d in to a p r o v ­ in c e b y P o m p e y , a fte r h e h a d d e fe a te d th e J e w s an d ta k e n J e r u s a le m , an d le ft to th e ju r is d ic t io n o f a g o v e r n o r . ( 1 4 . 8 . 1 1 - 1 2 )

Ammianus’ message could hardly be clearer.78 The significant cities o f fourth century Palestine are Caesarea, Eleutheropolis, Neapolis, Ascalon, and Gaza. He excludes from the list Jerusalem, which Constantine had turned into a rich, monumental city crowded with resident ascetics and visiting pilgrims,79 77 G . Dagron, Naissance d 'u n e capitale. Constantinople et ses institutions de 3 3 0 à 4 5 1 (Paris, 1974), 1 3 - 7 4 , 11 9 - 4 6 , 2 1 3 - 6 0 , 2 9 7 -3 14 , 3 6 7 -4 5 378 G. Stemberger, Ju d e n und Christen im H eiligen L an d. Palästina unter Konstantin und Theodo­ sius (M unich, 1987), 1 5 6 - 5 7 . 79 E. D . Hunt, H o ly L a n d Pilgrimage in the Later Roman Em pire, a .d . 3 1 2 - 4 1 0 (Oxford, 1982), 6 —27, 1 2 8 —79; P. M . L. Walker, H o ly C ity, H o ly Places? Christian Attitudes to Jerusalem and the H o ly L a n d in the Fourth C entury (Oxford, 1990), 19 9 -3 0 8 .

[93]

C

h r i s t ia n

L

a n g u a g e

a n d

A

n t i

- C

h r i s t ia n

P

o l e m ic

and which was already, when Ammianus wrote, challenging Caesarea for ec­ clesiastical primacy in the Roman province o f Palestine.80 Ammianus deliber­ ately closed his eyes to the importance o f contemporary Jerusalem, just as he closed his eyes to the central role that Christianity played in the politics, so­ ciety, and culture o f the Roman Empire after Constantine. 80

T h e political conflict between the sees ofjerusalem and Caesarea w ent back to circa 3 2 5 ,

w h en Macarius, the bishop o f Jerusalem, discovered w h at was believed to be the w o o d o f the cross on w hich Jesus was crucified: Z . Rubin, Jerusalem Cathedra 2 (19 8 2 ), 7 9 - 1 0 5 .

[94]

[IX] THINGS SEEN AND THINGS READ

Ammianus’ geographical and ethnographical excursus combine three very dif­ ferent types o f material.1 The framework usually comes from either official lists o f Roman provinces or lists o f peoples and places outside Roman territory that he may have found in Ptolemy’s Geography.12 He added historical notices from Festus’ Breviarium, drew on Solinus and Greek topographical works for particular sections, and used Sallust and Livy where they provided relevant material.3 Ammianus also sometimes drew on his own recollection o f what he had seen or heard. Hence the description visa vel lecta that he applies to the content o f his account o f the coasts o f Thrace, the Hellespont, and the Black Sea (22.8.1). The balance between visa and lecta varies from one excursus to another. The lost discussion o f the situs Africae (cf. 29.5.18) is likely to have been bookish and Sallustian, for Ammianus had read the Bellum Jugurthinum, but probably never set foot in Roman Africa. And the excursus on the Huns apppears to be almost entirely a tissue o f ethnographical commonplaces that owes nothing to 1 O n the quality o f these excursus and Ammianus* geographical know ledge, see n ow G . A . Sundwall,

A JP 1 1 7 (1996), 6 1 9 - 6 4 3 .

2 Am m ianus names Ptolem y with Eratosthenes, Hecataeus “ aliique huiusmodi cognitionum minutissimi sciscitatores” as asserting that the shape o f the coast o f the Black Sea is like a Scythian bow, w h ich is certainly untrue o f Ptolem y (2 2 .8 .10 ). Fo r Ammianus* use o f Ptolem y for oriental geography, see M . Schuffner, Ammianus Marcellinus in rerumgestarum libris quae de sedibus ac moribus complurium gentium scripserit (Prog. M einingen, 18 7 7 ), 1 4 - 1 7 ; T. M om m sen, Gesammelte Schrif­ ten 7 (Berlin, 1909), 4 0 2 - 9 ; L. Dillem an, Syria 38 (19 6 1), 1 3 5 - 4 2 . It is denied b y M . F. A . B ro k, Mnemosyne4 28 (19 7 5 ), 4 7 - 5 6 . 3 M o m m sen , Ges. Sehr. 7 (1909), 39 3 - 4 2 5 . Fo r an attempt to distinguish betw een w hat A m ­ mianus took from w ritten sources and w hat he added to them, A . M alotet, De Ammiani Marcellini digressionibus quae ad externas gentes pertineant (Diss. Paris, 1898), 1 - 4 6 .

[95]

T

h in g s

S

een

a n d

T

h in g s

R

e a d

close observation and hence possesses little historical value.45The excursus on the Black Sea, which may have been intended to challenge comparison with Sallust’s famous digression on the situs Ponti? is a mosaic put together from very diverse sources, mainly and perhaps exclusively Greek.6 The intellectual level is closer to that o f a periplus (such as the Periplus Maris Erythraei) than to serious scientific geography, and the Dutch commentators rightly doubt whether anything in it draws on personal observation, except when Ammianus writes about the Bosporus and its surrounds, which he had visited. Ammianus’ account o f the provinces o f Oriens, by contrast, though repeating phrases from Festus’ Breviarium o f 3 70,7 reveals significant facts about how this east­ ern believer in the traditional gods viewed the post-Constantinian Roman Empire in which he lived (Chapter VIII). The balance between what Ammianus had seen and what he had read is per­ haps most evenly held in his excursus on Gaul. The successive investigations o f Victor Gardthausen in 1873, Theodor Mommsen in 1881, and Walter Sontheimer in 1926, together with the evaluation o f their findings by R de Jonge in his commentary on Book XV, have identified with some precision the var­ ied elements that this excursus amalgamates.8 Nevertheless, it will be worth reviewing their conclusions and asking again what Ammianus has added from his own experience. Ammianus marks the importance o f his excursus on Gaul by quoting a fa­ mous passage o f “ the bard o f Mantua” where Virgil marks the transition from the “ Odyssey” o f the Trojans in the first six books to the “ Iliad” o f the second half o f the Aetieid, and he quotes more than a line with the words partly trans­ posed in order to avoid the full metrical pattern o f hexameter verse in prose.9 Ammianus names Timagenes as his principal authority for the origin o f the 4 C . K in g, A J A H 1 2 (19 8 7 , publ. 19 9 5), 7 7 - 9 5 (with reference to earlier, m ostly m uch more favorable analyses). In contrast, w h at Am m ianus says in his narrative about the Sarmatians and Q uadi seems to be based on accurate and reliable observation: U .-B . D ittrich, D ie Beziehungen Roms z u den Sarmaten und Q uaden ini vierten Jahrhundert n. Chr. (nach der Darstellung des A m m ianus Marcellinus) (Diss. Bonn, 1984). 5 Rufius Festus, O ra maritima, 3 6 - 3 7 : inclytam descriptionem. B . M aurenbrecher, C . S a l­ lustii C risp i Historiarum reliquiae (Leipzig, 18 9 1), 1 3 4 - 3 8 , reconstructed its outline from Pom po­ nius M ela, Chorographia 1 . 9 9 - 2 . 1 5 ; Pliny, N at. H ist. 4 . 7 5 - 9 1 , 6 . 1 —2 2 ; Valerius Flaccus, A rg ­ onautica 4 . 7 1 4 - 3 2 ; Am m ianus 2 2 . 8 .4 - 4 8 . T h e D utch commentators acidly note that “ only the odd phrase in Am m ianus’ resembles the validly attested fragments.** 6 I. Gualandri, Parola del Passato 23 (19 68), 2 1 1 . 7 M om m sen, G es. Sehr. 7 (1909), 3 9 6 - 4 0 0 . 8 V. Gardthausen, Jah rb. f ü r class. P h il., Supp. 6 (18 7 3 ) , 5 0 9 - 5 6 ; T . M om m sen, Hermes 16 ( 1 8 8 1) , 6 0 2 - 3 6 = G es. Sehr., (1909), 3 9 3 - 4 2 5 ; W . Sontheim er, K lio 20 (19 26 ), 1 9 - 5 3 . 0 1 5 .9 .1: ut Mantuanus vates praedixit excelsus, opus m oveo m aius/ m aiorque mihi rerum nascitur ordo (quoting A en . 7 . 4 4 - 4 5 : maior rerum mihi nascitur o rd o ,/ maius opus moveo).

[96]

T

h in g s

S

e e n

a n d

T

h in g s

R

e a d

Gauls: he commends him as a Greek in both scholarship and language, who collected information from a wide range o f earlier books and gave accurate in­ formation about matters on which his predecessors possessed only inaccurate opinions (15.9.2). Ammianus promises to relate the same facts as Timagenes “ clearly and plainly,” and both the structure and much o f the content o f the excursus reflect the fact that his main authority was writing in the reign o f Augustus.101 Ammianus took from Timagenes both the ethnography o f Gaul (15.9) and much o f what he says about the Alps (15.10). The provenance o f this material is clear,11 and Ammianus was fully aware that Gallic society had changed over the course o f four centuries: he carefully uses the aorist tense when he talks about the three categories o f men who represented learning and culture among the ancient Gauls (15.9.8). Comparison with the historian Diodorus Siculus and the geographer Strabo makes it clear that both they and Timagenes here reproduced the classification and in part the language o f Posidonius.12 All three Augustan writers named the poets and composers o f sacred hymns among the Gauls as bardi, the philosophic druids, and a third category, for whose Celtic name Ammianus may provide the best guidance, corrupt though the trans­ mitted text obviously is: “ euhagis vero scrutantes fservianif et sublimia nat­ urae pandere conantur.” 13 Diodorus describes these men as seers who prac­ tice divination by watching birds and through sacrifices, whereas Strabo states that they perform sacred duties and investigate nature and claims to give their Celtic name.14 The form in which the name was stated by Timagenes has been a matter o f erudite discussion, and the truth can only be divined by conjecture: nevertheless, it seems now to be agreed by competent judges that Ammianus repeats the Greek word euageis from Timagenes, who attempted to reproduce the Celtic term, whereas the ouateis offered by the manuscripts o f Strabo sounds uncomfortably reminiscent o f the Latin word vates (bard, poet).15

De Ira 2 . 2 3 . 4 - 8 ; Suda T 588 (4.54 9 Adler) = FG rH 88 T 2, 1. FGrH 88 F 1, 14 , 15 . 12 D iodorus 5 . 2 5 —3 2 = FG rH 87 F J 1 6 = Posidonius, frag. 16 9 Theiler, cf. J. M alitz, Die Historien des Poseidonios (Zetemata 79 , 19 8 3), 19 6, n. 2 1 7 . 13 I print the readings o f V : Seyfarth accepts the emendation euagcs for eultagis and registers a plethora o f m odern conjectures for scrviani, w h ich he rightly obelizes. 10 Seneca,

11 F. Ja c o b y prints Am m ianus 1 5 . 9 . 2 - 8 , 10 .9 , 1 2 . 1 - 4 as

14 D iodorus 5 . 3 1 . 2 - 3 ; Strabo 4 .4 .4 ( 1 9 7 C ) . 15 W . A ly, Strabonis Geographica, 4: Strabon von Amaseia. Untersuchungen über Text, Aufbau und Quellen der Geographie (Bonn, 19 5 7 ), 4 5 6 - 5 9 , cf. E . Bickel, Rh. Mus., N .F . 87 (19 3 8 ), 1 9 5 - 2 0 1 , 2 1 7 —20; Festschriftfür August Oxc (Darmstadt, 19 38 ), 1 6 4 - 6 6 . A ly argues that euagis accurately reproduces w h at both Posidonius and Tim agenes w rote and that Strabo found the w o rd already deform ed in his w ritten source: given the state o f his text (the Vatican palimpsest is not available for B o o k IV ), it m ight be preferable to emend the transmitted

[97]

ouateis to ouageis in Strabo.

T

h in g s

S

een

a n d

T

h in g s

R

e a d

Also from Timagenes must derive the surprising prominence given to King Cottius when Ammianus writes about the Alpes Cottiae (15 .10 .2 -8 ). To­ gether with the inscription on the triumphal arch that Cottius erected at Segusio,16 Ammianus provides the greater part o f what is known about him.17 Cottius was hardly o f much interest to anyone in the fourth century, but Am­ mianus had seen his tomb close to the city walls o f Segusio: he notes that Cot­ tius was received into friendship by Octavian, built roads to aid Alpine travel­ ers, governed justly, bequeathed permanent peace to his tribe, and still received cultic devotions in 355 (15.10.2, 7). What other literary sources did Ammianus use when writing about Gaul? He had read Sallust, he uses a Sallustian phrase when he mentions the people o f Saguntum,18 and he names Sallust at the end o f the excursus (15 .12 .5 -6 ). But there is no good reason to imagine that he derived any o f the material in the excursus from Sallust.19 Sallust’s Histories did not contain a digression on Gaul, for which his narrative o f the 70s b . c . provided no appropriate opportu­ nity. The passage that Ammianus quotes is known to come from Sallust’s pref­ ace, which observed that, within ten years, Caesar had conquered the whole o f Gaul except what was too boggy to be accessible.20 Mommsen derived part o f what Ammianus says about tribes, rivers, and the original provinces into which the Romans divided Gaul from Caesar’s Gallic War ( 1 5 . 1 1 . 1 —5).21 That hypothesis is unnecessary. For Ammianus can have taken what Caesarian material there may be in the excursus (if indeed there is any) from a later author, who himself Used Caesar, conceivably the lost por­ tion o f Livy.22 Sontheimer compared what Ammianus says about Hannibal’s route over the Alps (15 .10 .11) with Polybius and Livy and argued that the three historians all depend, directly or indirectly, on two early and somewhat discrepant accounts: Polybius (he held) used one o f these versions and Ammi­ anus the other, while Livy conflated the two.23 That could be true: if so, the two early or primary accounts were presumably those o f the Roman Fabius 16 C I L 5 . 7 2 3 1 =

ILS 94. Dessau reprints the text o f M onunsen, w h o verified readings from

ground level w ith the aid o f a telescope: for the first edition based on close inspection, see E . Ferrera,

L ’arc d'Auguste à Susa (Turin, 19 0 1), 30, w ith Planches X I V —X V I I .

17 P I R 2 J 2 7 4 . 18 1 5 .1 0 .1 0 : Saguntinis mem orabilibus aerumnis et fide. Am m ianus is co p yin g Sallust,

Hist.

2 .6 4 M : Saguntini fide atque aerumnis incliti prae mortalibus. 19 M o m m sen ,

Ges. Sehr. 7 (19 0 9 ), 39 4 , 400.

20 1 5 .1 2 .6 : om nes Gallias, nisi qua paludibus inviae fuere. T h e m odel is Sallust, om ni Gallia . . . , nisi qua paludibus invia fuit, perdomita. 21 M o m m sen , -Ges. Sehr. (19 0 9 ), 4 1 1 , n. 1. 22 Gardthausen, Ja h rb . fü r class. P h il., Supp. 6 ( 1 8 7 3 ) , 550 . 23 Sontheim er,

Klio 20 (19 2 6 ), 2 8 - 5 3 .

[98]

Hist. 1 .1 iM :

T

h in g s

S

een

a n d

T

h in g s

R

e a d

Pictor and Silenus.24 But what was Ammianus’ immediate source? Sontheimer identified it as “ an ancient annalistic account” and named its author as Claudius Quadrigarius.25 That is flatly impossible: although Livy used Quadrigarius ex­ tensively, no one shows direct acquaintance with his text after Aulus Gellius, who is the source o f most o f the apparent quotations o f Quadrigarius in Late Antiquity.26 Gardthausen and Mommsen more plausibly identifed Livy as the source o f Ammianus’ information about Hannibal.27 Ammianus also employed nonliterary written sources. He had before him ei­ ther a list o f the provinces and cities o f Gaul very similar to the extant, though later, Notitiae Galliamm or, more probably, a map or maps o f Gaul from the middle o f the fourth century: as de Jonge observed, he will presumably have used maps for his own journeys in the 350s. For the provincial divisions are correct for 355, the date where Ammianus places the excursus, not for the time o f writing.28 Moreover, Ammianus notes that distances in Gaul are mea­ sured not in miles, but in leugae (15 .11.17 ) , which were about one and a half Roman miles (2.2 kilometers), a practice that began under Septimius Severus.29 Whatever his written sources were or may have been, Ammianus draws on his own memories o f his journey over the Alps in 3 55.30 He emphasizes that he reports what inhabitants o f the region tell all travelers and has seen con­ firmatory inscriptions (15.9.6).31 He has also imposed form and order on the varied material that he incorporates. I f he has reproduced or used a written list o f Gallic cities, he has made a personal selection o f which cities to record, and he adds comments o f his own. He notes the ancient grandeur o f the walls o f Autun (15 . i i .1 1) and that Aventicum is deserted, its buildings half-ruined ( 15 .i i . 12). When Ammianus writes about the passes over the Alps, he writes as one who has himself stood at the top o f a mountain pass and admired the view he describes: “ From the top o f this Italian slope a plateau extends for seven miles as far as the post named after Mars; from there another loftier height, difficult to surmount, reaches to the peak o f Matrona, so called from an accident to a 24 P. G. Walsh, T. Livi ab urbe condita liber X X I (London, 1973). 38. For Fabius and Silenus, FG rH 809; 1 7 5 . Klio 20 (19 26 ), 49, 53. Historicorum Romanorum Reliquiae 1 2 (Leipzig, 1904)1 cccii—ccciii; L. H o lfo rd Strevens, Aulus Gellius (London, 19 8 8 ), 1 7 9 —80, 1 8 4 —85. 27 Gardthausenjahrb. für class. Phil., Supp. 6 (1873), 5 5 3 - 5 6 ; Mommsen, Ges. Sehr. 7 (1909), 25 Sontheim er,

26 H . Peter,

4 1 1 , n. I.

New Empire (19 8 2 ), 2 1 7 - 1 8 . Epigraphica 3 1 (19 6 9 ), 8 4 - 1 0 3 . 30 G . M . W o lo ch , Arctos 26 (19 9 2 ), 1 3 7 - 4 0 . 31 G . M . W o lo ch , Arctos 2 7 (1993), I49_ 53. attributes Am m ianus’ apparent mistakes m ainly 28

29 G . Walser,

to a fallible m em o ry and misleading maps.

[99]

T

h in g s

S

een

a n d

T

h in g s

R

e a d

noble lady. After that a steep but easier route lies open to the fort o f Brigantia” (15.10.6). The references to the statio named after Mars and the fort Brigantia subtly indicate the context in which Ammianus crossed the Alps: he had trav­ eled from Milan to Cologne when he helped to assassinate Silvanus in 355 (15 .5 .17 —31). A more personal and more revealing recollection surfaces in the passage where Ammianus describes the character o f the Gauls: A lm o s t all G a u ls a re o f im p o s in g sta tu re , fa ir - s k in n e d , w it h re d d is h h a ir, fe a rs o m e b e c a u s e o f th e ir sava ge e y es, e a g e r to q u a r r e l a n d e x c e s s iv e ly in s o le n t. F o r , n o t e v e n a w h o le g r o u p o f fo r e ig n e r s c a n sta n d u p to a s in g le o n e o f th e m if, in th e c o u r s e o f a d is p u te , h e s u m m o n s h is g r e y - e y e d w ife , w h o is m u c h s tr o n g e r th an h e is, e s p e c ia lly w h e n , w it h s w o lle n n e c k a n d g n a s h in g te e th , sh e s w in g s h e r e n o r ­ m o u s s n o w - w h it e a rm s a n d b e g in s to d e liv e r a r a in o f p u n c h e s m ix e d w it h k ic k s, lik e m issiles la u n c h e d b y th e t w is te d s trin g s o f a c a ta p u lt. ( 1 5 . 1 2 . 1 )

Ammianus had doubtless read about the physical strength o f Gallic women in Timagenes, since Diodorus observes that they are not merely as large as their menfolk, but their match when it comes to blows.32 But the phrase globus pere­ grinorum surely hints, if only unconsciously, at a personal experience. It evokes a dispute in a tavern— Ammianus and his fellow officers pitted against an inn­ keeper and his fearsome wife, who is depicted with a satirical intensity wor­ thy o f Juvenal.33 The passages that draw on personal recollection in the excursus on Gaul en­ courage a search for similar passages elsewhere in the Res Gestae in which Am­ mianus may indirectly reveal something significant about himself. The histo­ rian had seen much during his long and varied life. He gives an unforgettable account o f his adventures in 359 when he was besieged in Amida and escaped with difficulty when the Persians stormed the city (18.6—19.9).34 Two passages in this account are unintentionally revealing. As Ammianus and his compan­ ions hurried to take refuge in Amida in 359, they found a soldier in hiding in the woods o f Meiacarire and brought him before Ursicinus: S u s p ic io n w a s a ro u se d b y th e in c o n s is te n t a n s w e rs th a t h e g a v e in h is fr ig h t , a n d fin a lly th reats e x tr a c te d fr o m h im th e w h o l e tru th . H e t o ld us th a t h e w a s b o m

32 Diodorus 5 .3 2 .2 . 33 O n the viscerally visual nature o f Ju ven al’s poetical imagination, see R . Jenkyns,

Three

Classical Poets. Sappho, Catullus andJuvenal (London, 19 8 2 ), 1 7 4 - 2 2 1 . 34 O n the gripping quality (and literary artifice) o f these personal reminiscences, see Rosen,

Studien (19 70 ), 5 1 - 6 8 ; Bitter, Kampfichilderungen (19 76 ), 1 2 - 5 6 ; M atthew s, Ammianus (1989),

57- 65. [iOO]

T

h in g s

S

ee n

a n d

T

h in g s

R

e a d

at P a ris in G a u l: w h e n s e r v in g in a c a v a lr y r e g im e n t, h e h a d fle d in to e x ile in P e rsia b e c a u s e h e fe a r e d p u n is h m e n t f o r an o ffe n c e h e o n c e c o m m it te d . A f t e r th at, w h e n h e h a d p r o v e d th e e x c e lle n c e o f h is c h a r a c t e r b y a c q u ir in g a w if e a n d r a is in g c h ild r e n , h e w a s se n t to o u r t e r r it o r y as a s p y a n d o fte n b r o u g h t b a c k r e ­ lia b le in fo r m a t io n . O n th e p re s e n t o c c a s io n h e h a d b e e n se n t b y T a m s a p o r a n d N o h o d a r e s , g ra n d e e s w h o c o m m a n d e d r a id in g p a rtie s , a n d w a s r e t u r n in g to r e ­ p o r t w h a t h e h a d le a rn e d . A f t e r h e th e n a d d e d w h a t h e k n e w o f o p e r a tio n s o n th e o t h e r s id e , h e w a s k ille d , ( i 8 .6 .1 6 )

Killing at close quarters and being killed were a normal part o f life for an officer like Ammianus, who had spent a whole night in 359 pressed against a corpse: H e r e w e s to o d m o t io n le s s u n til th e su n ro s e , m ix e d w it h P e rsia n s w h o w e r e s t r iv ­ in g w it h e q u a l e ffo r t to r e a c h h ig h e r g r o u n d w it h us. W e w e r e p a c k e d t o g e t h e r so t ig h t ly th at th e b o d ie s o f th e d e a d p r o p p e d u p b y th e p ress c o u ld n o t fin d th e sp a c e to fa ll a n y w h e r e , a n d a s o ld ie r in fr o n t o f m e , w h o s e h e a d h a d b e e n sp lit in to t w o e q u a l h a lv e s b y a p o w e r fu l s w o r d - s t r o k e , w a s h e m m e d in o n all sid es a n d s to o d u p r ig h t lik e th e s tu m p o f a tre e . ( 1 8 . 8 . 1 2 )

Such experiences can hardly fail to mark a man unless he is unusually insen­ sitive. The fourth century was a brutal age, and the historian was capable o f brutality: indeed, one student o f Ammianus has argued that the cruel world in which he lived so brutalized the historian that he became insensitive to hu­ man suffering.35 Despite the obvious danger that the exegete may read into Ammianus’ text what is not really there, another passage requires consideration in this context: T h e p e o p le o f E g y p t a re f o r th e m o s t p a rt s w a r t h y a n d d a rk , r a th e r g lo o m y , le a n a n d w it h a d r ie d - u p lo o k , e a s ily r o u s e d to e x c it e d g e stu re s , q u a r r e ls o m e a n d m o s t p e rs is te n t in p u r s u in g a d e b t. A m o n g t h e m a m a n b lu s h e s w h o c a n n o t e x h ib it m a n y scars o n h is b o d y in c u r r e d b y r e fu s in g to p a y ta x e s. A n d it h as n o t y e t b e e n p o s s ib le to d e v is e a t o r tu r e h a rsh e n o u g h to c o m p e l an o b s tin a te r o b b e r fr o m th at r e g io n e v e n to state h is n a m e a g a in st h is w ill. ( 2 2 .1 6 .2 3 )

As in his description o f the muscular women o f Gaul, Ammianus is surely draw­ ing on his own experience. He writes as one who has tortured, or supervised 35

Blockley, A m m ianus ( 19 7 5 ), 15 6 : “ as he passes through various adventures, he is totally

alone— no friendship, no real relationship at all, no feeling.“

[ 101]

T

h in g s

S

een

a n d

T

h in g s

R

e a d

the torture of, an Egyptian who refused even to reveal his name. Ammianus may have enjoyed inflicting pain on others. For much in his narrative has a palpably sadistic quality. The problem o f the relation between erudition and observation in Ammianus, between book learning and the reporting o f fact is posed most acutely by his account o f an eclipse o f the sun in Book X X , which immediately precedes the proclamation o f Julian as Augustus: A t th e sa m e tim e , t h r o u g h o u t th e r e g io n s o f th e E a st th e s k y w a s se e n v e ile d in a d a rk m is t, a n d th e stars s h o n e th r o u g h it c o n t in u o u s ly fr o m th e first d a w n in g o f th e d a y u n til n o o n . A n d th e re w a s a d d e d to th e se te rro rs th e fa c t th a t, s in c e th e lig h t o f h e a v e n w a s c o v e r e d w it h its g lo b e h id d e n fr o m th e s ig h t o f th e e a rth , th e fe a rfu l m in d s o f m e n t h o u g h t th a t th e su n h a d b e e n e c lip s e d f o r lo n g e r th an u su al. ( 2 0 .3 .1 )

It has long been recognized that this passage is highly problematical: at the most obvious and elementary level, anyone who has observed a total eclipse o f the sun knows that it lasts but a very few minutes in any one place. Al­ though Szidat and the Dutch commentators have diligently investigated the sources o f the excursus on astronomy that follows this report (20.3.2—12),36 they have unfortunately failed to set out the full extent o f the problems in­ herent in the report itself. These were duly noted by M ax Biidinger, who drew an alarming conclusion: the reader and the researcher who wish to use Ammianus as a repository o f facts ought to be warned that either he has un­ critically repeated erroneous information or else he is indulging his imagina­ tion.37 The latter is the correct explanation o f three serious difficulties in what Ammianus says about the eclipse o f 360. The first concerns the date. Ammianus unambiguously dates the eclipse he describes to the late winter or spring o f the year. It follows the cashiering o f Ursicinus for the loss o f Amida in 359, even though the real responsibility (Am­ mianus asserts) lay with the stubbornly lazy Sabinianus (20.2), and it precedes the proclamation ofjulian as Augustus while he was still in winter-quarters in Paris (20.4, cf. 1.1). Now there were two solar eclipses in 360: one occurred on 4 March, at the date indicated by Ammianus, but was visible only in the 36 Also D . den H engst, M nem osyne4 39 (19 86 ), 1 3 6 - 4 1 . 37 M . Biidinger, A m m ian us Marcellinus und die Eigenart seines Geschichtswerkes (Denkschriften Wien, Phil.-hist. C I. 4 4 .5 , 18 9 5), 40.

[102]

T

h in g s

S

een

a n d

T

h in g s

R

ea d

southern hemisphere; the other, to which it has always been assumed that Ammianus referred, occurred on 28 August.38 The second difficulty is that Ammianus misstates the nature o f the eclipse. Modern astronomers distinguish between total eclipses o f the sun, when the in­ terposed disk o f the moon totally obscures that o f the sun, leaving only the solar corona visible to the observer on earth, and annular eclipses, when the moon, being further away from the earth, does not totally block out the di­ rect light o f the sun, but leaves a rim or ring o f the suns disk visible through­ out. (Some eclipses are total at some places on the earths surface, annular at others.) The distinction was known to ancient astronomers too: although Ptolemy effectively denied that annular eclipses could occur at all,39 the N eoplatonist Proclus states the distinction clearly and reports that Sosigenes, the teacher o f Alexander o f Aphrodisias, observed one, whose date can be fixed as 4 September 164.40 The solar eclipse o f 28 August 360 was an annular eclipse, not a total eclipse at any place on the surface o f the globe. Hence Ammianus cannot be reporting what anyone actually observed in 360 anywhere in the northern hemisphere. The third difficulty was stated with exemplary clarity by Julius Zech in a Leipzig prize essay o f 1853 : the eclipse was probably not visible at all in Roman territory.41 Although neither o f the two recent commentaries on Book X X cites Zech’s rather rare work directly, the basic facts are not in dispute: in Op­ polzers standard canon o f eclipses, the sunrise point for the eclipse o f 28 Au­ gust 360 was at a longitude o f about 54 degrees east o f the Greenwich merid­ ian and the midday point was about 124 degrees east, that is, the path o f annularity began at sunrise somewhere to the east o f Teheran and the eclipse was visible at midday in China.42

38 T . von O ppolzer,

Canon der Finsternisse (Denkschriften Wien, M ath.-naturwiss. C I. 52, Canon of Eclipses, trans. O . G ingerich ( N e w York, 19 6 2 ), 1 5 0 -

18 8 7 ), 1 5 0 - 5 1 : Blatt N r. 7 5 = 5 1 : C h art 7 5 .

T h e criticisms o f O p p olzer voiced b y J. M eeus, C . C . Grosjean, and W . Vanderleen, Canon of Solar Eclipses (O xfo rd , 1966), 3 3 - 3 9 , do not make any difference to the point at issue here: in their charts for eclipses between 18 9 8 and 2 1 6 1 (Charts 1 —24), the beginnings and ends o f eclipses are identical w ith O ppolzer’s ( Canon [19 6 2 ], Charts 1 4 9 - 6 0 ) . 39 Ptolem y,

Almagest 5 . 1 4 ( 1 . 4 1 7 H eiberg), cf. G . J. Toom er, Ptolemy's Almagest (London,

19 8 4 ), 2 5 2 , n. 53 . 40 Proclus, Hypotyposis 1 . 1 9 - 2 0 , 4.98 ( 1 0 . 1 3 - 2 3 , 1 3 0 . 1 6 - 2 3 Manitius), cf. O . Neugebauer, History of Ancient Mathematical Astronomy 1 (Berlin, 19 7 5 ), 104, n. 4. 41 J. Z e c h , Astronomische Untersuchungen über die wichtigeren Finsternisse (Leipzig, 18 5 3 ), 38,

53-5542 O ppolzer,

Canon der Finsternisse (18 8 7 ), 1 5 0 - 5 1 : Blatt N r. 7 5 = Canon of Eclipses (19 6 2 ),

1 5 0 - 5 1 : C h art 7 5 .

[103]

T

h in g s

S

een

a n d

T

h in g s

R

e a d

Nearly fifty years after Zech, Oppolzers pupil Friedrich Ginzel devised an apologetic strategy designed to save Ammianus’ credit. He argued that the most easterly Roman troops, who were in the vicinity o f Nisibis in 360, may have seen a small partial eclipse at sunrise and that this was combined with re­ ports that later arrived from Persia that there the sun had been completely darkened and stars observed to come out.43 Otto Seeck grasped the straw of­ fered by Ginzel without appreciating the real difficulty: he postulated that Ammianus based his “ description” o f the eclipse on a written source that drew on the report o f an eyewitness who had seen the eclipse in the Persian Em ­ pire.44 Although this theory became canonical,45 and Ginzel’s explanation has been hailed as probably definitive,46 it plainly will not do. First, Ammianus al­ ways uses the phrase Eoi tractus to designate the eastern parts o f the Roman Empire, not territory still further east (cf. 2 8 .1.1; 30.2.9, 4.8).47 Second, he states emphatically that the stars were visible for hours on end, from dawn to midday. In fact, even if some o f the planets and the brightest stars may have been briefly visible during the eclipse o f 28 August 360, the sky was never any­ where dark enough for “ the stars” to be visible at any time. Ammianus’ gross and crass misstatements about the date, nature, and duration o f the eclipse o f 360 show that what he says is no kind o f factual “ report” at all. What is the explanation? It must be sought, not through Quellenforschung, but in the historian’s imagination. Ammianus was in a position to know that there were two solar eclipses in 360, whether or not anyone in the Roman world saw either o f them. For ancient astronomers could predict both solar and lunar eclipses accurately: the Stoic interlocutor in Cicero’s De Natura De­ orum boasts that the exact dates and extent are known and predicted for all fu­

43 F. K . Ginzel, Spezieller Kanon iter Sonnen- und Mondßnsternisse für das Ländergebiet der klas­ sischen Altertumswissenschaften und den Zeitraum non 900 vor d ir. bis 600 nach d ir. (Berlin, 1899), 2 12 -13 . 44 O . Seeck,

Hermes 4 1 (19 0 6), 5 3 7 —38: “ so genau und mit so reichem Detail g esch ild ert. . . .

w ie es nur ein A ugenzeuge dieser unheim lichen Erscheinung konnte . . . D ie Beobachtungen, die er mitteilt, konnte man nur tief im Inneren des Perserreiches m achen” ;

Regesten ( 1 9 19 ) , 208:

“ D ie Schilderung bei A n im . 2 0 .3 .1 muss aus dem Perserreiche stam m en.” 45 H en ce the confident dismissal o f the problem s b y A . D em andt,

Abh. Mainz, G eistes- und

Sozialwiss. Kl. 19 7 0 , 5 0 1: “ beheben sich die Schw ierigkeiten . . . . w enn w ir einen B e o b ach ­ tungsort in M ittelpersien annehm en.”

dironology of Eclipses and Comets A . D. 1 —1000 (W oodbridge, 19 8 4 ), 58. T h e Ancient Astronomical Observations and the Acclerations of the Earth and the Moon (Baltim ore, 19 70 ), 1 1 8 ; Medieval Chronicles and the Rotation of the Earth (Baltim ore and Lo n d o n , 19 7 2 ), 5 3 7 ; The Moon’s Acceleration and its Physical Origins, 1 (Baltim ore, 46 D . J. Schove,

successive discussions by R . R . N e w to n ,

1 979 ), 3 8 5 - 8 6 , are second-hand and disregard the problem s raised b y Z e c h : nevertheless, N e w ­ ton feels com pelled to conclude that “ this record remains a m ystery” (469). 47 Sim ilarly

orbis Eous and partes Eoae: see the passages listed by Viansino, Lexicon 1.4 7 2 .

[104]

T

h in g s

S

een

a n d

T

h in g s

R

e a d

ture time;48 Marinus records that during Proclus’ lifetime liemcrographoi pre­ dicted an eclipse that occurred a year after the philosopher’s death;49 and Am­ mianus’ contemporary, the Alexandrian astronomer Theon, used the partial solar eclipse o f 16 June 364 to illustrate how observation confirmed the cal­ culation o f Ptolemy’s Handy Tables.50 Ammianus reproduces a standard literary account o f a total eclipse, which is an imaginative reconstruction, not a report derived from observation.51 The eclipse has a clear function in Ammianus’ narrative, for it intervenes between the fall o f Ursicinus and the ascent o f Julian. Hence it has been argued that, although the context may be false, the historian can be pardoned for follow­ ing a literary convention whereby solar and lunar eclipses accompany dramatic reversals o f fortune, important battles, and the deaths o f rulers.5253* In such cases, accurate chronology is often disregarded in order to synchro­ nize celestial and terrestrial events. The Battle o f Zama in 202 B.c., in which Scipio defeated Hannibal and won Mediterranean hegemony for Rome, was preceded by an eclipse o f the sun according to Zonaras, presumably repeat­ ing Cassius Dio, and followed by an eclipse according to Livy: for astro­ nomical reasons, Livy’s eclipse, which he includes among the prodigies o f the consular year 202/1 b .c ., must be that o f 6 May 203, Zonaras’ that o f 19 Oc­ tober 202.53 Eunapius, the contemporary o f Ammianus, produced a solar eclipse during the Battle o f the Frigidus on 3 —6 September 394:54 presum­ ably he simply transferred the total eclipse o f the sun on 20 November 393, which two chronicles register under the correct consular year,55 to a date at which no eclipse can possibly have occurred.56 The passing o f rulers, too, was 48 C ice ro , N D 2 . 1 5 3 : soli enim ex animantibus nos astrorum ortus obitus cursusque co g ­ novim us, ab hom inum genere finitus est dies mensis annus, defectiones solis et lunae cognitae praedictaeque in om ne posterum tempus, quae quantae quando futurae sint. 49 Marinus,

Vita Procli 37 . L - S - J 9 7 7 1 , w h ich cites only this passage for hemerographos, mis­

leadingly glosses as “ one w h o keeps a diary.” T h e w ord seems to o ccur elsewhere only in O ly m -

Mctcora 3 4 2 b 33 (ed. W . Stiive, Commentaria in Aristotelem Graeca 1 2 .2 [Berlin, 1900], 5 0 .2 1), w h ere hemerographoi seek an explanation for the phases o f the

piodorus’ com m entary on Aristotle’s planet M ercury.

Bulletin de 1’lnstitnt historique belge de Rome 4 6 - 4 7 ( 1 9 7 6 - 1 9 7 7 ) , 3 5 - 7 9 . R E 6 (1909), 2 3 4 9 - 5 0 : “ mehr schematisch als aus Beobach tung,” an annular eclipse

50 A . T ih o n , 51 F. Boll,

described as if it were a total exclipse; 2 3 6 3 : “ mancherlei Ü bertreibungen.” 52 A . Dem andt,

Abh. Mainz , Geistes- und Sozialwiss. Kl. 1970, 4 9 5 - 5 0 7 .

53 Zonaras 9 .1 4 ; L iv y 30 .38 .8 : C um is orbis minui visus est et pluit lapideo imbri, cf. F. K . Ginzel,

Handbuch der mathematischen und technischen Chronologie, 2 (Leipzig, 1 9 1 1 ) , 2 1 7 , 529.

34 Zosim us 4 .5 8 .3 .

55 Chr. min.

1.2 9 9 ; 2 .6 3, cf. Ginzel,

Kanon (189 9 ), 2 7 3 - 7 4 ; Boll, R E 6 (1909), 2 3 6 3 .

56 T h e nearest n e w m oon appeared near m idnight during the night o f 1 2 - 1 3 Septem ber: even the Elder Pliny could state clearly that a solar eclipse can occur only on either the first or the last day o f a lunar month

(Nat. Hist. 2 .13 ) .

T

h in g s

S

e e n

a n d

T

h in g s

R

e a d

announced in the heavens: according to Dio, the death o f Augustus was fore­ told by a solar eclipse, even though the only solar eclipse to which he can re­ fer occurred on 1 5 February a . d . i 7, two and a half years after Augustus died.57 Similarly, the evangelist Luke, who expressly puts himself in the tradition o f Greek historiography, produces an eclipse o f the sun while Jesus hung on the cross.58 The corresponding passages ofM ark and Matthew merely have a mys­ terious darkness, whose cause may be meteorological.59 Origen knew (and the pious held it to his discredit) that no solar eclipse can have occurred dur­ ing the crucifixion, since the gospels state the day as 14 or 15 Nisan, when the moon was on the opposite side o f the earth from the sun.60 The plea that Ammianus merely repeats a commonplace fails to exculpate an author who parades his technical knowledge o f astronomy in the excursus that immediately follows the notice o f the eclipse. The historian could have discovered from Ptolemy, whom he names (20.3.4), how to verify the correct date,61 or he could have consulted a list o f actual and predicted eclipses pro­ duced by a competent astronomer or professional astrologer.62 The hypothe­ sis that perhaps best explains what Ammianus has written is that he knew from the calculations o f others that a total eclipse occurred on 4 March 360: since this eclipse was visible only in the southern hemisphere, the whole o f his “ re­ port” is imaginative fiction. 57 D io 5 6 . 2 9 . 2 - 3 , cf. Ginzel, Handbuch, 2 ( 1 9 1 1 ) , 5 3 1 . 58 Luke 1 . 1, 2 3.4 4 . 39 M ark 1 5 .3 3 ; M atthew 2 7 .4 5 .

