East and West, Modes of Communication: Proceedings of the First Plenary Conference at Merida (Transformation of the Roman World)

EAST AND WEST: MODES OF COMMUNICATION

THE TRANSFORMATION OF THE ROMAN WORLD A SCIENTIFIC PROGRAMME OF THE EUROPEAN SCIENCE FOUNDATION Coordinators J A V I E R A R C E · E V A N G E L O S G H R Y S O S · IAN W O O D Team Leaders

Steering Committee

Miquel Barcelô M a r k Blackburn G i a n p i e t r o Brogiolo Alain Dierkens Richard Hodges M a r c o Mostert Patrick Périn W a l t e r Pohl Frans Theuws Leslie W e b s t e r

Gunilla A k e r s t r ö m - H o u g e n Volker Bierbrauer Niels H a n n e s t a d Przemysfaw U r b a n c z y k Mario Mazza H . H . van Regteren Altena Heid Gjöstein Resi L. C r a c c o Ruggini

Series Editor IAN W O O D VOLUME 5 EAST AND WEST: MODES OF COMMUNICATION

EAST AND WEST: MODES OF COMMUNICATION Proceedings of the First Plenary Conference at Merida

E D I T E D BY

EVANGELOS

CHRYSOS

AND

IAN

WOOD

' / 6 8 V

BRILL LEIDEN · BOSTON · KÖLN 1999

Library o f C o n g r e s s C a t a l o g i n g - i n - P u b l i c a t i o n D a t a East and West : modes of communication : proceedings of the first plenary conference at Merida / edited by Evangelos Chrysos and Ian Wood. p. cm. (Transformation of the R o m a n world, ISSN 1386-4165 ; v. 5) Papers presented at the first (Merida, 1994) and last (1997, Isernia) conferences of the European Science Foundation program on the transformation of the R o m a n world. Includes bibliographical references and index. Contents: T h e city of Mérida (Emerita) in the Vitas Patrum Emeritensium (Vlth century A.D.) / Javier Arce — Images as a substitute for writing / Peter Brown Images as a substitute for writing : a reply / Ian Wood — Administrative language and its public deployment / N. Oikonomedes — Some reflections on Mandarin language / Mayke de J o n g - Communicating holiness / Lennart Rydén T h e use and abuse of Latin hagiography in the early Medieval West / Ian Wood — Social language and its private deployment / Averil Cameron — Social language, identities and the control of discourse / Walter Pohl. ISBN 9004109293 (alk. paper) 1. Communication- History Congresses. I. Chrysos, Euangelos K. II. Wood, I. N. (Ian N.), 1950- . ' III. Series. P90.E23 1999 302.2Ό9—dc21 99-23080 CIP Die Deutsche Bibliothek - CIP-Einheitsaufnahme East and w e s t : m o d e s of c o m m u n i c a t i o n : proceedings of the first, plenary conference at Merida / ed. by Evangelos Chrysos and Ian Wood. - Leiden ; Boston ; Köln : Brill, 1999' (The transformation of the Roman world ; Vol. 5) ISBN 90-04-10929-3 ISSN ISBN

1386-4165 90 04 10929 3

© Copyright 1999 by Koninklijke Brill NV, Leiden, The Netherlands All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers MA 01923, USA. Fees are subject to change. PRINTED IN THF. NETHERLANDS

CONTENTS

Preface

vii

Introduction Evangelos Chrysos

ix

T h e City of Mérida (.Emerìta) in the Vitas Patrum Emeritensium (Vlth Century A.D.) Javier Ar ce

1

Images as a Substitute for Writing Peter Brown

15

Images as a Substitute for Writing: A Reply Ian Wood.

35

Administrative Language and its Public Deployment N. Oikonomedes

47

Some Reflections on Mandarin Language Mayke de Jong

61

Communicating Holiness Lennart Rydén

71

T h e Use and Abuse of Latin Hagiography in the Early Medieval West Ian Wood

93

Social Language and its Private Deployment Averti Cameron Social Language, Identities and the Control of Discourse Walter Pohl

111

....

127

Mit was für Mitteln kann einem physisch Anonymen Auctontas Verliehen werden? Beat Brenk H o w did rising Christianity cope with Pagan Sculpture? Niels Hannestad Le développement des langues regionales et l'introduction d'alphabets dans des communautés illettrées Christian Hannick Conflits et compromis langagiers en Occident latin: de la crise culturelle à l'invention linguistique (III C ‫־‬X C siècle) Michel Banniard

143

173

205

223

Transformation of the Roman World: Reflections on Current Research Paolo Delogu

243

The Transformation of the Roman World: Reflections on Five Years of Work Thomas F.X. Noble