Comm. ser. 1 3 4 ( P G 1 3 . 1 7 8 2 = E . Klosterm ann and U . Treu, Origenes Werke 1 1 . 2 2 Abh. Mainz, Geistes- und Sozialwiss. K l. 19 7 0 , 4 8 2 - 8 4 . 61 Besides com piling his Handy Tables, Ptolem y devoted B o o k V I o f the Almagest to eclipses: Toom er, Ptolemy’s Almagest (19 8 4 ), 2 7 5 - 3 2 0 . 60 O rigen,

[ G C S , 19 76 ], 2 7 4 - 7 8 ) ; Dem andt,

1,2 T h ree papyri relating to the prediction o f eclipses (P. O.vy. 4 1 3 7 , 4 1 3 8 , 4 13 8 a ) are to be

Astronomical Papyrifrom Oxyrhynchus (Memoirs of the American Philosophi­ cal Society, forthcom ing): the first has a list o f predicted lunar eclipses, including eclipses o f

published by A . Jones, a .d .

5 6 - 5 7 , the other tw o deal w ith the prediction o f eclipses in general.

[ 106]

[X] ENEMIES, ANIMALS, AND STEREOTYPES

Ammianus discloses his likes and dislikes o f individuals with unusual frankness and directness.1 When he held a strong opinion o f someone, he could express this opinion in vivid and unforgettable language that often evokes a precise visual image. Thus his account o f the arrest o f Peter Valvomeres strikes the reader as remarkable above all for its palpable visual qualities (Chapter II). Peter .was “ a very tall, red-haired man towering above the rest” who replied to the urban prefect’s question with truculence, was tortured and released, but later executed for raping a virgin o f good birth (15.7.4 —5). Many other in­ dividuals receive an equally sharp delineation from Ammianus, even if, like Peter Valvomeres, they appear in his narrative only once and briefly. Ammi­ anus usually delivers himself o f a firm moral verdict on their conduct or char­ acter; and when he condemns, as he does in the majority o f cases, he often employs animal imagery to reinforce his verdict. An excellent example o f the sharp characterization o f a historical actor who makes a single brief appearance in the Res Gestae is Ammianus’ presentation o f Mercurius, the “ Count o f Dreams” (comes somniorum), who pursued those who had supported the usurpation o f Magnentius. Mercurius would fre­ quently insinuate himself into dinner parties and other gatherings “ like a fierce dog which conceals its savage disposition by wagging its tail” : he used “ his poisonous skill” to turn reports o f dreams into confessions o f crimes that he then carried to the credulous emperor (15.3.5). Similar vignettes introduce many other characters who either appear once in a particular episode that 1 Fo r an anthology o f Anunianus’ portraits and an analysis o f the “ categories o f asessment” that he em ploys, see D rexler, AmmianslutUen (19 7 4 ), 3 8 - 6 5 , 1 3 - 1 9 .

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

Ammianus develops at length or play a part in his main narrative. Book X X V III describes a series o f trials for adultery and magic in Rome, in which Maximi­ nus o f Sopianae played a large role. He is introduced as follows: M a x im in u s . . . w a s b o r n in th e u tm o s t o b s c u r it y at S o p ia n a e , a t o w n in V a ­ le ria . H is fa th e r w as a c le r k in th e g o v e r n o r ’s o ffic e w h o w a s d e s c e n d e d fr o m th e C a r p i w h o m D io c le t ia n tra n s fe rre d to P a n n o n ia a fte r th e y h a d b e e n e x p e lle d fr o m th e ir o r ig in a l h o m e .2 A ft e r a m o d ic u m o f e d u c a t io n in lib e ra l s tu d ie s, an ig n o b le c a r e e r as a d e fe n se la w y e r , a n d a d m in is t e r in g C o r s ic a a n d also S a rd in ia , M a x im in u s b e c a m e g o v e r n o r o f T u sc ia . F r o m this h e w as p r o m o t e d to ta k e c h a rg e o f th e g r a in s u p p ly o f R o m e , b u t still re ta in e d h is p r o v in c ia l g o v e r n o r s h ip b e ­ c a u se his s u c c e s so r w a s d e la y e d e n r o u te f o r a lo n g p e r io d . A t th e start, h e c o n d u c te d h im s e lf c a u tio u s ly o u t o f th re e c o n s id e r a tio n s . F irst, th e re re s o u n d e d in h is ears th e p r e d ic t io n o f h is fa th e r, w h o k n e w e x t r e m e ly w e ll w h a t th e flig h t a n d s o n g o f s ig n ific a n t b ird s m ig h t p o r t e n d , to th e e ffe c t th a t h e w o u ld re a c h a h ig h p o s itio n o f state, b u t th e n p e ris h b y th e e x e c u t io n e r ’s s w o rd . N e x t , h a v in g a c q u ir e d a m a n fr o m S a r d in ia w h o w a s v e r y s k ille d in r a is in g e v il sp irits a n d e lic it in g p r o g n o s t ic a t io n s fr o m th e d e a d a n d w h o m h e la te r c r a ft ily d e c e iv e d a n d k ille d , so a r u m o r c ir c u la te d , h e w a s m ild e r a n d m o r e a m e n a b le as lo n g as th is m a n w a s a liv e , f o r fe a r o f b e in g b e tra y e d . L a stly , w h ile h e w a s still w o r m in g h is w a y th r o u g h in fe r io r p o sts lik e a sn a k e lu r k in g u n d e r g r o u n d , h e w a s n o t y e t s tr o n g e n o u g h to c au se th e d ea th s o f im p o r ta n t p e o p le . ( 2 8 . 1 . 5 - 7 )

The unfairness o f this characterization leaps o ff the page. What has not always been sufficiently realized is that Maximinus’ successful career was not unusual for a man o f his generation: it reflected the excellence o f his education and his talents as a lawyer, both o f which Ammianus denigrates, probably un­ justly and certainly as a result o f a prejudice with a personal basis.3 Ammianus repeats his description o f Maximinus as a “ serpent” twice more in his account o f the trials at Rome (28.1.33, 4 1 )- Other human serpents in the Res Gestae are the tyrannical Caesar Gallus (14.7.13), the odious pair Daniel and Barzimeres, who slandered Pap the king o f Armenia to Valens (3 0 .1.16 —17), the Roman mob confronting the prefect Leontius (15.7.4), and Isaurian brigands (19 .13.1). There is a certain repetitiveness about these comparisons to animals. For example, Ammianus detested Arbitio as much as Maximinus, and hence he uses identical language. Arbitio appears in Book X IV as an ardent intriguer among groups o f fickle courtiers who wished to destroy the Caesar Gallus 2 Apparently in 296: Phoenix 30 (19 7 6 ), 17 6 , 18 7 . •’ J. Szidat, Historia 44 (19 9 5), 4 8 1 - 8 6 .

[108]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

(14 . i 1.2). In Book XV, as he plots against Ursicinus, he is described as an ex­ cessively powerful man who was very skilled in entrapping honest men in fa­ tal snares: he was like a snake lurking underground (again subterraneus serpens) that watches from its hidden hole for individual passers-by and suddenly at­ tacks them (15.2.4). Such characterizations consistently indicate the historian’s dislikes, and they usually give voice to a deep personal animosity, as when Am ­ mianus speaks o f lawyers following the scent like Spartan or Cretan hounds until they reach their lairs (30.4.8).4 Analysis o f Ammianus’ use o f animal comparisons and similes produces some significant conclusions.5 Although the soldiers o f two legions recently trans­ ferred from Gaul bellow like wild beasts when they cannot make their way out o f the besieged city o f Ami da (19.5.2) and Valentinian conducts himself like a sagax bestia when he investigates the corrupt administration o f the prae­ torian prefect Probus (30.5.10), Ammianus’ animal comparisons usually have a highly negative connotation.6 When Valentinian fell ill in 367, there was talk o f Rusticus Julianus, the magister memoriae, as a possible successor, a man eager for human blood like beasts, as he showed when he became proconsul o f Africa (27.6.i).7 The future usurper Procopius hid himself awaiting an op­ portunity to seize power “ like a beast o f prey who will jump forth as soon as he sees something that can be caught” (26.6.10). Almost the same words are applied to Valentinus, a native o f Valeria in Pannonia, an arrogant fellow, the brother o f the wife o f Maximinus, “ that deadly vicarius and later prefect” (Am­ mianus alludes obliquely to the trials in Rome, which he had just described): exiled to Britain for a serious crime, the evil beast, unable to endure inactiv­ ity, plotted against Theodosius, who justifiably had him executed (28.3.4 —6). Leo, who assisted Maximinus in Rome, was just as bad as his superior: he was “ a grave-robbing brigand from Pannonia, slobbering cruelty with an animal grin and no less thirsty for human blood” (28.1.12). Groups and nations are also compared to wild animals. Both the Isaurian bandits who raided peaceful provinces in 353 and the Germans who were 4 Blockley, A m m ian us (1975), 183, lists a total o f fifty-three passages that use “ animal images.” 5 T. E. J. Wiedemann, Past Perspectives. Studies in G reek and Rom an Historical W riting , ed. I. S. M oxon, J. D . Smart, and A. J. Woodman (Cambridge, 1986), 1 8 9 - 2 0 1 , cf. Blockley, A m ­ mianus ( 1 9 7 5 ) . 1 8 3 - 8 4 .

6 For a list o f examples o f “ savagery and madness,” see R. Seager, A m m ian us M arcellinus: Seven Studies in H is Language and Thought (Columbia, Miss., 1986), 5 4 - 5 8 .

7 Julianus’ proconsulate is firmly dated to the biennium 3 7 1 - 3 7 3 : P hoenix 39 (1985), 150, 273. Hence the pluperfect ut ostenderat has a future reference and an aorist meaning.

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

about to invade Gaul in 357 were like wild beasts driven by hunger (14.2.2,® 16.5.17), and Ammianus states that the bodies o f the Huns were so prodi­ giously deformed and twisted that one would think that they must be twolegged beasts (31.2.2). The Saraceni o f the desert were in the habit o f swoop­ ing down like hawks from afar whenever they espied prey, seizing it and darting away at once (14.4.1). Similarly, the African desert tribe o f the Austoriani, ar­ rogant because o f success, swooped down upon the unfortunate Tripolitanians during the reign o f Valentinian “ like fierce birds o f prey roused to greater frenzy by the stimulus o f blood” (28.6.13): apart from substituting bestiae for rapaces alites, Ammianus applies virtually the same phrase to the triumphant Goths in 378, who had defeated Valens and were about to attack the city o f Adrianople (31.15.2 : ut bestiae sanguinis irritamento atrocius efferatae). Although the historian tends to regard barbarians as at least half bestial by nature, he does not apply animal similes indiscriminately to all non-Romans.9 Thus he never calls the Persians bestiae in his own voice, although he does allow a Roman emperor and the Persian king to trade animal insults. Julian parades starving Persian prisoners before his troops to improve their morale and calls them “ ugly she-goats disfigured with filth” (24.8.1): that passage is matched by a letter o f Shapur to Constantius that implicitly compares Romans to wild beasts (24.2.8). Christians, however, are not merely bestial, they are worse than animals. Julian invited Christian bishops who disagreed with one another and their congregations, who were also rent by schism, into the im­ perial palace and urged them to put their quarrels aside and allow everyone complete freedom o f religion. He did this so that, as freedom increased their internal dissensions, he would never in future need to fear a united church, since he knew that no wild beasts are so hostile to humans as the majority o f Christians are to one another (22.5.4). Specific proof o f the general charge comes in Book X X V II. When Vivendus was prefect, the see o f Rome fell va­ cant. A contest ensued between Damasus and Ursinus, who both had a pas­ sionate ambition to seize the position o f bishop that surpassed human bounds. The fighting between their supporters forced Vivendus to withdraw to a villa outside the city and left one hundred and thirty-seven corpses in the Basilica o f Sicininus on a single day (2 7 .3 .12 -13 ). The slaughter (if not Ammianus’ pre­ cise total for the number o f dead) is confirmed by a contemporary document.10 K The word bestiae , absent in both V and Seyfarth’s text, should be added before fa m e , as Valesius proposed: Ammianus is quoting Cicero, Pro Cluentio 67: ut etiam bestiae fame monitae plerumque ad eum locum ubi pastae sunt aliquando revertantur. ** Wiedemann, Past Perspectives (1986), 19 6 —201. 10 The writer o f the account o f Q uae gesta sunt inter U berium et Felicem episcopos (Collectio A v e l­ lana 1) alleges that the supporters o f Damasus invaded a basilica and killed 160 on the spot and

[n o ]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

As a vehicle for conveying his opinion o f individuals, Ammianus uses racial or cultural stereotypes as well as animal comparisons. The locus classicus for such stereotyping in a serious historian is Cassius Dio on Caracalla, the son o f Sep­ timius Severus and Julia Domna, who was born at Lyon: “ Antoninus be­ longed to three races, and he possessed none o f their virtues at all, but com­ bined in himself all their vices: the fickleness, cowardice and recklessness o f Gaul, the harshness and cruelty o f Africa, and the slyness o f Syria, from whence he came on his mother’s side.” 11 Ammianus is never so explicit. Yet regional and national characteristics were a staple o f ancient astrology and physiogno­ monies,12 and reasoning and prejudices similar to those o f Dio lie behind many o f his delineations o f character. The government o f Valentinian was largely an administration o f men from Pannonia, who were regarded as irretrievably un­ couth and uncultured.13 Ammianus notes that it was Pannonian officers who carried Valentinian’s election as Augustus in 364 and that the new emperor gave Pannonians preferment (26.1.5, 5.3). One o f Valentinian’s earliest ap­ pointments was Ursacius as magister officiorum: he was a crude Dalmatian, a cruel and irascible man (26.4.4, 5.7). The phrase Delmatae crudo reflects a cultural stereotype that was doubtless prevalent in Rome.14 But what o f Pannonians whom Ammianus for personal reasons admired? His presentation o f Viventius, who belongs to this category, is subtle. Viventius is introduced in Book X X V I as a native o f Siscia, which happens to be a town in Pannonia, but Ammianus here suppresses any mention o f his ethnic origin as a Pannonian (26.4.4). Viventius reappears later in Book X X V II as praefectus urbi, and Ammianus takes the opportunity to repudiate the stereotype explicitly by commending him as “ an upright and wise Pannonian, under whose quiet and peaceful adminis­ tration there was general plenty” (27.3.11). The phrase integer ct prudens Pan­ nonius was surely intended to sound paradoxical. In a text as poorly transmitted as that o f Ammianus, geographical names, like numerals,15 are particularly vulnerable to corruption. Hence the historian’s wounded many others, some o f whom later died— without suffering any losses themselves

(C S E L 3 5 . 3 .1 8 - 2 1 ) . 11 Dio 78(77).6.ia ( 3 .3 7 9 .1 7 -3 8 0 .3 Boissevain). Maurice devoted Book X I V o f his Strategikon to the “ Characteristics and Tactics o f Various Peoples” (trans. G. T. Dennis [Philadelphia, 1984], 1 1 3 - 2 0 ) . 12 T. Barton, Ancient Astrology (Lo n d o n /N ew York, 1994), 1 7 9 - 8 5 . 13 Alfoldi, Conflict (1952), 1 3 - 2 7 ; Matthews, Aristocracies (1975), 3 2 - 5 5 . 14 Alfoldi, Conflict (1952), 9 6 - 1 2 4 , a chapter with the title “ The Late Classical Ideal o f C u l­ ture in Conflict with the Illyrian Military Spirit.” 15 O n corrupt numerals (and names o f months) in Ammianus, see Appendix 7; on the la­ mentable state in which his text has been transmitted, Appendix 1.

[m l

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

presentation o f character as partly dependent on geographical origin cannot be documented properly. Two examples will illustrate how the condition in which the text o f the Res Gestae has survived clouds an exact understanding o f Ammianus’ presentation o f individuals. In Book XV, Paul “ the Chain” is introduced as “ a native o f Spain, with a smooth inscrutable face and an extraordinary capacity for scenting hidden perils” (14.5. i ). The paradosis has ortus in Hispania glaber quidam, and Richard Bentley emended to coluber quidam (“ a certain serpent” ).16 Clark printed Bent­ ley’s emendation, but Seyfarth rightly reinstates the transmitted reading.17 In Book XV, however, according to the manuscripts and editors, Paul and Mer­ curius are said to be instigators o f evil in the state, the one a Persian, the other from Dacia (15.3.4: hic origine Persa, ille natus in Dacia). In his critical appa­ ratus, Clark registered his own emendation o f Dacia to Baetica to bring this passage into conformity with the earlier one, since the Passion of Artemius, drawing on the lost ecclesiastical history o f Pliilostorgius, confirms that Paul was indeed a Spaniard.18 The emendation seems eminently reasonable. Yet the transmitted in Dacia is printed by all editors o f Ammianus, including Clark himself. What is significant is that no one would ever have questioned the reading in Dacia, had not the earlier passage and evidence external to Ammi­ anus stated that Paul’s origin was Spain, not Dacia. The fact that the reading in Dacia must be wrong unless Ammianus has contradicted himself suggests caution in dealing with other geographical names in his text.19 According to one fifteenth-century manuscript and Bentley, firmly sup­ ported by Adolf Kiessling,20 the quaestor Montius, who was lynched in Anti­ och in 354, was “ an African, but inclined to mildness” (14.7.12: Afer quidem, sed ad lenitatem propensior), unlike most o f his compatriots who were noto­ rious for their cruelty.21 But the ninth-centùry manuscript, which is here the sole independent witness to the text, has afen, and Seyfarth prints the emen­ dation acer, which he attributes to Gronovius and which the recent English translators accept. With this emendation, Ammianus characterizes Montius as “ a man o f excitable disposition, but not given to extremes,” which seems too 16 Bentley’s conjecture, in the margin o f his copy o f Gronovius’ edition (Leiden, 1693), was reported by K. Zangemeister, Rhein. Mus., N .F. 33 (1878), 470. 17 In favor o f glaber, L. Karau and I. Ulmann, Klio 48 (1967), 2 3 4 - 3 5 . IH Passio Arteniii 21 = Philostorgius, H E 7.6 3-7:1. It is relevant here that all seven personal names in 28.4.6 are indisputably corrupt (Ap­ pendix 1). 20 A. Kiessling, Jtilirb. für class. Phil. 103 (18 71), 499; Zangemeister, Rhein. Mus., N.F. 33 (1878), 4 71. 21 See, Dio 78(77).6.1a (quoted at n. 11); H A , Sev. 13.8: horum igitur tantorum ac tam il­ lustrium virorum . . . interfector ab Afris ut deus habetur.

[112]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

tame for a writer who is usually so pointed and incisive. Moreover, there is evidence outside Ammianus that Montius’ origin must be sought in the Greek East. The quaestor should probably be identifed with the cultivated “ Montius, best o f proconsuls” to whom Libanius dedicated his hypotheses o f Demos­ thenes’ orations, and whom he clearly implies to be a Greek speaker who also has an enviable command o f Latin.22 The friend o f Libanius was related to, perhaps even the father of, Calliopius who entered the Senate o f Constan­ tinople before 360.23 It could be, therefore, that the transmitted afen conceals an eastern provincial or ethnic designation: Montius could, for example, have been a Paphlagonian, like the philosopher and orator Themistius.24 For Pa­ phlagoniam had been a butt for literary ridicule ever since Aristophanes put the Athenian politician Cleon on the stage as a Paphlagonian slave.25 The origin o f the historian’s attitude toward a specific individual can often be discovered in or deduced from his text. Even though two are missing in a large lacuna in Book X X IX (Appendix 8), the praefecti urbis who are registered in the extant text o f the Res Gestae provide an appropriate sample for estimating how far personal and religious bias, which often may have been only semiconscious or even subconscious, hovers behind Ammianus’ verdicts on individuals.26 Either personal knowledge or an instinctive sympathy for pagans (some­ times both) can be discerned behind most o f the favorable protraits in this group. Leontius was an easterner and a Christian.27 Constantius appointed him prefect in 355 to enforce his religious policies in Rome, specifically to co­ erce the recalcitrant bishop Liberius (15.7.6—10).28 Leontius had come to An­ tioch as quaestor sacri palatii o f Gallus while Ammianus was there (14 .11.4 ). He may well be the son o f Domitius Leontius, the praetorian prefect o f Con­ stantius in the early 340s, who probably came from Berytus, and may have aided Ammianus’ career (Chapter VI). A sympathetic assessment o f his char­ acter introduces his prefecture. Leontius showed himself an excellent judge, who “ heard cases promptly and decided them impartially, and he was o f a 22 R. Förster, Libanii Opera omnia 8 (Leipzig, 1905), 600, cf. P L R E 1.535. 23 Libanius. Ep. 220, 2 14 , cf. O. Seeck, Die Briefe des Libanius (Texte und Untersuchungen 30, 1906), 99; P L R E 1 . 1 7 4 - 7 5 . 24 Vanderspoel, Themistius (1995), 3 1 - 3 8 . 25 Aristophanes, Knights 2 etc.; Lucian, Alexander 9, 17, 45; Libanius, Orat. 1.85 (“ three Paphlagonians, brothers in everything— character, ignorance, insolence and corpulence” ); Palladas,

Anth. Pal. 11.34 0 , where “ the Paphlagonian” means “ Cleon” and perhaps designates Themistius. 2h For full details o f the careers o f the prefects, see Chastagnol, Fastes (1962), 1 3 9 - 9 3 , with the modifications noted in Chapter IV and Appendix 8. 27 Epiphanius, Panarion 7 1 . 1 . 1 - 5 , cf. L ’Eglise et l ’empire au IV e siècle (Entretiens sur l ’Antiquité

Classique 34, 1989), 3 1 4 - 1 5 . 2H Phoenix 46 (1992), 2 5 6 - 6 1 .

[II3]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

kindly nature, though some thought him severe and too much inclined to condemn in order to preserve his authority” ( i 5.7.1). Ampelius, too, was personally known to Ammianus before he became prae­ fectus urbis under Valentinian. The notice o f his prefecture reviews his long ca­ reer. A native o f Antioch, Ampelius had been magister officiorum and had held two proconsulates (28.4.3): inscriptions and the Theodosian Code identify the proconsulates as those o f Achaea in 359-360 and o f Africa in 364—3Ö5.29 He was a pleasure seeker who, although capable o f occasional severity, preferred to court popularity: he decreed that taverns should not open nor cooked food be sold before the fourth hour o f the day, but allowed himself to be persuaded not to enforce it (28.4.4). Viventius, who is commended as an “ honest and wise Pannonian” when he became praefectus urbi in 365 (27.3.11), had previously, as quaestor sacri palatii in 364, conducted an examination o f friends o f the emperor Julian who had been accused o f using secret magic to make Valentinian and Valens ill (26.4.4). The honesty that Ammianus praises must have helped to produce the favor­ able outcome that he records: not a shred o f evidence was found for the ex­ istence o f a plot.30 Julian’s appointees, Maximus and Apronianus, naturally receive a favorable presentation. Julian appointed Maximus at Naissus in the autumn o f 361 in preference to Symmachus, who was his senior, because he was the son o f the sister o f Vulcacius Rufinus: in his term the food supply was abundant and the formerly frequent riots o f the populace ceased (21.12.24). Apronianus was ap­ pointed in Antioch in January 363 (23.1.4). On his journey to Rome he lost an eye through illness: blaming this loss on sorcery, he took stern measures against poisoners and sorcerers. He was an honest and severe prefect, under whom there was such a steady abundance o f supplies that, contrary to normal custom, no one even muttered a complaint (2 6 .3.1-3, 6). Equally favorable is the presentation o f praefecti urbis who were demonstra­ tively pagan. Under Tertullus there was a shortage o f grain that provoked se­ rious riots. But the shortage was not Tertullus’ fault, and he solved the prob­ lem: because the arrival o f the grain fleet from Africa was delayed by adverse winds and rough seas, he went to Ostia and sacrificed to Castor and Pollux, at which the storm subsided, the ships sailed into port, and the granaries were filled (19.10). Symmachus, who succeeded Apronianus in 364, was a man o f outstanding learning and modesty: during his prefecture, the holy city enjoyed unusual peace and plenty, and he restored a magnificent and solid bridge. Un­ 29 P L R E 1 . 5 6 - 5 7 , Ampelius 3; T. D. Barnes, Phoenix 39 (1985), 15 2 , 2 7 3; C P 82 (1987), 2 1 5 - 1 6 ; M . Heil, Z P E 108 (1995), 1 6 2 - 6 5 . ■1(1 On this episode, see now T. M . Banchich, Historia (forthcoming).

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

fortunately, the peoplè did not show proper gratitude: some years later (prob­ ably in 375) they burned down his beautiful house in Trastevere (27.3.3-4). Praetextatus also distinguished himself as prefect. His many acts o f honesty and uprightness, for which he had been famous since his youth, earned him the unusual distinction o f being feared by the citizens without losing their affection. His support o f truth and justice quieted the riots caused by Chris­ tian quarrels, and his exile o f Ursinus, the rival o f Damasus, produced pro­ found peace. Praetextatus was a scrupulously impartial judge, and he took a number o f salutary measures: he removed projecting galleries, he standardized weights to ensure fair dealing, and he demolished the walls o f private houses that had en­ croached on adjoining temples (27.9.8-10). Claudius, the last praefectus urbi to be registered in the Res Gestae, was also a zealous pagan: he was a quindecimvir sacris faciundis and he appears to have performed a taurobolium as prefect.31 Ammianus notes that his prefecture was marked by a flood o f the River Tiber. But no riots ensued and Claudius restored many old buildings, among them the Porticus Boni Eventus close to the temple o f the same deity (29.6.17—19). It may be significant that Ammianus makes no explicit allusion to the pa­ ganism o f either Symmachus or Praetextatus, which stands on record else­ where.32 In 384, shortly before Ammianus wrote, Symmachus’ son, the ora­ tor Symmachus, had led an embassy to the young Valentinian in Milan, which vainly requested the restoration o f the Altar o f Victory in the senate-house,33 and Praetextatus had died in December as consul designate for 385.34 In contrast to these praefecti urbis, two noble prefects receive a distinctly hos­ tile presentation. Ammianus introduces Memmius Vitrasius Orfitus, who was prefect twice (from December 353 to summer 355 and again from spring 357 to spring 359), as a man who conducted himself insolently as if his power were greater than it really was, and he stresses that frequent and serious riots marked his prefecture (14.6.1). Orfitus could doubtless trace his ancestry back to the Antonine aristocracy: 35 more relevant to his long tenure o f the urban prefec­ ture is the fact, which Ammianus does not mention, that Orfitus appears to 31 IL S 4 14 7 (Rome, dated 19 Ju ly 374). 32 Symmachus was a pontifex maior and a quindecimvir sacris faciundis (IL S 1257), whereas Prae­ textatus held many priesthoods (IL S 1259) and was later remembered as sacrorum omnium praesul (Macrobius, Sat. 1.1 7 .1) . 33 O n this famous episode, see Matthews, Aristocracies (1975), 203 —1 1 ; N . B. M cLynn, Am ­ brose ofAlilan. Church and Court in a Christian Capital (Berkeley, 1994), 1 5 1 - 5 2 , 16 6 —68. 34 Symmachus, Rel. 12.4. 35 As possible consular ancestors in the second century, note C . Memmius Fidus lulius A l­ bus, cos. suff. 19 1 or 192 (P IR 2 M 462), M . Gavius Orfitus, cos. ord. 165 (P IR 2 G 105), and T. Pomponius Proculus Vitrasius Pollio, cos. II 176, who married into the Antonine dynasty (P IR 1 p 558). Chastagnol, Fastes (1962), 140, makes Orfitus a novus homo.

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

have married into the imperial family.36 His predecessor in the management o f political affairs in Rome after the usurpation o f Magnentius was another imperial relative, Naeratius Cerealis, the uncle o f Gallus Caesar.37 The picture o f C. Ceionius Rufius Volusianus signo Lampadius, urban pre­ fect in 365—366, is particularly hostile (27.3.5—10). Ammianus begins with a stinging denunciation: he was so vain that he was upset if he was not praised for his unique skill in spitting. Although Ammianus admits that Lampadius was “ sometimes strict and honest,” the rest o f his notice is equally hostile. The most obvious manifestation o f Lampadius’ vanity, according to the his­ torian, was that as prefect he put his name on public buildings throughout the city claiming to have built rather than restored them. Moreover, frequent ri­ ots compelled him to withdraw to the Milvian Bridge, and he tried to avoid paying for the materials used in his construction and repair o f buildings. Am ­ mianus digresses to bring up a foolish act o f spectacular liberality that showed Volusianus’ contempt for the people o f Rome. When he was praetor long be­ fore, presumably late in the reign o f Constantine, he summoned some povertystricken beggars from the Vatican and gave them great wealth. The parallel complaint o f Eunapius, who alleges that Constantine wasted public money on useless buildings that soon collapsed,38 suggests that these beggars from the Vatican are Christian clerics who received a large donation for the great new church dedicated to Saint Peter.39 Volusianus’ wife Caecinia Lolliana is known to have been a priestess o f Isis, and their son commemorated his taurobolium in a dedication to Cybele and Attis on 23 May 390.40 Yet that does not ex­ clude the possibility that the prefect had been a Christian under Constantine, when it would have helped his career.41 Whether or not Ammianus’ intense dislike o f Volusianus had a religious di­ mension, it certainly had a personal basis. Ten years before Volusianus became praefectus urbi, he had been praetorian prefect in Gaul, a fact to which Ainmi■'h From Eusebius to Augustine (Aldershot, 1994), VII. 11 , n. 7; J R S 85 (1995), 144; Alan Cameron, J R /1 9 (1996), 2 9 5 - 3 0 1 . -,7 Ammianus’ obituary o f Gallus notes that Cerialis was his uncle ( 14 .11 .2 7 ) : his prefecture, which ran from 26 September 35 2 to 8 December 35 3 ( Chr: min. 1.67) will have been registered in the lost Book X III. Zosimus 2 .3 2 .1. 39 The church was under construction in the 330s: see C . Piétri, Roma Christiana: Recherches sur l ’église de Rome, son organisation, sa politique, son idéologie, de Miltiade à Sixte III (311-440) (Rome, 1976), 5 1 —64; T. D. Bames, Constantine (19 8 1), 310 , n. 6 1. 40 CIL 6 .5 12 = ILS 4 15 4 - O n the role ofLampadius and his family in the late-fourth-century revival o f the taurobolium, see N . N cLynn, Phoenix 50 (1996), 3 2 6 —28, w ho appears to assume that Ammianus alludes to the Vatican Phrygianum, not to St. Peter’s. 41 Constantine’s preference for Christians in the highest offices is documented in J R S 85 ( • 9 9 5 ). I 3 6 - 4 7 -

[ 1 1 6]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

anus draws explicit attention when introducing his urban prefecture (27.3.5: ex praefecto praetorio). During his tenure o f this earlier office, according to Ammianus, there were rumors that he, together with Eusebius, the former comes rei privatae, and Aedesius, the former magister memoriae, had suborned the actuarius Dynamius to forge the letter that caused the unfortunate Silvanus to rebel (15.5.3: subornatore et conscio, ut iactavere rumores incerti, Lampadio praefecto praetorio etc.). Ammianus presents Silvanus as a plain soldier and an honest man unable to cope with the intrigues at the court o f Constantius, just like Ursicinus, who also had been cheated o f the just reward o f his untiring labors (15.5.28). The fact that Ammianus was committed to hostility toward the enemies o f Silvanus and Ursicinus must be relevant to his treatment o f Volusianus. The longest and most detailed portrait o f an individual other than an em­ peror is that o f Sex. Petronius Probus, four times praetorian prefect and con­ sul in 37 1, who had died very recently at the time o f writing (2 7.11).42 Probus was a Christian o f noble lineage, whose father and paternal grandfather had both been ordinary consuls (in 341 and 322, respectively).43 Shortly after hold­ ing the ordinary consulate himselfin 371 (not recorded by Ammianus), Probus married the great heiress Anicia Faltonia Proba, thus becoming by default the head o f the clan o f the Anicii: hence he was not only nobilitatis culmen, but more specifically Anicianae domus culmen, a man whose power Ausonius proclaimed in 374 to be greater than that o f anyone except the three reigning Augusti.44 Probus, who came from Verona, was lauded in his lifetime by the grateful inhabitants o f the province Venetia and Histria for his learning, eloquence, skillful administration, moderation and devotion to the welfare o f those whom he governed.45 And after his death, Claudians panegyric on his young sons, who held the ordinary consulate together in 395, secured his reputation for posterity.46 In the chorus o f praise for Probus, there are only two dissenting voices. One is Jerome, whose Chronicle noted that as prefect o f Illyricum in 372 Probus “ destroyed the provinces which he governed by the most unfair 42 For Probus’ career, see P L R E 1.7 3 6 —40, Probus 5; Alan Cameron, J R S 75 (1985), 16 4 —82; on the date o f his death, J. F. Matthews, Historia 16 (1967). 388; Ammianus (1989), 22, 480 n. 27; T. D. Barnes, A JP 1 1 1 (1990), 4 18 : on Ammianus’ portrait o fh im , W. Seyfarth, Klio 52 (1970), 4 1 1 - 2 5 ; Drexler, Ammianstudien (1974), 6 5 - 7 8 ; M . B. M cC oy, Ancient World 11 (1985), 1 0 1 - 6 .

43 Averil Cameron, The Later Roman Empire

a

d

.

284-430 (Cambridge, Mass., 1993), 73,

bizarrely turns Probus into a “ pagan aristocrat” so that she can argue that Ammianus was “ rel­ atively unconcerned about religious matters, capable o f being equally scathing about pagans and Christians alike.” 44 IL S 126 5, 126 7; Ausonius, Ep. 1 6 .2 .1 6 —18. 45 IL S 1265 (Verona). 46 Excellently edited and elucidated by W . Taegert, Claudius Claudianus: Panegyricus dictus Olybrio et Probino consulibus (Zetemata 85, 1988), cf. A JP 1 11 (1990), 4 1 4 —19.

[IW ]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

exaction o f tribute before they were devastated by the barbarians.” 47 The other is Ammianus, who introduces Probus at great length on his appointment as praetorian prefect o f Illyricum to replace Vulcacius Rufinus, who died in office in 368: P r o b u s w a s k n o w n fo r h is h ig h b ir t h , p o litic a l in flu e n c e a n d v a st r ic h e s to th e R o m a n w o r ld , in a lm o s t e v e r y p a rt o f w h ic h h e o w n e d w id e l y sc a tte re d estates.

(I

re fra in fr o m e x p r e s s in g an o p in io n o n w h e t h e r th e y w e r e a c q u ir e d h o n e s t ly

o r n o t .) 4H S o m e as it w e r e d o u b le F o r t u n e , as th e p o e ts d e p ic t h e r, c a r r ie d h im o n h e r s w ift w in g s a n d d isp la y e d h im n o w as a d o e r o f g o o d w h o p r o m o t e d frie n d s to h ig h o ffic e , b u t s o m e tim e s as a d a n g e ro u s s c h e m e r w h o w r o u g h t h a rm b e c a u s e o f b lo o d th ir s t y g ru d g e s. A lt h o u g h h e e n jo y e d e n o r m o u s p o w e r as lo n g as h e liv e d , b y b e s t o w in g gifts a n d h o ld in g o n e o ffic e a fte r a n o th e r w it h o u t in ­ t e r r u p t io n , h e w a s a lte rn a te ly tim id to w a rd s th o se w h o re siste d a n d h a u g h ty t o ­ w a rd s th e t im id ,49 so th at w h e n h e fe lt c o n fid e n t, h e h e c to r e d in th e e le v a te d sty le o f tr a g e d y b u t, w h e n e v e r h e w a s fe a rfu l, h e su rp assed a n y c o m ic a c t o r in h is g r o v e lin g . A n d ju s t as th e ra c e o f fish es c a n n o t b re a th e fo r lo n g o n d r y g r o u n d , so h e b e g a n to d r o o p w it h o u t th e p re fe c tu re s w h ic h h e w as c o m p e lle d to h o ld b y th e q u a rre ls o f h is c o u n tle ss d e p e n d e n ts w h o w e r e a lw a y s e n g a g e d in c r im e b e c a u s e o f th e ir im m e n s e g r e e d an d w h o im m e r s e d th e ir m a s te r in p u b lic life so th at th e y m ig h t a c h ie v e th e ir m u ltifa r io u s en d s w it h im p u n ity . ( 2 7 . 1 1 . 1 - 3 )

Ammianus barely conceals his hatred o f Probus. He questions whether his vast inherited wealth {patrimonia sparsa) was honestly acquired or not: the phrase non iudicioli est nostri might sound judicious, but it raises gratuitous suspicions. Out o f office Probus was like a fish out o f water gasping its life away; in office, he lavished benefits on his criminally rapacious dependents. Ammianus could trust his readers to identify the familiae ingentes o f Probus as the Christian Anicii, whom he had blamed for the corruption that permeated the body politic in 47 Jerome, Chronicle 2 4 6 1 Helm. Some manuscripts substitute the name o f Equitius, consul in 374, who was magister militum in Illyricum while Probus was prefect ( P L R E 1.282, Equitius 2), for that o f Probus himself: the change is both deliberate and from a contemporary hand, though perhaps not from Jerome himself, cf. T. Mommsen, G es. Sehr. 7 (Berlin, 1909), 6 0 4 -0 5 . 48 The precise nuance o f the words iuste an secus non est iudicioli nostri is hard to define. Rolfe s “ whether justly or unjustly is not a question for m y humble opinion” certainly comes closer than Hamilton’s “ whether they were honestly come by or not is not for a man like me to say.” But it substitutes a cliché for the pretematurally rare iudiciolum, for which T L L 7.2.606 registers only this passage and the Fragmentum Bobiense de nomine et praenom ine (H. Keil, Grammatici L a ­ tini, 5 [Leipzig, 1868], 56 1.36 : absurde meo iudiciolo).

49 The transmitted text reads: “ interdum timidus ad audaces, contra timidos celsior” (27 .11.2 ). Since Ammianus is describing the typical behavior o f an arrogant bully, some correlative to in­ terdum (e.g., modo) probably needs to be supplied before contra , cf. T L L 7 .2 18 1.

[n s]

E

n e m i e s

, A

n i m a l s

,

a n d

S

t e r e o t y p e s

his own day (16.8.13). Ammianus mentions Probus himself only to besmirch his name.50 On the one hand, he omits Probus’ role in the proclamation o f the infant Valentinian in 375 (30.10.5).51 On the other, he emphasizes his role in the torture and execution o f Faustinus, the nephew o f Viventius, the “ hon­ est and wise Pannonian” (27.3.11), which he regarded as unjust (30.5.11 —12: Probo spectante negotium).52 Against this background, it must be suspected that, like so much else in the Res Gestae, Ammianus’ portrait o f Q. Clodius Hermogenianus Olybrius is subtly and deliberately polemical. His prefecture, Ammianus acknowledges, was quiet and peaceful (28.4.1—2). Olybrius was humane, concerned lest any word or act o f his be found harsh. He checked slander, pruned the profits o f the treasury, distinguished between right and wrong, and performed his duties ca­ pably and with mildness. These good qualities, however, were overshadowed by one grave fault, which, though harmless in itself, was discreditable in the holder o f a high office. Olybrius enjoyed luxury and devoted himself almost entirely to the stage and to love affairs, though not o f a criminal or incestuous nature. Since Olybrius was known to be a Christian,53 his luxus is likely to rep­ resent for Ammianus the truphe with which Greek-speaking pagans taunted Christians (Chapters VIII, XIV). 50 S e y fa rth , K lio 52 (1970), 4 2 1 - 2 5 . 51 Rufinus, H E 1 1 . 1 5 , records Probus’ role. It was illegal, and usually treasonable, to proclaim a new emperor without consulting any o f the surviving members o f the imperial college: J. Straub, Vom Herrsckerideal in der Spätantike (Berlin, 1939), 1 8 —20; Studien z u r H istoria A ugusta (Disserta­ tiones Bernenses 1.4, 1952), 1 4 4 - 4 5 .

52 O n this passage, see J. M o r e a u , A n n u aire de l'Institut de Philologie et d ’Histoire orientales 13 (19 53), 4 2 3 - 3 1 ; W . S e y fa rth , K lio 46 ( 1 9 7 5 ) . 374 ~ 78 M Collectio A vellan a 8 —10 ( C S E L 3 5 .5 0 —52); Prudentius, Contra Sym m achum 1 . 5 5 4 - 5 5 7 — a passage perhaps written c. 385, cf. T. D . Bames and R. W . Westall, Ph oen ix 45 (19 9 1), 5 8 - 6 1 .

[119]

[XI] EMPRESSES AND EUNUCHS

The Res Gestae move entirely in a mans world, and their author held entirely conventional views about the place o f women in it and about how they ought to behave. Modern studies have reconstructed the historian’s picture o f women and analyzed their political role, and one is promised on their social and moral significance.1 The result is distressingly predictable. Ammianus’ attitudes to­ ward women are so conventional for his time that they reveal little or noth­ ing about him as an individual. Yet something interesting may emerge if it is asked which women he chooses to include and which to omit in a definable category. For, just as with the male characters in his history (Chapters X , XII, XIII), Ammianus reveals his personal likes and dislikes without inhibition when dealing with the wives o f emperors. Constantina, a daughter o f Constantine and the wife o f the Caesar Gallus, is introduced at the start o f Book XIV, where Ammianus describes Gallus’ be­ havior as emperor in Antioch.12 The historian speaks o f her with undisguised hostility as one who incited her husband to cruelty and excess. Previously married by Constantine to his nephew Hannibalianus,3 she had the swollen pride o f one born into the imperial family: she was a “ Fury in mortal form, incessantly adding fuel to her husband’s rage, and as thirsty as he for human blood” (14. i . 2). Hence the obstinacy o f Gallus was increased by the influence 1 L. Jacob-Karau, Das Bild der Frau in den “Res Gestae” des Ammianus Marcellinus (Diss. Berlin, 19 71); G. Sabbah, Cognitio Gestorum (1992), 9 1 - 1 0 5 . 2 O n the political importance o f Constantina, especially in the crisis provoked by the usurpa­ tion o f Magnentius in 350, see B. Bleckmann, Chiron 2 4 (1994), 2 9 - 6 8 . 3 The marriage is also recorded in E at. Val. 1.35 ; Philostorgius, H E 3.22.