259

Modes of Communication: An Afterword

279

Ian Wood Index

283

List of Illustrations Illustrations

287

PREFACE

This volume of papers from the ESF programme on " T h e Transformation of the R o m a n World" contains the texts of lectures given at the first and last plenary conferences of the project: the one held in Mérida in 1994, the other in Isernia in 1997. T h e first conference took as its special theme the matter of "Modes of C o m m u nication", and was intended to provide an opportunity for comparing the experience of the Late and post- R o m a n West with that of the Byzantine East (a region which for logistical reasons was only given secondary consideration in the programme as a whole). T h e lectures from that conference constitute a broad and interdisciplinary assessment of the subject. T h e two plenary lectures of the final conference consider the question of the transformation of the Roman World itself, and the achievements of the ESF programme in dealing with that question. T h e first of them reflects on the subject as it looks now after five years of work: the second assesses that work and makes suggestions for the future. Although the papers in this volume were edited by two of the project's coordinators, the Mérida conference was very largely arranged by the third, Javier Arce, with expert help from members of the ESF office and the Escuela Espanola de Historia y Arqueologia in Rome, as well as from the city and museum of Mérida, and in particular from Pedro Mateos. T h e organisation of the Isernia conference fell to the safe hands of Federico Marazzi, who also had much appreciated support from the local commune and region, and also from the wonderfully enthusiastic community of Benedictine nuns of San Vincenzo al Volturno. T h e programme owes a great debt of gratitude to all involved in organising both conferences. T h e articles here are, to a very large extent, the texts of the lectures as delivered. I have not attempted to edit out the sense of their being spoken texts: that is how they were constructed and delivered, and as such they have the directness and appeal of the spoken word. Where necessary and possible I have added to the footnotes, but there are still occasions, particularly in the two lectures from Isernia, where there is allusion to a precise audience and auditorium, and where there is reference to work done within the project, but not

yet in print. T h e reader will, I hope, forgive me, if she or he has to wait for explanation from later volumes in the series. Editing such a volume of international and interdisciplinary papers obviously requires the linguistic and other academic skills of more than one specialist, and in preparing the manuscript for publication I have been helped by numerous friends and colleagues, who have read through texts and checked references. I would particularly like to thank Dr. Ann Christys, Dr. Walter Pohl, Mag. Helmut Reimitz and Professor Roger Wright for their help in getting the papers of this volume ready for the press. Thanks also to Liz Mason for compiling the index. Ian Wood

INTRODUCTION Evangelos Chrysos

In recent years a methodological trend in historical research has been to ask who the audience was for every single text, including literary as well as historiographical works, and even legal documents. For, as we know from our own experience, even legal documents, normally written in technical form and vocabulary, are at the same time phrased in a way which aims to gain consensus from the publie and to create the right "atmosphere" for their implementation. O n the other hand, a sort of "stratigraphie" approach to the sources, aimed at uncovering several layers of texture in the source material, has produced more subtle tools for interpreting their messages. This trend, however, has occasionally become an obsession, leading scholars to attribute the strangest or most malicious motives and intentions to authors in every idea and literary product that has come down to us. These are some of the reasons why the programme on " T h e Transformation of the R o m a n World" identified the topic of communication as an important question to deal with. O n the other hand, in a time of world communication nets and the "www" historians have become more alert to the question of how people in previous generations conveyed their messages; which were their modes and forms of communication. Those who have some experience in modern media production have learnt to focus first on who the "user" or the recipient of the information is expected to be; in other words, how the information will be disseminated to the "public". T o this end, a conference was held in Athens in 1991 on "Communication in Byzantium", 1 while in recent years colleagues have been attracted by such questions as the tools for publishing ideas and thus influencing public opinion 2 and the public as the recipient of messages.

1

Sec Η επικοινωνία στο Βυζάντιο, ed. Ν.G. Moschonas (Athens, 1993). G. Weiss, "Publizistik in Byzanz. Ein Beitrag zur Kommunikationsforschung in Byzanz", Byzantinische Zeitschrift 89 (1996), pp. 79-93. 3 K.P. Matschke, "Die spätbyzantinische Öffentlichkeit", Mentalität und Gesellschaft im Mittelalter. Gedenkschrift für Ernst Werner (Frankfurt, 1993), pp. 155 224. 2

This volume contains the papers of the first plenary conference of the European Science Foundation p r o g r a m m e on " T h e Transformation of the R o m a n World", which was held in Mérida in 1993. Technical reasons prevented the publication of the proceedings from being the first volume of the series. This unexpected delay has happily enabled us to include in the volume the two main papers of the third and final plenary conference of the programme, which was held in Isernia in J u n e 1997: a keynote address by Paolo Delogu and the very last paper at the conference by the American Professor T h o m a s Noble, who as an outside "reader" was asked to look critically from overseas and comment on the work accomplished by the programme. T h u s the volume represents some of the early steps of the programme, and at the same time some of the reflections on the whole theme at the end of our five years of research cooperation. At the beginning of the volume is an offering to the genius loci of our first plenary conference, the city of Mérida by Javier Arce, and especially on the local hagiographie text, the Vitas patrum emeritensium. At the same time, Professor Arce presents a good paradigm for communication in the period of transformation: Mérida in the sixth century hosted, according to the Vitas, merchants from the East and a physician who eventually became bishop of the city. T h e question of the reliability of this information is contested by Arce, who assigns it to the realm of hagiographie mythology, although evidence for other Syrians operating as merchants in the West at that time, such as Euphronius of Bordeaux and Eusebius the merchant, who eventually became bishop of Paris, leaves room for other explanations. 4 T h e case is of central importance for our programme, since it was used by Henri Pirenne as an ingredient in his theory of the unity of the Mediterranean world before the appearance of Islam. O n e could argue that to the audience of the Vitas, at least, the presence of orientals in Spain in the sixth century made sense and enhanced the importance in their eyes of the celebrated bishops of Mérida, Paul and Fidelis, although the n a m e Fidelis could hardly support an oriental origin. T h e main part of the conference was largely made up of pairs of papers, of which the first pair is devoted to the question of "images

4 Cf. D. Claude, "Der Handel des frühen Mittelalters", Untersuchungen zu Handel und Verkehr der vor- und frühgeschichtlichen Zeit in Mittel- und Nordeuropa, eds. K. Düwel et. al. (Göttingen, 1985), pp. 65 ff.

as substitute for writing". Peter Brown describes the late antique attitude towards images and the needs that produced them, in terms of their function as praesentatiae, closely related to the imperial images, which were circulated to facilitate the emperor's adoration by his subjects. For the time of the spätere Spätantike he sees images as functioning particularly as picturae made to be read by those who were not capable of getting the Christian messages in verbal form from reading the Bible and the Patristic texts. Professor Brown sees the shifting from praesentia to pictura as taking place at the end of the sixth century. As the most suggestive evidence he presents Gregory the Great's admonitions to Severus of Marseilles concerning iconic wall decoration in churches. In his paper Ian Wood checks the validity of this proposal by examining examples of monuments from Ravenna and the AngloSaxon world and concludes that this material does not evidence such clear-cut distinctions between late antique and early medieval Christian art. In the East, however, Pope Gregory's ideas on the usefulness of the images can easily be attested in seventh-century authors, like Anastasius Sinaïta (for which see Anna Kartsonis' seminal book on Anastasis). Thus there is no doubt that as the "ignorants" increased in the seventh century, due to the decline or even collapse of the antique educational system ‫״‬as literacy diminished throughout the Mediterranean world—other catechetical tools were needed urgently and the presentation in images of the figures (Christ, the Virgin and the Saints) and concepts of the Church became necessary. Seen in this cultural frame the date proposed by Peter Brown, the postJustinianic era, suits very well. However, the East by treating images as lectiones did not cease to see pictures as praesentiae; or to adore them. Brown's conclusion that "a watershed in the Christian imagination falls somewhere in the late sixth century" remains solid, even if the notion and function of praesentia-adoratio and pictura-lectio cannot be verified in all forms of Christian behaviour and Christian art. T h e next set of papers deals with the language(s) of administration. T h e administration of the Late R o m a n Empire was bilingual with Greek a n d / o r Latin being employed by large sections of the population. In the post-Justinianic period Greek was completely abandoned in the West and Latin was forgotten in the East. T h e language of the Byzantine court became more technical, and hence its public deployment restricted, while the use of some technical words in Latin merely demonstrated a conservative attitude and a lack of

originality without any practical benefit. T h e word ίουρισγέντιον as the Greek transliteration for ius gentium in Byzantine legal sources would suffice paradigmatically as a clear indication that not only the terms but also some important concepts and notions of the late antique State were fading away. Some bilingual inscriptions of functionaries serving in the western provinces on lead seals, as presented by N. Oikonomides, do not prove any practical interest in establishing modes of communication with the public who would read them. However, the language of the Byzantine court was never so remote from Greek-speaking citizens that we could apply the word "mandarin" to it. By contrast, the language used at the Carolingian court in the ninth century was much more restricted to a small group of functionaries, and therefore Mayke de J o n g tentatively labels it as a mandarin language. In both empires, however, more obviously in the Carolingian State, the substantial efforts to advance letters and arts— the renaissance—shows that arcane language was the product of necessity rather than the result of conscious effort. T h u s , again, administrative language and its public deployment was a matter of literacy. T h e decline of literacy created an audience incapable of understanding the written word going out from the court to the public. T h e two following papers deal in detail with the relationship of hagiography and historiography. Professor Rydén presents the results of his study of the prologues of some fifty Byzantine hagiographie texts and compares them to the prooemia of historiographical works up to the twelfth century. He analyses the aims of different authors, which, a m o n g other things, greatly influenced the style and language of the texts, at different periods up to the late-tenth century Menologium of Symeon Metaphrastes. Re-writing saint's Lives according to the new needs and tastes of the audience is of practical interest. T h u s Lcontius of Neapolis writing in the middle of the seventh century justifies his work on the Life of John the Almsgiver on the grounds that the two previous Lives were in a too high and lofty style, while Leontios' aim was to reach an uneducated audience. By contrast, Symeon Metaphrastes worked like a scholar in an academic way in order to upgrade texts according to the high taste of his late tenthcentury audience. Similar are the results of the detailed presentation of the Latin hagiography by Ian Wood, who states that function and audience vary, and they affect the form and the language of the