[ 1 20]

E

m p r e s s e s

a n d

E

u n u c h s

o f the empress,4 who was pushing her husband headlong to destruction, al­ though she ought to have tried “ by feminine mildness and sensible advice to bring him back to the way o f truth and humanity” (14.1.8). After the opening chapter o f Book XIV, Constantina disappears from Am­ mianus’ narrative until she dies o f a sudden fever in Bithynia while traveling to the court o f Constantius in the hope o f mollifying him (14 .11.6 ). But she reappears after her death in a significant context. According to Ammianus, Constantina was so deeply implicated in the misgovernment o f her husband that, when Gallus was interrogated after his deposition, he claimed that he had killed most o f his victims at her insistence. Ammianus disallows the plea with reference to a story about Alexander the Great, who refused to kill an inno­ cent man at the urging o f his mother (14 .11.22). But Alexander and his mother provide no real parallel, even by opposition. At the back o f Ammianus’ mind, it may be suspected, lie two biblical images. The most famous occasion when a man blamed his wife for his own transgression occurred in the Garden o f Eden: when questioned by God, Adam blamed Eve with the excuse that he had disobeyed him because the woman offered him the fruit o f the tree.5 But the words o f Alexander’s reply to Olympias (“ ask for another reward, my good mother; no favor can equal the life o f a man” ) evoke rather the execu­ tion o f John the Baptist to please Herodias.6 Eusebia was the second, or more probably the third, wife o f Constantius.78 She makes a number o f significant appearances in the Res Gestae* After the fall o f Gallus she protectedjulian (15.2.8), then opposed a proposal to send him to Gaul without the rank o f Caesar (15.8.3). Later, because o f her own steril­ ity, she induced Julian’s wife to drink a potion that ensured that she would mis­ carry whenever she conceived (16.10.18). Constantius renamed the diocese o f Pontus in her honor in 358 (17.7.7). She was the object o f jealousy from the wife o f Barbatio, who feared that if her husband became emperor he would abandon her for Eusebia who was a woman o f outstanding physical beauty (18.3.2). Finally, in the context o f Constantius’ subsequent marriage 4 Here, as elsewhere when speaking o f Constantina ( 14 .1.3 , 9.3) and Eusebia (15.2.8 , 8.3, 16 .10 .18 , 18 .3.2), Anunianus uses regina as the Latin equivalent o f basilissa, which had long been the normal Greek word for “ empress.” 5 Gen. 3 .12 . 6 Mark 6 . 1 4 - 2 9 ; Matthew 1 4 . 1 - 1 2 . 7 P L R E 1.3 0 0 —30 1. Constantius’ first wife, whom he married in 335 , was a daughter o f Julius Constantius, w ho was assassinated in 3 3 7 (Eusebius, V C 4.49; Athanasius, Hist. Ar. 6 9.1; Julian, Ep. ad Ath. 2726): since an emperor needed heirs, it is unlikely that he remained unmar­ ried for about fifteen yean until his marriage to Eusebia circa 35 2 (Julian, Orat. 3, ioçab, 1 lod). 8 T h e differing depictions o f Eusebia in ancient writers are compared by N . Aujoulat, Byzan-

tion 53 (1983). 7 8 - 1 0 3 , 4 2 1 - 5 2 .

[ 121]

E

m p r e s s e s

a n d

E

u n u c h s

to Faustina, Ammianus notes the death o f Eusebia, “ the sister o f the ex­ consuls Eusebius and Hypatius, a woman o f outstanding beauty and excellence o f character, whose kindness o f heart was unaffected by her high rank,” who had rescued Julian from danger before he was proclaimed Caesar (21.5.3). Eusebia was not only the protector o f Julian, but also the sister o f the man who was later Ammianus’ friend and patron. Hypatius was (it seems) the son o f Eusebius, a general who held the ordinary consulate in 347, but died be­ fore Constantius married his daughter.910Hypatius and his brother Eusebius, now the emperor’s brothers-in-law, were ordinary consuls together in 359.'° A dozen years later, while living in Antioch, the brothers were condemned for treason and exiled, though later pardoned. Ammianus praises Hypatius most warmly: A m o n g all th ese m y fr ie n d H y p a t iu s ( noster H y p a tiu s) s to o d o u t, w h o d e s e rv e s p ra ise fo r h is y o u t h fu l v i g o r a n d n o b le v ir tu e s . H e w a s a m a n o f c a lm a n d c o o l ju d g e m e n t , w h o s e h o n o r a b le a n d m ild c h a r a c te r n e v e r d e v ia te d f r o m th e u p r ig h t. H e a d d e d g lo r y to th e s p le n d o r o f h is a n c e s to rs a n d h im s e lf c o n fe r r e d h o n o r o n his d e sce n d a n ts b y his a d m ira b le a ctio n s in t w o p re fe ctu re s. ( 2 9 . 2 . 1 6 )

When Hypatius was appointed prefect o f the city o f Rome during the difficult winter o f 378—379, he was still in Antioch.11 But he must have gone to Rome to hold office, and, after a visit to Constantinople in 3 8 1,12 perhaps to pay his respects to Theodosius, he became praetorian prefect o f Italy and Illyricum in 3 8 2 -3 8 3.13 It was surely his friend Hypatius who encouraged Ammianus to come to Rome and who introduced him to Roman society:14 in what has al­ ways been regarded as a transparent piece o f autobiography (14.6.19), Ammi­ anus reveals that he was initially welcomed, but then scorned, presumably be­ cause he could not win social acceptance for himself, although he was the protégé o f an eastern nobilis. Ammianus’ praise o f Eusebia is due to his friend­ ship with her brother as much as to her help ofjulian when he was under sus­ picion in 354—355. 9 P L R E i. 3 0 7 - 8 , Eusebius 39. 10 For their careen, see P L R E 1 .3 0 6 - 7 , Eusebius 40; 4 4 8 —49, Hypatius 4. However, A m ­ mianus’ description o f them as patriciatus columina (29.2.9) does not suffice to prove that they were created patricii. 11 Libanius, Orat. 1 .7 9 - 8 0 . H e is attested in office on 5 April 379 ( C T h 11.36 .26 ). 12 Gregory o f Nazianzus, E p . 96. 13 For the evidence relating to this prefecture, see P L R E 1.449: it is securely attested from 9 December 382 ( C T h 1 1 .1 6 .1 5 ) to 28 M ay 383 ( C T lt 2.19.5). 14 S. Jannacone, A m m ian o Marcellino. Profilo storico-critico (Naples, 1970), 25. Ammianus could in theory have come to Rome with Hypatius in 3 7 8 - 7 9 and accompanied him on his visit to Constantinople in 38 1.

[122]

E

m p r e s s e s

a n d

E

u n u c h s

In contrast, the wife o f Ammianus’ hero Julian receives no praise at all. A few days after his proclamation on 6 November 355, the new Caesar received Helena “ the sister o f Constantius” in marriage (15.8.18). In 357 Helena went to Rome, where she suffered a miscarriage, having previously given birth in Gaul to a son whom the midwife (so it was alleged) had been bribed to kill by cutting the umbilical cord (16 .10 .18 -19 ). In 360, when Helena died, Julian sent her body to be buried on the Via Nomentana just outside Rome next to her sister Constantina (21.1.5). That is the sum o f what Ammianus has to say about Helena, but he notes that after her death Julian never had sexual rela­ tions with any other woman (25.4.2).15 Ammianus’ account o f the year 361 registers Constantius’ marriage to Faustina after the death o f Eusebia (21.6.4). Moreover, by the time the em­ peror left Antioch in November, Ammianus notes, his new wife was preg­ nant: the daughter bom after his death subsequently married Gratian (21.16.5). Ammianus never alludes to the wife o f either Jovian or Valens, although he names the father-in-law and the infant son o f both.16 Charito and Domnica exerted no political influence, and the name o f Jovian’s wife is first attested in writers o f the ninth century.17 But Ammianus records the attempt o f Pro­ copius in 365 to use the posthumous child o f Constantius to lend a spurious legitimacy to his usurpation.18 Since the widowed Faustina and her infant daughter happened to be in Constantinople, the usurper carried the child, then three years old, in his arms as he proclaimed his kinship to Constantius and Julian (26.7.10, 9.3). As soon as Constantia reached puberty, she was married to Gratian: Ammianus notes the marriage only because the bride was almost captured by barbarians as she was being taken to Gaul for her wedding (29.6.7). Valentinian is unique among Christian emperors o f Rome in that he married for a second time while his first wife was still alive. Christian writers found that embarrassing, and Valentinian’s first wife is named by only four authors o f late antiquity, who form a very motley collection. Three late chronicles (John Malalas, the Paschal Chronicle, and John o f Nikiu) call Valentinian’s first wife Marina and allege that she was banished for participating in a fraud­ 15 Julian refers several times in the Misopogon to sleeping alone and to his lack o f interest in sex (340c, 34_scd, 348d, 367b). On Ammianus' praise o f his castitas, see Chapter VIII, at nn. 2 6 - 2 9 . 16 Lucillianus, whom Jovian appointed magister equitum et peditum (25.8.9, 1 0 .11, cf. P L R E 1 . 5 1 7 - 1 8 ) , and Varronianus, consul with his father in 364 ( 2 5 .10 .11, 17); and Petronius, whom Valens made a patricius (2 6 .6 .7 -9 , P L R E 1.6 9 0 - 9 1 ) , and Valentinianus, nicknamed Galates, who was born in 366, consul in 369 and died in 3 7 1 (P L R E 1.38 1). 17 Nicephorus, Chronicon compendiosum 10 4 .14 de Boor, translated by Anastasius Bibliothecarius, Chronographia tripertita 47.6 de Boor. Domnica, in contrast, is named by the ecclesiasti­ cal historians o f the fifth century (P L R E 1.265). 18 O n Procopius’ desperate search for legitimacy, see N . J. E. Austin, R w ista storica d e ll’antichità 2 (1972), 1 8 7 -9 4 .

[ 123]

E

m p r e s s e s

a n d

E

u n u c h s

ulent property deal.19 The story looks like an attempt to exculpate Valentinian from any blame attached to discarding her, although it could contain ele­ ments o f truth. For the ecclesiastical historian Socrates, writing about 440, calls Valentinian’s first wife Severa and offers a very different version o f why Valentinian married Justina “ even though his former wife Severa was still alive.” Valentinian (he reports) fell in love with Justina after Severa introduced her to her husband, then married her without expelling Severa from the palace. Moreover, he issued a law, which he published in every city, allowing anyone who wished to have two lawful wives.20 The late chronicles have usually been preferred to Socrates and his version o f events dismissed as unbelievable.21 But Socrates does not say that Valentin­ ian issued a law permitting bigamy, as has sometimes been assumed:22 he not only fails to say that Valentinian’s law allowed a man to marry two women at the same time, but he also explicitly calls Severa the emperor’s “ former wife.” What Socrates says is that Valentinian issued a legal ruling that permit­ ted him (and incidentally others) to remarry after divorce.23 On general grounds, Socrates deserves preference over much later writers. Moreover, the legal bar to his marriage to Justina that Valentinian needed to remove can be identified.24 There was no impediment to the marriage on the bride’s side. Justina was the daughter o f one Justus, who had governed Picenum under Constantius.25 She has been convincingly identified as a descendant o f Constantine,26 and she had been married to Magnentius as a young girl, so that she had been a widow since 353.27 Valentinian may have been more interested in her imper19 Malalas 3 4 1 . 1 - 7 Bonn; Chronicon Paschale 5 5 9 .7 —13 Bonn; John o f Nikiu, Chronicle , trans. R. H. Charles, p. 82. 20 Socrates, H E 4 . 3 1 . 1 3 —17 , repeated by Jordanes, Rom . 310 . 21 O. Seeck, Geschichte des Untergangs der antiken Welt 5 (Berlin, 19 13 ), 4 3 1; W . Ensslin, R E 14 (1930), 1 7 5 6 —57; A . Nagl, R E 7 A (1948), 2 14 8 . D. Woods, A ncient Society 2 6 (1995), 2 8 1 —82, 288, n. 53, conjectures that Valentinian may have met Justina at the court o f Magnentius. U n­ fortunately, his argument that Valentinian served under Magnentius, but salvaged his career by a timely defection to Constantius rests on the false assumption that Martin o f Tours also served under Magentius because he left the army “ by about August 35 6 at the latest” ( 2 8 2 ,2 8 4 - 8 8 ) : for the correct chronology o f Martin’s military career (enlistment under Constantius autumn 352, discharge autumn 357), see A n a l. Boll. 1 1 4 (1996), 2 5 —32. 22 As by Theophanes 5 6 .2 3 - 3 1 de Boor.

21 A . D . M anfredini, A tti d ell'Accadentia Romanistica Costantiniana 8 (Perugia, 1990), 5 2 2 - 2 8 ; A . A g ava , Women and Lau> in Late A n tiq u ity (O xfo rd , 1996), 18 0 , n. 70. 24 A similar dispensation allowed Justinian to marry the former stage performer Theodora, who was technically infam is : C J 5.4.23 ( 5 2 0 - 2 3 ) ; Procopius, Anecdota 9 .5 1, cf. D . Daube, C ath ­ olic U niversity o f Am erica L a w R ev ie w 16 (19 6 6 —67), 380 —89. 25 Socrates, H E 4 . 3 1 .1 1 - 1 2 . 26 N e w E m pire (1982), 44. 27 Zosimus 4 .1 9 .1, 4 3 .1, cf. P L R E 1 .4 8 8 -8 9 .

[124]

E

m p r e s s e s

a n d

E

u n u c h s

ial pedigree than her beauty or sexual allure, and the recent precedent o f C on­ stantina, who had been married first to Hannibalianus (who was killed in 337), then to Gallus in 351 (14.1.2), indicates that her status as a widow did not constitute a legal impediment to her remarriage. Socrates clearly implies that it was Valentinian’s first marriage that consti­ tuted the impediment to his union with Justina. How could that be? Unfor­ tunately, the section o f the Theodosian Code in which the title De Repudiis occurs has not survived in full, and the Breviarium o f Alaric preserves no law on divorce between a constitution that Constantine addressed to his praeto­ rian prefect Ablabius in 331 and one o f Honorius that modified it in 4 2 1.28 Constantine’s law is very relevant to Valentinian. Constantine ruled in 331 that a man could divorce his wife only for adultery, sorcery, or procuring, and that a woman could divorce her husband only for murder, sorcery, or robbery o f tombs, not for drunkenness, gambling, or whoring. A man who divorced a wife who was not guilty o f any o f the three specified offenses could not re­ marry: if he did, the discarded wife could enter his house and seize all the dowry o f the second wife to compensate for the wrong done to her.29 The law, it appears, said nothing about divorce by mutual consent, which contin­ ued to be valid without restriction until the sixth century.30 But the empress Severa was doubtless reluctant to agree to a divorce that entailed the loss o f her social position. It follows that Valentinian needed to extend the grounds for nonconsensual divorce to accommodate himself. The writer known as Ambrosiaster complains that an edict o f Julian gave women the power to divorce their husbands that they had previously lacked, with the result that they were now (in the 380s) divorcing their husbands every day.3132Before Valentinian, therefore, Julian had relaxed the restrictions on divorce that Constantine had introduced, but Ambrosiaster’s generalized complaint does little to clarify exactly how he modified the law o f 331 and does not exclude the possibility that Jovian partly reinstated Constantine’s law after Julian had repealed it or that Julian allowed women as well as men to di­ vorce their spouses for sexual infidelity. Now Augustine refers to a secular law (lex Imius saeculi) which freely allowed remarriage after divorce (interveniente repudio)?2 It seems hard to deny the relevance o f this observation, which im­ 28 C 77 » 3 . 1 6 . 1 - 2 . 2 uses th e h e ro e s w h o m h e sets b e fo r e us lik e ty p e s in a tr a g e d y in o r d e r to te a ch us th at k in g s o u g h t n e v e r to in flic t in sults, n o r use th e ir p o w e r in d is c rim in a te ly , n o r b e c a r r ie d a w a y b y a n g e r lik e a b o ld h o rse ru s h ­

4

P. Athanassiadi-Fowden, Ju lia n am i H ellenism . A n Intellectual Biography (Oxford, 19 8 1), 1 4 -

22, 2 7 - 2 9 . s Julian refers to his Batavian campaign o f 358 ( O rat . 3(2], 56b, cf. Ammianus 1 7 .8 .3 - 5 ) , but Rome and Persia are still at peace (66 d -6 7a). 6 Julian, Orat. 1 conforms to the prescriptions for a basilikos logos in the rhetorical handbook attributed to Menander: W . C . Cave (Wright), T h e Em peror Ju lia n 's Relation to the N e w Sophistic and N eo-Platonism , with a Stu d y o f H is Style (Diss. Chicago, pub. London, 1896), 2 6 - 2 9 ;

C . Gladis, D e Them istii L ib a n ii Ju lia n i in Constantium orationibus (Diss. Breslau, 1907), 2 0 - 2 9 ; I. Tantillo, L a prim a orazione di G iu lian o a C ostan zo (Rome, 1997), 1 6 - 3 1 , w ho argues convinc­ ingly that Julian composed the speech for Constantius’ visit to Rome in 3 5 7 ( 3 6 -5 0 ). 7 Bowersock, Ju lia n (1978), 4 3 - 4 4 ; Athanassiadi-Fowden, Ju lia n (19 8 1), 6 3 - 6 6 . K Julian, Orat. 3(2), 4 9 0 -5 0 3. M y translations from Julian are based on the Loeb version by Wilmer Cave W right (London/Cam bridge, Mass., 19 13).

[144]

T

he

N

e w

A

c h il l e s

ing along without bridle or rider, and to advise generals not to resent the contempt of a king, but to endure his censures with self-control and serenity, so that the life of both may not be filled with remorse.9 The very first sentence o f the speech is very revealing in another way, too. Julian explicitly invokes Achilles, who in his wrath put aside spear and shield to praise the deeds o f the demi-gods, just as in his winter quarters the Caesar has turned aside from war to praise Constantius. At a deeper level, however, Julian’s language evokes a still more provocative comparison. Achilles took up his psaltcrion and his cithara. Now Julian never uses the word psaltcrion elsewhere, and the combination o fpsaltcrion and cithara is found nowhere in Greek literature except in the Septuagint and in authors who are quoting from or alluding to the Greek Bible.101Julian had read both the Old and the New Testament either entire or at least in large part.11 Moreover, his allusion is probably a very precise one. The phrase “ psaltery and harp” occurs in a passage that appears in two psalms attributed to King David.12 The head­ ing to the earlier o f the two psalms in which it occurs states that David com­ posed the poem while he was hiding in the cave to escape king Saul. Although the headings are neither original nor authentic, they were accepted as accu­ rate and authoritative by Christians o f the fourth century, who used them as hermeneutical guides for interpreting the text.13 I f Julian subconsciously thought o f himself as David pitted against Saul, the implications were dan­ gerous— and treasonable. Moreover, whether deliberately or not, several pas­ sages o f the speech reveal its author to be a pagan and a Neoplatonist who be­ lieved that the true ruler should be the prophet and servant o f “ the king o f the gods.” 14 Unlike Saul, Constantius died in his bed. His last act was one o f ultimate statesmanship: he named Julian his heir and successor. Ammianus subtly calls 9 Julian, Oral. 3(2), 5obc. 10 E.g., Josephus, A J 1.64, in reference to Genesis 4 .21. In that passage, however, the Hebrew text has 31W1 H 3D, i.e., “ harp and pipe” (N EB ): in the passage from the Psalms adduced here (n. 12), the Hebrew has 11331 *?3 Jn, where both nouns do indeed refer to stringed instruments. 11 See the quotations, allusions, and parallels noted by R. Asmus, Julians Galiläerschrift ini

Zusammenhang mit seinen übrigen Werken (Prog. Freiburg i. B., 1904), 8 - 3 9 ; E. Masaracchia, Giuliano Imperatore: Contra Galilaeos (1990), 3 4 9 - 5 1 , 3 7 7 - 7 9 , 3 8 9 - 9 1 . 12 Psalm 57 (56). 7 - 1 1 = 108 (107). 1 - 5 . 13 Gregory o f Nyssa wrote a whole treatise In Inscriptiones Psalmorum, ed. J. M cDonough, Gregorii Nysseui Opera 5 (Leiden, 1962), 1 - 1 7 5 . For fourth-century application o f the heading o f Psalm 57(56) to exegesis o f the psalm, see Eusebius, P G 2 3 .5 0 1 - 0 8 ; Athanasius, P G 2 7 .2 5 7 —

6 1; Gregory o f Nyssa, Insc. Ps. 2 .1 4 (15 4 .15 — 15 8 .21). 14 Oral. 3(2), 90a, cf.J. Bidez, L'EmpereurJulien: Oeuvres complètes 1.1 (Paris, 1932), 1 1 3 —15, who detected pagan notions o f ritual purity and Neoplatonic ideas o f the great chain o f being in the speech ( 7 o d -7 ia , 8 o a -c , 8id, 8 2 d -8 3 a , 9obc).

[145]

T

he

N

e w

A

c h il l e s

this disinterested and patriotic act into question by representing it as merely an uncertain rumor (22.15.2, 5). As a result, many modern historians o f the fourth century, including the most critical, have dismissed the rumor as base­ less.15 But that is to place too much trust in a biased historian. The dying Constantius decided to prevent civil war. What wiser action could he take be­ fore facing his creator? Ammianus’ denial (for that is what his words convey) is an extreme example o f his partisanship for Julian, although paradoxically also o f his independence o f judgment. Once Constantius was safely dead, Julian dropped the overt hostility that he displayed in the summer and autumn o f 361 and saluted him as “ the blessed Constantius.” 16 For to continue to de­ nounce him as a tyant would have cast doubt on the legitimacy o f both his initial appointment as emperor and the dying Constantius’ conferral on him o f the title o f Augustus. Libanius employed comparison to Achilles to characterize Julian’s treament by Constantius both while he was in Antioch in 362 and after his death.17 As a skilled orator, Libanius knew how to quote Homer for his own purposes. After Julian left Antioch in anger at the city, he composed a plea for recon­ ciliation: everyone (he stated) was talking about the Persian expedition, not about the Trojan War or Alexander, but the wrath o f Achilles needed urgently to be assuaged.18 Explicit comparison to Achilles was always a double-edged compliment to an emperor who was impetuous and prone to anger. Hence it plays a negligible part in Julian’s conception o f himself as Augustus, in his propaganda while he ruled and in retrospective panegyrics after he died. Yet implicit evocation o f Achilles often lurks beneath the surface o f Ammianus’ text. When he reached Ctesiphon, Julian drew up his battle line secundum Homericam dispositionem and there ensued a display o f valor comparable to the combats o f Hector, the brav­ ery o f Achilles and the crowning glories o f the Persian wan o f the early fifth century b . c . (23.6.9—14). Ammianus’ obituary commences by stating that Julian was “ vir profecto heroicis connumerandus ingeniis, claritudine rerum et coalita maiestate conspicuus” (25.4.1). The Latin conveys much more than “Julian must be reckoned a man o f heroic stature, conspicuous for his glori­ ous deeds and his innate majesty” (as Hamilton and Wallace-Hadrill render 15 Bowersock, Julian (1978), 65. 16 Julian, Epp. 33, 3^9d“ 39oa; 40 (45.27 Bidez-Cum ont); 59, 443b: contrast Ep. ad Ath.. 2 7 0 c -2 7 4 b , where Julian depicts him as a murderer. However, official canonization as divus Constantius is neither celebrated on the Roman imperial coinage nor, apparently, attested by any inscription. 17 Libanius, Orat. 12.49, 73; 13.6; 18.66, cf. 32. ,H Libanius, Orat. 15 .1, 8, 3 1 - 3 5 .

[146]

T

h e

N

e w

A

c h il l e s

it). Ammianus is not using a dead metaphor: he means that Julian belongs with the heroica ingenia, the heroes o f great talent.19 He is thinking primarily o f Achilles, whose glorious deeds ultimately failed to bring him success because o f a fatal weakness. The concealed image o f Achilles, it may be suggested, underlies the whole o f Ammianus’ account o f Julian. After the public cere­ mony at which he was proclaimed Caesar on 6 November 355, Ammianus’ Julian goes with Constantius to the palace in Milan constantly whispering to himself a line o f the Iliad: “ wrapped in death’s purple by all-powerful fate.” 20 Seven and a half years later, as Julian willfully disregards the warnings sent by the gods as he marches into Mesopotamia, his blindness and folly are those o f a tragic hero.21 Julian himself proclaimed two famous rulers as his principal models. They are Alexander and Marcus Aurelius. Julian names them together at the start o f his letter to Themistius, which he probably wrote in 3 56,22 and the same pairing is implicit in the Caesars, which he composed in Antioch in December 362. This work offers an assessment o f Roman emperors from Julius Caesar to Constantine and his sons. Julian selected the literary form o f a dialogue de­ scribing a banquet that Romulus gave for the gods to celebrate the Saturnalia. This allowed him both to omit emperors whom he did not wish to discuss and to introduce Alexander as one o f the competitors for the title o f best em­ peror.23 The winner o f the contest is Marcus Aurelius, who receives the ma­ jority o f the gods’ votes in a secret ballot.24 However, after the Roman emper­ ors have been welcomed and seated, but before the competition commences, Heracles insists that Alexander be invited as superior to all o f them,25 which might seem to suggest that he is superior to Marcus. For Heracles had a spe­ cial meaning for Julian, who proclaimed that he was the savior o f the world 19 Apart from using heroici versus o f the poems o f the Gallic bardi (15.9.8), Ammianus restricts the epithet heroicus to mythical heroes in general (28.4.12) or to individuals like Mopsus (14.8.3) and the heroes w ho fought at Troy (22.8.3).

20 Iliad 2.10 4 . 21 Rosen, Studien (1970), 14 9 —78, cf. G . W irth , Julianus Apostata (1978). 4 5 5 —07. 22 Julian, Ep. ad Them. 253a. For the hypothesis that Julian composed most o f the letter in 356 , but completed and sent it for his own political advantage between the spring o f 360 and Novem ber 3 6 1, see T. D. Barnes and J. Vanderspoel, G R B S 22 (19 8 1), 1 8 7 - 8 9 . Further argu­ ments for dating the composition o f the letter to the winter o f 3 5 5 —356 are produced by R. Smith, Julian’s Gods: Religion and Philosophy in the Thought and Action ofJulian the Apostate (L o n d o n /N ew York, 1995), 2 7 - 2 9 . S. Bradbury, G R B S 28 (1987), 2 3 5 - 5 1 , argues that Julian wrote the whole letter then, including its final page (266c—267b). 23 G . W . Bowersock, Y C S 2 7 (1982), 1 5 9 - 7 2 . 24 Julian, Caes. 335c. 25 Julian, Caes. 3i6 b c.

T

h e

N

e w

A

c h il l e s

and that he walked on water as though it were the dry land.26 Since the time o f Dio Chrysostom, who assimilated Trajan to him, Heracles had stood as a model for Roman emperors, who imitated their divine avatar by strenuous and unceasing efforts to better the lot o f mankind:27 hence the short-lived dy­ nasty o f Herculian emperors in the days o f Diocletian.28 But there is an un­ usual warmth and an unusual polemical edge to what Julian says about Her­ acles, whom he makes the son o f Athena, a virgin mother on the model o f Mary, the mother o f God.29Julian goes on, however, to emphasize Alexander’s cruelty, drunkenness and lack o f self-control: Alexander could not, therefore, serve as his exemplar except on the actual field o f battle.3" After Julian’s death, comparison with Alexander, like comparison with Achilles, became a double-edged compliment. Like Alexander, Julian won glo­ rious victories in war and died young. But, whereas Alexander had defeated the Persian king and conquered the Achaemenid Empire, Julian failed and perished at the head o f an army in retreat. Ammianus distracts his readers from this damaging comparison by a series o f other comparisons to great generals o f Greece and Rome. Book X V I opens with a miniature panegyric o f the new Caesar. He was a second Titus in wisdom,31 most like Trajan in his military successes, clement like Antoninus Pius, and “ in his striving after truth and perfection the equal o f Marcus Aurelius, on whom he endavoured to model his own actions and character” (16.1.4). Later, as he prepares for war against Persia, Julian encoun­ ters criticism, but no more lets himself be deflected from his purpose than Hercules did when assailed by the Pygmies or the wild Lindian Thiodamas (22.12.4). In the course o f his narrative, Ammianus compares Julian to a wide range o f historical exemplars. The earliest in time is the Athenian general Cimon who, despite his victory over the Persians at the River Eurymedon, was— a lacuna has presumably removed a reference to his ostracism (17 .11.3 ). 26 Julian, Orat. 7, 219c! (from Julians attack on the C ynic Heraclius, composed in Constan­ tinople in the winter o f 3 6 1 —362). 27 R. Hoi'stad, C yn ic Hero am i C yn ic K ing. Studies in the C yn ic Conception o f M an (Uppsala, 1948), 5 0 - 7 3 ; C . R Jones, The World o f D io Chrysostom (Cambridge, Mass., 1978), 1 1 6 - 2 0 . 2M Maximian ( 2 8 5 -3 0 5 ), Constantius (2 9 3 -3 0 6 ), Severus ( 3 0 5 -0 7 ) and Constantine, until he abruptly dropped the title after his conversion (N ew Em pire [1982], 24, cf. Lactantius, Mort. Pers. 52.3).

29 Julian, Orat. 7, 220a, cf. Orat. 5, i66ab. Th e assimilation o f Heracles to Jesus was acutely noted by Asmus, Ju lia n s Galiläerschrift (1904), 3 4 - 3 5 . 30 Julian, Caes. 32 1c , 3 3 0 C - 3 3 1 C . On Julians necessarily ambiguous attitude toward Alexan­ der, see now R. J. Lane Fox, C Q , N .S. 47 (1997), 2 4 7 - 5 2 . 31 Although Titus' high reputation was ensured by the praises o f both Tacitus and Suetonius, Julian disapproved and presented him as a promiscuous playboy: Caes. 3 11a , cf. J. F. Gilliam, A J P 88 (1967), 2 0 3 - 8 ; Bowersock, Y C S 27 (1982), 167.

[148]

T

h e

N

e w

A

c h il l e s

Next in time is the Theban commander Epaminondas who was fatally wounded in 362 b . c . at the Battle o f Mantinea (25.3.8). When Ammianus compares Julian to Alexander the Great, it is not for victories in battle. Like Alexander, Julian divided the night into three equal parts for rest, for business, and for the Muses, but unlike Alexander he needed no artificial devices to pre­ vent him from falling asleep (16.5.4). And like Alexander and many other skilled generals, Julian sent ahead a few troops spread out to create the im­ pression that they formed a large army (21.8.3). Alexander and Scipio both declined to touch beautiful captive girls who were offered to them for their pleasure. So, too, did Julian, who avoided even lusting after them.32 Scipio Aemilianus is invoked twice. He was accused o f sleepiness, although his vigi­ lance destroyed two most powerful cities (17 .11.3 ). And he acted bravely, with Polybius at his side, during the siege o f Carthage (2 4 .2 .16 -17 ). Even Sulla is brought in. During the Battle o f Strasbourg, Julian imitated Sulla, who, when deserted by his men in battle against Archelaus, strode forward, seized the standard, and advanced against the enemy (16 .12.4 1). Like Julian, Pompey was unfairly criticized, despite his bravery and patriotism, even though only two utterly trivial complaints could be brought against him (17 .11.4 ). And Julius Caesar provided a model for writing in a tent during the night (25.2.3). This list contains some unexpected names. From Sallust onward, Sulla was remembered more for slaughtering his political opponents than for his victo­ ries over the enemies o f Rome.33 So, too, was Marius, whose victory over the Cimbri and Teutones Ammianus attributes to unnamed “ excellent generals” (31 . 5.12: per duces amplissimos superati). Ammianus names Marius twice. One mention is conventional and unfavorable: Marius and Cinna allowed the plebs o f Rome to loot the houses o f the proscribed (30.8.9). The other is most surprising and relates to Julian. Ammianus includes both men among the select company o f famous Romans who had genii familiares (21.14.5): three o f the other names are predictable (king Nunia, the elder Scipio and Augustus), but the inclusion o f Marius is puzzling. Despite the valiant efforts o f the Dutch commentators, the precise inpiration o f Ammianus’ list is uncertain. But the passage reinforces the comparisons o f Julian to famous generals, which set Julian as a military commander in a long and noble tradition, both Greek and Roman. -12 O n the implications o f this observation, see Chapter V III, at nn. 2 6 - 2 9 . -u Sallust, Cat. 5.6, 1 1 . 4 —5, 2 ^-4. 37.9, 5 1 . 3 2 - 3 4 ; Hist. 3.49.9 Maurenbrecher. W hen Virgil lists victorious Roman generals, he includes Marius, but not Sulla (Georgies 2 .1 6 9 - 7 0 ) . For Sulla as an exemplum o f cruelty: Pliny, Nat. Hist. 7 .1 3 7 ; Suetonius, Tib. 59.2; Pan. Lat. 12(9).20.4, 2 1 . 1 ; 2 (i2 ).4 ö .i; Claudian, Ruf. 1.26 3. hi view o f his reputation, the emendation Sullas in Pan. Lat. 2 (i2 ).7 .4 must be regarded as questionable.

[149]

T

he

N

e w

A

c h il l e s

Ammianus structures his narrative o f Julian’s actions as emperor in four main chronological sections. Julian is proclaimed Caesar by Constantius and sent to Gaul in Book XV, which closes with a long excursus on Gaul designed to em­ phasize the importance o f Julians arrival in Gaul (Chapter IX). Four books then cover Julian’s actions as Caesar in Gaul from his arrival in the winter o f 355—356 to his return to his winter quarters in Paris after the campaign o f 359-360 (X V I—X IX ). The next two books describe Julian’s proclamation as Augustus early in 360, his subsequent dealings with Constantius, and his sei­ zure o f the Balkans in the summer o f 361 (X X —X X I). Book X X II describes Julian as sole Augustus in Constantinople, crossing Asia M inor and in Anti­ och: it covers the period from Constantius’ death on 3 November 361 (2 2 .1.1, 2.1) to December 362 (cf. 2 3 .1.1). Then an elaborate (and significant) series o f notices tied to the Kalends ofjanuary 363 (23.1) introduces Julian’s ill-fated Persian expedition (Chapter V). The expedition itself, with the reign o f Jo ­ vian added as a mere footnote or epilogue to that o f Julian (Chapter XII), oc­ cupies a full triad o f books (X X III—X X V ). Within this narrative structure, Ammianus includes both formal excursus and extensive discussion o f Julian’s political, administrative and religious reforms. If the historian’s presentation o f Julian is to be assessed at all adequately, analysis must concentrate on those aspects o f it where independent evidence exists and where significant inferences can be made.34 Three topics are here chosen for detailed consideration where both external and internal criteria are available. They do not include Julian’s administrative reforms, which are likely always to remain matters o f controversy.35 For the administrative history o f the Roman Empire contains many obscurités that will probably never be elu­ cidated because the type o f archival material used by historians o f more recent periods is too often lacking. In the case o f Julian there are two additional 34 For an elegant general assessment o f Ammianus’ depiction o f Julian, see J. Fontaine, L ’E m ­ pereur Ju lie n . D e l ’histoire à la légende, 3 3 1 - 1 7 1 5 (Paris, 1978), 3 1 - 6 5 .

35 For discussion o f these, see W. Ensslin, K lio 18 (1923), 1 0 4 -9 9 ; R. Andreotti, Nttova R ivista Storica 1 4 (1930), 3 4 2 - 8 3 ; Bowersock, Ju lia n (1978), 7 1 - 7 8 ; Athanassiadi-Fowden, Ju lia n

(19 81), 9 6 —120; Elliott, A m m ianus (1983), 1 0 4 - 2 1 ; Pack, Städte und Steuern (1986), 5 7 —300; Matthews, Am m ianus (1989), 1 0 4 - 1 3 . Ammianus’ account ofjulian’s actions in response to the food shortage in Antioch (2 2 .13 .4 — 14.2) has been scrutinized with especial intensity: see P. d ejon ge, M nem osyne 4 1 (1948), 2 3 8 45; G. Downey, Studies in Rom an Social and Economic H istory in H onor o f A . C . Jo h n so n (Prince­ ton, 19 5 1), 3 1 2 - 2 1 ; P. Petit, Lihanius et la vie municipale à Antioche au I V e siècle a p r è s J.-C . (Paris, 1955). 10 9 —18; J. H. W. G. Liebeschuetz, A ntioch: C ity and Im perial Adm inistration in the Later Rom an Em pire (Oxford, 1972), 1 2 7 - 3 1 ; Pack, Städte und Steuern (1986), 3 0 1 —37 7; P. Garnsey, Fam ine and Food Su p p ly in the Graeco-Rom an World. Responses to R isk and C risis (Cambridge, 1988), 2 4 7 - 4 8 ; Matthews, A m m ianus (1989), 4 0 8 - 1 3 .

T

he

N

e w

A

c h il l e s

difficulties. First, the full text o f most o f the original documents enacting his administrative reforms is lost: hence their precise purport must be deduced ei­ ther from brief and potentially misleading extracts in the Theodosian Code or from partisan, biased, and partial reports o f authors whose main aim is ei­ ther to praise or to blame. Hence there is, for example, serious uncertainty whether Julian’s law that prohibited Christians from teaching forbade Chris­ tians from all teaching, even in private houses, as is implied by both Ammianus, who deprecates the law (22.10.7, 25.4.20) and Christian writers,36 or applied only to the holders o f municipal chairs, whose appointment is the subject o f the extract preserved in the Theodosian Code.37 Second, the general tenor o f all the literary sources, favorable and hostile, Christian and pagan alike, is to emphasize a contrast between the administrative policies o f Julian and C on­ stantius even where the primary documents prove basic continuity.38

Julian in Gaul When Constantius proclaimed Julian Caesar and sent him to Gaul, he intended him (like Gallus) to be the figurehead o f an administration appointed by and obedient to himself.39 In 356, Julian and Constantius acted in tandem: the Caesar advanced through Gaul to the Upper Rhine, then sped north to relieve Cologne (16 .2 -4 ), while the Augustus pinned the Alamanni down by ad­ vancing through Raetia. But Ammianus has obscured the cooperation o f the two emperors operating in different theaters o f war: his narrative o f the cam­ paigning season o f 356, at least in the form in which it survived the Middle Ages, omits any mention o f Constantius, although his account o f the next year refers to it (16 .12 .15 - 16 ) . 36 Th e evidence is collected by J. Bidez and F. Cum ont, Im p. Caesaris F la v ii C la u d ii Iuliani Epistulae Leges Poematia Fragmenta varia (Paris/London, 1922), 6 9 - 7 5 no. 6 1. 37 C T h 13 .3 .5 . cf. T. M . Banchich,_/HS 107 (1987), 1 6 4 - 6 5 ; A ncient World 2 4 (1993), 5 - 1 4 . 38 W . A. Goffart, C P 65 (1970), 1 4 5 - 5 1 . However, C T h 2.29 .1 shows that Julian radically changed the rules pertaining to payment for lobbying at court (suffragium): T. D . Barnes, C P 69 (1974), 2 8 8 - 9 1 ; P h o en ix 39 (1985), 3 7 1 - 7 2 .

39 Julian, E p . ad A th . 2 7 7 d -2 8 o c , cf. R. P. C . Blockley, Latomus 3 1 (1972), 4 4 5 - 5 5 . Julian’s claim that Constantius gave him real command over the Gallic army in spring 3 5 7 (278d) is con­ tradicted by Ammianus (16 .12 .14 ). The praetorian prefects in Gaul, Honoratus ( 3 5 5 - 3 5 7 ) and Florentius (3 5 7 —360), were appointed by Constantius and took their orders from him rather than from Julian: the former had been comes O rientis under Gallus and became the first praefec­ tus urbis Constantinopolitanae in 359, while the latter operated independently ofjulian, whom he felt entitled to disobey, and fled to Constantius as soon as Julian was proclaimed Augustus (P L R E 1.4 3 8 —39, Honoratus 2; 365, Florentius 10).

[i 5 1 ]

T

he

N

e w

A

c h il l e s

In 357, too, Julian was supposed to cooperate with the army o f Constantius, which was under the command o f the magister peditum Barbatio, not o f the Au­ gustus himself (16 .11).40 It was in this year that Julian won the smashing vic­ tory near Strasbourg that was (and is) the basis o f his reputation as a good gen­ eral. Ammianus gives a long and detailed account o f the battle that uses every literary trick and rhetorical device at his disposal to enhance Julian’s glory (16.12). Literary analysis has effectively brought out its two main features.41 First, Ammianus throughout contrasts the confidence, arrogance, and feroc­ ity o f the barbarians with Julian’s calm resoluteness and his divine support: the caelitis dei favor and a salutaris genius help to bring the battle about (16 .12 .13); a standard bearer utters the exhortation “ Follow, Caesar, the guidance o f your lucky star: you are Fortune’s darling” (16 .12.18); and the propitiati numinis arbitrium aids the Romans (16.12.52). Second, the Roman victory at Stras­ bourg in Book X V I forms a diptych with the Roman defeat at Adrianople in Book X X X I. For in 378 the Goths display the virtues that the Romans had shown in 357 while the Romans o f 378 have acquired the vices o f Chnodomarius and his Alamanni.42 Ammianus nevertheless provides the fullest and most reliable account o f the Battle o f Strasbourg. The Germans had been crossing the Rhine for three days and the Roman army had encamped twenty-one miles from their entrench­ ment. After his infantry had advanced slowly from dawn to midday, Julian ad­ vised caution: Ammianus provides him with a speech in which he proposes that his tired troops halt, rest, refresh themselves, and attack the following day (16 .12 .9 -12 ). When the troops clamor to fight, the decision to do so is taken by the celsae potestates and the praetorian prefect Florentius— that is, by the men whom Constantius had appointed to conduct the campaign with Julian as their titular commander. Ammianus reports the reason urged by Florentius, who argued that, although fighting always involved some risk, battle must be joined while conditions were favorable and while the barbarians were in one mass. I f they dispersed, an unmatched opportunity would be lost, and the Ro­ man troops, who were impetuous by nature, might mutiny if they felt that an inevitable victory had been snatched from their grasp (16.12.14). Had Julian really been in command, and had his plan been put into effect, there would have been no glorious victory at Strasbourg, where Roman tactics were to 40 Ammianus denigrates Barbatio unfairly: Rosen, Studien (1970), 8 4 - 9 5 ; Elliott, A m m ianus (1983). 7 6 - 7 9 41 Rosen, Studien (1970). 9 5 —* 3 1 ; Bitter, Kampfschildcrungen (1976), 5 6 - 1 0 1 ; R. R C . Blockley, Ph oen ix 3 1 (i977). 2 1 8 - 3 1 . 42 Blockley, Ph oen ix 31 ( i 9 7 7 )> 2 2 4 —28, cf. G. B. Pighi, N u o i’i studi am m ianei (Milan, 1936), 6 1-12 7 .

[152]

T

he

N

ew

A

c h il l e s

push the largely unprepared Germans into the River Rhine and the marshes along its western bank.43 The victory thus belonged to others. But Julian soon claimed the credit that was at the time assigned to Constantius, and he began to use “ his” victory as justification for acting with a greater degree o f independence. Julian under­ stood the importance o f propaganda: he wrote his own account o f the Battle o f Strasbourg.44 It has heavily influenced all extant accounts, including that o f Ammianus, who had read it (16.5.7). The historian, however, had access to information independent ofjulian through his contacts in the high command o f the Roman imperial army.45 His knowledge o f the discussion by Julian and his advisers over whether to attack the Germans before they finished crossing the Rhine must come from fellow-officers who had been there and was pre­ sumably acquired shortly after the battle. Thus, despite his panegyrical tone and presentation, Ammianus underminesjulian’s claim to credit for the victory by exposing Julian’s version o f the battle as misleading on the crucial issues o f strategy and tactics46— which calls his reputation as a good general into doubt. Moreover, the subversive detail that Ammianus supplies both confirms his claim to have questioned participants in historical events (15 .1.1) and implies that he was contemplating a history o f his own time long before 378. After the campaign o f 357, Julian turned his attention to the Lower Rhine, where he prosecuted war with less enthusiasm. He was emancipating himself from the tutelage o f Constantius’ appointees and beginning to contemplate a challenge for supreme power. A classic analysis o f his strategy established long ago that Julian stopped serious campaigning after 358 to prepare cleverly and covertly for the proclamation as Augustus that occurred in Paris in the early months o f 360.47 The argument inevitably relies heavily on Ammianus’ de­ tailed account ofjulian’s campaigns in 358 and 359 (17 .8 - 10 , 1 8 .1—2), though rejecting his interpretation o f the actions he narrates. Ammianus presents Julian throughout as a consistently loyal subordinate who was taken by surprise when his troops unexpectedly proclaimed him Au­ gustus and compelled him to agree to accept the title against his wishes (20.4.16: assentire coactus est). The letters and speeches that Julian composed as Caesar 43 O n the site and course o f the battle, see esp. G . A. Crum p, Ammianus Marcellinus as a M il­ itary Historian (Historia Einzelschriften 27, 1975), 8 5 - 8 9 ; Matthews, Ammianus (1989), 2 9 6 -3 0 0 ; H. Elton, Warfare in Roman Europe, A.D. 3 5 0 -4 2 5 (Oxford, 1996), 2 5 0 - 5 6 . 44 Eunapius, frag. 9 Müller =

17 Blockley: other allusions are collected by Bidez and C u -

mont, Iuliani Epistulae Leges (1922), 2 1 2 - 1 3 no. 160. 45 Crum p. Ammianus (19 75), 20. 46 Matthews, Ammianus (1989), 9 1 —92. 47 I. Müller-Seidl, H Z 180 (19 55), 2 2 5 - 4 4 . There is a good diagrammatic map o f Julians campaigns in Matthews, Ammianus (1989), 82: Map 2.

[153]

T

h e

N

e w

A

c h il l e s

reveal aspirations and ambitions that the historian has suppressed. B y invok­ ing Achilles openly and King David unconsciously, the second panegyric on Constantius lays bare a deep resentment and reveals dangerous hopes.48 In a letter to the doctor Oribasius, the Caesar spoke more freely still. He described a dream whose political meaning is transparent. Julian saw a very tall tree bending over to the ground under its own weight and height and beside it a small shoot growing out o f its roots. He feared for the safety o f the young plant, and when he drew nearer he saw the large tree falling to the ground. The small tree was still standing; its roots remained in the earth; and an un­ known person advised the dreamer not to fear for its safety. Julian commented disingenuously: “ God knows what this means.” 49 Julian was thus already thinking or dreaming o f declaring himself the equal o f Constantius as early as the winter o f 358—3 59.50 A year later he was ready for usurpation. The fall o f Amida in the late summer o f 359 had damaged Constantius’ prestige, and his request that Julian send some o f his best troops to reinforce the eastern frontier provided an occasion too good to miss.51 Ac­ cording to the official story, which Julian circulated in the letters he wrote to Greek cities in 361, Constantius was preparing to remove Julian from power out o f jealousy over his successes. Hence he wrote him insulting letters and gave orders for the transfer o f Caesar’s best troops from Gaul to the East. The troops were distressed at the prospect. When an anonymous letter was circu­ lated accusing Constantius o f betraying Gaul, those around Julian, who were all loyal to Constantius, urged him to dispatch the troops at once. Julian as­ sembled the men in Paris. The night before their departure they unexpect­ edly surrounded the palace shouting. Julian prayed to Zeus, who gave him a sign that that he should not oppose their will. Finally, although he continued to decline their salutation as Augustus and to refuse the diadem, he allowed a soldier to place his collar on his head, after which he returned to the palace groaning in his heart.52 This version o f events was false, and contemporaries knew it to be false.53 48 Above, pp. 1 4 4 - 4 5 . According to Ammianus, Julian was saluted Augustus by his victori­ ous troops immediately after the Battle o f Strasbourg in 357, but swore on oath that he would never take the title (16.12.64). 49 Julian, E p . 10. 50 Bowersock, Ju lia n (1978), 17 ; J. F. Drinkwater, Studies in Latin Literature and Rom an History 3 ( Collection Latomus 180, 1983), 3 7 0 —75. 51 seen, 52 53

O n the background to this request, whose consequences Constantius cannot have fore­ see Drinkwater, Studies in Latin Literature 3 (1983), 3 7 6 - 7 8 . Julian, E p . ad A th . 2 8 2 - 8 5 ; cf. Libanius, O ral. 18 .9 5 —105; Zosimus 3 . 9 . 1 - 3 . Müller-Seidl, H Z 180 (1955), 2 4 1 —44; K. Rosen, Acta Classica 12 (1969), 1 2 1 —49; B o w ­

ersock, J m/m « (1978), 4 6 - 5 4 , whose analysis o f the episode is not invalidated by the minor flaws detected in it by R. S. O. Tomlin, Ph oen ix 34 (1980), 2 6 6 —70.

[154]

T

h e

N

e w

A

c h il l e s

When the praetorian prefect Mamertinus, who had been in Gaul at the time, delivered a gratiarum actio for his ordinary consulate on I January 362 in Con­ stantinople, he passed quickly and with evident embarrassment over his bene­ factor’s proclamation as Augustus. To this obligatory topic he allotted only a few vague, general, and tendentious remarks. Could Julian (he asked rhetor­ ically) have surrendered Roman cities to the enemy just to avoid offending his cousin?54 More than twenty years later, Ammianus did not need to be so coy. M ore­ over, Eunapius had probably already presented Julian as a justified rebel against tyranny. For his later Lives of the Philosophers, composed in 399, states that Julia'n “ summoned the hierophant from Greece, and after performing with his aid certain rites known to them alone, he was emboldened to abolish the tyranny o f Constantius.” 55 Eunapius is presumably speaking o f Julian’s actions after his proclamation at Paris, which he probably depicted as the spontaneous elevation o f an unwilling Augustus.56 Although Ammianus adds details not reported elsewhere, such as the name o f the hastatus o f the Petulantes who placed on Julian’s head the torques that he was wearing as bearer o f the dragon standard (20.4.18: Maurus nomine quidam), he adheres closely to the official version o f his proclamation as Augustus that Julian circulated in 360 and 361 by depicting an unsuspecting innocent who yielded to extreme ne­ cessity and pressing danger (20.4-6). Yet he concurs with Zosimus in in­ cluding something that conclusively proves Julian’s complicity in the proc­ lamation: after he had addressed the troops in the outskirts o f the city, but before they surrounded the palace, he entertained his officers to dinner (20.4.1 3).57 They lamented their lot— and presumably decided to do some­ thing about it.

Julian’s Religion For six years as Caesar and then as Augustus, Julian comported himself as a Christian, even though he had been converted from the Christianity o f his 54 Pan. Lat. 3(11).5 .2 . M Eunapius, Vit. phil. 7 .3.7 (476). He names Oribasius o f Pergamum and the African Euhe­ merus as his accomplices: Bowersock, Julian (1978), 50. For disproof o f the existence o f a “ pa­ gan underground” in the East covertly aiding Julian before November 36 1, see Drinkwater,

Studies in Latin Literature 3 (1983). 3 4 8 - 6 0 . 56 D . F. Buck, A H B 7 ( i 9 9 3 ). 73 - 8 0 . s7 O n the damning implications o f this dinner party, see Bowersock, Julian (1978), 4 9 - 5 1 . Zosimus 3.9 has a parallel, but less detailed, account.

[■ 55]

T

he

N

e w

A

c h il l e s

upbringing to paganism in 3 5 1.58 At first he enforced Constantius’ ecclesias­ tical policies in Gaul, sending Hilary o f Poitiers into exile in Asia M inor in 356 after he had been condemned by the Council o f Baeterrae.59 After he was proclaimed Augustus, however, Julian began to support Constantius’ ecclesi­ astical opponents and allowed the “ Catholic” bishops o f Gaul to hold a coun­ cil in Paris during 360.6" The death o f Constantius removed all external con­ straints, and he commenced open avowal o f paganism at once.61 It may, or may not, be significant that Ammianus delays the moment o f change, which oc­ curred in Naissus, until Julian has entered Constantinople (22.5.1—2). Julian’s religious policies as sole emperor are clear, at least in outline.62 He defined the purpose o f his religious reforms as the preservation o f traditional religion.63 Yet Julian was a neopagan rather than an restorer o f authentic tra­ ditional religion. When he described what a city ought to be, he offered a pe­ culiar mixture o f the ancient city and a puritanical ideal. A happy city, on Julian’s definition, was one fu ll o f m a n y s h rin e s a n d m a n y s e c re t r ite s , w it h c o u n tle s s h o ly p rie s ts w it h in its w a lls d w e llin g in h o ly e n c lo s u re s . F o r th e s a k e o f th is, I m e a n k e e p in g e v e r y ­ th in g in th e c it y p u r e , th e y h a v e e x p e lle d w h a t is s u p e r flu o u s , s o r d id , a n d v ic io u s fr o m th e c i t y — p u b lic b a th s, b r o th e ls , s h o p s , a n d e v e r y t h in g o f th at s o rt w it h o u t e x c e p t io n .64

Julian’s ideal pagan city observes Christian standards o f purity. For Julian’s pa­ ganism was not o f the original, earthy variety: it was a puritanical and ascetic neopaganism that owed a great deal to Christianity and was therefore markedly untraditional.65 A perceptive analysis o f Julian’s personality, which firmly re­ 58 Julian, E p . h i , 434d. 59 Hilary, C S E L 6 5 .1 9 8 .5 - 1 5 , cf. Vig. Chr. 46 (1992), 1 2 9 - 4 0 ; Athanasius (1993), 14 1 , 1 7 1 ; P. Smulders, H ila ry o f Poitiers’ Preface to his ''O p u s Historicum ” (Leiden, 1995), 1 1 9 —3 1. 611 C S E L 6 5 .4 3 - 4 6 , cf. Athanasius (1993), 1 5 3 - 5 4 .

61 Julian, E p . 2 6 , 4 15 c , informs the philosopher Maximus that he is openly worshipping the gods and has sacrificed many hecatombs o f oxen, apparendy while still in Naissus (cf. 415a).

02 J. Bidez, Vie de Ju lie n (1930), 2 1 9 —35, 2 6 1 - 7 6 , 2 9 1 —3 1 5 ; Bowersock, Julian (1978), 7 9 83; T. D. Barnes, Athanasius (1993), 1 5 3 - 5 5 . T h e present discussion deliberately eschews any attempt to define the precise intellectual nature o f Julian’s paganism. 63 Julian, E p . 89, 453bc. 64 Julian, Orat. 6, l86d. 65 H. Koch, R evu e belge de philologie et d ’histoire 6 (1927), 1 2 3 - 4 6 ; 7 (1928), 4 9 - 8 2 , 5 1 1 - 5 0 , 136 3-8 5 . For a conspicuous example o f Julian’s ignorance o f authentic tradition in a different area, see V. A . Maxfield, The M ilitary Decorations o f the Rom an A r m y (London, 19 8 1), 2 4 8 —5 1. Julian dec­ orated those who showed the greatest bravery in the capture o f Maozamalcha with obsidionales coronae in old-fashioned style (24.4.24: veterum more), and before Ctesiphon he decorated sol­

[156]

T

he

N

e w

A

c h il l e s

jects the quasi-hagiographical approach o f much modern writing about the emperor, recognizes his fanaticism and classifies him as a “ revolutionary as­ cetic.” 66 Ammianus acknowledges but plays down the fanaticism, and he se­ riously misrepresents both Julian’s religious policies and his religious beliefs. Ammianus gives his readers a reign o f Julian in which no Christian suffered for his faith (Chapter V), he declines to record his attempt to found a neo­ pagan counterchurch, and he completely omits two central elements o f the emperor’s personal creed: theurgy and Mithraism. Julian had embraced Iamblichus’ version o f Platonism wholeheartedly, and he regarded this Neoplatonist heir o f a long philosophical tradition, from whom he had learned his philosophy, as an equal o f the great founder o f the school. For Julian, Plato was o f course a great philosopher, but Iamblichus o f Chalcis was by no means inferior to him in genius: Iamblichus was a glorious hero, divine, and the friend o f God.67 Iamblichus accepted the authority o f the so-called Chaldaean Oracles, which were associated with Julianus the Chaldaean and his son Julianus the Theurgist, who had revealed them to the world in the reign o f Marcus Aurelius.68 Iamblichus integrated the Chaldaean Oracles into his philosophical system, so that the practice o f theurgy became central to it: although he presented his system as the logical development o f classical Greek philosophical thought, he transformed what had been primär-

diers with navales coronae et civicae et castrenses (24.6.15). Such decorations had fallen completely out o f use circa 200, so that Julians revival was an act o f conscious antiquarianism. But he made a significant mistake. Aulus Gellius, who has a chapter on coronae militares , defines the corona obsidotialis as one given by a besieged city to the general who has liberated it by lifting the siege and

states that Q . Fabius Maximus received one for rescuing Rome from Hannibal (Noctes Atticae, 5 .6 .8 -10 ) . 66 Bowersock, Ju lia n (1978), 1 2 - 2 0 . Th e merits o f this brief, but incisive, study were sadly underestimated by J. F. Matthews, T L S 3 November 1978, 128 3, who failed to see that neither its brevity nor some minor factual errors invalidate its analysis o f Julians tortured psyche: for a juster appreciation, see F. Paschoud, R E L 58 (1980), 1 1 7 - 2 3 . 67 Julian, Orat. 4, I58cd, cf. i5od. On Iamblichus, see now the collection o f essays edited by H. J. Blumenthal and E. G. Clark, T he D ivin e Iamblichus. Philosopher and M an o f G ods (London, 1993)» with the review article by P. Athanassiadi, J R S 85 (1995), 2 4 4 - 5 0 . 68 S u d a I 4 33, 4 34 ( 2 .6 4 1 - 4 2 Adler). It is possible that Porphyry incorporated some o f the Chaldean Oracles in his youthful Philosophy from Oracles , viz., frags. 2 1 9 - 2 5 des Places, which are quoted from Porphyry in Eusebius, Praep. E v . 5 . 8 4 - 7 , 9 .1., 12 .1. However, their prove­ nance is disputed by E. R. Dodds, H T R 54 (19 61), 2 6 4 - 6 7 . Porphyry later questioned the au­ thority o f the Chaldean Oracles in a work that may have had the composite title “ O n the re­ turn o f the Soul, in Reply to the Writings o f Julian the Chaldean.” However, although the direction o f the change in Porphyry’s opinions is clear (J. Bidez, Vie de Porphyre le philosophe néo­ platonicien [G h en t/Leipzig, 19 13 ], 8 8 - 9 7 ) , it is normally assumed that two separate works are in

question: A . Smith, Porphyrii Philosophi Fragmenta (Leipzig, 1993), 3 1 9 - 5 0 (frags. 283 - 3 0 2 ) , 4 3 5 40 (frags. 3 6 2 -6 8 ) .

[157]

T

h e

N

e w

A

c h il l e s

ily a philosophy with mystical overtones into something closer to a religion with both a theology and cult acts.69 Julian saw himself as a disciple o f Iamblichus, but he gave a unique and in­ dividual twist to his teachings by combining theurgy with Mithraism in a novel blend o f personal mysticism.70 In his Hymn to the Mother of the Gods, composed at Pessinus in 362, he vaunts his participation in the Mithraic mys­ teries and claims Julian the Chaldaean as a Mithraist: And if

I

s h o u ld also t o u c h o n th e s e c re t t e a c h in g o f th e m y s te r ie s w h ic h th e

C h a ld e a n r e v e a le d in d iv in e fr e n z y c o n c e r n in g th e G o d o f th e S e v e n R a y s , lift in g u p th e so u ls o f m e n t h r o u g h h im , I s h o u ld b e s a y in g w h a t is in c o m p r e h e n s ib le , in d e e d c o m p le t e ly in c o m p r e h e n s ib le to th e c o m m o n h e r d , b u t w e ll k n o w n to th e b le sse d th e u rg is ts .71

Julian practiced a conscious syncretism which integrated Roman tradition into his Iamblichan philosophical and religious thought world. At the army camp in Carnuntum in 308, the former emperor Diocletian, the senior reign­ ing emperor Galerius, and the newly proclaimed Augustus Licinius had re­ stored a shrine o f the Deus Sol Invictus Mithras whom they saluted as the susvtainer o f their imperial rule.72 Julian went much further. The Hymn to King Helios, which Julian addressed to his praetorian prefect Salutius, envisages He­ lios as a Neoplatonic triad: the sun is transcendental and as such indistin­ guishable from the Good in the intelligible world; the sun is identical with Mithras and the ruler o f the intellectual gods; and the visible sun illumines the physical world.73 The worship o f Mithras and the quadriennial games o f Sol 69

O n this development, see E. R. Dodds, The Greeks and the Irrational (Berkeley, 19 5 1), 283 -

3 1 1 ; H. Lewy, Chaldean Oracles and Theurgy: Mysticism, Magic and Platonism in the Later Roman

Empire (Cairo, 1956; photographic reprint with corrigenda and extensive addenda, ed. M . Tardieu, Paris, 1978); F. W. Cremer, Die chaldäischen Orakel und Jamblich de mysteriis (Diss. Cologne, 1967; publ. Meisenheim am Glan, 1969); R. T. Wallis, Neoplatonism (London, 1972), 1 0 0 - 1 0 , 1 1 8 - 2 3 ; A. Smith, Porphyry's Place in the Neoplatonic Tradition. A Study in Post-Plotinian Neoplatonism (The Hague, 1974), 8 3 - 1 4 1 ; B. Nasemann, Theurgie und Philosophie in Jamblichs De mysteriis (Diss. Cologne, 1989; pub. Stuttgart, 1991); B. E. Pearson, Neoplatonism and Gnosticism, ed. R. T. Wallis and J. Bregman (Albany, 1992), 253 - 7 5 ; G. Shaw, Theurgy and the Soul. The Neo­ platonism of Iamblichus (University Park, Pa., 1995), 16, 6 1 - 2 4 2 . 70 O n the importance o f Mithraism in Julians thought, see Athanassiadi-Fowden, Julian (19 81), 3 8 - 4 1 , 1 1 3 - 1 4 , 1 3 3 - 3 5 , 15 3 , 1 9 7 - 9 8 . It is denied by Sm ith, Julian's Gods (1995), who attempts to demonstrate that Julian was at heart a traditional polytheist. 71 Julian, Orat. 5, I72d-i73a. 72 IL S 657: d. S. i. M . fautori imperii sui Iovii et Herculii religiosissimi Augusti et Caesares sacrarium restituerunt. For the political context, see, briefly, Constantine (19 81), 32. 73 Julian, Orat. 4(10), 132C-133C, cf. P. Athanassiadi-Fowden,J T S , N .S. 28 (1977), 360-71.

[158]

T

h e

N

e w

A

c h il l e s

Invictus may be recent (the solis agon was instituted by Aurelian in 274),74 but Julian claims that the Romans have worshipped the sun since the days o f King Numa.75 Julian names Mithras only once in the Hymn and only once else­ where.76 Such reticence was required o f a Mithraist by the secret nature o f the cult, which was less a religion than an ancient version o f freemasonry for sol­ diers and government servants.77 Nevertheless, the equation o f Mithras with Helios was fundamental to Julians religious beliefs, as the other passage that names Mithras makes clear. In the concluding paragraph o f the Caesars, Her­ mes informs Julian that he has granted him knowledge o f his father Mithras and adds the following injunction: “ Keep his commandments, securing for yourself a secure stem-cable and mooring during your lifetime, and making with good hopes a guardian god who will be favorable to you when you need to depart from here.” 78 There is independent confirmation o f the importance o f Mithras for Julian. The sophist Himerius, a native o f Bithynia, had been teaching in Athens as the holder o f an official chair o f rhetoric since circa 353. When he heard o f Julians successful advance to Naissus, he abruptly left Athens, attached him­ self to the imperial court, and eagerly took part in a celebration o f the mys­ teries o f Mithras in the Mithraeum which Julian constructed in the imperial palace in Constantinople.79 Himerius proclaimed the fact in an address to the city that was also a panegyric o f the emperor. It opens as follows: H a v in g p u r ifie d o u r so u l to H e lio s - M it h r a s a lre a d y b y th e aid o f th e g o d s in th e c o m p a n y o f an e m p e r o r w h o is a fr ie n d o f th e g o d s , c o m e n o w ! le t us lig h t u p a s p e e c h r a th e r th an a la n te rn to th e e m p e r o r an d th e c ity ! A la w o f A th e n s o rd e rs th e m y s t ic in itia te s to c a r r y lig h t a n d h a n d fu ls o f g ra in to E le u s is as to k e n s o f c iv ilis e d life . F o r o u r m y s t ic in itia te s le t a sp e e c h b e p re s e n te d as a t h a n k o ffe r in g , i f in d e e d A p o llo is, as I th in k , id e n tic a l w it h H e lio s , a n d s p e e c h e s a re th e c h ild r e n o f A p o l lo .80

An issue o f bronze coins from most o f the imperial mints except Rome com­ bines the reverse legend Securitas rei pubflicae) with a bull and two stars (some­ 74 Chr. m in. 1.14 8 ; Jerome, Chronicle 223 b Helm, cf. E. Cizek, L ’empereur A urélien et son temps (Paris, 1994), 1 7 8 - 8 2 . 75 Julian, O ral. 4(10), I 5 5 a - i5 6 b . 76 Julian, Orat. 4(10), 155b ; Coes. 336c. 77 For Mithraism as a club, see R. L. Beck, Voluntary Associations in the Rom an World, ed. J. S. Kloppenborg and S. G. Wilson (London/Nevv York, 1996), 176 —85. Its affinités with freemasonry were noted by N . Swerdlow, C P 86 (19 91), 48, 6 2 - 6 3 . 78 Julian, C aes. 336c. 79 Libanius, O ral. 18 .12 7 , cf. C P 82 (1987), 2 2 1 - 2 2 . 80 Himerius, Orat. 4 1.1.

T

he

N

e w

A

c h il l e s

times also an eagle).81 The type has never been convincingly explained.82 But two stars associated with a bull ineluctably suggest Cautes and Cautopates, the helpers o f Mithras, who were associated with the stars Aldebaran and Antares,83 and there can be no doubt o f the centrality o f Mithras in Julian’s mind, de­ spite both his reticence and the silence o f Ammianus.84 Like the rites o f Mithras, theurgy was practiced in secret, not in public. Julian engaged in theurgy with the assistance o f his philosophical mentor and guide, Maximus o f Ephesus, who accompanied the Persian expedition and was present when the emperor died (25.3.23).85 Ammianus’ treatment o f Julian’s relationship with Maximus is idiosyncratic. He presents Maximus as a genuine philosopher, not as the arrogant charlatan depicted in Eunapius’ Lives of the Philosophers, who took advantage o f an unsuspecting Julian.86 But he criticizes the emperor sharply for excessive display o f his friendship for him. One day, when he was he was hearing legal cases, it was reported that M ax­ imus had arrived from Asia. The emperor leapt up in undignified haste, ran out, kissed Maximus, and brought him back with him into court. Ammianus condemns this “ untimely ostentation” that showed Julian up as “ excessively anxious for empty glory” (22.7.3). Ammianus knew that Julian had been con­ verted in 35 1, since he makes him pray secretly to Mercury, that is, to Her­ mes, whom Julian often invokes as the god o f eloquence and his protector,87 shortly after his arrival in Gaul (16.5.5). But he so deeply disapproved o f Julian’s blend o f paganism that he ridiculed the emperor’s religion.

81 R I C 8 .19 5 Lyons, 2 3 6 —38; 229 Arles, 3 1 3 —23; 3 3 7 Aquileia, 2 4 2 - 4 3 ; 380 Siscia 4 1 1 - 1 3 ,

4 1 7 —19; 392 Sirmium, 1 0 5 - 0 7 ; 423 Thessalonica, 2 2 2 —26; 438 Heraclea, 1 0 1 —04; 4 6 2 —63 Constantinople, 1 6 1 - 6 4 ; 4 8 3 - 8 4 Nicomedia, 1 1 8 - 2 2 ; 500 Cyzicus, 1 2 5 —28; 532 Antioch, 2 1 6 —18. Julian, M is. 353d, alludes to the Antiochenes’ mockery o f his coinage: Socrates, H E 3 . 1 7 . 5 - 6 , states that it was this issue that they ridiculed. It is also mentioned by Ephrem, C o n ­ tra Iuliamtm 1.1 6 —19.

82 J. P. C . Kent, R I C 8 (19 81), 47. Ephrem, Contra Iuliamtm 1 . 1 6 —19 (cd. E. Beck, C S C O = Syr. 78 [19 57], 7 4 —75: English translation by K. E. M cVey, Ephrem the S y ria n : H ym ns (N ew York, 1989], 2 3 1 —32) inveighs against the bull coinage and compares the bull on the coins 174

to the golden calf o fE xo d u s 3 2 . 1 - 3 5 : Bowersock, Ju lia n (1978), 104, deduces that “ the mean­ ing o f Julian’s strange new coinage was quite unknown even then.” 83 R. Beck, Jo u rn a l o/M ithraic Studies 2 (19 77), 6 - 8 . 84 The contrary is confidently asserted by R. Turcan, M ithra et le mithraicisme (Paris, 1993), 1 18: ‘Tien, ni dans ses discours ni dans son monnayage ou sa politique, ne comporte le moindre indice probant de mithraicisme.” He appears not to mention Himerius. 85 O n whom , see P L R E 1 .5 8 3 - 8 4 , Maximus 2 1. 86 Eunapius, Vit. ph il. 7 . 4 . 1 - 4 (477). 87 Julian, O ral. 4, 132a; 6, 182c; 7, 225b, 2 3 0 c - 2 3 ia , 232d, 234b, 237c: in the Caesars, Her­ mes is named as the author’s informant (307a), who has also given him knowledge o f Mithras (336c, quoted above, p. 159).

[160]

T

he

N

e w

A

c h il l e s

Ammianus belonged in the philosophical tradition o f Porphyry (Chapters VII, X IV ), who was intellectually opposed to Iamblichus on many issues.88 The two philosophers perhaps differed most sharply over the issue o f animal sacrifice. Porphyry condemned it roundly as irrational in his Letter to Anebo; Iamblichus replied to him under the pseudonym o f Abammon, using the schol­ arly genre o f “ problems and solutions” to pose objections one by one and an­ swer them in the tone o f a master correcting a wayward pupil.89 Ammianus fastened on this aspect o f Julians religion: H e d r e n c h e d th e altars w it h th e a b u n d a n t b lo o d o f s a c rific ia l v ic t im s w it h e x ­ c e ssiv e fr e q u e n c y . O n se v e ra l o c c a s io n s h e s a c r ific e d a h u n d r e d b u lls at a tim e , c o u n tle ss flo c k s o f v a r ie d liv e s to c k a n d w h it e b ird s s o u g h t a fa r b y la n d a n d sea. A s a re su lt, a lm o s t e v e r y d a y so ld ie rs , w h o g o r g e d th e m s e lv e s u n c o n t r o la b ly w it h m e a t till th e ir b e llie s w e r e d is te n d e d a n d th e y w e r e d e m o r a lis e d b y t h e ir c r a v in g fo r d r in k , w e r e c a r r ie d to th e ir q u a rte rs o n th e s h o u ld e rs o f p a s s e rs -b y t h r o u g h th e stre e ts f r o m p u b lic te m p le s w h e r e th e y in d u lg e d in w ild p a rtie s w h ic h o u g h t to h a v e b e e n p u n is h e d r a th e r th a n p e r m itte d . ( 2 2 .1 2 .6 )

Ammianus proceeds to complain about the unparalleled expense that Julian’s religiosity entailed and that divination was practiced by many who had no gen­ uine expertise in interpreting oracles, inspecting entrails, or understanding the 88 For the deep disagreement between the two philosophers over basic issues, see H. D. Saffrey, Philom athes : Studies and Essays in the H um anities in M em ory o f P hilip M ed an (The Hague, 19 71), 2 2 7 - 3 9 : Smith, Porphyry's Place (1974), x v ii-x v iii, 8 1 - 1 4 1 . A long tradition makes Iamblichus a pupil o f Porphyry and Porphyry “ the historical and ideological mediator between Plotinus and Iamblichus” : B. D. Larsen, Jam bliq u e de Chalcis. Exégète et philosophe (Diss. Aarhus, 1970: pub. Aarhus, 1972), 3 6 —40;}. Dillon, Iamblichi Chalcidensis in Platonis Commentariorum Frag­ menta (Philosophia A n tiq u a 23, 19 73), 8 - 1 1 ; Shaw, Theurgy (1995), 2 - 3 , 1 2 - 1 3 . But h is highly doubtful whether Iamblichus ever studied with Porphyry, as is stated by Eunapius, Vit. pltil. 5 . 1 . 2 -

3, (458). Th e profound intellectual divergences between the two men are noted by G . M au, R E 9 (19 16 ), 645; H .-I. Marrou, Tlte Conflict between Paganism and Christianity in the Fourth C entury, ed. A . Momigliano (Oxford, 1963), 1 3 2 - 3 3 ; Smith, Porphyry's Place (1974), x v ii-x v iii. And Iamblichus, who had a son who was married by 300 (Porphyry, Vita Plotini 9 .3 - 5 ) , must have been born before 250, probably c. 2 4 2 - 2 4 3 : } . Bidez, R E G 32 (1919), 32 (“ né vers 250 au plus tard” ); Alan Cameron, Hermes 96 (1968), 3 7 4 - 7 6 , so that he was only ten to fifteen years younger than Porphyry. Accordingly, since Iamblichus is not known ever to have left Syria, it is probable that, if he studied with Porphyry, he did so, not “ at Rom e” (as P L R E 1.450; T. D. Barnes, G R B S 19 [1978], 105), but before the latter went to Rome in 263 and met Plotinus.

89 Both the conventional title D e M ysteriis , bestowed on it by Marsilio Ficino, and the con­ ventional division into ten books are seriously misleading: M . Sichert, D ie Handschriften, A u s ­ gaben und Übersetzungen von Iamblichus D e M ysteriis. E in e kritisch-historische Studie ( Texte und U n ­ tersuchungen 62, 19 57); B Z 53 (i960), 9; Saffrey, Philomathes (19 71), 2 2 7 - 3 9 ; D iv in e Iamblichus

(1993), 1 4 4 - 5 8 -

[161]

T

he

N

e w

A

c h il l e s

flight o f birds (22.12.7). Subsequently, Ammianus’ obituary ofjulian delivers a verdict on his religion that both mocks and condemns: H e w a s t o o d e v o te d to d iv in a t io n , so th a t h e s e e m e d in this re s p e c t to riv a l th e e m p e r o r H a d r ia n . H e w a s s u p e rs titio u s r a th e r th an a g e n u in e o b s e r v e r o f r e li­ g io u s rite s, a n d h e s a c r ific e d in n u m e r a b le a n im a ls re g a rd le ss o f e x p e n s e , so th at it w a s r e c k o n e d th at, i f h e h a d r e tu r n e d f r o m P a rth ia , th e re w o u ld s o o n h a ve b e e n a d e a rth o f ca ttle . H e r e s e m b le d th e fa m o u s C a e s a r M a r c u s , a b o u t w h o m w e are to ld th at th e f o llo w in g w a s said : G r e e tin g s to M a r c u s f r o m th e o x e n w h it e . W e ’re d o n e f o r i f y o u w i n a n o t h e r fig h t.

(25.4-I7)

Julian was superstitiosus magis quam sacrorum legitimus observator: Ammianus be­ lieved that he perverted traditional religion as badly as Constantius perverted Christianity. Yet he filters out Julian’s Mithraism, his devotion to Helios, and his practice o f theurgy. Instead, he emphasizes traditional forms o f divination, such as haruspicina and augury (21.2.4, 2 2 .1 .1 —2), notes that haruspices accompaniedjulian on the Persian expedition (23.5.10, 25.27-28), and states repeat­ edly that Julian ignored the warnings that the gods gave him as he marched into Mesopotamia (23.2.7, 3.3, 5.6, etc.).90 At the time, in contrast, Julian had written to Libanius from Hierapolis to inform him that all the omens had so far been propitious.91

T he Persian Expedition Much has been written about Ammianus’ detailed account o f Julian’s invasion o f Mesopotamia, and especially about what written sources he may have used.92 The present exposition, therefore, can concentrate on Julian’s strategy, 90 O n these omens and portents, see W . Liebeschuetz, Roma Renasceris, ed. M . Wissemann (Frankfurt, 1988), 19 8 --2 13 . D. Conduché, Latomus 24 (196$), 3 6 4 - 7 9 , argues that the ad­ verse omens reflect contemporary disapproval o f Julian’s plans voiced by officers serving on the expedition. 91 Julian, Ep. 98, 399d, 401b. 92 For a critical survey o f both the problems and the solutions offered by m odem scholar­ ship, see M . F. A . Brok, Die perzische expeditie van Keizer Julians volgens Ammianus Marcellinus (Groningen, 1959); F. Paschoud, Zosime: Histoire Nouvelle i (Paris, 19 7 1), xliii-lvii; 2 .1 (Paris, 1979), x ii-x ix , 1 0 0 - 2 2 7 . Am ong recent discussions, note E. Bliembach, Libanius Oratio 18

(Epitaphios): Kommentar (§§ î i i - j o S ) (Diss. Würzburg, 1976), 1 2 9 - 2 2 5 ; Bowersock, Julian (1978), 1 0 6 - 1 9 ; Elliott, Ammianus (1983), 1 2 2 - 3 4 ; Matthews, Ammianus (1989), 1 3 0 - 8 3 ; C . W . Fomara ,J H S h i (1991). 1 - 1 5 ; A JA H 10 (1985, publ. 1994), 2 8 - 4 0 . R. T. Ridley, His -

T

h e

N

e w

A

c h il l e s

which has often been misunderstood and which Ammianus misrepresents. Ammianus states that Julian took care that no rumor o f his march preceded him, so that he might seize Babylonia before he was detected (23.2.7: u t . . . improvisus Assyrios occuparet). Modern students o f the campaign usually as­ sume that Julian looked to Alexander the Great not only as his hero, but also as his model.93 Some have even asserted that he regarded himself “ as a rein­ carnation o f Alexander.” 94 Hence they have supposed that his strategy must have been one o f strategic surprise— to march rapidly on Ctesiphon and take it by storm before Shapur, diverted by a second Roman army in northern Mesopotamia, could return to defend his capital.95 Julian failed and thus re­ vealed that he was “ no second Alexander.” 96 But strategic surprise, in the sense o f mounting an unexpected invasion against an unprepared foe, was im­ possible in ancient conditions,97 and few leaders in history (with the conspic­ uous exception o f Josef Stalin in 1941) have shut their eyes and ears to obvi­ ous signs o f an impending invasion. When Julian arrived in Antioch on 18 or 19 Ju ly 362 (22.9.1 5),98 it soon became obvious that he was preparing to in­ vade Persia. Shapur knew what Julian intended and sent an embassy to Antioch to try to avert the impending invasion. Libanius refers to its arrival in a speech delivered on 1 January 363, and after Julian was dead, he lamented his brusque refusal to negotiate.99 Ammianus omits both the embassy and Julian’s display o f obstinacy. He also omits the fact, on casual attestation in a letter o f Liban­ ius, that Julian hoped to replace Shapur with Hormisdas.100

toria 22 (1973), 3 1 7 - 3 0 , demonstrated that the account in Zosimus 3 . 1 3 - 3 0 , despite manifold shortcomings, is in important respects superior to that o f Ammianus. 93 N . H. Baynes, B yzan tin e Studies and O ther Essays (London, 1955)» 3 4 6 - 6 7 . 94 Athanassiadi-Fowden, Ju lia n (19 81), 19 3, repeating verbatim the allegation o f Socrates, H E 3 .2 1.7 .

95 W . E. Kaegi, Athenaeum , N . S. 59 (1981)» 2 0 9 - 1 3 . 96 B. Strauss and J. Ober, The A natom y o f Error: A ncient M ilitary Disasters and T h eir Lessons fo r M odern Strategists (N ew York, 1990), 2 1 6 - 4 3 , a perceptive and penetrating essay, despite being

mistaken on this point. 97 A . D. Lee, The Eastern Frontier o f the Rom an Em pire , ed. D. H. French and C . S. Lightfoot ( B A R , Supp. Ser. 553, 1989), 2 5 7 - 6 5 ; Information and Frontiers. Rom an Foreign Relations in Late A n tiq u ity (Cambridge, 1993), 1 1 2 - 2 8 , 1 4 9 - 6 1 . 98

As interpreted by F. Cum ont, Syria 8 (1927)» 3 3 9 -4 0 , whose inference is confirmed by

Jerom e’s commentary on Ezekiel (3.8 .13, 14 [ P L 2 5 .8 2 - 8 3 = C C L 75-99])- Older editions o f Jerome misguidedly emended the transmitted Iulio mense to Iunio: from this false emendation de­ rives the erroneous date o f June for Julian’s arrival in Seeck, Regesten (1919), 2 10 , repeated in Matthews, A m m ian us (1989), 540, n. 13, although his text has the correct month (409). 99 Libanius, Orat. 1 2 . 7 6 - 7 7 ; 1 7 1 9 ; 1 8 .1 6 4 - 6 5 . 100 Libanius, E p . 1402.3. Ammianus records several actions by Hormisdas as a general dur­ ing the Persian expedition (24.1.2, 1.8, 2.4, 2 .1 1 , cf. Zosimus 3 . 1 5 . 4 - 6 , 18 .1, 23.4, 29.2).