hagiographie texts. T h u s again the needs and the taste of the audience, their cultural level and ability to enjoy a higher or lower style determined the outcome. As Alcuin's two versions of the Life of Willibrord demonstrate, an essential factor was also the way in which texts were absorbed, by reading or by listening. Characteristically Alcuin wrote the prose version to be read in the monastic community, while his poetic version was meant for private reading. T h u s in both Greek and Latin hagiography we realize how authors were always conscious of the possibilities and dangers of rhetoric in this highly propagandistic literary genre. In the next section Averil C a m e r o n defines culture as a social language and observes its diffusion in the period of the transformation despite the "totalitarian control" exercised over the people either by the state or by the dominant ideology. She introduces the notion of permeability, which she sees functioning as a stable channel of communication between the horizontal forces of a society which remains civil, and therefore somehow free and able to a c c o m m o d a t e the uncertainties of the day by applying defensive attitudes to political, social and ideological crises. Applying this approach specifically to the complex of ethnic and political identities in a period of war and settlement Walter Pohl demonstrates, with the help of four remarkable cases from the sixth century, the character of the late antique deployment of social discourse. T h e elements at work in the process of constructing identities were constantly re-adapting within the social discourse. Archaeology and history of art were rather under-represented in the agenda of the Mérida conference. However, two substantial presentations, by Beat Brenk and Niels Hannestad, helped to enlarge the scope of assessing the modes of communication in the period of transformation. T h e development from sculpture to wall-painting and mosaic marks the period of transformation par excellence. Professor Brenk analyses the discussion a m o n g Christians of the presentability of the h u m a n being, necessarily in its physical form, and the typology of expressing the spiritual authority of Christ and of physically anonymous saints and martyrs in late antique portraits in wall decoration and in icons. T h e function of the vast a m o u n t of pagan sculpture in the Christian environment of Late Antiquity, its use or rather re-use in public and in private places is the topic of Professor Hannestad. T h e two papers pose the fundamental question of how

important was the decline of sculpture, this characteristically Greek form of art creation, for the transformation of classical culture in the Christian world. T h e proceedings of the Mérida conference conclude with papers on the development of new languages: Christian Hannick discusses how Armenian and Georgian in the East and Slavonic in the North, with their newly created alphabets, produced new tools of communication in direct connection with the expansion of Christianity in these lands. Michel Banniard, on the other hand, summarizes his research on the development of Latin and the emergence of Romance in the west from the third to the tenth centuries. At the last plenary conference in Isernia a keynote paper by Paolo Delogu served as introduction and at the same time as a summing up of discussions that took place in different working groups, and especially in the group which was devoted to questions of production and distribution of goods. Professor Delogu operates with certain particular aspects and notions, such as disruption, disaggregation versus re-integration and, above all, transformation, as ways out of the old clichés of decline, death etc. H e supports the notion of dividing the period of transformation into two periods: the first up to the seventh century being still under the influence of R o m a n structures, and the second increasingly dominated by the early medieval potential of the Carolingian political arrangement and Carolingian culture. As a way out of Henri Pirenne's shadow Delogu offers a "policentered model 5 ', leaving adequate room for both the Arab and the Byzantine domains and cultures. A Byzantinist cannot but welcome this new exegetical option. At the end of this volume we publish the paper given by T h o m a s F.X. Noble at the Isernia conference. O u r American colleague, who followed the work of the programme over its years of existence, albeit from considerable geographical distance, was invited to give a critical reflection on the five-year project: the organization, the method of work, the topics selected or favoured for discussion, and the achievements. Occasionally Professor Noble is too polite or generous with his praise; and we are of course very thankful to him. Very welcome also, however, are his critical remarks, pointing out omitted areas of research a n d / o r aspects unwisely neglected.

T H E C I T Y O F M É R I D A (EMERITA) I N T H E VITAS PATRUM EMERITENSIUM ( V I T H C E N T U R Y A.D.) Javier Arce

D u r i n g the so-called G r e a t Persecution, following the E m p e r o r Dioceletian's edictum of 303, a governor of Lusitania, tortuosissimus draco, Datianus—who might have been an imaginary figure—contributed, perhaps without realising it, to the creation of the most powerful of all myths and legends of the city of Augusta Emerìta, when he condemned to death the sweet and innocent virgin Eulalia, for refusing to offer sacrifices to the Emperor. 1 Almost a century after this episode, a poet also born in Hispania, Q.A. Prudentius, using his full poetic power in the service of the new religion to extol the glories of the martyrs of Spain in dactylic trimetres, became the panegyrist of the supernatural defender of the city.‫ ״‬This redeeming and protective role of Eulalia over the city is already expounded to his listeners or readers by the historian Hydatius in the year 469, when he relates in his Chronica that in the year 429 the Sueve rebel Heremegarius, audaciously offending the city (perhaps in besieging it), also dishonoured the saint's tomb. 3 Hydatius' admonition carries Biblical overtones: the all-powerful Lord God took revenge by drowning the offender in the river Anas. 4 Years later, in 456, Hydatius again tell us that the king of the Goths, Theoderic, daunted by the supernatural powers of the martyr, did not dare to ravage the city (of Mérida). 3 In a few years, Eulalia had become the

1

For the edict see T. Barnes, Constantine and Eusebius (Cambridge, Mass., 1981), pp. 22 ff. See also, J . Arce, "Prudencio y Eulalia", Extremadura Arqueolôgica III (1992), pp. 9 15: id., El ultimo siglo de la Espafia romana (284 409), 2nd ed. (Madrid, 1986), p. 39, n. 18. Recently, M. Roberts, Poetry and the Cult of the Martyrs, The Liber Penstephanon of Prudentius (Michigan, 1993). 2 Peristephanon, III, 1.5‫־‬ 3 Hydatius, Chronicle, a. 429, ed. R.W. Burgess, The Chronicle of Hydatius and the Consularia Constantinopolitana (Oxford, 1993): "Not far from Emerita which Heremegarius had scorned, thereby causing an affront to the holy martyr Eulalia . . ." The term spreverat does not necessarily imply a siege of the city, or, at least, is ambiguous. 4 Hydatius, ibid. : Ana divino bracchio praecipitatus inteúit. 5 Hydatius, Chr. a. 456-7: "Theoderic was preparing to pillage Emerita, but was

supernatural defender of the city, as the Virgin Mary was that of Constantinople. In the words of N o r m a n Baynes, referring to saints as defenders of cities in the Byzantine period: "God had given her to be the acropolis, the fortress, the wall of defence for the capital; the city was her city; the appeal through the centuries is always the same: as in the past, defend thy city" 6 —an obvious refence to the Virgin Mary and Constantinople, but which can be applied equally to Eulalia and Emerita. It seems that Eulalia's shield ceased to preside over Emerita (in the same way as the Akathistos had no longer served its purpose against the Turks in Constantinople), when Arab troops under the c o m m a n d of Musa b. Nusayr, conquered the city in 712 after an ardous siege, and its inhabitants were made to "pay the gizya, the properties of those who had fled to Galicia, the properties and the treasures of the C h u r c h " , while preserving their lives and possessions. T h e strong walls of Mérida had withstood the siege for six months, and the Arabs had had to build a tortoise (dabbaba) in order to destroy one of its towers—which they were, however, unable to do. T h e city was taken by the only possible means: surrender. T h e Arabs had no alternative but to resort to the old tactics of R o m a n legions to threaten the walls—which still remained intact in the eighth century. T h e value of the booty was suitably exaggerated by Arab historians: the jacinth belonging to Alexander the Great, and a gigantic topaz which lit up the altar of the Church; a lamp and a pitcher, which was subsequently placed in the mosque at Damascus. 7 But, long before, Eulalia had become widely known beyond the frontiers of her native land. T h e echoes of Prudentius' verses reached Africa, Gaul, Italy, even the East. Augustine, in Africa, dedicated a sermon to her, converting this fragile girl into a fortis femina.8 Gregory of Tours and Venantius Fortunatus evoke her miracles or her image; in the mosaics of Ravenna, Eulalia takes place next to Agnes; and

deterred by warnings from the blessed martyr Eulalia": moliens depredari bealae Eulaliae. . . terretur (ed., Burgess). 6 N. Baynes, "The Supernaturais Defenders of Constantinople", Byzantine Studies and Other Essays (New Haven, 1974), pp. 254-5. 7 According to the Bayan at Maghrib, by Ibn-Idhari, fl. a. 1313: on the siege and surrender of Mérida in the Arab sources see P. Chalmeta, Invasion e Islamizacion (Madrid, 1994), pp. 176 ff. 8 de provincia Hispaniae, sancta et fortis femina quae per qffectu vincit sexum: Augustine, Sermo, 212.

in the Synaxarium in Constantinople a virgin called Eulalia is devoutly referred to in Tessalonica civitate.9 Eulalia is the centre-piece of that singular work, the Vitas Patrum emeritensium, which I will deal with in this paper. In this small book, the city and the saint are one and the same, and the entire Christian world of the city revolves around her shrine and her presence, around her worshippers and miracles. Towards the end of the sixth century, the period to which the Vitas refers, Emerita is the city of Eulalia. And the public, whether learned or uneducated, discovers through the hagiographer's sermo humilis the palpable reality of the city's monuments and every day life, presided over by the untouchable and perpetual presence of Eulalia. T w o main problems dominate research on the Vitas. First, problems of edition, interpolations, date, authorship; and second, problems of the historical reliability of the text. Thanks to the recent edition of Antonio Maya in the Corpus Christianorum, the text of the Vitas is now well established and many philological problems have been solved: the Vitas survives in two recensions: the first dating from 6 3 0 - 6 3 8 , written by an anonymous deacon from Mérida; the second, probably completed between the years 670 680, is dedicated to Bishop Festus, to w h o m the author addresses himself with a publie confession of faith. This new recension had a new title, several additions and interpolations, and was the work of Paul, also deacon of Mérida. 1 0 I will here concentrate on the historical problems of the work, paying attention only to evidence relating to the city of Mérida and to the historical reliability of the document, while trying, also, to put the Vitas into its proper context as a literary work. It is somewhat difficult to pinpoint the historical context in which the work was written. We can only establish the general background. It was an age of relative calm in Visigothic Spain: the period of King Sisenand—a Godiic aristocrat from Septimania, who had usurped from Suintila, being made king by the army in Caesaraugusta. This is a period for which conciliar documents take the place of chronicles.

!1

On the subject see the interesting. remarks by Λ. Recio Verganzones, "La Martir Eulalia en calendarios y martirologios, en la devociôn popular y en su iconografia (siglos IV VII)", Extremadura Arqueotogica III (1992), pp. 81-1 10. T h e standard, and perfectly good edition of J. Garvin, The Vitas Sanctorum Patrum Ementensium (Washington, 1946), lias been superceded philologically by the recent (and excellent) edition by A. Maya in Corpus Christianorum, Series Latina CXVI (Turnhout, 1992): a new historical commentary is still needed.