T

he

N

e w

A

c h il l e s

The only strategic surprise that Julian could achieve was the precise route o f invasion. However, his range o f options was very limited: there was no pos­ sibility o f a sudden breakthrough in a lightly defended sector o f the sort that the German tank divisions achieved in France in 1940. Julian must advance either down the Euphrates or close to the Tigris. His plan was to keep Shapur guessing by advancing to Carrhae before splitting his forces. Ammianus notes that he went from Antioch to Hierapolis “ by the usual route” (23.2.6), then to Carrhae: Julian wrote to Libanius from Hierapolis describing his reception in Litarba, a village o f Chalcis, Beroea, and Batnae on the way there.101 The main expeditionary force then swung south, crossed the Euphrates, and marched downriver along its right bank in tandem with the fleet. What did Julian intend to do with this fleet after the outward march? This is the crux interpretationis for understanding his strategy. Ammianus goes badly wrong, and he is dutifully followed by most modern writers. Ammianus was puzzled by Julian’s destruction o f the fleet in front o f Ctesiphon: he depicts it as an outright error that Julian tried too late to rectify (24.7.4-6). In reality, Julian must have intended to destroy his fleet as soon as it ceased to be useful and became an encumbrance.10210 3Julian advanced down the Euphrates, used the Naarmalcha canal to cross to Ctesiphon, planning to return to Roman terri­ tory by the end o f the campaigning season (23.2.5). His intended return route was along the left or northern bank o f the Tigris, then through the foothills o f the Zagros Mountains, where he expected to find sufficient supplies o f food.10-' He could not take a fleet either up the Tigris (it flowed too strongly) or away from the river into the foothills: therefore, because he must either abandon his boats to the enemy or destroy them, his chosen strategy dictated their destruction. Did Julian intend to capture Ctesiphon? Probably not: he had neither a large enough army nor adequate siege equipment to take the city by storm— and a long siege was unimaginable. If he had a rational strategy, it was rather to menace Ctesiphon, thereby compelling Shapur to come to its rescue. Julian surely intended to fight Shapur in pitched battle, in which the Romans had 101 Julian, Ep. 98, cf. Zosimus 3 . 1 2 . 1 - 2 . O n his route, see F. Cum ont, Etudes syriennes (Paris, 19 17 ), 1 - 3 3 , who noted that the absence o f any mention o f the famous shrine o f Atargatis in­ dicates that a Christian copyist has omitted an important part o f the letter ( 2 3 - 2 4 ) . 102 The destruction is presented as a planned and deliberate action by Libanius, Orat. 1 8 . 2 6 1 63; Zosimus 3 .2 6 .2 -3 . N . J. E. Austin, Athenaeum, N .S. 50 (1972), 3 0 1 - 9 , plausibly suggests that Ammianus’ narrative combines two different and partly contradictory versions o f Julian’s actions at Ctesiphon— that o f the high command and that o f officers and men who were not privy to their deliberations and decisions. 103 Lee, Information (1993), 8 7 - 8 9 .

[164]

T

he

N

e w

A

c h il l e s

often prevailed over Persian forces. He would have his back to Ctesiphon, but he would be protected from attack in the rear by the river. Shapur frustrated this plan by declining to march quickly to the relief o f his capital. Instead, he remained in northern Mesopotamia and dealt with the army under Procopius and Sebastianus, who were instructed to join Arsaces, the king o f Armenia (23.3.5, cf. 2.2), and to advance through Corduene to facilitate Julian’s with­ drawal to Roman territory. Libanius complained that Arsaces played Julian false and that Procopius and Sebastianus failed to carry out their orders be­ cause o f personal quarrels with each other.104 Military action by Shapur is an equipollent explanation. Unfortunately, Ammianus’ account o f the activities o f this army, whatever it was, is lost in a lacuna (24.7.3). A rational strategy can be discerned behind Julian’s actions when they are matched with maps o f Mesopotamia.105 After advancing down the Euphrates, he planned to return up the northern bank o f the Tigris. But, if rational, his strategy was excessively optimistic, and Shapur countered it effectively. Al­ though Ammianus was with the army, he failed to understand Julian’s strat­ egy— or that his generals had retrospectively foisted on his dead hero re­ sponsibility for a catastrophe that could still have been avoided at the time o f his death. For disaster followed on their decision to attempt a shorter and more direct route back to Roman Mesopotamia rather than following the longer, but safer, route north o f the Tigris into the foothills o f the Zagros Mountains, where Julian expected to procure adequate supplies o f food and water. Sha­ pur starved the Roman army into submission only after the death o f Julian. A generation later, Ammianus could not recapture the genuine atmosphere o f the late summer and autumn o f 363, when Jovian’s subjects genuinely wel­ comed his accession as emperor (Chapter XII). The imperial coinage cele­ brated the safe return o f most o f the expeditionary force that Julian had led into Persia as if it were a Roman victory won by the new Augustus.'06 And Jovian was saluted by the Syriac poet Ephrem for bringing solace to the afflicted and terrifying their oppressors.107 104 Libanius, Orat. 18.260. los Ridley, Historia 2 2 (197 3), 3 1 7 - 3 0 , cf. M . A. Beek, Atlas o f Mesopotamia, trans. D. R. Welsh (London/Edinburgh, 1962), 48 (physical map and soil types); Atlas o f the M id d le E ast , ed. M . Brawer (N ew York/London, 1988), 82 (agriculture in modern Iraq). 106 K. Ehling, K lio 78 (1996), 1 8 6 - 9 1 . 107 Ephrem, C arm ina N isibena 2 1.14 , 2 1 - 2 3 (ed. E. Beck, C S C O 2 18 = Syri 92 [19 6 1], 57, 5 8 - 5 9 , with a German translation in C 5 C O 2 19 = Syri 93 [19 61], 70, 7 2 - 7 3 ) . Ephrem appears to be writing before he realized that Jovian had agreed to surrender Nisibis to the Persians (Beck, ibid, iii-iv).

[XIV] PAST, PRESENT, AND FUTURE

An officer and a gentleman is not normally expected to be deeply versed or profoundly interested in philosophy. By the same token, however, impreci­ sion or inconsistency does not prove that such a man’s apparent interest in phi­ losophy is merely superficial or second-hand, as has been assumed in most modern discussions o f Ammianus’ thought world and his “ religious and moral universe.” 1 Ammianus’ philosophical views (so it is normally held) are not the product o f independent reflection, for they derive from handbooks, from pop­ ularized Neoplatonism, from Cornelius Labeo: the historian had never read any Plotinus or Porphyry in the original Greek, but knew them only through the translations or adaptations by Nicomachus Flavianus.12 Such assumptions have never been proved and ought to be rejected. John Matthews, to his credit, has broken with scholarly tradition: he argues that Ammianus’ allusion to Plo­ tinus “ suggests more careful reflection on certain issues” and treats Ammianus’ “ Neoplatonic learning” as something that he acquired for himself.3 But it is R. L. Rike who has provided the proof that Ammianus has a personal creed that is the product, not o f borrowed learning, intellectual confusion, or apa­ thy, but o f conscious decision and the exercise o f thought.4 1

P.-M . Camus, A m m ien M arcellin. Témoin des courants culturels et religieux à la f i n du I V e siècle

(Paris, 1967)» 266: “ l’historien est resté sur le seuil de la philosophie sans y pénétrer réellement.” Camus* book is a remarkable achievement: he wrote it at the age o f twenty-three, but was killed in a traffic accident on 18 August 1964 before he could amplify and deepen his researches: they would (I suspect) have led him to conclusions very similar to my own. 2 For such belittling o f Ammianus’ erudition, see Chapter VII, at nn. 5 1 - 6 1 . 3 Matthews, Am m ianus (1989), 4 3 4 - 3 5 . 4 Rike, A p e x O m nium (1987), 8 - 3 6 . His conclusions are rejected by Matthews, A m m ianus (1989), 545, n. 10.

[166]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

Ammianus has a fundamentally consistent metaphysical scheme, which he applies to his evaluation o f historical events and historical actions. The two cru­ cial passages occur in the obituary o f the Caesar Gallus in Book X IV and the excursus on divination in Book X X I. Both passages present a divine hierar­ chy or triad. In the first, the just celestial godhead (superni numinis aequitas) watches over events, while subordinate to it Adrasteia or Nemesis, whom the ancient theologians depicted as the daughter ofjustice, administers the fates and elements (14 .11.2 5 -2 6 ). In the second, an omnipresent force o f undiminished vigor, Jupiter, shares with men the gifts o f divination, while Themis, whom the ancient theologians made his bedfellow, presides over those powers and publishes the decrees fixed by fate (21.1.8). These two hierarchies are not only related, but are in fact similar and com­ plementary: they are “ two aspects o f a single theological scheme, which the historian first describes from the viewpoint o f aequitas, the descent o f Fate upon men, then from that o f benignitas, the religious and intellectual ascent o f men to the gods through a generously shared knowledge o f divine will.” 5 In both passages, Ammianus carefully aligns himself with the theologi veteres: he belongs to an ancient and noble tradition, and he did not intend to say any­ thing new or original. The second passage is the key to Ammianus’ philo­ sophical viewpoint. What he says about Themis and tetheimena is strikingly similar to a passage o f Porphyry known only because Eusebius’ Preparation for the Gospel quotes it.6 Recent exegetes assume that Ammianus cannot have read Porphyry: hence Joachim Szidat posits an intermediary as his direct source,7 while the Dutch commentators observe that “ Porphyry was a polymath who gathered his data from many sources, which may also have been accessible to Ammianus.” The latter observation may be factually correct. But it ignores something both startling and relevant about the word tethimena in the text o f Ammianus. In the corpus o f Greek literature, the perfect participle passive o f the Greek verb tithemi occurs just over one hundred times, but has the meaning “ fixed decrees o f fate” only in this one passage o f Porphyry quoted by Eusebius.8 It is probable, therefore, that Ammianus knew Porphyry di­ rectly9, and there is no good reason to deny that he read him in the original Greek (Chapter VII).

5 Rike, Apex Omnium (1987), 15. 6 Eusebius, P E 3 . 1 1 . 1 4 = Porphyry, frag. 3 5 8 .3 4 - 3 7 , Smith = De cultu simulacrorum, frag. 7, p. 10 * Bidez. 7 J. Szidat, Mus. Helv. 39 (1982), 1 3 6 - 4 4 . From similarities between Servius and Macrobius, he infers that Ammianus knew Porphyry only through his western interpreters. 8 I rely on a search in the Thesaurus Linguae Graecae.

i} For other points o f similarity, see Szidat, Mus. Helv. 39 (1982), 1 3 9 - 4 5 .

[167]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

Ammianus puts before his readers’ eyes a Roman Empire in which, despite everything, the old religion is still alive.101 At the shrine o f Mopsus in Cilicia cures are still performed: “ his heroic manes cure a variety o f diseases” (14.8.3). That is not merely an antiquarian observation: in his panegyric on Theodorus, Gregory o f Nyssa claims that the martyr is now in heaven “ seeking from God on our behalf what will benefit us” and that he has made the church at Euchaita a place o f healing for various diseases.11 An Apis bull was found in Egypt in 362. Ammianus describes its distinctive markings (a crescent moon on the right flank) and the customs attending its discovery as if they still per­ sisted with undiminished vigor: an Apis bull is brought to Memphis, “ a city famous for the frequent visits o f the god Aesculapius” and “ the inhabitants o f those parts believe” that the finding o f an Apis bull “ portends a good harvest and various other blessings” (22.14 .6 -8). One episode at Rome is especially significant. When Tertullus was praefec­ tus urbi, a severe shortage o f food threatened. Riots ensued, in which the pre­ fect was frequently mobbed by menacing crowds, even though it was obvious that he could do nothing about the rough weather at sea and contrary winds that were holding up the grain fleet. Finally, in despair, a weeping Tertullus produced his small sons and addressed the rioters in tears:

“ L o o k at y o u r fe llo w - c it iz e n s , w h o w i l l sh a re y o u r f a t e — m a y th e g o d s in h e a v e n a v e rt th e o m e n !— u n less a h a p p ie r fo r t u n e s h in e s u p o n u s. S o , i f y o u t h in k th at n o d isa ste r c a n h a p p e n i f t h e y a re p u t o u t o f th e w a y , h e r e th e y a re at y o u r d is­ p o s a l.” S o o t h e d b y th is p itifu l s p e e c h , th e m o b , w h ic h is in c lin e d to c le m e n c y b y its v e r y n a tu re , fe ll s ile n t, a n d c a lm ly a w a ite d t h e ir fu tu r e lo t. S o o n a fte rw a rd s, at th e w is h o f th e d iv in e p o w e r w h ic h has a tte n d e d R o m e f r o m its c ra d le a n d g u a r a n t e e d th at it w ill b e e te rn a l, w h ile T e rtu llu s w a s s a c r ific in g in th e te m p le o f C a s t o r a n d P o llu x at O s tia , th e sea b e c a m e s m o o t h a n d th e w in d c h a n g e d to a lig h t s o u th e r ly b r e e z e : th e sh ip s e n t e r e d h a r b o u r u n d e r fu ll sail a n d r e p le n is h e d th e w a r e h o u s e s w it h g ra in . ( 1 9 . 1 0 . 3 - 4 )

Whether this story is true or not does not matter. The significant fact is that Ammianus chose to include something that has been construed as an aretal-

10 P. Brown, Authority and the Sacred: Aspects o f the Christianisation o f the Roman World (Cam ­ bridge, 1995), 3-2: Ammianus failed or refused to see “ the writing on the wall for a whole, nonChristian way o f life and worship.” 11 Gregory o f Nyssa, De sancto Theodoro, ed. J. P. Cavarnos, G regorii Nyssetti Opera (Leiden, 1990), 6 9 .2 3 - 2 4 (P G 46.745)-

[168]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

ogy o f Castor and Pollux.12*It is a story that fits the well-known pattern o f di­ vine response to sacrifice or prayer. Ammianus included it because he firmly believed that the gods intervene actively in human affairs. Ammianus held that ancient learning was not yet dead despite its deca­ dence. When he digresses on earthquakes, he appeals to the veteres who ap­ proached closer to the truth than “ our common ignorance” or the writers about physics who always disagree with one another (17.7.9: haec nostra vul­ garis inscitia . . . sempiterna . . . physicorum iurgia). The tradition to which Ammianus regarded himself as heir went back beyond Greco-Roman antiq­ uity to ancient Egypt before the alphabet was invented: he commends hiero­ glyphics and he quotes Hermapion’s Greek translation o f those on the obelisk in the Circus Maximus, in which the Sun gives King Rameses dominion over all the earth and salutes him as Apollo and Ares ( 17 .14 .17 —23). In Ammianus’ own day, ancient learning continued to flourish, at least in Alexandria: A lt h o u g h m o s t, in c lu d in g th o s e w h o m

I have

m e n tio n e d , flo u r is h e d l o n g a g o ,

e v e n n o w v a r io u s b ra n c h e s o f le a r n in g are n o t sile n t in th e city. F o r th e te a c h e rs o f d iffe r e n t d is c ip lin e s still s o m e h o w sta y a liv e ; h id d e n tru th is b r o u g h t to lig h t b y th e g e o m e t e r ’s r u le ; a m o n g th e m m u s ic has n o t w h o l ly d r ie d u p n o r h a r m o n y re la p se d in to sile n c e ; s o m e , t h o u g h n o t m a n y , still k e e p w a r m th e s tu d y o f th e m o v e m e n ts o f th e e a rth a n d th e stars; th e re are o th e rs w h o are s k ille d in a r ith ­ m e tic ; an d in a d d itio n to th ese a f e w are s k ille d in th e s c ie n c e w h ic h re v e a ls th e w a y s o f th e fates. T h e s tu d y o f m e d ic in e t o o , w h o s e assistan ce is o fte n n e e d e d in th is life o f o u rs , w h ic h is n e it h e r fru g a l n o r s o b e r, a d v a n c e s d a ily , so th at f o r a d o c t o r m e r e ly to say th a t h e w a s tra in e d in A le x a n d r ia is s u ffic ie n t r e c o m m e n ­ d a tio n o f h is sk ill. ( 2 2 . 1 6 . 1 7 —18 )

Ammianus appears to allude to the mathematician Pappus and to Theon, the father o f Hypatia, who composed arithmetical and mathematical works, a commentary on Ptolemy’s Almagest, and much else besides, including a trea­ tise on divination.1-' Phrase after phrase either states or implies a sad decline in the ancient tradition— quodam modo; nondum . . . penitus exaruit; apud quosdam licet raros; pauci. The one exception is Alexandrian medicine, which appears to have flourished more in Ammianus’ day than it had in the third century.14 12 J. Vogt, Ablt. Mainz, Geistes- und sozialwiss. Kl. 1963, 814. Suda II 265, 0 205 (4.26, 2.702 Adler). On Pappus and Theon, see esp. A. Jones, Pappus

o f Alexandria: Book 7 o f the Collection (Sources in the History of Mathematics and the Physical Sciences 8, 1986), 1 . 1 - 1 5. 14 V. Nutton, Clio Medica 7 (1972), 1 6 5 - 7 6 .

[169]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

Ancient learning was especially valid and relevant where it provided an in­ sight into the divine order o f things, that is, in divination. Ammianus intro­ duces an excursus on divination to defend his hero Julian against the malevoli who accused him o f practicing the black arts to discover the future. The his­ torian protests that it is legitimate for a wise man to master this important branch o f learning: T h e sp irit th at ru le s th e e le m e n ts in a s m u c h as th e y a re e te rn a l b o d ie s a n d is a lw a y s a n d e v e r y w h e r e a c tiv e in fo r e s e e in g e v e n ts sh ares w it h us th e g ifts o f d iv ­ in a t io n th r o u g h o u r p u rs u it o f v a r io u s d is c ip lin e s , a n d th e p o w e r s o f re a lity , w h e n th e y are p r o p itia te d b y d iv e rs e rite s , s u p p ly w o r d s o f p r o p h e c y to m o rta ls as i f fr o m th e v e in s o f n e v e r - fa ilin g s p rin g s . O v e r th e se th e d iv in e p o w e r T h e m is is said to p re s id e , w h o is so n a m e d b e c a u s e sh e m a k e s k n o w n b e fo r e h a n d w h a t is o r d a in e d b y th e f ix e d d e c re e s o f d e a d ly fa te ( w h ic h in G r e e k a re c a lle d tethimena) a n d to w h o m th e a n c ie n t t h e o lo g ia n s a ssig n e d a p la c e in th e b e d a n d o n th e th r o n e o f J o v e , th e life - g iv in g fo r c e . A u g u r ie s a n d a u sp ic e s a re n o t c o lle c te d b y th e w ill o f b ird s, w h o h a v e n o k n o w le d g e o f th e fu tu re . ( N o t e v e n a to ta l f o o l w o u ld assert th at.) O n th e c o n ­ tra ry , G o d d ire c ts th e flig h t o f b ird s, so th at th e n o is e fr o m th e ir b e a k s o r th e m o v e m e n t o f th e ir w in g s , w h e t h e r v io le n t o r g e n tle , m a y s h o w th e fu tu r e in a d v a n c e . F o r , b y th ese arts t o o , th e g r a c io u s d e it y lo v e s to r e v e a l im p e n d in g e v e n ts, e it h e r b e c a u s e m e n d e s e rv e to k n o w o r b e c a u s e h e is t o u c h e d b y a ffe c ­ tio n fo r th e m . In th e sam e w a y th o se w h o in sp e c t th e p r o p h e t ic e n tra ils o f a n im a ls k n o w th e c o u rs e o f e v e n ts fr o m th e c o u n tle s s sh ap es w h ic h th e y ta k e . T h e d is c o v e r e r o f th is d is c ip lin e , a m a n c a lle d T a g e s , a c c o r d in g to fa b le , w a s s e e n to s p r in g s u d ­ d e n ly o u t o f th e g r o u n d s o m e w h e r e in E t r u r ia . C o m i n g e v e n ts are also re v e a le d w h e n m e n s h e a rts a re o n fir e a n d u tte r d i­ v in e w o r d s . F o r th e su n , w h ic h is th e m in d o f th e u n iv e rs e , as n a tu ra l p h ilo s o ­ p h e rs say, sen d s o u t o u r m in d s fr o m it s e lf lik e sp ark s: w h e n it k in d le s t h e m in to fla m e , it m a k e s th e m a w a re o f th e fu tu re . H e n c e th e S ib y ls o fte n say th at th e y b u r n w it h g re a t fla m es ru s h in g v io le n t ly t h r o u g h th e m . M u c h is also s ig n ifie d b y th e s o u n d o f v o ic e s a n d sign s w h ic h w e e n c o u n te r , as w e ll as b y t h u n d e r , lig h t ­ n in g a n d th u n d e rb o lts a n d e q u a lly b y th e tra c k s o f m e te o rs . ( 2 1 . 1 . 8 - 1 1)

This excursus has a clearly articulated logical structure. First comes the cos­ mological basis o f divination. Then its four domains, divided between the artificial (augury and haruspicina) and the natural (prophecies and dreams). Fi­ nally, Ammianus defends divination against scepticism. He carefully combines Roman and Greek methods o f predicting the future, invoking Tages and the

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

Sibyls and quoting Aristotle and Cicero.15 There is no mention here, how­ ever, o f astrology. The reason cannot be that Ammianus denied its validity, which he asserts elsewhere (23.6.25),16 but because in this sphere he accepted complete determinism. I f our fate can be completely predicted from the po­ sition o f the stars at our birth, knowledge o f our horoscope can be o f no prac­ tical use. The types o f divination that Ammianus discusses relate rather to di­ vine warnings that allow the individual some scope for freedom o f action within a framework immutably laid down by Themis. Ammianus seems to be inconsistent here. But so (it may be argued) are all determinists, like the disciples o f Karl Marx who strove, plotted, even died endeavoring to bring about a revolution that their creed proclaimed to be in­ evitable even if they never lifted a finger.17 For, if all our actions are deter­ mined by fate, heredity, or any other pervasive forces beyond our control, or predestined by God, then it makes no sense to exercise choice, since freedom o f the will is mere illusion, our apparent choices due to no act o f our will. But none o f us can avoid conducting our lives on the assumption that we do in fact make choices. The compatibilist position espoused by Ammianus had a long and respectable pedigree in Greek philosophy.18 “ The future is dark, the present burdensome; only the past, dead and finished, bean contemplation.” Ammianus would heartily have agreed with this ver­ dict o f a modem practitioner o f the historian’s art.19 Growing up in the newly Christian Roman Empire, Ammianus could not avoid contemplating the course o f human history. The dominant ideology o f his youth was Christian triumphalism. In his Chronicle, in his Ecclesiastical History, in his General Ele­ mentary Introduction, in his vast double polemic against Porphyry, the Prepara­ tion for the Gospel and the Demonstration of the Gospel, and in his more popular 15 O n the different types o f divination practiced in the Greco-Roman world, see still A . Bouché-Leclercq, Histoire de la D ivination dans l'A n tiq u ité 1 (Paris, 1879), 1 1 1 - 3 7 4 ; 2 (1882), I — 3 17 . Discussion o f this excursus has too often concentrated on identifying its hypothetical written sources: e.g., Ensslin, Am m ianus (1923), 8 3 —96. 16 Ensslin, A m m ian us (1923), 7 7 - 8 2 . A. Bouché-Leclercq, L ’Astrologie grecque (Paris, 1899), 570, adduced Juvenal as a precedent and possible model for Ammianus’ ridicule o f Roman aris­ tocrats who consult the stars before going out to dine or to bathe (28.4.24). 17 For three classic modem discussions o f the problem, see K. Popper, D ie O pen Society and Its Enem ies (London, 1945), chap. X X II; I. Berlin, Historical Inevitability (London, 1954); E. H. Carr, W hat Is H istory? (London, 1961), 8 5 - 9 3 .

18 R. Sorabji, Necessity, Cause and Blam e: Perspectives on Aristotle’s D ieo ry (Ithaca, 1980), traces it back through the Stoics to Aristotle. 19 G. R. Elton, D ie Practice o f H istory (Sydney, 1967), 1. W hen that writer turns to questions o f appropriate size and scale, he notes that “ some 500-page books turn out on further inspec­ tion to be inflated articles” (123).

[171]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

Theophcmy, Eusebius o f Caesarea set out an idiosyncratic view o f history. Christianity, so Eusebius maintained, was the primeval religion o f the human race, identical with the religion o f the patriarchs o f the Old Testament, from which the traditional religions o f the Greco-Roman world were mere o ff­ shoots or declensions. Divine providence had prepared the world for the com­ ing o f Christ by means o f the temporary dispensation o f Judaism and had then ensured the worldly success o f Christ’s church, which constituted a race or nation o f a unique type with a special relationship to God.20 Even in Eusebius’ day, few Christians can have accepted his theories about religion in the Old Testament. No contemporary Christian, however, doubted the central propositions o f his overall view o f human history: God had sent his Son to save men, and divine providence either guided or intervened in hu­ man history to produce the success o f the Christian church. In the East, the church’s victory was very recent when Ammianus was born, less than ten years after Constantine defeated “ the last o f the persecutors.” Christians had come to power in the cities o f the East through war and violence, as the regime o f Licinius first harassed the church, then collapsed when Constantine invaded his territory. In the revolutionary situation that existed in 324 —325 the new master o f the East acted decisively to establish Christianity as the official reli­ gion o f the Roman Empire.21 Christianization proceeded apace after 325. Constantine confiscated the treasures laid up in temples over the centuries and used the proceeds to finance an enormous program o f church building throughout the East, but especially in Palestine and above all in Jerusalem.22 Under Constantine, few cults or temples were suppressed by imperial action, but bishops began to attack shrines and temples with increasing freedom as time passed.23 The brief reign o f Julian suddenly interrupted the process, and Christians in the reigns o f Jovian and Valens showed more restraint than un­ der Constantius. But with the accession o f Theodosius in 379, the suppres­ sion o f pagan cults recommenced in earnest. N ow it was not merely bishops and monks attacking local cults: the praetorian prefect Cynegius toured the East twice (in 384—386 and again in 388), encouraging and supervising the demolition o f temples.24 In 391, the Serapeum in Alexandria succumbed,

211 For a full exposition, see J. Sirinelli, Les Vues historiques d ’Ensèbe de Césarée durant la période prênicéennc (Dakar, 1961); more briefly, T. D . Barnes, Constantine (19 8 1), 1 2 6 —36, 1 7 9 —88.

21 For this interpretation, which is by no means universally shared, see Constantine (19 81), 2 0 8 - 1 2 , 245 —60; From Eusebius to A ugustine: Selected Papers (Aldershot, 1994), nos. I V —IX , all composed with the modern parallel o f Iran after the fall o f the Shah and the return o f Ayatol­ lah Khomeini very much in mind.

22

Constantine (19 81), 2 4 7 - 5 0 .

22 G . Fowden,_/TS, N .S. 29 (1978), 5 3 - 7 8 . 24 Matthews, Aristocracies (1975), 1 4 0 - 4 5 .

[172]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

which Ammianus had recently advertised as one o f the wonders o f the world (22.16.12).25 Against this background, any pagan o f the later fourth century needed to formulate a coherent interpretation o f history if he wished to challenge the dominant ideology. Julian adumbrated one in his Caesars by placing per­ sonified Luxury and Wantonness next to the Jesus who welcomed seducers, murderers, the sacrilegious, and infamous, including Constantine.26 Eunapius soon developed these hints into an interpretation o f the fourth century in which Rome was strong and successful until the conversion o f Constantine, who refused to celebrate the Secular Games when they fell due in 314 and then totally weakened, demoralized and corrupted the Roman Empire. The original text o f Eunapius’ history is lost except for sporadic fragments, but Zosimus who wrote at the start o f the sixth century faithfully reproduced his views (and often his words) in the central section o f his history: hence Eunapius could with some plausibility be styled “ the first historian o f Rome’s fall.” 27 If the first edition o f Eunapius’ history ended with the Battle o f Adrianople (as seems probable),28 then Ammianus could consult it, and he may well have Eu­ napius in mind when he complains that those who are ignorant o f ancient times (antiquitatum ignari) allege that the defeat o f Valens in 378 was the dark­ est moment in Rome’s history (3 1.5 .11). Although he was Greek, Ammianus took the whole history o f Rome from its founding by Romulus within his purview. At the start o f his excursus on Rome in Book XIV, he states his belief in the eternity o f Rome and compares the history o f the city to a human life (14.6.3 -5 ). In its infancy and childhood, a period o f three hundred years, the Roman people waged local wars. In ado­ lescence it expanded beyond Italy; in manhood it conquered the world. Then, when it was verging on old age and able to triumph sometimes through rep­ utation alone, it retired (ad tranquilliora vitae discessit).29 Accordingly, the city 2z> Th e date is certain, despite recent attempts to argue for 392 (C P 88 [1993], 6 1 - 6 2 ) . 26 Julian, Caes. 336a, cf. E p . 89, 453c. 27 W . A . Goffart, A H R 76 (19 71), 4 1 2 - 4 1 , applied the sobriquet to Zosimus, but every single one o f his proof texts comes from the Eunapian section o f Zosimus ( 1.4 7 -5 .2 6 ). 28 Eunapius produced two editions o f his history, one before he wrote his Lives of the Philoso­ phers and Sophists in 399 and one that ended in 404 and was used by Zosimus. The traditional date o f the first edition (circa 395) is ably defended by F. Paschoud, Bon­ ner Historia-Augusta-Colloquium 1977-1978 (Bonn, 1980), 1 4 6 - 6 2 ; Bonner Historia-AugustaColloqninm 1982-1983 (Bonn, 1985), 2 5 3 - 8 4 ; Zosime 3.2 (Paris, 1989), 9 0 - 9 1 . In favor o f an earlier terminus, see C P 7 1 (1976), 2 6 5 - 6 7 ; The Sources o f the Historia Augusta (Collection Latomus 155» 1978)» 1 1 4 - 2 3 ; Constantine (19 81), 403, n. 5; T. S. Banchich, G R B S 2 1 (1986), 3 1 9 - 2 4 . A . Baker, G R B S 29 (1988), 3 8 9 -4 0 2 , argues that the “ new edition” that Photius inspected was a bowdlerized version by a later hand, not one o f the two authentic editions. 29 Ammianus uses the phrase tranquilliora vitae only once elsewhere, to refer to the death, not the retirement, o f Varronianus, the father o f Jovian (25.5.4).

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

behaved as a rich parent who is thrifty and wise behaves to his children: it ceded control o f its estate to the emperors so that they could administer it (Cciesaribus tamquam liberis suis regenda patrimonii iura permisit) ,30 Very similar comparisons survive in three other Latin authors besides Am­ mianus: Florus and Lactantius, who wrote before Ammianus, and the Historia Augusta after him.31 This Lebensaltervergleich has been much discussed.32 But much o f the discussion has been misguided as far as Ammianus is concerned. For Ammianus, who echoes him verbally, took the idea from Florus, who seems in turn to have adapted Seneca.33 Moreover, the evidence o f Cicero es­ tablishes the equation “ old age equals weakness” 34— which some exegetes o f Ammianus have doubted.35 It follows that Ammianus’ message is a pessimistic one. Florus asserted that under the emperor Trajan, in whose reign he was probably writing, the Roman Empire had bestirred itself to action and that, contrary to the expectation o f all, its old age was revived as if it had regained 30 Despite the consensus o f modern editors, the paragraph should end with these words: the next sentence, which refers to the cessation o f popular elections and the Senate as reflecting the prestige o f Rome (14.6.6), belongs with what follows and introduces the long denunciation o f the vices o f contemporary Rome (14 .6 .7 —26); it does not conclude the comparison o f Romes history with four stages in a human life. Similarly, the next two paragraphs should be 1 4 .6 .9 10 and 1 1 - 1 4 (not 9 - 1 1 and 1 2 - 1 4 ) , and 1 4 .6 .1 8 - 2 0 should form a single paragraph (not two, as in Seyfarth s text). 31 Florus, pr. 4 - 8 ; Lactantius, D iv . hist. 7 . 1 5 . 1 4 - 1 6 ; H A , Cams 2 - 3 . Th e Seneca whom Lactantius quotes is usually assumed to be the Elder Seneca, who died in 39. 32 See esp. R. Haussier, Hermes 92 (1964), 3 1 3 - 4 1 ; Demandt, Zeitkritik (1965), 1 1 8 - 4 2 ; Metaphern fü r Geschichte: Sprachbilder und Gleichnisse im historisch-politischen Denken (Munich, I 9 7 8 )> 3 7 —45; Gymnasium 87 (1980), 1 7 8 - 8 9 . The latter argues that Ammianus depicts the old age o f Rome as the apogee o f the city, a state o f eternal, unaltering prosperity from which all ele­ ments o f decadence have disappeared (Metaphern [1978], 39). 33 H. Finke, Ammianus Marcellinus und seine Quellen zur Geschichte der römische Republik (Diss. Heidelberg, 1904), 3 8 - 4 0 ; Drexler, Ammianstudien (1974), 15 5 —Ö5. 34 Drexler, Ammianstudien (1974), 1 6 0 - 6 2 , adducing Cicero, A d Quintum fratrem 2 .1 3 .5 ; De re publica 1.58, 2.3, 2 .2 1, 3.34. 35 S o J. F. Matthews, The Inheritance of Historiography (Exeter, 1986), 22: “ Ammianus is not talking, as a cursory reading might suggest, o f a decline but o f a rejuvenation o f Rome in the persons and office o f the emperors; it is through their eflorts in conducting wars in the name o f Rome (nomine solo aliquotiens vincens) that Rome will live— or conquer: victura— as long as there are men.” This interpretation fails to respect the order o f the text: Ammianus attributes the vic­ tories won nomine solo to the late prime o f the Roman people, before it ceded control to the em­ perors. Moreover, the phrase victura dum erunt homines Roma does not contain a pun on two meanings o f the future participle victura. For it occurs at the very beginning o f the excursus be­ fore any mention o f victories: it means simply “ Rome will live as long as the human race.” But in what sense will Rome live on? Even if Ammianus is not using vivere in this and a later passage (26 .1.14 : victura cum saeculis Roma) metaphorically, as Ennius had in his famous funerary epi­ gram (Varia 18 Vahlen: volito vivus per ora virum), it need mean no more than “ exist” or “ live on in an increasingly squalid existence” : the passage is correctly interpreted by G. B. Conte, His­

tory of Latin Literature, trans. J. Solodow (Princeton, 1994), 649.

P

a s t

, P

r e s e n t

,

a n d

F

u t u r e

its youth.36 Ammianus has removed the optimistic conclusion o f his immedi­ ate model. It must be a deliberate choice. Ammianus refers to the heroes o f Republican Rome in many passages, but usually as examples o f behavior or in geographical excursus.37 When Julian, about to invade Persia, invokes M. Curtius, Mucius Scaevola, and the noble Decii (23.5.19), these are standard exempla o f patriotism and devotion to Rome, whose function is primarily rhetorical. There appears to be only one passage in the Res Gestae that offers serious reflections on the course o f Roman his­ tory over the centuries, but it is a passage o f immense significance. Ammianus compares the disaster o f Adrianople to earlier Roman disasters, both Repub­ lican and imperial:38 T h o s e w h o a re ig n o r a n t o f a n c ie n t tim e s say th at th is w a s th e d a rk e st d isa ste r w h ic h e v e r fe ll u p o n th e state, b u t th e y a re le d a stra y b e c a u s e th e y a re o v e r ­ w h e lm e d b y t h e ir h o r r o r at r e c e n t d isasters. A r e v ie w o f e a rlie r o r e v e n q u ite r e ­ c e n t h is t o r y w ill s h o w th at s u c h m e la n c h o ly e v e n ts h a v e o fte n h a p p e n e d . T h e T e u t o n e s a n d C i m b r i f r o m th e fu rth e s t s h o re s o f th e O c e a n s u d d e n ly flo o d e d in to Ita ly , b u t a fte r t h e y h a d in flic t e d e n o r m o u s d e fe a ts o n th e R o m a n sta te , th e y w e r e o v e r c o m e b y o u r e x c e lle n t g e n e ra ls in th e ir last b a ttle s, a n d th e y le a rn e d fr o m t h e ir d e s t r u c t io n r o o t a n d b r a n c h in t h e ir last tria ls w h a t m a rtia l p o w e r c a n a c h ie v e w h e n c o m b in e d w it h g o o d ju d g e m e n t . A g a in , w h e n M a r c u s m ie d th e e m p ir e , th e m a d f u r y o f d is c o rd a n t trib e s a c t in g as o n e , a fte r th e c la n g o r o f v a st w a rs, a fte r th e s u ffe rin g s o f c itie s w h ic h w e r e c a p tu re d a n d s a c k e d , p a id th e p e n a lty fo r k illin g a g o v e r n o r < . . . > a n d o n ly a t in y p a rt o f th e ir n u m b e r w e r e le ft u n s c a th e d .39 B u t a fte r c a la m ito u s losses th e s itu a tio n w a s re s to re d in d u e

36 Florus i pr. 8. Th e main clause reads: a Caesare Augusto in saeculum nostrum haud multo minus anni ducenti, quibus consenuit atque decoxit. Hence it is normally deduced that Florus was writing circa 135 : see, e.g., P. Jal, Florus: Oeuvres 1 (Paris, 1967), Lxix-cxi. But a date under Trajan seems historically more plausible. Accordingly, I would reconstruct Florus’ chronological schema as follows: (1) prima aetas sub regibus = infantia lasting 250 years; (2) adulescentia o f 250 years from 509 to 264 B. C .; (3) maturitas o f 200 years quibus totum orbent pacavit plus the reign o f A u­ gustus; (4) nearly 100 years o f senility between Augustus and Trajan (emending ducenti to centum). 37 Th e passages are conveniently collected and discussed by Finke, Ammianus Marcellinus (1904), 6 0 - 9 1 . 38 O n the intellectual context o f this passage, see N . Lenski, TAPA 12 7 (1997), 1 2 9 - 6 8 , who collects and discusses the contemporary reactions to the Battle o f Adrianople that survive. In his official panegyrics o f Theodosius, Themistius first minimized, then magnified the effects o f the Roman defeat; Libanius and Jerome succumbed to despair when they heard news o f the disas­ ter; and Ambrose indulged in apocalyptic musings about the end o f the world. 39 The text printed by Seyfarth is defective in both logic and grammar, and what I offer is not a real translation: the fact that eorum refers to the gentes whom Marcus defeated leads me to posit a lacuna.

[175]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

c o u rs e . T h e re a so n w as th at th e o ld , s o b e r m o r a lit y h a d n o t y e t b e e n in fe c te d b y th e e ffe m in a c y o f a la x e r w a y o f life , a n d th at th e re w a s n o c r a v in g fo r o s t e n ­ ta tio u s b a n q u e ts a n d ill- g o t t e n g a in : h ig h a n d lo w a g r e e in g w it h o n e a n o t h e r w ith u n a n im o u s e n th u sia sm h a ste n e d to w a rd s a g lo r io u s d e a th f o r th e sta te as i f < s t r i v i n g to r e a c h > so m e c a lm an d p e a c e fu l h a rb o r. A lt h o u g h th e h o rd e s o f S c y t h ia n trib e s , w h o in v a d e d b y b u r s tin g t h r o u g h th e B o s p o r u s an d th e sh o re s o f th e P r o p o n tis w it h t w o th o u s a n d sh ip s, w r o u g h t b it ­ te r m assacres b y la n d an d sea, th e y lo st th e g re a te st p a rt o f th e ir n u m b e r a n d re tre a te d . T h e t w o e m p e ro rs D e c iu s , fa th e r a n d s o n , fe ll fig h t in g a g a in st th e b a r­ b a ria n s; th e citie s o f P a m p h y lia w e r e b e s ie g e d , m a n y isla n d s la id w a s te , a n d th e w h o le o f M a c e d o n ia set o n fire ; a vast t h r o n g b e s ie g e d T h e s s a lo n ic a a n d C y z i ­ cu s fo r a lo n g tim e ; A n c h ia lu s w as ta k e n , an d at th e sa m e p e r io d N ic o p o lis , w h ic h th e e m p e r o r T ra ja n fo u n d e d as a m e m o r ia l o f h is v i c t o r y o v e r th e D a ­ cian s. A ft e r m u c h c ru e l sla u g h te r h a d b e e n r e c e iv e d a n d in flic t e d , P h ilip p o p o lis w a s d e s tro y e d w it h o n e h u n d r e d th o u sa n d , u n less th e h is to rie s lie , b u tc h e r e d w ith in its w alls. F o r e ig n fo e s ro a m e d fr e e ly o v e r E p ir u s , T h e s s a ly , a n d th e w h o l e o f G r e e c e . B u t a fte r th e g lo r io u s le a d e r C la u d iu s h a d b e e n m a d e e m p e r o r a n d th e n c a r r ie d o f f b y a n o b le d e a th , th e y w e r e d r iv e n o u t b y A u r e lia n , a n e n e r g e t ic m a n a n d a s e v e re a v e n g e r o f in ju r ie s , a n d r e m a in e d sile n t a n d im m o b ile f o r lo n g ag es, e x c e p t th at o n ra re la te r o c c a s io n s b a n d s o f ro b b e r s r a id e d are as n e a r to t h e m — an d w e r e k ille d . ( 3 1 . 5 . 1 1 - 1 7 )

What is the great change to which Ammianus here alludes? and when did it occur? The answer to these questions is easy. For Ammianus transposes the vocabulary o f denigration used by Julian, Libanius, and Eunapius from Greek into Latin (Chapter VIII). His solutioris vitae mollities corresponds to the Lux­ ury and Wantonness whom Julian depicted as the boon companions o f Jesus as he extended pardon to murderers and adulterers.40 Ammianus blames the decline o f the Roman Empire on Christians and Christianity in a manner very similar to Eunapius,41 though with greater intellectual sophistication. Ammi­ anus asserts that sobria vetustas has been destroyed by Christianity, which is in­ trinsically corrupt. This connection emerges most clearly in Ammianus’ com­ ments on Julian’s reform o f the palace after the death o f Constantius: T h e g re a te r p a rt o f th e m [sc. th e p a la tin i o f C o n s ta n tiu s ] h a d g r o w n w i d e l y s u c h a se e d b e d o f all th e v ic e s th at th e y c o r r u p te d th e state b y th e ir d e p r a v e d lu sts a n d h a rm e d m a n y m o r e b y th e ir b a d e x a m p le th an b y th e im p u n it y w it h w h ic h th e y 40 Julian, Caes. 329a, 336a. 41 Zosimus 2 .14 .2 (failure to celebrate the Ludi Saeculares in 314 ), 2 .3 0 - 3 8 (Constantine’s innovations). Zosimus sums up Constantine s achievement as the defiling o f the state (2.39.1).