T h e troubled reign of Leovigild, the unifying conversion of Recared, the Third Council of Toledo and the decisive victory of the Catholics over the Arians are all in the past. Byzantines are no longer present in Hispanic territory—having been expelled for good by Suintila in 624. At last the Visigothic king governed the whole of Hispania. Toledo was the undisputed capital of the kingdom, the sedes regia, the place where the principal councils met. It is also the period in which the influence of Isidore's powerful personality lies across all the decisions that were taken in ecclesiastical as well as political matters. C h u r c h authority has been victorious everywhere. Proof of this is provided by the Fourth Council of Toledo, held in the year 633, in which we are told that King Sisenandus humiliated himself cum lacrimis et gemitibus in the newly constructed Church of Saint Leocadia, before more than fifty-nine bishops." T h e bishops then announced their intention of intervening in politics: usurpation of power became anathema, and the method of succession to the throne was settled— the bishops were to be present at the ceremony. They, also, finally, proclaimed what should be the virtutes of the future rulers: humilitas, moderatio, iustitia, pietas erga subditos, indulgentia. T h e idea was, certainly, inspired by Isidore. Despotism would be anathematised. From now onwards Church and State would run together, the second being controlled and supervised by the first. 12 Such was the situation when the Liber Vitas Patrum emeritensium, written by a deacon from the C h u r c h of St. Eulalia, was first read out. Mérida was still a city of m a j o r importance at the beginning of the seventh century—indeed, it had been the capital of the dioecesis Hispaniarum since the reforms of Diocletian. And Emerita was still Toledo's rival. This was especially the case in the period of Athanagild, who was responsible for moving the capital of the kingdom to Toledo. Mérida played a prominent role in opposing Leovigild's arianism during the second half of the sixth century, through its patres, its bishops—something which is manifestly clear in the Vitas. It had a wealthy and powerful C h u r c h and, above all, it had Eulalia who was its guardian. Eulalia gave the city its prestige, and was, also,

" IV Con. Toi. (proerriium): cf. Concitios Visigoticos e hispano-romanos, ed. J . Vives (Barcelona-Madrid, 1963), p. 186: rex Sisenandus primum coram sacerdotibus Dei humo prostratus. . . per se interviniendum Deo postulavit. 12 IV Con. Toi. c. LXXV, III. For the historical context see the excellent introduction of M. Diaz y Diaz to the edition of the Etimologias I, ed. J . Oroz y M.A. Marcos Casquero (Madrid, 1993), pp. 7-95.

the foeus for pilgrimages, and hence the source of revenues for the Church. Toledo needed all this, protectors, saints, martyrs. And King Leovigild used every means to give the new capital of his kingdom the same status. T h a t is why he tried to seize Eulalia's robe for his C h u r c h — t o gain a protector. T h e fight for the robe was a fight for supremacy between the two cities. And this is clearly evidenced by the Vitas. Every reader could understand it. In his work the deacon from Mérida continously underlines this aspect of the role which had been played by his city thirty years earlier. People should know the story. T h e ultimate purpose of the Vitas is to claim for the city prestige a n d supremacy over other capitals, from its being the epicentre of orthodoxy. Emerita deserved this position because of the patronage of a famous saint, because of the exemplary conduct of its bishops and because of its defence of Catholicism. T h e work amounts to much more than this. It is aimed at the people of Mérida, straightforward and simple people, and also at the mixed Romano-Gothic society which still inhabited the city—as we shall see. It is, in fact, an anti-Arian pamphlet. It is hagiography; but not exclusively. It is also biography. Among the literary productions of Hispania in the seventh century, it is an unususal piece, because it is neither ecclesiastical history, nor is it a chronicle, nor an account of the lives of distinguished men and w o m e n . " T h e a u t h o r is a deacon who had probably been educated in Mérida itself, in the building next to the basilica of St. Eulalia— domus ecclesiae—which he describes so vividly, with its cellulae for one or two youths—pueri—like Augustus, or Quintilian or Veranius, subjects of one part of the Vitas. Like all deacons, the education he received was similar to that which Bishop Paul, whose biography he writes, gave to his nephew Fidelis, to prepare him for his succession to the epicopacy: strenue erudivit in omnem officium ecclesiasticum.14 O u r deacon's library (bibliothecam) contained the Dialogues of Gregory the Great, the Vita Martini by Sulpicius Severus, part of the works of Prudentius, and the Vita Desiderii, recently written by the learned king

l:i For the literature of the period see M.C. Diaz y Diaz, Index Scriptorum Latinorum Medii Aevi Hispanorum (Salamanca, Madrid, 1959); J . Fontaine, Isidore de Seville et la culture classique dans l'Espagne wisigothique (Paris, 1959) and Carmen Codoner's contribution to Historia de Espana dirigida por Menéndez Pidal, III (Madrid, 1991). 14 Vitas Sanctorum Patrum Emeritensium (henceforth VPE), IV, 4, 1-7.