[176]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

o ffend ed. S o m e o f them , w h o had g ro w n fat on the plunder o f temples and sm elled o ut p ro fit at e v e ry op portunity, had been raised at a b o u n d from grin d ­ in g p o ve rty to en o rm o u s w ealth: b e c o m in g habituated always to appropriate the p ro p erty o f others, they bribed , robbed, and squandered w ith o u t restraint o f any kind. H e n c e there sprouted in abundance the seeds o f a laxer w a y o f life ( flu x ­

ioris vitae initia) and p erju ry; there was no regard for reputation, and an insane p rid e polluted its credit th rough the crim inal pursuit o f gain. A t the same tim e, glu tto n y g r e w and reckless expen diture on banquets: trium phs in battle w ere re­ placed b y trium ph s at table, there was lavish use o f silk and the art o f textiles was enh anced, the k itch en b ecam e the o bject o f ever k eener attention, and preten­ tious sites for lu xu rio u s houses w ere sough t o u t— sites so large that i f C in cin n a ­ tus had possessed so m u ch farm land, he w o u ld have forfeited the g lo ry o f poverty after his dictatorship. ( 2 2 . 4 . 2 - 5 ) T h e p lu n d e r o f te m p le s w a s n o t m e r e ly a n o p e r a t io n o f p r iv a t e e n te r p r is e : C o n s t a n t i n e h a d s e n t s p e c ia l c o m m i s s i o n e r s t o t o u r e v e r y e a s t e r n p r o v i n c e a n d c o n f i s c a t e t e m p l e t r e a s u r e s a n d o r n a m e n t s , i n c l u d i n g (it is r e p o r t e d ) d o o r s m a d e o f p r e c i o u s m e t a l s . 42

D i o c l e t i a n a n d C o n s t a n t i n e h a v e o f t e n b e e n r e g a r d e d as p o l a r o p p o s i t e s . 43 T h a t w a s t h e o p i n i o n o f e v e r y o n e in t h e f o u r t h c e n t u r y , w h e t h e r C h r i s t i a n o r p a g a n , a n d a n e l o q u e n t m o d e r n r e s t a t e m e n t c o n t r a s t e d D i o c l e t i a n as “ t h e la s t a n d m o s t r a d ic a l d e f e n d e r o f t h e R o m a n i d e o l o g y o f r e s t o r a t i o n ” w i t h C o n s t a n t i n e , t h e r e v o l u t i o n a r y w h o b r o k e w i t h a ll e x i s t i n g t r a d i t i o n s a n d a b a n d o n e d e v e r y t h i n g R o m a n . 44 Y e t r e c e n t r e s e a r c h h a s d i s c o v e r e d a b a s ic c o n t i n u i t y o f a d m i n i s t r a t i o n a n d a d m i n i s t r a t iv e r e f o r m t h a t , e v e n i f it b e g a n in 2 8 4 , p r o c e e d e d a l o n g e s s e n t i a l l y t h e s a m e lin e s a f t e r 3 0 5 , 3 1 2 , 3 2 4 , a n d 3 3 7 . O n e e x a m p l e w i l l a p t l y il l u s t r a t e . Z o s i m u s s a w t h e e s t a b l is h m e n t o f r e g i o n a l p r a e t o r i a n p r e f e c t u r e s as a s u d d e n a n d d e l e t e r i o u s i n n o v a t i o n b y C o n s t a n ­ t i n e 45— w h i c h s o m e m o d e r n h i s t o r i a n s o f R o m e u s e d t o t r a n s f e r t o D i o c l e ­ t i a n . 46 I n f a c t , t h e r e g i o n a l p r e f e c t u r e s o f t h e la t e f o u r t h c e n t u r y c a m e in t o e x ­ 42 Eusebius, Panegyric o f Constantine 8 . 1 - 7 ; V C 3 .5 4 —58, cf. L ’Église et l ’empire au I V e siècle (Entretiens sur l ’Antiquité Classique 34, 1989), 3 2 2 - 3 3 . 43 Fora survey o f ancient opinions, see N . Baglivi, Orpheus, N .S. 12 (1991), 4 2 9 - 9 1 . M . T. W. Arnheim gives successive chapters o f his study o f The Senatorial Aristocracy of the Later Roman Em­ pire (Oxford, 1972) the titles “ Diocletian, Hammer o f the Aristocracy” and “ Constantine the Reformer” ( 3 9 -7 3 ) . 44 A . Schenk von StaufFenberg, Das Imperium und die Völkerwanderung (Munich, 1948), 10 8 -11. 45 Zosimus 2 .3 3 . 1 - 2 . 46 M ost influentially, O. Seeck, Rhein. Mus., N.F. 69 (1914), 1 - 3 9 . Epigraphical discoveries long ago rendered this view untenable (New Empire [1982], 1 2 3 - 3 9 ) .

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

istence gradually in a process that lasted several decades. There were regional prefects in Africa under Constantine, but only as a temporary measure; the lasting transformation o f the prefecture began in the West in 343, spread to the East in 354, and was not completed before the 360s.47 Historians now speak rather o f “ the new empire o f Diocletian and Constantine” and o f “ the political, social and economic evolution o f the Roman world from Diocletian to Julian.” 48 Constantine was not the indiscriminate disturber o f ancient laws and tradi­ tional custom whom Ammianus makes Julian denounce (21.10.8: novator tur­ batorque priscarum legum et moris antiquitus recepti). His radical inno­ vations were largely confined to the religious sphere, but here they were fundamental and far-reaching. Although there was no “ Constantinian revo­ lution,” as some have incautiously imagined, there was a “ Constantinian re­ formation”— with all that that phrase implies.49 Unfortunately, Ammianus’ account o f the reigns o f Diocletian and Con­ stantine has perished with the rest o f the first half o f his history. The few ex­ plicit references to Diocletian in the extant books reveal little about the his­ torian’s attitude.50 He knows that Diocletian settled Carpi in Pannonia (28.i . 5)51 and that the Caesar Galerius walked on foot in front o f the chariot o f the Augustus when the two entered Antioch: like other writers o f the fourth century, he construes this as a deliberate humiliation rather than a sym­ bolic gesture o f subordination (14 .11.10 ).52 He repeats the canard that Dio­ cletian was the first emperor to introduce the ceremony o f adoratio (15.5.18).53 It is true that it was Diocletian who made adoratio a normal part o f court cer­ emonial, obligatory for imperial relatives as well as officials.54 But Ammianus’ 47 Bonner Historia-Augusta-Colloquium, 1984-1985 (1987), 1 3 - 2 3 ; Z P E 94 (1992), 2 4 9 -6 0 . On the still obscure question o f the number and function o f praetorian prefects in the early years o f the fourth century, see J R A 9 (1996), 5 4 6 - 4 8 . 48 A . Chastagnol, VEvolution politique, sociale et économique du monde romain de Dioclétien àJulien: La mise en place du régime du Bas-Empire (284-365) (Paris, 1982): the title o f my New Empire (1982) was chosen in implicit homage to Gibbon, Decline and Fall, Chapter X III (1.38 3 [B] = 1.390 [W]). 49 From Eusebius to Augustine (1994), nos. V, VIII, IX. 50 S. A. Stertz, Studies in Latin Literature and Roman History 2 (Collection Latomus 168, 1980), 5 1 1 - 12, argues that Ammianus gives a negative picture o f Diocletian, while himself happily conflating Maximian and Galerius. 51 For proof that it was Diocletian and not, as often assumed, Galerius, see Phoenix 30 (1976), 187. 52 As argued by W . Seston, R E A 42 (1940)» 5 1 5 - 1 9 . cf. J. W . Eadie, The Breviarium of Festus (London, 1967), 1 4 6 - 4 8 . ' 5,1 Also Victor, Cues. 3 9 .2 - 4 ; Eutropius, Brev. 9.26; Jerome, Chronicle 2 2 6 e Helm. All three authors derive the story from the lost “ Kaisergeschichte.” 54 Lactantius, Mort.Pcrs. 18.9, makes it clear that Maxentius was expected to “ adore” both his father, the Augustus Maximian, and his father-in-law, the Caesar Galerius, cf. A. Alfoldi, Die monarchische Repräsentation im römischen Kaiserreiche (Darmstadt, 1970), 2 5 - 1 1 8 .

[178]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

moral indignation reveals a significant gap in his Latin reading: had he read Suetonius, he would have discovered that L. Vitellius performed adoratio o f Caligula.55 And when he invokes Diocletian in 285 as the most recent prece­ dent for Julian’s sharing o f the ordinary consulate o f 363 with a private citi­ zen, he either forgets or ignores the fact that Maximian did so in 288, three years after Diocletian (Chapter V). It is nevertheless clear that for Ammianus, the reign o f Diocletian belonged to Rome’s glorious and ancient past. He reveals this significant assumption obliquely in a passage criticizing Constantius for entering Rome in triumph in 357 as if all the enemies o f the Empire had been laid low and as if he were coming to close the Temple o f Janus. According to Ammianus, Constantius never defeated any foreign tribe either himself or through his generals, and he never appeared in person in the thick o f battle. He preferred empty ceremo­ nial, ignorant perhaps o f the personal courage shown by some earlier emper­ ors: one entrusted himself to a small fishing-boat at the height o f a raging storm, another vowed to sacrifice his life for the state, and a third had per­ sonally made his way into the enemy’s camp with some common soldiers (i6 .ro .i-3 ). The three emperors whom the degenerate Constantius failed to emulate are easily identified. Caesar’s attempt to cross the Adriatic in a fishingboat at night was a famous episode even before Lucan wrote it up.5657Two other pagan writers o f the fourth century depict the death o f Claudius in 270 as a formal devotio.51 And it was Galerius who reconnoitred the camp o f Narses during his Armenian campaign o f 297.58 Christians had a very different opin­ ion o f Galerius as the main instigator o f the “ Great Persecution” : 59 by calling Galerius glorious and heroic, even though he does not name him, Ammianus marked a sharp break between Rome’s heroic past, which lasted until the reign o f Diocletian, and the corrupt and degenerate empire o f Constantine and his Christian successors. Ammianus took it as proven fact that “ Constantine was the first to whet the appetite o f his courtiers, but it was Constantius who fattened them with the marrow o f the provinces.” For under Constantius the leading men o f the state 55 Suetonius, Vitellius 2.5: primus C . Caesarem adorare ut deum instituit. The passage was duly adduced by Alfbldi, Repräsentation (1970), 39. 56 Lucan 5 .5 0 4 - 6 7 7 , cf. Valerius Maximus, Mein. 9.8.2; Pluarch, Caes. 38; Florus 2 . 3 5 - 3 7 ; Appian, B C 2 .57 ; D io 41.4 6. 57 Victor, Caes. 3 4 .2 —3; Epitome 34,3. SÄ Eutropius, Brev. 9 .2 5 .1; Festus, Brer. 25, cf. N . J. E. Austin and N . B. Rankov, " Explo­

ratio. ” Military and Political Intelligence in the Roman World from the Second Punic War to the Battle of Adrianople (N ew York, 1995), 6 2 - 6 3 . w Lactantius and Eusebius consistently depict Galerius as the moving force behind the antiChristian legislation o f Diocletian: in favor o f accepting their testimony, see now Usurpationen

in der Spätantike (Historia Einzelschriften n i , 1996), 9 9 - 1 0 9 .

[179]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

were consumed by a boundless lust for wealth and totally disregarded justice and right. Ammianus names the worst and most prominent offenders in a la­ cunose passage: Rufinus the praetorian prefect,60 Arbitio the master o f cavalry, the grand chamberlain . . . , the quaestor [. . .]nus,61 and in Rome the family o f the Anicii, whose contemporary descendants rival their forebears in their inability to rest content even with much greater possessions (16 .8 .12 —13).62 Ammianus sees a process o f corruption rather than the sudden change de­ nounced by the Greek historians Eunapius and Zosimus. He blames Con­ stantine for starting it and Constantius for carrying it much further. But in this passage, whatever he may have said in the lost books, he concentrates his odium on the leading Christian family in Rome. The Anicii had risen to prominence with an African supporter o f Septimius Severus, and they were to play an important historical role in the fifth and sixth centuries.63 When Ammianus was composing his Res Gestae, the culmen Aniciorum was the great Sextus Petronius Probus, consul in 371 and four times praetorian prefect, who had married into the clan, and to whom Ammianus devoted a long and hos­ tile character sketch (Chapter X). Ammianus makes it clear that he regarded Christians as thoroughly corrupt in every way. The emperor Julian was o f central importance to Ammianus. But he, no less than Constantine, belonged to a past that was dead. Could Julian have re­ versed the changes in Roman society that had occurred under Constantine and his sons? On the one hand, Ammianus has so constructed his narrative o f the Persian expedition as to lead readers to infer that Julian could have avoided disaster if he had heeded the long series o f omens and portents that the gods sent to warn him: for example, the Sibylline books forbade the emperor to set 00 Vulcacius Rufinus, consul in 34 7 and maternal uncle to the Caesar Gallus ( 14 .11.2 7 ) , was a pontifex maior (IL S 1237) and hence a pagan ( P L R E 1.7 8 2 —3): Ammianus later praises him as “ omni ex parte perfectus,” although he never let slip an opportunity to enrich himself if he thought that he might remain undetected (27.7.2). 61 V has a lacuna o f 19 letters between praepositusque cnbiclui laps and amis quaestor, the latter cannot be identified (P L R E 1.997, -anus), and Eusebius, whose name is often supplied (as in the Penguin translation), is not attested as praepositus sacri cubiculi before the death o f Constantine (Chapter X I, at n. 42). 62 Elsewhere Ammianus makes Julian denounce Constantine as a subversive innovator (21.10 .8 ) and holds him responsible for the defeat o f Julian’s Persian expedition (25.4.23). But eleven other references to Constantine seem to be neutral ( 14 .1.2 , 14 .11.2 0 , 1 4 .1 1 .2 7 , 15 .5 .19 , 1 5 *5 -3 3 »

1 5 1 3 - 2 , 16.7 5. 1 7 -4 - 1 3 » 2 1 1 2 . 2 5 , 2 5 .3 .2 3 , 26.6.14). 3 O11 Q . Anicius Faustus, cos. suff. circa 198, see A. R. Birley, The African Emperor Septimius

Severus2 (London, 1988), 1 4 3 - 4 4 , '4 7 . 15 T - 5 2 ; for his descendants in the third century, see M . Corbier, Epigrafia e ordine senatorio (Tituli 5, 1982, pub. 1984) 7 4 0 - 4 1 ; for Anicii in the fourth, fifth, and sixth centuries, P L R E 1 . 1 1 3 3 : Stemma 7; 2.130 9 : Stemma 3. T h e family came from Uzappa in Africa Proconsularis, as did the Memmii (cf. Chapter X , n. 37).

[180]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

foot outside Roman territory during the consular year 363 (23.1.7).64 On the other hand, Julian disregarded the prudent advice o f Sallustius, the praetorian prefect o f Gaul, to postpone his invasion o f Persia until the gods signified their approval “ because no human strength or virtue has ever availed to forestall what has been ordained by the decree o f destiny” (23.5.5: quod praescripsit fatalis ordo). Yet the obituary o f Julian argues that moral responsibility for his failure rested with Constantine. For it was Constantine, Ammianus argues, who started the long war against Persia as a result o f believing the lies o f Metrodorus. Hence the long series o f disasters under Constantius— Roman armies cut down, units captured, cities destroyed, fortresses seized or demol­ ished, the provinces exhausted by the burdens imposed on them, with the Persians claiming Roman territory up to the shores o f the Propontis. In the West, the insolent Germans were on the point o f breaking through the Alps to ravage Italy, while the provincials were in panic until Julian, sent to Gaul as a Caesar only in name, retrieved the situation with miraculous rapidity. It was with the same eagerness to restore the East that he attacked the Persians, whom he would have conquered, “ had his plans and glorious deeds been con­ sonant with the decrees o f heaven” (25.4.23-26). After the death o f Julian, the Roman Empire was a different place.65 The Alamanni recovered from the losses and damage inflicted by Julian and be­ came a permanent threat to the Gallic frontier (27.1.1). The Persian king not only imposed a shameful treaty on Jovian, but did not even respect its terms and pursued his designs on Armenia, with no small success (27.12.1). There was constant trouble on the Danube, where the admission o f the Goths to Ro­ man territory in 376 led inexorably to the defeat o f Valens in 378. Even Valentinian, though endowed with virtues that Ammianus celebrates in his obituary (30.7), had deleterious vices: he was cruel, he allowed greed to reach unpar­ alleled heights (30.8), and, despite his reputation as a careful commander, he could act with foolish recklessness in the field (2 7.10 .1—i i ) .66 In Ammianus’ eyes, the salient features o f the society in which he lived after 363 were disor­ der, repression, and corruption. After Book X X V I Ammianus abandons a connected year-by-year narra­ tion o f events centered on the activities o f the eastern and western emperors each campaigning season and each winter (Chapter IV). Books X X V II to X X X have a more thematic arrangement and a looser chronology: Valens’ 64 O n the omens o f January 363, see Chapter V. 65 O n “ il ‘dopo Giuliano,’ ” see now N . Baglivi, Annnianea (Catania, 1995), 1 7 7 —243. 66 O n the damaging implications o f this passage, where Ammianus seems to be deliberately undermining Valentinian’s own favorable depiction o f his actions, see Sabbah, Methode (1978), 209; R. Seager, Papers o f the Leeds International Latin Seminar 9 (1996), 1 9 1 - 9 6 .

[I8l]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

three Gothic campaigns o f 367—369 are grouped together in a single chapter (27.5) , and there is room for dispute over what date Ammianus indicates for the campaign o f the elder Theodosius in Britain (27.8, 28.3).67 The historian also has a thesis to argue. He narrates in some detail, and thereby emphasizes, episodes such as the havoc caused in Rome by Maximinus (28.1) and in An­ tioch by a treasonable seance (2 9 .1-2 ),68 the miseries o f Tripolitania caused by nomadic raids, and the corrupt conduct o f the comes Africae Romanus (28.6) .69 Ammianus gives especial prominence to the campaigns o f the elder Theodosius, whom he praises warmly several times: he is introduced as a man known for his military successes (27.8.3: officiis Martiis felicissime cognitus), he bears a heroic name (28.3.1: dux nominis inclyti),70 he is a magnificent leader o f armies (28.6.26: ductor exercituum ille magnificus), and he is com­ pared to Domitius Corbulo and Lusius Quietus, the generals o f Nero and Tra­ jan (29.5.4). Yet the fact that Theodosius needed to campaign in Britain (27.8, 28.3), in Raetia (28.5.15, 29.4.5), and in Africa (28.6.26, 29.5.4-56) shows the prevalence o f disorder71— and Firmus the M oor rebelled because Romanus dishonestly used his influence at court to turn Valentinian against him (29.5.2). In Ammianus’ eyes, the disaster o f 378, to which he devotes his last book, was the product o f greed and corruption. Valens admitted the Goths into the Roman Empire because he wished to reinforce his army while enriching his treasury by commuting the obligations o f his subjects to provide recruits into cash payments (3 1.4 .3—4). The Goths became hostile because o f the sinister greed o f the two generals Lupicinus and Maximus, who supervised their crossing o f the Danube. Predictably, there was insufficient food on hand in Thrace to feed the Goths. Hence “ these loathsome generals” collected all the dogs that they could find and demanded a slave in exchange for each dog: among those enslaved for food were some Gothic nobles (3 1.4 .11). And finally, 67 R. S. O. Tomlin, Britannia (1974), 3 0 3 —09, puts the campaigns in 36 7 and 368, R. P. C . Blockley, Britannia (1980), 2 2 3 - 2 6 , in 368 and 369. 68 O n 2 8 .1, see Appendix 9; on 2 9 . 1 - 2 , W . Seyfarth, Klio 46 (1965), 3 8 1 —382; H . Funke, J A C 10 (1967), 1 6 5 - 7 0 ; Elliott, Ammianus (1983), 1 5 4 —58; T. Zawadski, Labor Omnibus Units. Gerold Walser zum 70. Geburtstag dargebracht uon Freunden, Kollegen und Schülern, ed. H . E. Herzig and R. Frei-Stolba (Wiesbaden, 1989), 2 7 4 - 8 7 . 69 O n the career o f Romanus, see B . H. Warmington, B Z 49 (1954), 5 5 - 6 4 . That Anim ianus’ negative portrait o f him is highly biased is demonstrated by A . Demandt, By z antion 38 (1968), 3 5 4 —60. Th e allegation that Romanus demanded no few er than 4,000 camels before he would help Lepcis (28.6.5) ought to have aroused suspicion long ago. 70 Ammianus styles two other men dux inclytus, viz., Scipio (23.5.20) and Epaminondas (25.2.8). 71 O n these campaigns, see A . Demandt, Afrika und Rom in der Antike, ed. H .-J. Diesner, H. Barth and H .-D . Zimmermann (Halle, 1968), 2 7 7 - 9 2 ; Hermes 100 (1972), 8 1 - 1 1 3 ; J. Matthews, Aspects o f the Notitia Dignitatum (B A R , Supplementary Series 15 , 1976), 15 7 —86; Ammianus (1989), 3 6 7 - 7 6 .

[182]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

on 9 August 378, Valens made the catastrophic decision to fight rather than to wait for the arrival o f the western emperor Gratian, who was hastening to reinforce him, because o f his own obstinacy and the flattery o f his courtiers, who urged him not to share with his nephew a victory already as good as won (31.12.7 ).72 The last six books o f Ammianus construct a subtle and complex argument to prove that the weakness o f the Roman Empire after 378 was caused by the corruption that began with Constantine, flourished under Constantius, and reached a peak under Valentinian and Valens, when it infected not just the emperors and their proximi but the whole administration o f the empire. The times were out o f joint, but that is Ammianus’ historical thesis, not objective reporting.73 It is bad historical method to use Ammianus’ narrative as if it con­ stituted straightforward evidence that corruption so increased in the later fourth century that it became a major cause o f the “ decline o f the Roman Em ­ pire.” 74 Moralists have always tended to depict their own age as more degen­ erate than that o f their parents.75 Such conventional opinions are not always false, for some governments are truly corrupt, even when measured by the normally lax standards that prevail in political life.76 Yet a general charge o f corruption needs to be illustrated by specific and convincing examples, even though, in the nature o f the case, documentary proof will usually be unob­ tainable. For in any age corruption is more likely to be the characteristic o f specific structures o f power than to represent the temper o f a whole society. Ammianus respected convention by stopping his consecutive narrative o f western events at the death o f Valentinian in November 375 and his consec­ utive narrative o f eastern events at the death o f Valens in August 378. For he knew the cardinal rule o f imperial historiography, that history could only be written about dead emperors since the living expected (and deserved) to re72 Ammianus’ account o f the battle has often been discussed: among recent studies, note N . J. E. Austin, Acta Classica 15 (1972), 7 7 - 8 3 ; T. S. Burns, Historia 22 (1973), 3 3 6 —45; Bitter,

Kampfschilderungen (1976), 1 0 2 - 1 8 ; Matthews, Ammianus (1989), 2 9 6 - 3 0 1 ; M . P. Speidel, Klio 78 (1996), 4 3 4 - 3 7 ; Austin and Rankov, Exploratio (1995), 2 4 1 - 4 3 . R. S. O. Tomlin, The Roman World, ed. J. Wacher 1 (Lo n d o n /N ew York, 1987), 118 , aptly observes that Ammianus’ narra­ tive “ ekes out some vital facts with masses o f ‘colour.’ ’’ 73 On the demonstrable dangers o f trusting partisan statements by contemporaries, see J. H. Hexter, Reappraisals in History (Evanston, 111., 1961), 1 1 9 - 2 2 , 1 5 3 - 6 1 . 74 As does R. MacMullen, Corruption and the Decline of Rome (Lon do n /N ew Havèn, 1988), 1 4 6 - 4 7 , cf. the classic analysis o f “ corruption and its antidote, terrorism’’ in the fourth century by Alfoldi, Conflict (1952), 2 8 - 4 7 . 75 As so often, Horace gives classic expression to a banal sentiment: aetas parentum peior avis tulit/ nos nequiores, mox daturos/ progeniem vitiosiorem ( Odes 3.6 .4 6 -4 8 ). 76 For a striking and well-documented example, see S. Cameron, On the Take: Crime, Cor­

ruption and Greed in the Mulroney Years2 (Toronto, 1995).

[183]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

ceive panegyric. Hence, in his farewell to his readers in his last paragraph, he invites younger and abler men to continue the story— in panegyrical style (31.16.9: maiores stili).77 Although Gratian was safely dead, his activities as western emperor, with the exception o f military campaigns relevant to east­ ern affairs (31.10 , 11.6 , 12.4, 7), could not be described without describing the embarrassing execution o f the elder Theodosius, the father o f the ruling emperor, at Carthage in the winter o f 375~376.78 Ammianus, therefore, con­ cluded Book X X X with the proclamation o f the younger Valentinian as Au­ gustus at Aquincum on 22 November 375 (30.10.5)79 and devoted his final book to events connected with the Battle o f Adrianople. The emperor Theo­ dosius himself comes into the narrative once as a young man, where he re­ ceives a conventional compliment (29.6.15: princeps postea perspectissimus). Yet the close o f Ammianus’ work confirms Erich Auerbach’s diagnosis that “ his manner o f writing history nowhere displays anything redeeming, nowhere anything that points to a better future.” 80 For the final paragraph discloses a deep underlying pessimism that gives the lie to those who glibly speak o f “ the fundamental optimism o f Ammianus.” 81 A brief postscript to the Battle o f Adrianople hints at what Ammianus ex­ pected o f the future. First, Ammianus describes how the Goths unsuccessfully raided Constantinople, were repulsed by Saracen troops, and dispersed over the Danubian provinces (3 1.16 .3 -7 ). He then turns to another episode: D u r i n g th e se d a y s, J u liu s , m a s te r o f th e so ld ie rs b e y o n d th e T a u ru s , d is t in ­ g u is h e d h im s e lf b y a s w ift a n d s a lu ta ry d e e d . L e a r n in g o f th e d isasters in T h r a c e , h e se n t s e c re t le tte rs to th e c o m m a n d e r s , w h o w e r e all R o m a n s , a n u n u su a l t h in g

77

Demandt, Zeitkritik (1965), 6 1 - 6 2 . Similar statements are made in similar contexts by

Eutropius, Bren. 10 .18 .3; Jerom e, Chronicle 7 . 3 - 7 Helm. It was conventional for panegyrists to declare themselves unequal to their great theme: Libanius, Orat. i t . 6, c(. R. Helm, Rh. Mus., N .F. 76 (1927), 305 —6. T h e commonplace is parodied in H A , Quad. Tyr. 15 .10 . 7H Jero m e, Chronicle 248 e H elm , has the execution as the first item under the twelfth year o f Valens (i.e., 376), and he registers the slaughter o f 30,000 Alam anni near Strasbourg b y the arm y o f Gratian in the follow ing year (2 4 8 ' ).

7“ Th e date and place are certified by Descr. cons. 3 7 5 .3 ; Socrates, H E 4 .3 1.7 . O n this proc­ lamation, see recently G . de Bonfils, Ammiano Marcellino e l ’imperatore (Bari, 1986), 3 6 —47; J. Szidat, Historia Testis, ed. M . Piérart and O. C urty (Fribourg, 1989), 1 7 5 - 8 8 . Ammianus’ em­ phatic statement that the infant was “ imperator legitime declaratus” (30.10.5) confirms that he was writing before the death o f Valentinian II (15 M ay 392): J. Straub, Vom Herrschen deal in der

Spätantike (Berlin, 1939), 19, 220 n. 12 2 ; Alan Cameron, J R S 61 (19 7 1), 2 6 1. 80 Mimesis (19 53), 5 9 - 6 0 (quoted and discussed in Chapter II). 81 C . P. T. Naudé, Gnomon 41 (1969), 484. Matthews turns the historian into a historical op­ timist whom he explicitly compares to Olympiodorus o f Thebes (Ammianus [1989], 472, cf. Aristocracies [19 75], 3 8 2 —86). That seems also to be a mistaken estimate o f Olympiodorus: A. Gillett, Traditio 48 (1993), 1 - 2 9 . [184]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

at th is p re s e n t p e r io d , a n d g a v e o rd e rs th at th e G o t h s w h o h a d b e e n r e c e iv e d e a r­ lie r a n d d is p e rs e d in v a r io u s c itie s a n d fo rtre s se s w e r e to b e c o lle c te d w it h o u t a r o u s in g t h e ir s u s p ic io n s o u ts id e th e w a lls in th e e x p e c ta tio n o f r e c e iv in g th e p a y th a t t h e y h a d b e e n p r o m is e d an d all w e r e to b e k ille d o n o n e an d th e sa m e d a y as i f th e s ig n a l fo r b a ttle h a d b e e n g iv e n . W h e n th is w is e p la n w a s c a r r ie d o u t w it h ­ o u t fuss o r d elay, th e p r o v in c e s o f th e E a s t w e r e r e s c u e d f r o m g re a t p e rils . ( 3 1 . 1 6.8 )

Eunapius and Zosimus have a version o f the same events that differs from Am ­ mianus in three crucial respects: first, they place it in the spring o f 379, after the accession o f Theodosius, not in the autumn o f 378; second, they report that the Goths were causing widespread damage in Asia Minor; and third, they ex­ plicitly identify the Goths who rioted and were killed as teenage hostages sur­ rendered to the Roman authorities in 376.82 Now a homily o f Gregory o f Nyssa refers to an attack by “ Scythians” on the shrine o f Theodorus at Euchaita on the Black Sea, which it dates to the Julian year 379.83 Since the first two particulars in the account o f Eunapius and Zosimus are thus confirmed by contemporaneous evidence, a cogent case has recently been made for be­ lieving them against Ammianus on the third too.84 At the very least, it appears to be undeniable that Ammianus has knowingly transposed an episode that occurred in the spring o f 379 to the immediate aftermath o f the Battle o f Adrianople in the autumn o f 378. Moreover, nothing in Ammianus’ text ex­ plicitly contradicts Eunapius’ statement that the Goths slaughtered on the or­ ders ofjulius were mainly unarmed young men. Ammianus’ presentation must thus be pronounced to be deliberately misleading. That is a disturbing con­ clusion on at least two levels. First, it disproves the claim that immediately fol­ lows and on which so many modern interpretations o f Ammianus are based— that the historian has never knowingly distorted the truth. Second, it reveals how Ammianus believed that “ the barbarian crisis” should be tackled. Like Synesius a few years later,85 Ammianus believed that Goths on Roman soil were a plague that could be eradicated only by the most drastic measures. He 82 Eunapius, frag. 42 Müller = 42 Blockley; Zosimus 4.26. 83 Gregory o f Nyssa, De sancto Theodoro 6 1 . 1 5 - 6 2 . 2 Cavarnos. Since Gregory was speaking on 7 February 380, his reference to “ the year which has passed” (61.16) establishes 379 as the year in which the raid occurred: see C . Zuckerman, Travaux et Mémoires i i (19 91), 480, 4 8 2 84. Th e argument rests indirectly on the assumption that Basil o f Caesarea died on 1 January 379: for proof that this date, which has recently been challenged, must be correct, see Studia Patristica 29 (1997), 6 - 1 3 . Another homily o f Gregory records a raid by “ Scythians” on Comana, but gives no indication o f the date (De Baptismo [P G 46.424]). 84 Zuckerman, Travaux et Mémoires 11 (1991), 4 7 9 - 8 6 . 85 Synesius, De Regno, which is conveniently summarized by Alan Cameron and J. Long,

Barbarians and Politics at the Court o f Arcadius (Berkeley, 1993), 1 0 3 - 6 , cf. A . Garzya, II mandarine e il quotidiano. Saggi sulla letterature tardoantica e bizantina (Naples, 1983), 1 7 1 - 9 8 .

[1 8 5 ]

Pa

s t

, P

r e s e n t

,

a n d

F

u t u r e

marks the didactic import o f the episode that he narrates. The efficacia ofjulius was salutaris, he devised a consilium prudens: no reader could miss or mistake the lesson being imparted.86 Ammianus recommended massacre as politically expedient— and even, if necessary, genocide.

K.

86 S. Elbern, Hermes 1x5 (1987), 106. O n the didactic nature o f Book X X X I as a whole, see Rosen, Cognitio Gestorum (1992), 8 s -9 0 .

[186]

[ XV] TACITUS, AMMIANUS, AND MACAULAY

Ammianus and Tacitus both wrote history in a way that can be described as dramatic. Yet they write history, depict the actors in the historical drama that they narrate, and represent reality in very different ways. It is worth the ef­ fort, therefore, to attempt to analyze how precisely, or in what precise sense, each o f these two historians is dramatic. For important studies o f both histo­ rians have obscured the individual genius o f each by assimilating it to that o f the other. Edmond Courbaud began his classic book on the artistic proce­ dures o f Tacitus by quoting Racine’s paradoxical opinion that Tacitus was “ the greatest painter o f antiquity,” 1 then analyzed the Histones in visual terms that would have been more appropriate for Ammianus: Tacitus (Corbaud asserted) divided every book into a series o f tableaux and always subordinated narra­ tive to the tableau, because it was the composition o f tableaux that gave him artistic satisfaction.12 On the other side, it has been claimed that Ammianus presents the confrontation between the urban prefect Leontius and Peter Valvomeres “ almost like a scene from a play” with “ the contrasting emotions and postures” leading to a “ dialogue between the central characters” resolved by “ ritual violence.” 3 1 In the second preface to his Britannicus: Racine, Théâtre complet, ed. M . Rat (Paris, i960), 236. 2 E. Courbaud, Les Procédés d ’art de Tacite dans les "Histoires” (Paris, 1918), x, 28, 66, 96, 120, 1 2 1 , 130, 16 1. Tw o specific examples o f Courbaud’s method: he analyzes Hist. 3 .10 as “ un ex­ emple caractéristique,” where “ deux tableaux dans un même chapitre sont encadrés par trois fragments de récit” ( 1 2 5 —27), and he finds in Tacitus, as in Racine, a strong preoccupation with describing gesture (134). 3 J. F. Matthews, Homo Viator. Classical EssaysforJohn Bramble (Bristol/Oak Park, 1987), 279. Also Ammianus (1989), 460: “ his descriptions o f persons and situations resembling (sometimes by deliberate choice o f language) scenes from a play or masque.”

[ 187]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

In fact, it is Ammianus who concentrates on the visual aspect o f an event and narrates it as a series o f tableaux, Tacitus who presents an episode as a se­ ries o f scenes from a play in which the historical actors display emotions that issue in action. Johann Wilhelm Siivern seems to have been the first to sug­ gest that Tacitus deliberately imitated tragic form both in the overall structure o f the Annals and Histories and in his presentation o f individual episodes.4 But it was a famous address by Friedrich Leo in 1895 that placed the idea at the center o f learned discussions o f Tacitus as a literary artist. Leo provocatively asserted that Tacitus was more a poet than a historian in the same way as Plato was more a poet than a philosopher, then analysed Tacitus’ presentation o f the emperors Tiberius, Nero, Galba, Otho, and Vitellius as tragedies. Leo meant this in a strong, virtually literal sense, not a weak metaphorical one. He di­ vided the tragedy o f Tiberius into five acts: the first is dominated by the heroic figure o f Germanicus, the second by the avenging o f his death; in the third act the evil Sejanus enters and prepares the ruin o f Tiberius’ only son and the family o f the dead Germanicus; the climax o f the drama comes in the fourth act with the fall o f Sejanus; there follows the horrible end o f the lonely tyrant, unrolling quickly, like the fifth act o f one o f Shakespeare’s histories.5 Leo’s analysis soon became a cliché o f Tacitean scholarship. Pieter Everts ar­ gued that the structure o f various episodes was dramatic in a formal sense: he analyzed Tacitus’ accounts o f the mutiny in Pannonia in 14 and o f Germani­ cus’ campaign in Germany in 16 as dramas in five acts, o f the mutiny on the Rhine in 14 and the march o f Caecina back to safety in 15 as dramas in four acts, and o f the plot and punishment o f Scribonius Libo as a tragedy in three acts.6 Similarly, Kenneth Quinn divided Tacitus’ account o f the murder o f Agrip­ pina into four distinct scenes and an epilogue.7 In more general terms, Leo’s pupil Eduard Fraenkel contended that Tacitus’ narrative comprises a series o f scenes structurally linked together as if forming a drama;8 Clarence Mendell that the only artistic unity in the Annals is its dramatic construction;9 Einar Löfstedt that Tacitus is a “ tragic poet” whose style and technique were shaped

J. W . Siivern, Abh. Berlin, Hist.-phil. Kl. 1 8 2 2 - 2 3 , 9 5 - 9 6 , 107, 12 1. s F. Leo, Tacitus, ed. V. Pöschl (Darmstadt, 1969), 1 3 - 1 4 . 6 P. S. Everts, De Taciti historiae conscribendae ratione (Diss. Utrecht, pubi. Kerkrade, 1926): re­ spectively, Ann. 1 . 1 6 - 3 0 , 2 .8 - 2 2 , i . 3 1 - 4 4 , 1 . 6 3 - 5 8 , 2 . 2 7 - 3 1 . 7 K. Quinn, Latin Explorations: Critical Studies in Roman Literature (London, 1963), 1 1 5 , 123 (on Ann. 1 4 .1 - 9 ) . 8 E. Fraenkel, Kleine Beiträge zur klassischen Philologie 2 (Rome, i960), 329: “ die reifste Kunst des Tacitus in szenenhaftem Aufbau gipfelt.” y C . W . Mendell, Y C S 5 (19 35), 53: “ he accepted the annalistic form as a framework but gave it the semblance o f unity by means o f dramatic technique.”

[188]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

by the fact that he “ occupied himself with the composition o f tragic poetry in his youth.” 101 Leo’s analysis has not commended itself to all recent students o f Tacitus: it has been pronounced “ too schematic” and Tacitus’ technique has been de­ fined as “ pictorial-dramatic,” more indebted to rhetoric and rhetorical histo­ riography than to drama.11 That is to miss a central aspect o f the way in which Tacitus depicts historical reality. Margarete Billerbeck has rightly protested and reasserted the validity o f Leo’s approach: division into five acts can be seen even in such a minor episode as the murder o f his mistress by the tribune O c­ tavius Sagitta,’2 and Tacitus habitually uses a variety o f dramatic devices, in­ cluding the heightening o f expectation, emotional death scenes, and sudden reversals.13 In the present context, it suffices to show how Tacitus presents two important episodes in the Annals as a series o f separate acts or scenes and how he signals transitions carefully, if often only implicitly, by indicating a change o f scene or setting— a feature o f Tacitus’ narrative technique that editors too often obscure by making their paragraph divisions coincide with the mislead­ ing chapter divisions introduced in the Renaissance.14 The first book o f the Annals begins with a rapid sketch o f Roman history, a brief author’s preface, and an analysis o f the political situation at the end o f the reign o f Augustus.15 Tacitus then narrates the accession o f Tiberius as a drama in five acts.16 The first covers the final illness and death o f the old em­ peror: there was a rumour that he had visited his exiled grandson Agrippa Pos10 E. Löfstedt, Roman Literary Portraits, trans. P. M . Fraser (Oxford, 1958), 153. 11 F. R. D. Goodyear, T he Annals of Tacitus, Books 1 - 6 i (Cambridge, 1972), 195; A .J. W ood­ man, Rhetoric in Classical Historiography. Four Studies (London and Sydney, 1988), 16 0 -9 6 . U n ­ fortunately, the latter scholar has often treated Tacitus as a novelist who invents freely, not a his­ torian interested in discovering the truth: thus his analysis o f Tacitus1 account o f the collapse o f the amphitheatre at Fidenae under Tiberius reaches the astounding conclusion that the histo­ rian had at his disposal “ no more source material than is provided by Suetonius11 (A .J. W ood­ man, C Q , N .S. 22 [1972], 156), ignoring totally the fact that Tacitus refers to and paraphrases the senatus consultum condemning the contractor Atilius, who is otherwise unknown (Ann. 4.62.1 - 6 3 . 3 , cf. P IR 2 A 1293. 12 Tacitus, Ann. 13.44. 13 M . Billerbeck, A N R W 2 .33.4 (19 91), 2 7 5 2 - 7 1 . O n the theatricality o f Tacitean death scenes, see also R. Barthes, A Barthes Reader, ed. S. Sontag (N ew York, 1982), 1 6 2 - 6 6 ; A . Malissard, Tlteater und Gesellschaft im Imperium Romanum, ed. J. Blänsdorf (Tübingen, 1990), 2 1 3 —22. Billerbecks approach is now adopted by Tony Woodman in his illustration o f how Tacitus por­ trays the Pisonian conspiracy in Ann. 1 5 .4 9 - 7 4 as belonging to “ a world o f unreality11 (Tacitus and the Tacitean Tradition, ed. T. J. Luce and A .J . Woodman [Princeton, 1993], 1 0 4 - 1 2 ) . 14 For instances o f the same phenomenon in Ammianus, see Chapter IV. 15 Tacitus, Ann. 1 . 1 . 1 - 2 -I- 1.3 + 2 - 5 . 16 Tacitus, Ann. 1 . 5 - 1 5 = 1.5 + 6 + 7 . 1 - 8 . 5 + 8 .6 - 1 0 .7 + 1 0 . 8 - 1 5 . O n the need to be­ gin new paragraphs at 1.3, 8.6, and 10.8, see Goodyear, Annals 1, 88, 1 5 1 , 169.

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

tumus and was intending to restore him, but he fell seriously ill and died, and Tiberius immediately seized power. The second act presents the killing o f Agrippa Postumus and the reaction o f Tiberius and his mother. The scene moves to Rome for the third act, in which oaths o f loyalty are taken to the new emperor, who summons the Senate, which meets and discusses arrange­ ments for the funeral o f Augustus. The fourth act is set on the day o f the fu­ neral: Tacitus evokes the sentiments o f those who could recall the burial o f Julius Caesar, alludes briefly to the Res Gestae Divi Augusti, and offers his ver­ dict on Augustus in the form o f the opinions for and against him that men o f wisdom expressed on the occasion. Finally, the Senate meets: it consecrates the deceased emperor as Divus Augustus, discusses the legal powers o f Tiberius (it would be anachronistic to call it his constitutional position), and settles var­ ious other matters.17 The episode o f the “ marriage” o f Messalina to her young lover C. Silius (which was in reality no such thing)18 has a similar structure, (i) Tacitus begins with a prologue that vouches for the truth o f the story, however far-fetched it may appear to be, then analyzes the situation by imagining a debate among the imperial freedmen, who felt themselves imperiled by the empress’s new amour.19 (2) Despite the hesitations o f Callistus and Pallas, Narcissus decides to act: he persuades a concubine to reveal his wife’s infidelity to Claudius while he is at Ostia and confirms the truth o f all she says, after which Claudius consults his consilium. (3) In Rome Messalina and Silius continue their wild orgy until news arrives that Claudius is returning to the city, whereupon Mes­ salina sets out to meet him, taking a Vestal virgin with her. (4) Claudius re­ turns to Rome: Narcissus prevents Messalina from gaining access to him, shows him the house o f Silius, and takes him to the camp o f the praetorian 17 Tacitus marks the main transitions in Ann. 1 . 5 - 1 5 carefully: (1) 5 .1: haec atque talia agi­ tantibus gravescere valetudo Augusti; (2) 6 .1: primum facinus novi principatus; (3) 7 .1 : at R o­ mae, after which new paragraphs should begin at 7.5 (sed defuncto Augusto) and 8.1 (nihil primo senatus die); (4) 8.6 (die funeris), with new paragraphs at 9.3 (at apud prudentes) and 10 .1 (dicebatur contra); (5) io.8: ceterum sepultura more perfecta templum et caelestes religiones de­ cernuntur, i.e., in the Senate on 17 September 14. It ruins the articulation o f the text to divide 8 .6 —9.2 into two paragraphs, as Goodyear does. Th e subsequent Teubner edition o f H. Heub­ ner (1983) faithfully follows the conventional chapter divisions (which are o f renaissance origin) and begins new paragraphs at 9 .1, but not at 8.6 or 9.3, and at 1 1.1 instead o f 10.8. ,H B. Levick, Claudius (London, 1990), 6 4 - 6 7 , cf. J. Colin, Les Études classiques 24 (1956), 2 5 - 39-

19

Tacitus, Ann. 1 1 .2 7 —28. After the action proper commences (29.1), changes o f scene are

marked by the following phrases: at Messalina (31.2 , where Heubner correctly begins a new paragraph); trepidabatur nihilo minus a Caesare (33.1); interim Messalina (37.1). Reparagraphing is also necessary in the final chapter o f the book: although the traditional chapter division oc­ curs immediately before Messalina’s death, the logical break surely comes after it with the tran­ sition from the suppression o f the “ conspiracy” to its aftermath ( 3 8 .1 b: corpus matri concessum).

[190]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

guard, where he authorizes the execution o f all the alleged plotters except two. (5) Messalina seeks out her mother and is killed in her house. By way o f epilogue Tacitus describes the reaction o f Claudius, the condemnation o f Messalina by the Senate, and the honors voted to Narcissus. In this episode, as in his account o f the accession o f Tiberius, Tacitus breaks his narrative into separate and clearly defined sections, each o f which begins with an initial phrase indicating a change o f place or temporal progression. The technique is dramatic: Tacitus uses modes o f presentation and expression familiar to him from the recitation drama that flourished in the first century.20 Hence, when he describes the mutiny o f the praetorians under Otho, Tacitus focuses his readers’ attention on the mutinous soldiers and their motives, passing quickly over the decisive confrontation between the emperor and the mutineers.21 Ammianus’ method o f narrating similar episodes is vastly different: as with the arrest o f Peter Valvomeres (Chapter II), it is the decisive confrontation it­ self that he chooses to emphasize. The two historians also differ vastly in their presentation o f physical detail and in their readiness to use ordinary vocabu­ lary for everyday things. Tacitus maintains a highly elevated tone throughout. Tiberius was bald and had a blotchy face, but Tacitus avoids the vulgar word calvus: the emperor had a “ summit denuded o f hair.” 22 The emperor Vitellius was a glutton who gorged himself like a pig: Tacitus avoids the normal Latin word .for pig, for which he substitutes a periphrasis o f extreme preciosity: “ lazy animals which lie down and go to sleep if you offer them food.” 23 Ammianus has no such stylistic inhibitions. He uses colloquialisms and words whose stylistic overtones are far from elevated, such as crapulentus (29.5.54),24 palpamentum (27.12.6),25 suctus (31.6.7),26 and vomito (14 .11.24 ).27 The stylistic gulf that separates the two historians is perhaps most obvious in their treat­ ment o f spitting. Ammianus can say straightforwardly that when Constantius 20 O n Tacitus* relation to recitation drama, see H S C P go (1986), 2 3 2 - 4 4 . 21 Tacitus, Hist. 1.8 2 .1, cf. H. Heubner, Studien zur Darstellungskunst des Tacitus (Hist. 1, 1 2 II,5i) (Würzburg, 1935), 2 3 - 2 4 ; E. Hohl, Klio 32 (1939), 3 0 7 - 2 4 . Comparison with Plutarch, Otho 3 . 6 - 1 1 , shows that the emphasis is Tacitus’ own. 22 Tacitus, Ann. 4.57.2: nudus capillo vertex, ulcerosa facies et plerumque medicaminibus in­ terstincta, cf. R. Syme, Tacitus (Oxford, 1958), 343. 23 Tacitus, Hist. 3 .3 6 .1: umbraculis hortorum abditus, ut ignavia animalia, quibus si cibum suggeras, iacent torpentque, praeterita instantia futura pari oblivione dimiserat. Tacitus uses the word sus twice: once in registering Jewish abstention from pork (Hist. 5.4.2: sue abstinent), once in the notice o f a prodigy in 62, which he doubtless transcribed from the Elder Pliny (Ann. 12 .6 4 .1: suis fetum editum, cui accipitrum ungues inessent). 24 O n the form o f the adjective, see Blomgren, Quaestiones (1937), 1 1 9 - 2 0 : all the deriva­ tives o f crapula appear to be vulgar or biblical ( T L L 4.1098). 25 T h e word occurs nowhere else, but its cognates are non-literary (T L L 10.160). 26 O L D 1889 cites only Varro and the Elder Pliny. 27 O L D 2 10 3 cites only Seneca, Epp. 18.4, 10 8.37; Suetonius, Vit. 13 .1 .

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

entered Rome no one saw him spit or wipe his mouth or nose (16 .10 .10 ),28 and he uses both the verb spuo in three other passages (21.16.7; 23.6.80; 27.3.5) and the noun sputamen (14.9.6). When Tacitus described Nero performing on stage with a lyre, he avoided the normal Latin words for spit and saliva: the emperor wipes his face with the garment that he was wearing “ so that no em­ anations from his mouth or nose were seen.” 29 Another linguistic difference between the two historians is equally significant.30 As far as he can, Tacitus avoids official terminology and the technical terms for administrative posts.31 Ammianus in contrast has no objection to using bureaucratic and official phraseology.32 H ow much does Ammianus owe to Tacitus? The question has received greatly divergent answers. Long lists o f borrowings and imitations have been pro­ duced,33 and Ammianus has been saluted as “ the heir o f Tacitus, in every sense.” 34 Yet it must be doubted whether it is illuminating to regard Ammi­ anus as the successor o f Tacitus in any real sense:35 the world about which Ammianus writes had undergone a profound transformation, his historical techniques and methods o f presentation are vastly different from those ofTac2H Ammianus’ words “ nec spuens aut os aut nasum tergens aut fricans” appear to echo Xenophon, Cyropaedeia 8 .1.42, 8.8: see C . J. Classen, Rh. Mus., N .F. 1 3 1 (1988), 1 7 7 - 8 1 . 2S) Tacitus, Ann. 16.4.3: ingreditur theatrum, cunctis citharae legibus obtemperans, ne fessus resideret, ne sudorem nisi ea, quam indutui gerebat, veste detergeret, ut nulla oris aut narium excrementa viserentur. I avoid the obvious English translation o f excrementa in order to preserve Tacitus’ elevated stylistic level: he uses the word elsewhere o f the blind man who came in quest o f healing and asked Vespasian “ ut genas et oculorum orbes dignaretur respergere oris excre­ mento” (Hist. 4 .8 1.1). A similar avoidance o f vulgar words is evident at Ann. 11.3 2 .3 : “ vehiculo quo purgamenta hortorum excipiuntur.” 30 For archaisms, often with a colloquial flavor, in Ammianus, see H. Fesser, Sprachliche Beobachtungen zu Ammianus Marcellinus (Diss. Breslau, 1932), 2 9 - 6 1 . Tacitus wrote before ar­ chaism became fashionable under Hadrian. 31 Thus, although legati o f provinces are frequent in his text, none is ever styled legatus Augusti, legatus pro praetore or legatus Augusti pro praetore, as they normally are on inscriptions (H. Dessau, IL S 3 .1, pp. 3 6 8 -8 2 ) . 32 For some examples, see A. Helttula, Studia in honorem Uro Kajanto (Arctos, Supplement 2, 1985), 4 1 - 5 6 ; Studies on the Latin Accusative Absolute ( Commentationes Humanarum Litterarum 81, 1987), 4 6 - 4 7 , 1 0 5 - 1 1 . Th e topic merits a systematic investigation (C P 88 [1993J, 66). 33 H. Wirz, Philologus 36 (1877), 6 3 4 - 3 5 ; Fesser, Beobachtungen (1932), 2 3 - 2 7 ; G . B. A. Fletcher, Rev. phil. 63 (1937), 3 9 0 - 9 2 ; Rosen, Ammianus (1982), 9 5 - 9 6 . 34 Syme, Tacitus (1958), 503, n. 8, cf. Ammianus (1968), 129. O n this issue at least, Arnaldo Momigliano was in complete agreement with his arch-enemy: he opined that Ammianus “ con­ tinues Tacitus as the Historia Augusta continues Suetonius” (Essays in Ancient and Modern Histo­

riography [Oxford, 1977], 131). 35 D. Flach, Historia 32 (i973), 333 ” 50; L. E. Wilshire, C J 68 ( 1 9 7 2 - 7 3 ) , 2 2 1 - 2 7 ; R. P. C . Blockley, Latomus 32 (19 73), 6 3 - 7 8 ; I- Borszak, Acta Antiqua 24 (1976), 3 5 7 - 6 8 ; Matthews, Ammianus (1989), 3^, 456, 468, 4 8 2 - 8 3 , n. 35, 549, n. 4, 550, n. 1 1 ; G. W . Bowersock, J R S 80 (1990), 246.

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

itus, his style is characterized by opulence and abundance where Tacitus strives for terse brevity,36 and he stands in the Greek rather than the Latin tradition o f historiography (Chapters VII, IX).37 Moreover, many o f the alleged imita­ tions o f Tacitus will not withstand critical scrutiny.38 A careful assessment is needed, therefore, o f the true nature and real extent o f Ammianus’ debt to Tacitus. There can be no doubt that Ammianus had read widely and deeply in Latin. The Res Gestae quote Cicero many times, not infrequently from works that are now lost, and long lists o f convincing verbal echoes have been compiled.39 Throughout his work, Ammianus repeats and adapts phrases from other Latin authors, above all Terence, Sallust, Virgil, Livy, and the antiquarian Aulus Gellius.40 In addition, echoes o f a wide variety o f other Latin authors can be documented, including the prose writers Valerius Maximus, Curtius Rufus, Seneca, Florus, and Apuleius, and the poets Horace, Ovid, Lucan, Valerius Flaccus, and Silius Italicus.41 Indeed, so pervasive are echoes o f earlier Latin writers that possible allusions must always be considered when emending cor­ rupt or supplementing lacunose passages.42 36 H. Hagendahl, Eranos 22 (1924), 1 6 1 - 2 1 6 : he defines the three main traits o f Ammianus’ style as “ color poeticus, variatio, abundantia.’’ 37 For comparison o f Ammianus’ “ narrative sentence’’ to that o f other Latin historians, see A . Debru, Rev. phi!.} 66 ( 1992), 2 6 7 - 8 7 . 38 For example, W . Richter, Historia 23 (1974), 3 6 6 - 6 7 , confidently claims that 3 1 .7 .1 6 (reli­ qua peremptorum corpora dirae volucres consumpserunt/ assuetae illo tempore cadaveribus pasci,/ ut indicant nunc usque albentes ossibus campi) cannot reflect Ammianus’ personal in­ spection o f the battlefield at Ad Salices because the passage copies what Tacitus says about the remains o f the slaughtered legions o f Quintilius Varus (Atm. 1.6 1.2 : medio campi albentia ossa, ut fugerant, ut restiterant, disiecta vel aggerata). In fact, Ammianus is closer to the model that he and Tacitus have independently imitated: sanguine adhuc campique ingentes ossibus albent (Vergil, Aen. 12.35). 39 H. Michael, De Ammiani Marcellini studiis Ciceronianis (Diss. Breslau, 1874); G. B. A. Fletcher, Rev. phil.3 1 1 (1937), 3 7 7 - 8 1 . However, the three derivations alleged by W irz, Philo-

logus 36 (1877), 6 33, are not persuasive. 40 See the passages adduced by Valesius on 2 5.4 .14 , 22; M . Hertz, De Ammiani Marcellini studiis Sallustianis dissertatio (Breslau, 1874); Hermes 8 (1874), 265 n. 1, 2 7 1 , 2 7 5 - 3 0 2 ; Wirz, Philo­ logus 36 (1877), 6 2 8 - 3 5 ; H. Finke, Ammianus Marcellinus und seine Quellen zur Geschichte der römischen Republik (Diss. Heidelberg, 1904), 1 0 - 1 4 , 2 0 - 3 0 ; H. Hagendahl, Studia Ammianea (Diss. Uppsala, 19 2 1), 1 - 1 4 ; Fesser, Beobachtungen (1932), 3 - 2 3 ; Fletcher, Rev. phil. ,3 11 (1937)1 37 7, 3 8 2 - 8 6 , 3 9 0 - 9 3 ; Ut pictura poesis (Leiden, 1955), 8 5 - 8 6 ; Bitter, Kam pf Schilderungen (19 7 6), 2 0 8 - 1 0 ; C . W . Fornara, Historia 41 (1992), 4 2 7 - 3 8 . 41 Hertz, Hermes 8 (1874), 2 7 2 - 7 3 ; E. Schneider, Quaestiones Ammianeae (Diss. Berlin, 1879), 12 -

13 , 3 4 _ 3 5 » 38; C . Weyman, Sb. München, Phil. u. hist. C I. 18 9 3.2 (1894), 3 6 1; Finke, A m ­

mianus (1904), 1 5 - 2 0 , 3 0 - 4 0 ; E. Bickel, Gott. G el. A n z . 180 (19 18 ), 2 8 2 - 9 2 ; H. Hagendahl, Studia Ammianea (Diss. Uppsala, 19 21), 15; Fletcher, Rev. p h il. ,3 11 (1937), 383, 3 8 6 - 8 9 ; Bitter, Kampfschilderungen (1976), 2 0 8 - 9 . 42 C . Brakman, Ammianea et Annaeana (Leiden, 1909), 1, 1 5 - 1 6 ; Hagendahl, Studia (19 21), 1 3 - I4-

[193]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

On general grounds, therefore, it is probable that Ammianus read Tacitus. One certain borrowing will constitute proof positive, and two separate echoes o f the same passage o f Tacitus have been identified, which are clearly intended to be significant and consequently must be deliberate.4 1*43 Tacitus opens the sec­ ond book o f his Histories with a striking expression: struebat iam fortuna in diversa parte terrarum initia causasque imperio.44 Ammianus repeats Tacitus’ phrasing very closely in the opening words o f Book X X II: d u m h a e c in d iv e rsa p a rte t e r r a r u m fo r t u n a e s tr u u n t v o lu b ile s c asu s. ( 2 2 . 1 . 1 )

Tacitus then proceeded to describe the journey o f Titus, the son o f Vespasian, who was on his way to Rome in the hope o f being adopted as Galba’s heir. At Corinth Titus received news that Galba was dead and deliberated whether to continue to Rome and risk becoming a hostage in the hands o f Otho or to re­ turn to his father: v his a tq u e ta lib u s in te r sp e m m e t u m q u e ia c ta tu m sp es v ic it .45

Ammianus consciously imitates this sentence in a similar context. At the end o f Book X X , the eastern emperor Constantius contemplates the prospect o f war with Julian as the winter o f 3 6 0 -3 6 1 approaches: his ac talibus imperator inter spem metumque iactabatur. (20.11.31) Since the two situations are similar, the literary imitation is surely both con­ scious and deliberate.46 Another echo o f Tacitus may be detected in Ammi­ anus’ account o f the arrest o f Peter Valvomeres. Peter’s enormous body and red hair (15.7.4: vasti corporis rutilique capilli) are so characteristic o f Ger­ mans that they could be considered a commonplace rather than inspired by 41 See, respectively, Fesser, Beobachtungen (1932), 23; Fletcher, Ut pictura poesis (1955), 85. 44 Tacitus, Hist. 2 .1 .1. 45 Tacitus, Hist. 2 .1 .1 - 2 . 1 . 46 Rosen, Ammianus (1982), 9 5 - 9 6 . Similar arguments can be deployed to prove that A m ­ mianus’ account o f the proclamation ofjulian as Augustus (20.4.9—5.7) deliberately copies Tac­ itus’ account o f the mutiny o f the Rhine armies in 14 (Ann. 1 .3 1 - 4 9 ) : M . F. Williams, Phoenix 5 1 (1997). 6 3 - 6 8 .

[194]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

the rutilae comae o f Tacitus’ Germans.47 But Peter is also a turbarum acerrimus concitor (1 5 .7 .5 ). Tacitus too has an acerrimus belli concitor, he is Antonius Primus at the beginning o f the third book o f the Histories, a convicted forger who is about to seize command o f the troops in Pannonia who have declared for Ves­ pasian.48 Hence it becomes likely that phrases such as opum contemptor (25.4.7) and expediti consilii (3 1 . 5 .7 ) are unconscious echoes o f Tacitus,49 which implies that Ammianus had read him carefully and thoroughly. Ammianus thus imitates Tacitus both in his presentation o f individuals and in his vocabulary.50 Moreover, the Res Gestae began with the accession o f the emperor Nerva, exactly where Tacitus ended, and Ammianus’ original plan was probably to cover the period from Nerva to Julian in thirty books, which corresponds exactly to the number o f books o f Tacitus’ Annals and Histories (Chapter III). Hence it is legitimate to believe that, despite the fundamentally Greek nature o f his work, Ammianus also thought o f himself as a historian who continued, imitated, and emulated Tacitus. Indeed, the two aspects are complementary, not contradictory. For it was precisely his knowledge o f Greek historians o f the second, third, and fourth centuries that equipped Ammianus to write a history intended to continue and supplement Tacitus for Roman readers who were in the habit o f reading the scandalous and salacious imperial biographies o f Marius Maximus (28.4.14). The Historia Augusta stands as con­ temporary testimony that “ the lonely historian” Ammianus failed to change the literary tastes o f Roman aristocrats.51 The modem historian whom Ammianus most resembles is not the one with whom he has traditionally been compared.52 Edward Gibbon blended the his­ torical insights and cynicism o f Tacitus with a sentence structure that owes most to the perfectly balanced periods o f Isocrates: although he regarded Am -

47 Tacitus, Germ. 4 .1, cf. A gr. 11.2: rutilae Caledoniam habitantium comae, magni artus. A c­ cording to Jerome, Hilarion healed a Frankish “ candidatus Constantii imperatoris, rutilus coma, et candore corporis indicans provinciam” (V ita Hilarionis 13.2). 48 Tacitus, H ist. 3 .2 .1. Editors o f Tacitus usually print concitator, but the Medicean manuscript has conciator, which Orelli emended to concitor. 49 Respectively, Tacitus, H ist. 4.5.2: opum contemptor (sc. Helvidius Priscus); H ist. 4.42.5: expediti consilii (o f Aquillius Regulus); cf. Fletcher, Rev. pltil .3 1 1 (1937), 3 9 1; U t pictura poesis

(1955), 86. 50 Löfstedt, Roman Literary Portraits (1958), 1 5 9 —60, drew attention to a shared linguistic preference that may be unconscious, and thus all the more revealing: like Tacitus, Ammianus heavily prefers the archaic preposition ob, which he uses 126 times, to its commoner synonym propter, which occurs only thrice ( 1 5 .1 1 .4 , 18 .2 .15 , 23.6.87). 51 Syme, Ammianus (1968), 1 2 6 - 2 1 0 . 52 E.g., by M . L. W . Laistner, The Greater Roman Historians (Berkeley, 1947), 1 4 3 —44.

[I95j

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

mianus highly as a historical source, Gibbon recoiled instinctively from his stylistic excesses.53 Ammianus’ modern congener is not Gibbon, but Thomas Babington Macaulay. Although his essay o f 1828 on “ history” considers the classic historians Herodotus, Thucydides, Polybius, Sallust, Livy, and Tacitus,54 the young Macaulay had nothing to say about Ammianus, who did not form part o f a normal classical education (then or now):55 he first read Ammianus in India and pronounced his history “ the worst written book in ancient Latin,” whose “ style would disgrace a monk o f the tenth century.” 56 Ammi­ anus had no direct influence on Macaulay, who derived his historical approach and his conception o f what history ought to be from the novels o f Sir Walter Scott and his literary style largely from the Evangelical preaching to which he was exposed when very young.57 Nevertheless, there are deep affinities be­ tween Ammianus and Macaulay, who, despite his revulsion at the ancient writer’s style, instinctively recognized in him “ many o f the substantial quali­ ties o f a good historian.” 58 Both the Res Gestae and the History of England strive for formal symmetry, both display strong partiality for or unfairness toward the actors in their narrative, great and small, both allow these personal preju­ dices to determine their evaluation o f evidence, and both employ composi­ tional modes more suited to the novel than to history. In his journal for 6 February 1854, after a day spent largely on his History of England, Macaulay wrote: “ I worked hard at altering the arrangement o f the first three chapters o f the third volume. What labour it is to make a tolerable book, and how little readers know how much trouble the ordering o f the parts has cost the writer!” 59 Macaulay strove for strict symmetry between, if not within, volumes.60 The first two volumes, published together in 1848, have five long chapters each. The first describes England under King James II

53 Decline and Fall 2.264, n. 18 (B) = 1.687, n. 16 (W ) (quoted in Chapter I). 54 T. B. Macauley, Edinburgh Review 47 (1828), 3 3 1 - 6 7 . O n the intellectual context o f the essay, see G. P Gooch, History and Historians in the Nineteenth-Century1 (London, 19 13 ), 2 9 4 304: J. Clive, Macaulay. The Shaping o f the Historian (N ew York, 1973), 9 8 —14 1. 55 There seems to be no mention at all o f Ammianus in Tlte Legacy o f Rome: A New Appraisal, ed. R. Jenkyns (Oxford, 1992), despite extensive discussion o f Augustine. 56 G. O. Trevelyan, Life and Letters o f Lord Macaulay, 1 (Oxford, 19 6 1: first published in 1876), 423: a letter written in Calcutta on 30 November 1836. 57 See esp. G. Levine, The Boundaries o f Fiction: Carlyle, Macaulay, Newman (Princeton, 1968), 7 9 - 1 6 3 ; W . A . Madden, The Art o f Victorian Prose, ed. G . Levine and W . Madden (N ew York, 1968)

, 1 2 7 - 5 0 ; H. R. Trevor-Roper, The Romantic Movement and the Study o f History (London,

1969) , 8, 1 6 - 2 1 ; Clive, Macaulay (19 73), 119 , 12 2 , 1 3 5 - 3 6 , 248, 498. 58 Trevelyan, Life 1 (19 61), 423. 59 Trevelyan, Life 2 (19 61), 305. 60 M odern critics o f Macaulay have had virtually nothing to say about the formal structure o f his History : the omission strengthens the central contention o f Chapter 111.

[196]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

and ends with “ the first faint indications o f a great turn o f fortune,” and the second begins with James at the height o f his power and ends with the inter­ regnum after his flight. The third and fourth volumes, also published as a pair in 1856, each have six long chapters. The third volume begins with the stark sentence “ The revolution had been accomplished” ; it concludes with Wil­ liams departure from London in January 1691 after a plot to assassinate him. The fourth begins with William’s embarkation for Holland and ends with the treaty o f Ryswick in 1697, as a result o f which “ the cheerful bustle in every seaport and every market town indicated, not obscurely, the commencement o f a happier age.” The fifth and final volume o f his History, which was pub­ lished by Macaulay’s sister after his death, concludes with the death o f W il­ liam in 1702 and contains three chapters, probably the number intended by the deceased author.61 Like Ammianus, Macaulay has strong likes and dislikes, which he expresses without inhibition in his delineation o f the characters whose actions he de­ scribes. Like Ammianus too, Macaulay allows his biases to shape and mould his narrative. His hero is William o f Orange, who must be exculpated from any blame for the Massacre o f Glencoe, his villains Kingjames II and any who made a successful career under him, whatever their subsequent service to Wil­ liam, so that William Penn and John Churchill, Duke o f Marlborough, re­ ceive very unflattering portaits.62 Macaulay is as unfair to James as Ammianus is to Constantius, even more inclined to disbelieve ill o f William than Am­ mianus is to suppress Julian’s intolerance, fanaticism, and devotion to religious practices o f which he disapproved (Chapter XIII). Ammianus’ account o f the history o f the Roman Empire from 353 to 378 has provided the basis and the framework for almost all modern accounts o f the period to almost the same degree as that o f Thucydides has for the Pelopon­ nesian War and that o f Polybius for Rome’s dealings with the Greek world from 220 to 146 B.c. Yet his presentation o f events depends on antecedent prejudices that almost always have a personal basis. The biases damage Am ­ mianus’ objectivity just as much as Macaulay’s prevent his History of England from being an impartial account o f British history and politics in the late sev­ enteenth century. Macaulay’s History is now regarded as a work o f literature, imagination, and eloquence that reveals more about its author and the nine­ 61 That appears to be the clear implication o f the letters o f October and 14 December 1859 quoted by Trevelyan, Life 2 (19 61), 395. 62 C . Firth, A Commentary on Macaulay's History o f England, ed. G . D avies (London, 1938), esp. 8 8 - 9 0 , 2 0 1 - 4 , 2 6 5 - 7 3 , 2 7 7 - 3 0 3 , 3 3 4 - 6 7 -

[197]

T

a c i t u s

, A

m m i a n u s

,

a n d

M

a c a u l a y

t e e n t h c e n t u r y t h a n a b o u t t h e “ G l o r i o u s R e v o l u t i o n ” o f 1 6 8 8 . 63 T h e f a c t t h a t as a h i s t o r i a n A m m i a n u s s h o w s s u c h s i m i l a r i t i e s t o M a c a u l a y o u g h t t o e n ­ c o u r a g e a s i m i la r e s t i m a t i o n o f h is R e s G e s ta e as a w o r k o f i m a g i n a t i v e l i t e r a ­ t u r e . 64 T h a t w i l l in n o w a y d i m i n i s h t h e a c h i e v e m e n t o f o n e w h o h a s b e e n s a lu t e d as t h e g r e a t e s t l i t e r a r y g e n i u s t h a t t h e w o r l d p r o d u c e d b e t w e e n T a c i ­ tu s a n d D a n t e . 65 M a c a u l a y d e c l a r e d t h a t “ t o b e a r e a l l y g r e a t h i s t o r i a n is p e r ­ h a p s t h e r a r e s t o f i n t e l l e c t u a l d i s t i n c t i o n s . ” 66 A m m i a n u s h a s s e c u r e d a p e r m a ­ n e n t p la c e in t h e s e l e c t g r o u p o f r e a l l y g r e a t h i s t o r i a n s p r e c i s e l y b e c a u s e , l i k e M a c a u l a y ’s H is t o r y o f E n g la n d , h is R e s G e s ta e e x h i b i t t h e c r e a t i v e a n d i m a g i n a ­ t iv e p o w e r s o f a n o v e l is t .

63 On the manifold ways in which Macaulay resembles the great novelists o f Victorian E n ­ gland, see Levine, Boundaries (1968), 7 9 - 1 6 3 . It can even be claimed with some plausibility that Macaulay thought o f the ideal historian as both novelist and poet: R. Weber, Papers on Language

and Literature 3 (1967), 2 1 6 - 1 9 . For a general discussion o f “ History and the N ovel,” see D . LaCapra, History and Criticism (Ithaca, 1985), 1 1 5 - 3 4 . 64 For a spirited defense o f imagination as necessary for the historian, see H. R. TrevorRoper, History and Imagination (Oxford, 1980). Ronald Syme s Roman Revolution (Oxford, 1939), will long stand as an example o f how to combine detailed and accurate research with the imag­ inative reconstruction o f a past era. 63 E. Stein, Geschichte des spätrömischen Reiches 1 (Vienna, 1928), 3 3 1 = Histoire du Bas-Empire i, trans, and rev. J.-R . Palanque (Paris/Bruges, 1 9 5 9 ). 2 15 . fi6, Edinburgh Review 47 (1828), 365.

[198]

APPENDICES

A P P E N D IX

[I] THE TEXT OF AMMIANUS

The salient facts about the manuscript tradition o f Ammianus’ Res Gestae have been stated with elegance and economy by L. D. Reynolds.1 The text is trans­ mitted by two Carolingian manuscripts o f the ninth century: M

Fragmenta Marburgensia, the six surviving leaves o f a manuscript once at Hersfeld that were rediscovered at Marburg in 1875 and are now in the Landesbibliothek at Kassel (Philol. 2 27). Vatican lat. 1873, from Fulda.

V

There are also fourteen manuscripts o f the fifteenth century that all derive di­ rectly or indirectly from V and hence are o f value only as repositories o f hu­ manist conjectures except where a leaf o f V has been lost (31.8.5 paulatim3 1 . 10 .1 8 incredibile dictu est). Although the two Carolingian manuscripts are very close in date, at least on paleographical grounds, it was demonstrated beyond possible doubt in 1936 that V was copied from M .12 Since M , which broke o ff at 30.9.6, was employed by Gelenius for his edition o f Ammianus (Basel, 1533) before it was dismem­ bered, Gelenius’ edition is an important witness to the text. However, con­ siderable care must be exercised when trying to deduce the reading o f the lost 1 Texts and Transmission. A Survey of the Latin Classics, ed. L. D. Reynolds (Oxford, 1983),

6 - 8. 2 R. P. Robinson, Philological Studies in honor o f Walter Miller (Columbia, Missouri, 1936), 1 1 8 40; W . Seyfarth, Der Codex Fuldensis und der Codex E des Ammianus Marcellinus (Abh. Berlin , Klasse fiir Sprache, Literatur und Kunst 1962, Abh. 2). H. Nissen, Ammiani Marcellini fragmenta Mar­

burgensia (Berlin, 1876), 1 8 - 2 4 , had argued that M was the “ archetype” o fV , but L. Traube, Vor­ lesungen und Abhandlungen 3 (Munich, 1920), 3 3 - 3 8 , dated M slightly later than V on palaeographical grounds and hence held M and V to be independently descended from a lost archetype.

[201]

A

p p e n d ix

i

manuscript from Gelemus’ text: where Gelenius reproduces the edition o f 15 18 , it cannot simply be assumed that M had the same reading; where he deviates, he has not necessarily corrected from M rather than resorted to conjecture.3 The first edition o f Ammianus to meet proper critical standards was pro­ duced by C. U. Clark, who had previously written a doctoral dissertation on the textual tradition.4 Although Clark’s edition was never properly completed (it lacks both an index and the discussion o f the manuscripts promised in the preface), it remains fundamental. What will long be the standard edition, the Teubner o f Wolfgang Seyfarth, has brought improvements to Clark’s text mainly in matters o f punctuation, typography, and the presentation o f the paradosis and renaissance conjectures in the apparatus criticus.5 Other texts sub­ sequent to Clark, it must be said with regret, but firmly, have marked a re­ gression to pre-critical standards by too often supplementing lacunae with unsatisfactory fifteenth-century conjectures (the Loeb edition by J. C. Rolfe [19 35 -19 3 9 ], the unfinished Budé, whose first volume appeared in 1968, and Roger Blockley’s selections.6 A complication recently brought to light calls into question the strictness o f the evaluation o f the paradosis on which Clark and Seyfarth base their con­ stitution o f the text. The early humanist Flavio Biondo claimed, in a marginal note found in a manuscript o f Ammianus that belonged to Cardinal Bessarion, that he had seen a manuscript that contained the equivalent o f a folium in the lacuna at 16.10.4: h ic d e e st u n iu s fo lii s c r ip tu r a e x iis q u e in e x e m p la r i v e t u s t o le g isse m e m in i, e t est p ars m u lti fa c ie n d a : a m e in Italia a p u d O c r ic u l u m p o s ita . . B . 7 •’ G . Pasquali, Storia della tradizioue e critica del testo2 (Florence, 1962), 81 —83, I 2 34 C . U. Clark, The Text Tradition of Ammianus Marcellinus (Diss. Yale, 1904). Clark had studied with Ludw ig Traube in M unich and his work was both encouraged and assisted by Mommsen. 5 O n Seyfarth’s edition, see R. Cappelletto, Rio. fil. 109 (19 8 1), 8 0 - 8 5 . Clark printed a double apparatus below his text, with the readings o f V and M presented separately from those o f the derivative manuscripts and editors; he italicized emendations printed in the text even if the emendation affects only part o fa word (e.g., 16 .8 .1: perpetrabantur); and he added a comma after every clausula within a sentence or a clause whether the sense requires it or not, sometimes even where the added comma was likely to confuse a modern reader (e.g., 14 .2 .18 : congesta undique saxa, telaque habentes in promptu). Seyfarth elegantly marks internal clausulae by two typographical devices: an extra space within the line and a hasta when the clausula coincides with the end o f a printed line (see Appendix 6). 6 R. Blockley, Ammianus Marcellinus: A Selection (Bristol, 1980). This scholar fails, for ex­ ample, to mark lacunae in 1 5 . 5 . 1 4 - 1 5 , 30, even though an important part o f Ammianus’ story o f how Silvanus’ troops were persuaded to desert him appears to be missing and V demonstra­ tively leaves gaps o f 15 and 20 or 22 letters in the two passages. 7 Venice, Biblioteca Nazionale Marciana, lat. Z 388 [= 1850] = W , fol. 48V: best published by R. Cappelletto, Recuperi ammianei di Flavio Biondo (Rome, 1983), 19; Tav. 1,2, w ho also re-

[202]

T

he

T

e x t

o f

A

m m i a n u s

Rita CappeUetto, who has republished this marginal note and demonstrated that the attribution to Biondo in another manuscript o f Ammianus is correct (N = Paris, Bibliothèque Nationale, lat. 6120, fol. 30v), identified the passage to which Biondo refers: it is indeed found in the entry for Ocriculum in his Italia illustrata, which appears to quote from the missing portion o f text: S iq u id e m A m m ia n u s M a r c e llin u s , C o n s t a n t ii C a e s a r is C o n s t a n t in o p r im o n ati a d v e n t u m a C o n s t a n t in o p o li R o m a m lib r o s e x t o d e c im o d e s c r ib e n s , d ic it ip s u m d u x is s e in c o m it a t u O r m is d a m , P e r s a r u m g e n tis a r c h ite c tu r a e p e r itis s im u m , iu s sisse q u e illi u t p r im a r ia q u a e q u e d ig n io r a u rb is R o m a e a e d ific ia d ilig e n t e r in sp e c ta o r d in e sib i o s t e n d e r e t . E t c u m O c r ic u l u m d e it in e r e esset v e n t u m , P e rs a m , im ­ p e r a t o r e iu b e n t e o m n iu m c o llo q u io d e s t itu tu m , ab O c r i c u l o R o m a m p riu s in ­ g re ss u m fu isse q u a m q u o in lo c o u rb s in c h o a s s e t d is c e r n e r e ac in te llig e r e n o v e r it .*8

Moreover, Biondo’s Roma triumphans quotes a statement o f Ammianus about Christian clergy riding in carriages not found in the extant manuscripts: m o r e m v e r o h u n c , q u e m T a c itu s s a c e r d o tib u s c o n c e s s u m fu isse d ic it , A m m ia n u s M a r c e llin u s n o s tr o s s e rv a sse C h r is t ia n o s s a c e rd o te s u t m u lti q u a m o rn a tis s im i u n ic o v e h e r e n t u r c a rp e n to s c r ib it .9

CappeUetto argued that the missing passage that Biondo claims to have read described the flight o f Hormisdas from Persia (cf. 16 .10 .16 : cuius e Perside discessum supra monstravimus): she deduced that Biondo had seen a frag­ mentary manuscript o f Ammianus that probably contained Books X IV and XV, and she identified this fragmentary manuscript as part o f the lost manu­ script from Hersfeld, separated from the main body o f the text in the same way as Book X X X I certainly was.10 Alan Cameron has drawn a conclusion o f much greater consequence from the evidence discussed by CappeUetto: he argues that, because Gelenius’ edi­ tion o f 1533 repeats from the Froben edition o f 1518 the feeble supplement for 16.10.4 that Castelli had added in his edition (Bologna, 1517), the lost M had precisely the same lacuna in this passage as the extant V, not the superior

publishes the note in N (28). Clarks apparatus to 16 .10 .4 quotes the annotation in both W and N ; Seyfarth s does not. 8 Quoted from CappeUetto, Ricuperi (1983), 35, who notes Biondo’s handwritten alterations to the first edition o f Italia illustrata in Vatican, Ottobonianus latinus 2369, which she dates c. 1 4 5 3 - 1 4 5 5 (3 4 ). l} Biondo, Roma Triumphans 9.200F, quoted from CappeUetto, Ricuperi (1983), 74. 10 CappeUetto, Ricuperi (1983), esp. 4 0 - 4 3 (Hormisdas), 8 6 - 9 1 (arguing that the complete Hersfeldensis had been brought to Italy by Enoch o f Ascoli).

[203]

A

p p e n d ix

i

supplement known to and used by Biondo." Hence Cameron deduced that a third complete manuscript o f Ammianus survived the Middle Ages to be­ come available in Italy in the fifteenth century. That is a bold claim to base on such slender evidence. For, even if the inference could be accepted as plausible a priori, the material assembled by Cappelletto comprises only one apparent quotation in Biondo’s Italia illustrata that can be tied to a specific passage (16.10.4) and two from his later Roma triumphans (of 14 57—1459), which can­ not: the passage quoted above and an apparent adaptation o f an extant passage (17.4.5 ° n Cornelius Gallus).1112 The question is an intricate one that hardly admits o f a certain answer. E x­ treme caution is therefore necessary, and the seventeenth-century editor Valesius may have been correct to claim that Biondo was lying about what he had read. In the lacuna at 16.10.4 between apparatu and secunda Orfiti praefec­ tura Valesius supplied merely IV Kal. Maias, which is attested as the day on which Constantius entered Rome in 357 (Descr; cons. 357.2).1314 5At the place where Biondo claimed a lacuna o f a leaf, V marks a lacuna o f seventeen let­ ters. Whether or not Biondo should be believed, there is no reason to believe that the differing sizes o f the many blank spaces that the scribe o f V left to in­ dicate lacunae or illegible passages in his exemplar are in any way accurate in­ dications o f how much text has been lost in each case. Lacunae are extremely frequent in the transmitted text o f Ammianus. How frequent can perhaps best be illustrated by the following list o f passages in Books X V -X V II, where V marks a lacuna: 1 . P assages w h e r e little se e m s to b e lo st, s o m e tim e s o n l y a s in g le w o r d o r p a rt o f a w o rd : XV

3 . 1 0 , 5 .5 (tw o ), 7 . 3 , 1 1 . 1 8

XVI 2.3, 2.8, 8.6, 8.IO, 10.21, 1 1.6, 12.65, 12.70 XVII 1.3, 2.2, 13.21, 13.24 2 . P assag es w h e r e m o r e has b e e n lo st, p e rh a p s a su b sta n tia l a m o u n t o f te x t:

XV 4.1-3, 4 6-7. 4 1 1.14 5-2 - 3. 5 1 4 - 1 5 . 5-30 XVI 4.1-2 (42 letters), 8.i 3 15 XVII 3.6, 12.7 11 Alan Cameron, HSCP 92 (1989). 423-36. 12 Cappelletto, Ricuperi (1983), 6 6 - 7 4 . 12 Quoted by Cappelletto, Ricuperi (1983), 29, n. 40. Valesius quotes the note in N , reports Biondo’s own note, then adds “ sed ego Blondio non credo.” 14 Seyfarth marks thirteen separate lacunae in 15.4: most comprise a few letters, but V marks one as more than two lines (2 —3) and one as 30 lines (11). 15 V marks two lacunae, o f 19 and 2 7 letters, respectively, the latter between the words Atii-

ciique and vonnn, where Seyfarth accepts Clark’s emendation to Aniciorumque without any lacuna at all. [204]

T

h e

T

e x t

of

A

m m i a n u s

Whatever the cause, Book X X I X seems to be especially lacunose: the account o f the campaign o f Theodosius against Firmus in Mauretania and Numidia has disappeared (at 5 .3 —4, 5.30, where V marks a lacuna o f two and a half lines), and the loss must be postulated o f a substantial passage that registered the prae­ fecti urbis who held office between the summer o f 372 and the spring o f 374 (at 5 .1, where V again marks a lacuna o f two and a half lines, cf. Appendix 8). There are also some undeniable lacunae in passages where V does not mark any discontinuity in the text. The most important occurs in 24.7.3, where V reads “ hinc opulenta sed ille avidae semper ad ulteriora cupiditatis etc.” with no hint that anything is amiss. Grammar alone proves the existence o f a la­ cuna between opulenta and sed ille, and Valesius saw that a substantial passage has been lost here, which originally contained an account o f the activities o f Procopius and Arsaces. The forces that they commanded played a vital part Julian’s overall strategy in 363 (cf. 23.2.2, 2.5): in Ammianus subsequently states that he explained why they failed to carry out the tasks assigned to them (24.7.8: quod nec adminicula, quae praestolabamur, cum Arsace et nostris ducibus apparebant ob causas impedita praedictas), but this explanation has vanished without trace from the transmitted text. The paradosis o f Ammianus is often corrupt as well as lacunose. Over the cen­ turies editors have brought countless improvements to the text. But many corrupt readings may still stand unquestioned in Seyfarth’s text.16 A pair o f examples from the very first page o f Book X IV will illustrate. In the very first sentence o f the extant text, according to V and most editors, the Caesar Gal­ lus was “ ad principale culmen insperato cultu provectus” (14 .1.1). What does cultu mean here? Seyfarth has defended the transmitted text, taking insperato cultu as an ablative o f quality attached attributively to principale culmen, and renders “ mit einem Gepränge, auf das er nie zu hoffen gewagt hätte.” 1718But logic and Ammianus’ normal use o f the word cultus,18 together with two closely parallel passages, commend the conjecture saltu proposed by A. Keller­ bauer and A. Kiessling.19 Gallus was unexpectedly proclaimed Caesar as a re­ sult o f the usurpation o f Magnentius: Ammianus elsewhere uses saltus o f the unjustified promotion o f Agilo to replace Ursicinus (20.2.5: immodico saltu 16 See H. Gärtner, H enncs 97 (1969), 3 6 2 - 7 1 , whose arguments are not always accepted by Seyfarth. 17 In his note on 1 4 .1 .1 , Seyfart also argues that Gallus’ appointment as Caesar was “ kein Sprung” (1.256 , n. 3). 18 Viansino, Lexicon 1 .3 4 6 - 4 7 . 10 A . Kellerbauer, Blätter für das bayerische Gymnasialschulwcsen 7 (18 71), 12; A . Kiessling, Jah rb. für dass. P hil. 103 (18 71), 496. Seyfarth’s defense o f cultu is refuted by K. Bringmann, A n ­ tike und A ben d lan d 29 (1973), 5 8 - 5 9 n. 62.

[205]

A

p p e n d ix

i

promotus) and o f Valens’ conferring o f the title patricius on his undistinguished father-in-law Petronius (26.6.7: promotus saltu repentino patricius). In the second sentence o f Book X IV Ammianus states, according to V and all editors, that the Caesar Gallus p r o p in q u ita t e e n im r e g ia e stirp is g e n t ilit a t e q u e e tia m tu m C o n s t a n t ii n o m in is e f­ fe r e b a tu r in fastu s, si p lu s v a lu isse t, a u su ru s h o s tilia in a u c to r e m su a e fe lic ita tis u t v id e b a tu r . ( 1 4 . 1 . 1 )

The word etiamtum (= “ even then” ) 20 appears to be misplaced and should perhaps, therefore, be transposed to where it logically belongs, after fastus. If this transposition were made, then Ammianus would be saying that Gallus was puffed up by sharing the name o f Constantius, and even then (in the summer o f 353. before the deaths o f Montius and Domitianus) “ was disposed, as it seemed, had he possessed any greater power, to venture on a hostile course o f action against his benefactor.’’ An unacknowledged corruption should also be diagnosed in the famous passage describing Constantius’ entry into Rome when he sees lavacra in modum provinciarum exstructa (16.10.14). Admittedly, translators o f Ammianus find no difficulty in rendering the phrase as “ the baths built up in the manner o f prov­ inces” (Rolfe), “ des thermes aux constructions grandes comme des provinces” (Galletier), “ die Bäder, so gross wie ganze Provinzen” (Seyfarth) or “ the buildings o f the baths as big as provinces” (Hamilton and Wallace-Hadrill), and de Jonge claims that “ the exaggeration is very attractive and typical o f Ammianus.” But what can be the point o f comparing baths with whole provinces? Hadrianus Valesius conjectured piscinarum (“ bathrooms like swim­ ming pools” ). Seyfarth is right to reject that emendation (with appeal to G. Lumbroso),21 but the lack o f a convincing emendation does not establish provinciarum as what Ammianus wrote. Seyfarth was compelled to confess that he could not explain what the text he prints can mean.22 What is needed is surely a comparison o f baths in Rome to some larger structure or structures in the provinces: as a diagnostic conjecture, the emendation provincialium may be offered. The poor state o f the text in general is very relevant to the passage in the sec­ ond formal excursus on the city o f Rome in which Ammianus derides certain 20 O n etiamtum in Ammianus, see Viansino, Lexicon 1 .4 8 7 - 8 8 . 21 Lumbroso s comments on the passage are reported by G . B. Pighi, Aegyptus 13 (1933), 279. 22 H e confessed in the notes to his Latin-German edition that “ Der Sinn ist nicht eindeutig” (1 [1968], 296, n. 1 14).

[206]

T

he

T

e x t

of

A

m m i a n u s

Roman nobles for sporting pretentious names (28.4.6). Seyfarth, who repro­ duces the readings o f Clark but modifies his punctuation, prints the passage as follows: p r a e n o m in u m c la r it u d in e c o n s p ic u i q u id a m , u t p u ta n t, in im m e n s u m s e m e t to l­ lu n t, c u m R e b u r r i et F la v o n ii e t P a g o n ii G e r e o n e s q u e a p p e lle n tu r ac D a lii c u m T a r r a c iis et F e ra siis a liis q u e ita d e c e n s s o n a n tib u s o r ig in u m in s ig n ib u s m u ltis.

One difficulty in this text is only an apparent one. Ammianus cannot have re­ ferred to praenomina, since none o f the names quoted is a praenomen, and the only praenomina in use in the late fourth century were unpretentious ones such as Quintus.23 But praenomina here could mean “ family names,” 24 as it does in other writers contemporary with Ammianus.25 However, it might be prema­ ture to exclude the possibility that the transmitted praenominum may be a cor­ ruption o f an original that read either prae nominum or prae nominum.26 The names themselves are problematical: every one o f the seven as printed by Clark and Seyfarth is improbable, even those that are the result o f modem emendation. Hence discussion must begin from the form in which each o f the names is actually attested in V, which presents them as follows: c u m r e b u r r i e t fla b u n ii e t p a g o n ii g e r e o n is q u e a p p e lle n tu r ac d a lii c u m ta rra c u s et p e rra s u s q u e .

Every one o f these names is corrupt, including the first. For, although Rebur­ rus is a relatively well-attested Spanish cognomen,27 which has been restored by plausible emendation as the cognomen o f the praefectus urbi in 283 (Chr. min. 1 .66),28 it was a low-class name totally devoid o f the quality o f claritudo.29 23 E.g., L. Aurelius Avianius Symmachus, praefectus urbi in 3 6 4 - 3 6 5 ( C IL 6.1698 =

IL S

1698), and his son Q . Aurelius Symmachus, consul in 39 1 (C IL 6.1699 = IL S 2946; IC U R 1.39 5, 18 35 ; C IL 8.24584; A E 19 6 6.518 ), cf. O. Salomies, Die römischen Vornamen. Studien zur römischen Namengebung ( Commentationes Humanarum Litterarum 82, 1987), 4 0 4 - 1 3 . 24 B. Salway, JR S 84 (1994), 140. 25 W . Hartke, Römische Kinderkaiser (Berlin, 19 5 1), 1 3 0 —3 1 ; P. Veyne, Rev. pliil.3 38 (1964), 2 5 3 - 5 7 , citing Epigrammata Bobiensia 8b.9 (“ praenomen ductum ex atavis” ); H A, Macr. 3 . 4 - 7 ; Diad. 6 . 2 - 1 0 . Also C IL 3 .12 2 8 (Apulum: perhaps third century). 26 Ammianus uses the phrase prae ceteris in three other passages to contrast an individual with the group to which he belongs (25.6 .14 ; 2 8 .1.19 ; 29.5.4).

27

I. Kajanto, The Latin Cognomina ( Commentationes Humanarum Litterarum 36.2, 1965), 238,

registers a total o f 69 Reburri, o f whom 38 are Spanish. Salomies, Vornamen (1987), 1 3 1 - 3 2 , pro­ duces a single example o f Reburrus as a praenomen (C IL 2.59 1). 28 Th e name is transmitted as Titucius Roburrus: H. Dessau proposed Reburrus (P IR 1 T 209). 29 Syme, Ammianus (1968), 1 5 1 .

[207]

A

p p e n d ix

i

Adapting and incorporating earlier attempts at correction, Anthony Birley has now restored sense, point, and relevance to the passage by emending the whole series o f names (originally eight, o f which the last is lost) as follows: c u m P r o b i et F a lto n ii et R a g o n ii C e io n iiq u e a p p e lle n tu r A lb in i c u m P a m m a c h iis et V i t r a s i is . . . q u e .30

In each case, at least one contemporary bearer o f the name can easily be identified: Sex. Petronius Probus, consul in 3 7 1; Faltonius Probus Alypius, praefectus urbi in 391; Ragonius Vincentius Celsus, praefectus annonae before 391; Ceionius Rufius Albinus, praefectus urbi from 389 to 391; Pammachius, whom Jerome saluted as “ consulum pronepos et Furiani germinis decus” (E p. 66.6); and Memmius Vitrasius Orfitus, praefectus urbi in 353 —355 and again in 357

- 3 5 9 - 31

30 A. R. Birley, Historiae Augustae Colloquia, N .S. 111: Colloquium Maceratense (Bari, 1995), 5 9 -6 0 . 31 Whether or not Birley’s precise proposals are all correct, the persons to w hom Ammianus alludes are surely to be sought in PU R E 1 . 1 1 3 3 , 1 1 3 7 - 3 8 , 114 0 : Stemmata 7 (Anicii), n - 1 3 (Caesonii and Ceionii), 16 (Flaviani).

APPENDIX

[2 ] THE STRUCTURE OF LIVY’S A B URBE C O N D ITA

Livy embarked upon his history o f Rome from its origins before the Battle o f Actium: ' originally, therefore, he must have intended to conclude the work at some point earlier than 31 B.c., and it seems virtually certain that his orig­ inal intention was to compose a total o f one hundred and twenty books end­ ing either with the death o f Cicero in 43 b . c . or approximately at that point.123 As the work progressed, however, Livy decided to go beyond his original ter­ minus, and he continued his narrative as far as the death o f Drusus in 9 B.c., probably crowning the whole work with an epilogue that looked forward to the disaster o f Varus in a .d . 9.-1 Since the preface to Book C X X I referred to the death o f Augustus, this continuation cannot have been given to the world before September 14 .4 It may be inferred, therefore, that Livy, who was the last o f the Republican historians and who was twitted by Augustus for being 1 For proof o f this date (which undermines most o f what has been written about Livy’s at­ titude to Augustus), see T. J. Luce, TAPA 96 (1965), 2 0 9 -4 0 ; A . J. Woodman, Rhetoric in Clas­ sical Historiography (London/Sydney, 1988), 1 2 8 - 3 4 . 2 P. G . Walsh, Livy. Greece and Rome: New Surveys in the Classics 8 (1974), 9. 3 T h e words clades Quinctilii Vari at the end o f Periocha C X L II are known only from the read­ ings o f a lost manuscript that P. Pithou copied into the margins o f his copy o f Sigonius’ edition o f Livy, but the fact that Julius Obsequens ends with the sentence “ multitudo Romanorum per insidias subiecta est” provides apparent confirmation o f the report: accordingly, O. Rossbach printed the words in his Teubner edition o f the Periochae and Julius Obsequens (Leipzig, 1910): for subsequent discussion o f their authenticity, see F. Lautenbacher, BPW 30 (1910), 1 1 8 6 - 9 3 ; O. Rossbach, ibid., 1 3 9 6 - 9 8 ; P. Jal, Tite-Livc: Histoire Romaine 34 .1: Abrégés des Livres de l'H is­

toire Romaine de Tite-Live (Paris, 1984), i.c iii-c v , 2.54; J. Briscoe, Gnomon 57 (1985), 4 2 0 - 2 1 . 4 For proof that the heading “ E x libro C X X I qui editus post excessum Augusti dicitur,” is authentic, see M . D. Reeve, C Q , N . S. 38 (1988), 4 7 7 - 9 1 . O n the correct interpretation o f the words editus . . . dicitur, see A . Klotz, R E 13 (1927), 8 1 9 - 2 0 , 823; R. Helm, Hieronymus* Zusätze

in Eusebius* Chronik und ihr Wertfü r die Literaturgeschichte (Philologus, Supp. 2 1.2 , 1929), 5 2 - 5 3 . There is no substance whatever in the widely accepted theory that Livy died in

[209]

a

.d .

12 , two

A

p p e n d ix

2

a “ Pompeian” (Tacitus, Ann. 4.34.3), observed the cardinal rule o f imperial historiography: anyone who wished to write serious history could write only about dead emperors, since the living emperor expected (and deserved) pane­ gyric o f the sort that Velleius Paterculus lavished on Tiberius.5 Livy did not advance beyond his original terminus o f 43 B.c. as long as Augustus was still alive, and although he published a continuation after Augustus died, he stopped at the death o f Drusus in 9 b . c . s o that he could avoid the embarrassing pe­ riod when the husband o f Julia and son-in-law o f Augustus withdrew from public life and lived in exile while his political enemies prospered.6 The structure o f the surviving portion o f Livy’s history (including the lost Books X I- X X ) is in little doubt. Although he may have written and pub­ lished the first book separately, Livy arranged his vast narrative in groups o f five, ten, and fifteen books. The first pentad covered the regal period (I) and the early days o f the Republic down to the capture o f Rome by the Gauls (II—V). The second and third pentads together described Rome’s conquest o f peninsular Italy, perhaps with a hardly perceptible break between Books X and X I. The next fifteen books centered on Rome’s struggle with Carthage. Book X V I began with an account o f the origins and early history o f Carthage, and Book X X closed on the eve o f the Second Punic War: Livy thus clearly marked o ff a pentad that covered the years 264—220 b . c . and whose main themes were, first, the First Punic War (X V I—X IX ), then Rome’s conflicts with the Gauls in North Italy and the so-called First Illyrian War (XX). The ten books devoted to the Second Punic War (X X I—X X X ) begin with a for­ mal character sketch o f Hannibal and conclude with the triumph o f Scipio in 201: they comprise two carefully balanced pentads that present first Car­ thaginian successes and Rome’s refusal to accept defeat, then Rome’s gradual and inexorable advance to total victory.7 The theme o f the next fifteen books is Rome’s conquest o f the Greek world in three pentads that concentrate suc­ cessively on the wars against Philip V o f Macedon (X X X I—X X X V ), Anti­ ochus III o f Syria (X X X V I—X L), and Perseus (XLI—XLV).8 Among the ex­ tant books, I has a preface to the whole work, II to the history o f the Roman

yean before Augustus, which was developed (though not invented) by R. Sym e, Roman Papers, i (Oxford, 1979), 4 1 4 - 1 6 : see R. Jeffreys, C Q , N .S. 35 (1985). 1 4 0 - 4 8 . s H. Peter, Die geschichtliche Utteratur über die römische Kaiserzeit bis Theodosius I und ihre Quellen, i (Leipzig, 1897), 2 9 3 - 9 4 , cf. T. Janson, Latin Prose Prefaces (Stockholm, 1964), 1 0 0 - 6 . 6 O n this obscure period, see esp. R. Syme, Roman Papers, 3 (Oxford, 1984), 9 1 2 —36; The

Augustan Aristocracy (Oxford, 1986), 8 2 —103. 7 O n the structure o f X X 1- X X X , see esp. E. Burck, Einführung in die dritte Dekade des Livius (Heidelberg, 1950), 7 - 5 6 ; U vy, ed. T. A. Dorey (London, 19 71), 2 1 - 4 6 ; P. G. Walsh, A N R IV 2.30 .2 (1983), 1 0 5 8 - 7 4 . * In favor o f analyzing Books X X X I - L as two decades, see P. Jal, Rev. p h il? 49 (1975), 2 7 8 -8 5 .

[2 1 0 ]

T

h e

S

t r u c t u r e

of

L

i v y

’s A b

U

rbe

C

o n d it a

Republic, VI to the period after the Gallic capture when Rome was reborn and the historian’s materials become clariora certioraque, X X I to the Second Punic War, and X X X I to the wider canvas o f Rome’s dealings with the Hel­ lenistic world (“ iam provideo animo . . . in vastiorem me altitudinem ac velut profundum invehi” ), while the presumed preface to X LI is lost in a lacuna. What o f the almost one hundred lost books from XLVI onward? Certainty is unattainable. The Periochae are unreliable indicators o f the emphases that Livy gave his material where they can be checked against the surviving books: therefore, there is no valid reason to suppose them more accurate and trust­ worthy for the lost books.9 Until recently, the consensus o f scholarship has been that the weight and complexity o f his material compelled Livy to aban­ don his original plan o f writing in pentads and decades.101However, the analy­ ses o f the structure o f the whole work by P. A. Städter and T. J. Luce, even though they disagree in detail, have proved that, if Livy ever abandoned pentadic and decadic structure, he did so only toward the very end o f his his­ tory.11 Moreover, G. Wille and R. M . Ogilvie have argued that Livy never abandoned composition in groups o f five, ten, and fifteen books: had he lived long enough, they suggest, Livy would have produced a total o f 150 books.12 Although it is implausible, for reasons that have nothing to do with the struc­ ture o f his history, to suppose that Livy intended to press on beyond 9 B .c . or that he could have written about Tiberius’ exile, close examination o f the contents o f Periochae C X X X I —C X L II renders it highly probable that Livy maintained a structure o f pentads and decades through to the very end, so that his account o f the reign o f Augustus was arranged in two groups o f five books. For there is something seriously amiss with the transmitted book numbers o f the last few Periochae, which ascend to the aesthetically implausible total o f one hundred and forty-two. According to the Periochae, Livy included the follow­ ing temporal periods or events in each o f the books from C X X X I onward: CXXXI

E v e n t s f r o m la te 3 6 to th e e n d o f 3 5

C X X X II

F r o m O c ta v ia n ’s D a lm a tia n c a m p a ig n o f 3 4 to th e b a ttle o f A c t iu m in A u g u s t 3 1

C X X X III

F r o m A n t o n y ’s flig h t fr o m A c t iu m to O c ta v ia n ’s tr ip le t r iu m p h o f 1 3 - 1 5 A u gu st 29

9 P. Jal, Tite-Live 3 4 .1: Abrégés (1984), i.lv ii-lx x ix . 10 See the conspectus o f analyses by scholars from H. Nissen (1872) to R. Syme (1959) tab­ ulated by T. J. Luce, Livy. The Composition o f his History (Princeton, 1977), 14. 11 P. A . Städter, Historia 2 1 (1972), 2 8 7 - 3 0 7 ; Luce, Livy (1977), 3 - 2 4 , cf. P. Jal, Tite-Live: Histoire Romaine 1 1: Livre X X I (Paris, 1988), v ii-ix . 12 G. Wille, Der Auflau des Livianischen Geschichtswerks (Heuremata 1, R. M . Ogilvie, P C P S 2 10 (1984), 1 1 6 - 2 5 .

1973),

114 -19 ;

A

C X X X IV

p p e n d ix

2

P resum ably fro m 2 9 and at least as far as A u g u stu s’ visit to G a u l in sp rin g 2 7

CXXXV

T h e w a r w a ge d b y Crassus in T h r a c e (presum ably cam paigns o f 2 8 - 2 7 ) ; A u g u stu s’ con qu est o f Spain (officially co m p leted in 2 5 ); the subjugation o f the Salassi (25)

CXXXVI

Lost, but C en so rin u s reports that it included the L u d i Saeculares o f i 7 13

C X X X V II

Lo st

C X X X V III

T h e con qu est o f Raetia ( 15 ) ; D ru su s’ census in G a u l ( 1 3 ) ; the death o f A g rip p a ( 12 )

C X X X IX

Even ts o f 13 and 12 : D rusus crosses the R h in e, suppresses troubles co n n ected w ith the G allic census and dedicates the altar o f R o m e and A ugustus at L y o n .14

CXL

T h e cam paigns o f Piso in T h r a c e ( 1 2 - 1 0 ) and D rusus in G e r ­ m an y; the death o f O cta via , the m o th er o f M arcellu s ( 1 1 )

CXLI

D ru su s’ cam paign o f 1 1 ; T ib e riu s ’ defeat o f the D alm atians and Pannonians; peace w ith the Parthians, that is, the h an d in g o ver o f the fo u r sons o f Phraates in 1 0 13

C X L II

D ru su s’ cam paign o f 9; his death, funeral, and p osth um ous honors.

From 29 down to 12 b . c ., each book o f Livy covered at least three years. Ac­ cording to the transmitted numbers, the last three books covered one year each. That is a priori most improbable. It should be inferred that the trans­ mitted total is erroneous.16 After the Periochae o f C X X X V I and C X X X V II had been lost, it seems that someone, in a futile attempt to repair the loss, ex­ panded the summary o f Book C X L to cover three books (viz., the Periochae o f C X L —CXLII). In Livy’s original scheme, five books centered on Augustus as a military leader, from his first campaign in Dalmatia to his last active cam­ paign in Spain (C X X X I—C X X X V ), with which Augustus had concluded his autobiography,17 were followed by five books on the campaigns o f Drusus and Tiberius, the reigning emperor and his late brother. 13 Censorinus, De die natali 17.9, whence Livy, frag. 65 Weissenborn-Müller = frag. 63 Jal. 14 On the multifarious problems concerning the altar, R. Turcan, A N R W 2 .1 2 .1 (1982), 6 0 7 -4 4 ; D. Fishwick, The Imperial Cult in the Latin West. Studies in the Ruler Cult o f the Western Provinces o f the Roman Empire 1.1 (Leiden, 1987), 9 7 - 1 3 7 . 15 Res Gestae 32.2, cf. J. Gagé, Res Gestae D ivi Augusti3 (Paris, 1977), 142. 16 J. Briscoe, Gnomon 57 (1985), 420, conjectured that “ the author o f the Periochae” may have had only 140 books in his text o f Livy. 17 Suetonius, Div. Aug. 8 5.1: the fragments are collected by E. Malcovati, Imperatoris Caesaris Augusti Operum Fragmenta5 (Turin, 1969), 8 4 - 9 7 .

[212]

APPENDIX

[3 ] AMMIANUS’ REFERENCES TO THE LOST BOOKS

J. F. Gilliam published a useful discussion o f the passages in the Res Gestae that concern the period from Hadrian ( 1 1 7 —138) to Carinus (283 —285), including those that make no explicit reference to the lost part o f the work.1 He noted several passages where use o f Herodian is clear; in one o f them Ammianus fol­ lows Herodian against the superior authority o f Cassius Dio and the Historia Augusta (22.9.6 < Herodian i . n . 1 - 2 ; 26.8.15 < Herodian 3 .4 .1-6 ; 2 9 .1.17 < Herodian 1.8.5, cf. Dio 72 4 4, H A, Comm. 4 .2 -4 ; 3 1 .1 0 .1 8 - 1 9 < Hero­ dian 1.15.6). The present appendix confines itself to quoting Ammianus’ ex­ plicit references to what he had written in the lost books in general support o f the thesis about the total number o f books advanced in Chapter III.12 Schol­ ars who accept the transmitted and traditional numeration must perforce ar­ gue that most o f the lost books contained a relatively brief narrative with few excursus.3 14 .1.8

u t in G o r d ia n o r u m a c tib u s factita sse M a x im in i t r u c u le n t i illiu s im p e r a t o r is re ttu lim u s c o n iu g e m

1 4 .4 .2

s u p e r q u o r u m [sc. th e S a ra c e n i] m o r ib u s lic e t in a c tib u s p r in c ip is M a r c i et p o ste a a liq u o tie n s m e m in i re ttu lisse

1 J. F. Gilliam, Bonner Historia-Augusta-CoHoquium 1970 (Bonn, 1972), 1 2 5 —47, c f A. M . Em ­ mett, History and Historians (1983), 45, 52 n. 32.

2 For a catalogue o f the phrases that Ammianus uses for cross-references, forward as well as backward, see H . Cichocka, Eos 64 (1976), 2 0 3 - 2 2 . Mistaken cross-references and historical al­ lusions misidentified as cross-references are discussed by R. M . Frakes, Phoenix 49 (1995), 2 3 4 43, whose list o f “ definite cross-references” misleadingly includes 3 1.16 .9 ( 2 4 3 -4 5 ). 3 H. Cichocka, Eos 63 (1975), 340.

[213]

A 1 4 . 7 .7

p p e n d ix

3

Serenianus e x d u ce, cuius ignavia populatam in P h o e n ice C elsein ante rettulim us

P resum ably b y Je w ish rebels in 3 5 1 o r 3 5 2 d u rin g the in su rrection in G a lile e .4 1 4 .7 .2 1

M esop o tam ia iam digesta, cu m bella Parthica narrarentur

P ro bably in several excursus.5 1 4 .1 0 .2

H erm o g e n is ex m agistro eq uitum filius apud C o n sta n tin o p o lim , ut supra rettulim us, po pu larium q u o n d am turbela discerpti

In early 3 4 2 (Je ro m e , Chronicle 2 36 *; Descr. cons. 3 4 2 .2 ). 15 .5 .16

Lanogaiso vetante, tunc tribu n o , q u em d u m m ilitaret candidatus solum affuisse m o ritu ro C o n sta n ti supra rettulim us

In early 3 5 0 , shortly after 18 Ja n u a ry.6 1 5 .6 .4

Poem enius raptus ad suppliciu m interiit, qui, ut supra rettulim us, c u m T re veri civitatem clausissent D e c e n tio , ad defenden dam p leb em electus est

A p p aren tly in 3 5 3 , shortly before the final defeat o f M a g n e n tiu s.7 16 .6 .2

D o ru s q uidam e x m ed ico scutariorum , q u e m n iten tiu m rerum cen tu rio n em sub M a g n e n tio p ro vectu m rettulim us accusasse A d e lp h iu m urbi p raefectu m

C lo d iu s C elsinus A delph ius was praefectus urbi fro m 7 J u n e to 18 D e c e m b e r 3 5 1 ( C hr ; min. i .66). 1 6. io . 1 2

q uo d autem per o m n e tem pus im perii nec in consessum veh icu li q u em q u am suscepit nec in trabea so ciu m p rivatu m ascivit, ut fecere principes consecrati, et similia m ulta elatus in ardu um supercilium tam quam leges aequissimas observavit, praetereo m e m o r ea m e ret­ tulisse cu m incidissent

16 .10 .16

H orm isdas cuius e Perside discessum supra m onstravim us

P resum ably circa 3 2 4 : Jo h n o f A n tio c h states that it was the e m p ero r Licin iu s 4 T h e nature and extent o f the rebellion are discussed b y M . A vi-Y o n ah , The Jews o f Palestine. A Political Historyfrom the Bar Kokhba War to the Arab Conquest (O xford, 19 7 6 ), i 76 - 81 ; J. Geiger, Scripta Classica Israelica 5 ( 1 9 7 9 - 8 0 ) , 2 5 0 - 5 7 ; P. Schäfer, Tradition and Re-interpretation in Jewish and Early Christian Literaturey ed. J. W . van H enten (Leiden, 1986), 1 8 4 - 2 0 1 ; J. A rce , Athenaeum, N . S. 65 (1987)» 1 0 9 - 2 5 ; G . Stem berger .Juden und Christen im Heiligen Land. Palästina unter Kon­ stantin und Theodosius (M unich, 19 8 7), 1 3 2 - 5 0 ; M . M o r, The Eastern Frontier of the Roman Em ­ pire, ed. D . H . French and C . S. Lightfoot (O xford, 1989), 3 3 5 —53. N o n e o f them appears to discuss this passage o f Am mianus. Th om pson, Ammianus (19 4 7 ), 58, dated the rebellion to 3 5 3 - 3 5 4 and deduced that A m m i­ anus deliberately omitted it. 5 Reference to a single excursus, w h ich occurred in Am m ianus’ narrative o f the period 2 4 0 298, is argued by R. M . Frakes, A H B 7 (19 9 3), 1 4 3 - 4 7 . 6 For the evidence, see Athanasius (19 9 3), 2 2 5 - 2 6 . Lanaogaisus is otherwise unknow n.

7 R IC 9 . 1 6 4 - 5 , Trier 3 2 8 - 3 7 (coins minted in the name o f Constantius as sole Augustus). Poemenius is not in P L R E 1.

[214]

A

m m i a n u s

’ R

e f e r e n c e s

t o

t h e

L

o s t

B

o o k s

w h o r e c e iv e d H o rm is d a s (frag. 1 7 8 ) , w h ile Z o s im u s n arrates th e s to r y o f H o r m is d a s ’ flig h t to A r m e n ia an d su b se q u e n t re c e p tio n b y C o n s ta n tin e b e ­ tw e e n th e d efea t o f L ic in iu s an d his su rre n d e r o n 2 9 S e p t e m b e r 3 2 4 ( 2 .2 7 , cf. 3 . 1 3 . 3 - 4 ; Z o n aras 1 3 . 5 . 1 3 ) . 8 18 .9 .3

T r ic e n sim a n i D e c im a n iq u e F o rten se s et S u p e rv e n to re s atq u e P ra e ­ ve n to re s c u m A e lia n o iam c o m ite , q u o s tiro n es tu m e tia m n o vello s h o rtan te m e m o ra to a d h u c p ro te c to re erupisse a S in g a ra P ersasque fusos in so m n u m rettu lim u s trucidasse c o m p lu re s

19 .2 .3

e le p h a n to ru m agm in a rugosis h o rre n d a c o rp o rib u s le n ite r in c e d e ­ b a n t arm atis onusta, ultra o m n e m d irita te m taetri sp ectacu li fo r m i­ danda, ut rettulim us saepe

2 0 .1.1

(sc. Ju lia n ) v e re b a tu r ire sub sidio transm arinis, u t rettu lim u s ante fecisse C o n sta n te m

In late Ja n u a r y o r F e b ru a ry 3 4 3 . 9 2 0 .6 .5

reseratam u rb e m o b sid io su p e rio re d o c u im u s

T h e “ p re c e d in g sie g e ” o f Sin g a ra w as p re su m a b ly raised b y the fam o u s n ig h tbattle, w h ic h (it is n o w clear) m u st be dated to 3 4 4 . 10 2 0 .1 1 .3 2

[sc. C o n sta n tiu s] v in c e re saltem p e r du ces optabat, q u o d aliqu otiens m e m in im u s con tigisse

2 1.8 .1

c o m m isit . . . et Io v io quaestu ram , cuius in actibus M a g n e n ti m em in im u s

2 1.16 .7

q u o d au tem n e c os tersisse u n q u am vel nares in p u b lico n e c spuisse n ec transtulisse in p artem alterutram v u ltu m aliqu an do est visus n e c p o m o ru m q u o ad v ix e r it gustaverit, ut dicta saepius p raeterm itto

2 2 .9 .6

cuius [sc. the statue o f C y b e le fro m Pessinus] super adventu p auca cu m aliis h u ic m ateriae co n gru en tib u s in actibus C o m m o d i p rin ­ cipis digessim us p er e x c e s s u m 11

2 2 .13 .3

A sclep iades p h ilosoph us, cuius in actibus M a g n e n tii m em in im u s

2 2 .15 .1

res A e g y p tia c a e tangantur, q u aru m n o titiam in actibus H ad rian i et Sev eri p rin cip u m digessim us late, visa pleraque narrantes

2 3 .5 .7

G o rd ia n i im p e ra to ris . . . cuius actus a p ueritia prim a e xe rcitu u m q u e felicissim os ductus et insidiosum in teritu m digessim us tem pore com p eten ti

8 For what else is known about Hormisdas, P L R E 1.443; on the significance o f what A m ­ mianus makes him say to Constantius, see R. O. Edbrooke, Mnemosyne4 28 (1975), 4 12 —17. 9 Athanasius (1993), 225. 10 Athanasius (1993), 220, 3 1 2 n. 19; R. W . Burgess, Continuatio Antiochiensis Eusehii: A n A n ­ tiochene Continuation of Eusebius' Chronici canones (forthcoming), no. 37. 11 For excessus = “ digression,” see 2 3.6 .1, 2 7 .4 .1, 28.4.6; for its commoner synonyms, J. M ar­ tin, Antike Rhetorik. Technik und Methode (Munich, 1974), 8 9 - 9 1 .

[215J

A 2 3 .6 .2

p p e n d ix

3

h o c r e g n u m [sc. P ersia] . . . o b cau sas, q u a e sa e p e re ttu lim u s , c u m a p u d B a b y lo n a M a g n u m fata ra p u isse n t A le x a n d r u m , in v o c a b u lu m P a rth i c o n c e s s it A rsa c is

2 3 .6 .2 4

q u a [sc. S e le u c ia ] p e r d u c e s V e r i C a e s a r is , u t a n te re ttu lim u s , e x p u lsa ta

2 3 .6 .5 0

u b i [sc. in H y r c a n ia ] e tia m t ig r id u m m ilia m u lta c e r n u n t u r fe r a e q u e b e stia e p lu re s, q u a e c u iu s m o d i s o le n t c a p i c o m m e n t is , d u d u m n o s m e m in im u s rettu lisse

2 3 .6 .8 3

m ilita ri c u ltu ac d is c ip lin a [sc. o f th e P ersian s] p r o lu d iis q u e c o n ­ tin u is re i c astren sis e t a rm a tu ra e , q u a m sa e p e fo r m a v im u s

2 5 .4 .2 3

sc ia n t d o c e n t e v e rita te p e rs p ic u e n o n Iu lia n u m sed C o n s t a n t in u m a rd o re s P a rth ic o s s u c c e n d iss e , c u m M e t r o d o r i m e n d a c iis a v id iu s a c q u ie s c it, u t d u d u m re ttu lim u s p le n e

2 5 .8 .5

q u o d [sc. th e t o w n o f H a tra ] e r u e n d u m a d o r ti t e m p o r ib u s v a riis T ra ia n u s et S e v e ru s p rin c ip e s b e llic o s i c u m e x e r c itib u s p a e n e d e le ti su n t, u t in e o r u m a c tib u s has q u o q u e d ig e s sim u s p a rte s

2 7 .8 .4

q u o n ia m c u m C o n s ta n tis p r in c ip is actu s c o m p o n e r e m , m o t u s a d o ­ le sc e n tis et s e n e sc e n tis o c e a n i s itu m q u e B r it a n n ia e p r o c a p tu v ir iu m e x p la n a v i, ad ea q u a e d ig e sta su n t s e m e l r e v o lv i s u p e r flu u m d u x i

v C le a r ly in h is a c c o u n t o f C o n s t a n t iu s ’ fa m o u s w in t e r c r o s s in g o f th e E n g lis h C h a n n e l in J a n u a r y - F e b r u a r y 3 4 3 . 12 2 8 .3 .8

a rc a n o s g e n u s h o m in u m a v e t e r ib u s in s titu tu m , s u p e r q u ib u s a liq u a in a c tib u s C o n s t a n tis re ttu lim u s

A m m ia n u s is h e r e d e s c r ib in g th e e x p lo its o f th e e ld e r T h e o d o s iu s in B r it a in : h e m u s t h a v e e x p la in e d w h o th e arcani w e r e in th e c o n t e x t o f C o n s t a n s ’ B r it is h e x p e d it io n o f 3 4 3 . 2 8 .4 .6

p r im o n o b ilita tis [sc. o f R o m e ] , u t a liq u o tie n s p r o l o c o r u m c o p ia fe c im u s ,13 d e in p le b is d ig e r e m u s e rra ta

2 9 .5 .1 6

Ic o s iu m o p p id u m , c u iu s su p ra d o c u im u s c o n d it o r e s

2 9 .5 .1 8

C a e s a r e a m . . . u r b e m o p u le n ta m q u o n d a m et n o b ile m , c u iu s it id e m o r ig in e m in A f r ic a e situ d ig e s sim u s p le n e

2 9 .6 .1

o b se ssa q u e ab is d e m [sc. th e Q u a d i] M a r c o m a n n is q u e A q u ile ia d iu O p it e r g iu m q u e e x c is u m et c ru e n ta c o m p lu r a p e r c e le re s acta p ro c in c tu s v i x re siste n te p e rru p tis A lp ib u s Iu liis p r in c ip e P io q u e m an te d o c u im u s M a r c o

12 F irm icu s M atern u s, D e errore profanarum religionum 2 8 .6 ; L ib an iu s, Orat. 5 9 . 1 3 9 , cf. A thana -

sius ( 1 9 9 3 ) , 2 2 5 . 13 Frakes, P hoenix 4 9 ( 1 9 9 5 ) . 2 4 2 > 24

232 13 2 -3 « 8 8 -9 0 139

1.1 4 .2 - 6 8 9.7 9.12

194 91, 177 i , 76, 96 7m 5 8 -5 9

H 12.6 12.8 14 .6 -8 1 6 .1 7 -1 8 16.23

5 3 » «5 , 8 7 ,

i

XXVII

X X V III 2 34 -36

2.6 5.5 6.24 2.16 4.27

66 84

6.15 8.1

156 -5711 205 no

3.2

69

4

5 8 - 5 9 , 84, 1 4 6 - 4 7 ,

5

52, 1 4 0 - 4 1

is c u s s e d

6.10 6 .1 1

X X IX

109 249

35-10

1 16

3 .11 10 .111 .1 -

h i,

12 3

i 1.57 3 .4 - 6 4 .14.6 6

161 85 168 169 101 4 6 -5 1, 18 0 -8 1, 2 19 -2 0 , 234 232 181 28

7.3

XXV

XXVI

D

2

119 18 1 118 -19

108, 109, 182, 2 3 3 34, 237, 2 4 1-4 6 2711,41 109 119 2 0 5-7 n o , 182

2.16

122

3-4

3.7 5 6 .1 7 - 1 9

83 70 41, 205, 2 3 7 -4 0 115 , 2 3 7 - 3 8

XXX

4.3 - 2 2 9.2

7 -8 ,10 9 126

XXXI

5 .11 —17 7-i6 1 2 .8 - 9 16.8 16.9

149, 173, 1 7 5 - 7 6 19311 83 18 5 - 8 6 i, 26, 54, 6 5 -6 6 , 79, 185

162, 181

[290]