Sisebut. His stories, his aims, his narrative are based on these works. 15 T h e author insists, suspiciously, on the credibility of his writing, and declares that everything he writes he has heard from reliable witnesses. 16 This was, therefore, an oral culture, transmitted and adorned by word of mouth, although in its zeal to appear authentic the Vitas may come across as a parody of the Lives of saints and hagiography in general. T h e book has a catechetic purpose and aspires to instruct by pleasing. Its audience is the casual reader and listener (ut omnium legentium vel audientium) and it claims to be directed to simple folk: simplicibus simpliciter veraciterque nanamus. T h e author, faithful to his teachers, has no time for arid and complicated rhetoric: one should not speak 0bscurìs sermonibus because it is not intructive, but dull (sed faticamus). Gregory of T o u r s had said: philosophantem rhetorem intellegunt pauci, loquentem rusticum, multi}' We do not know if there was somebody behind this work; we do not know whether a bishop incited his deacon to write the Vitas of his predecessors, or if it is the spontaneous work of the author himself. T h e r e is one undeniable fact: the author has a thorough knowledge of his city and makes it the stage for supernatural events. T h e people of Mérida lived with these beliefs, and the roads, basilicas, walls, monasteries and spiritual leaders of the city possessed miraculous powers. Again quoting Baynes—referring to the Byzantines, but wholly valid for Mérida in the seventh century: "there can be no doubt that the Byzantines lived in a world where miracles could happen and did happen, and that belief in miracle is itself a fact of history which the student ignores at his peril". 18 My principal concern here is the historical "reality" of the evidence of the Vitas. I only partially share the doubts that Roger Collins set out in an excellent article on the Vitas: "I confess myself", says Collins, "reluctant to use a work of hagiography, a sophisticated literary genre with its own rules and conventions, merely as a quarry 15

Cfr. A. Maya VPE, pp. LXXV VI; VPE, Praef. 3-5. que non relatu atiorum agnovimus nequefinctisfabutis didicimus, sed que ipsi, eos referentes, auribus nostris audivimus: VPE, praef. 16-18. The same claim is found in much hagiographical literature: cf. Gregory's Dialogues. 17 Gregory of Tours, Histories, Praef., ed. B. Krusch and VV. Levison, Monumnenta Germaniae Historica, Scriptores Rerum Merovingicarum I 1 (Hannover, 1951). On this topic see the excellent pages of E. Auerbach, Lenguaje Literarì0 y publico en la Baja latinidad y en la Edad Media (Barcelona, 1969), pp. 82 ff. See also VPE, Epil. 18 Baynes, "The Supernatural Defenders of Constantinople", p. 248. 16

for useful circumstantial information; there are grave dangers inherent in using material in a way it was never intended for and out of original context". 19 Collins' scepticism is wholly justified. I will try to separate what might be fact and what might be fiction in those topics in the Vitas, which provide evidence for the life of a city in mid to late sixth-century Spain.

Emerita, a christian city T h e main body of the Vitas covers the history of the city's bishops between 530 and 605, from the episcopate of Paul to that of Masona, with a brief reference to Innocentius and Redemptus. A question that should be addressed in the first place is why the author chose precisely this period in the life of the Church in Mérida and no other. It seems that it was because this was a brilliant period in the administration of the see and, especially, because it marked its economic climax. But I would like to suggest also that it is because those bishops were able to maintain the i n d e p e n d e n c e a n d the supremacy of their diocese against Toledo and the Arians. T h e Vitas is therefore, as I have said, an anti-arian pamphlet. And as such, the catholic aspects of the city are powerfully stressed. As in much other literature of the late R o m a n period, the bishop in the Vitas is the main figure of city life, and the centre of all activity. H e is the leader of the community; he confronts the king, organises charitable activity, is generous repairing or building up monuments; he is protected by the local dux (Claudius); he receives an embassy of merchants; he gives gifts and presents. All city life goes around him, his person and his prestige. This is normal in the late R o m a n period and has been appropriately emphazised. 2 0 But although he has such an important role in the Vitas, the bishop is not actually the administrator of the city or the military authority. 21

19

R. Collins, "Mérida and Toledo: 550-585", Visigolhic Spain: new approaches, ed. E.James (Oxford, 1980), pp. 189 219; and id. Espana en la Atta Edad Media (Barcelona, 1986), pp. 116-137. 20 See P. Brown, Power and Persuasion in Late Antiquity (Wisconsin, 1992): J.H.W.G. Liebeschuetz, Barbarians and Bishops (Oxford, 1990): Rita Lizzi, Vescovi e strutture ecclesiastiche nella città tardoantica (Como, 1989). 21 The dux of the city was Claudius; VPE, V, 10, 29; V, 11, 32; 40; 50; 90; 44. For important remarks on the eastern part of the Empire, see S. Mitchell, Anatolia:

In the Vitas, Augusta Emerita (in fact the city is always mentioned as Emerita, never with the title Augusta) is a city without a single pagan element: there is no mention at all of her pagan monuments. An ancient R o m a n colony, founded by Augustus, with a splendid marble forum, dating from the Julio-Claudian period, with sumptuous and splendid pagan monuments—theatre, amphitheatre, temples, circus, trophies, etc., the city was in decline in the third century; but in the Constantinian period it saw a revival as an urban centre, as a result of its new role as capital of the diocesis hispaniarum.22 W e do not know when the urban transformation of the city began. Recent excavations have shown that the middle of the fifth century is a turning point in the city's urban structure. 23 In the sixth century christian topography invaded the urban space and completely changed the main pagan traits of the city; but city-life continued until the eighth century, and the transformation did not imply interruption or abandonment. T h e city was a strategic point in the communications of western Spain, being a crucial crossroads in the axis Hispalis-Legio-Gallaecia.24 Mérida is, perhaps, the most active city in the period from the fourth to the eighth centuries in the whole of Hispania23 T h e city has produced a considerable number of architectural fragments, most of them finely decorated, which art historians and archaeologists date between the sixth and seventh centuries. 26 It is certain that most of this architectural decoration belonged to christian buildings, given the ornamental motifs; but the problem is that we can not assign them to any specific m o n u m e n t . However, the Vitas mentions quite a large n u m b e r of basilicae or churches scattered throughout the city. From the Vitas we can count them but hardly identify their location. T h e y are (i) the basilica of land, men and gods in Asia Minor I, The Celts in Anatolia and the impact of Roman rule (Oxford, 1993), pp. 73 ff. 22 Javier Arce, "Mérida tardorromana (284- 409)", Espana entre el mundo antiguo y el mundo medieval (Madrid, 1988), pp. 190 ff. 23 Pedro Mateos, "La arqueologîa urbana en Mérida: el proceso arqueolôgico", in Merida. Excavaciones arqueolôgicas 1994-1995, Memoria (Merida, 1997), pp. 215 ff. In general the whole volume contains the results of the latest archaeological research in the different sectors of the city. Cf. also P. Mateos, "La cristianizaciôn de la Lusitania (ss. IV VII): Extremadura en época visigoda", Extremadura Arqueolôgica IV (Merida, 1995), pp. 239 ff. (especially pp. 248 ff). 24 Hydatius provides first hand evidence of this. 25 Javier Arce, "Mérida tardorromana", passim. 26 M. Cruz Villalôn, Mérida visigoda. La escultura arquitectonica y litúrgica (Badajoz, 1985).

St. Eulalia, situated extra-muros, as is appropriate to a martyrial basilica; 27 (ii) the basilica of St. M a r y — t h a t was known in the middle of the sixth century as St. Jerusalem 2 8 (this basilica was considered to be the oldest of all the churches in the city and therefore as being first in rank); 29 (iii) there is specific mention of churches dedicated to the martyrs: one to Ciprianus and another to Laurentius; 3 0 and (iv) there is mention of a basiliculam sancti Iohannis,3' and another dedicated to St. Lucretia; 32 (v) there are general references to basilicas plures, founded by the bishops; 33 (vi) extramuros we find, apart from St. Eulalia, the ecclesia Sancti Fausti,34 that was at a distance of one mile from the city wall. 35 T h e r e were also monasteries: one attached to the basilica of St. Eulalia, 36 and another in the vicinity of the city—Cauliana; 3 7 and there were also monasteries inside the city. 38 O t h e r christian buildings are also mentioned: a xenodochium for pilgrims, poor and the sick;39 the episcopium, or bishop's palace; 40 and a baptistery. 41 T h e treasure of the C h u r c h (thesaurus) was kept in the most important church, that of St. Mary, and it included the treasure of St. Eulalia. 42 As mentioned above, St. Eulalia's basilica included a monastery in which there was a library (bibliotheca), and it was the residence of youths (oblati) consecrated to the service of the saint. T h e y received instruction and education in this place. 43 Such was, then, the architectural p a n o r a m a of Mérida in the middie of the sixth century. T h e r e is no mention of an arian basilica (but surely one existed). 44 T h e intensive archaeological activity of recent years has not been of very much help in identifying these

27

VPE, V, 5, 33; V, 5, 49. VPE, IV, 9, 7. 29 VPE, V, 6, 60. 30 VPE, IV, 10, 13-14. 31 VPE, IV, 9, 14-15. 32 VPE, IV. 7, 19. 33 VPE, V, 3, 11. 34 VPE, IV, 7, 18. 35 VPE, IV, 7, 18 19: ab urbe fere miliario distat. 36 VPE, IV, 4, 1-6. 37 VPE, II. 38 VPE, V, 3, 10. 39 VPE, V, 3, 13. 411 VPE, V, 13, 52. 41 VPE, IV, 9, 15. 42 VPE, V, 6, 5 9 - 6 0 . 43 VPE, I and IV, 1 , 7 . 44 There is mention of an arian basilica in V Con. Toi., 6, 49-50. 28

buildings, with the exception of the basilica of St. Eulalia. 45 Archaeologists have identified the unquestionable remains of a christian basilica, now in the substructure of the C h u r c h of Santa Eulalia, as the foundations of the original church of the saint. T h e r e is, however, still some doubt. For example: during excavation some tombs were found, but none that could correspond with the tombs of the bishops of Mérida of the sixth century—Paul, Fidelis, Masona, all of them buried, as stated by the Vitas, in the basilica "very close to the saint's memorial". 4 6 Instead, excavation has produced a handsome funerary inscription relating to the t o m b of one Eleutherius, archdeacon. 47 It is possible, but hardly credible, that the archdeacon Eleutherius mentioned in the Vitas would have been buried in saint's Eulalia basilica, given the author's near-condemnatory attitude towards him. 48 T h e r e has also been a proposal to identify one building in Mérida with the xenodochium mentioned in the Vitas', but more research is needed. 4 9 In any case, the coincidence between archaeology and the Vitas is not my concern here. W h a t is important at this juncture is to establish that the urban christian topography of the city as described by the author of the Vitas cannot be an invention, because the audience, or the reader, knew his own city and its m o n u m e n t s perfectly well. Therefore the historical conclusion is that Mérida changed profoundly between the middle of the fifth century and the sixth, and was transformed into a christian capital during this period. Another, marginal, problem is whether these churches were "Visigothic" or, rather, christian architecture of the Visigothic period (and there has been m u c h discussion a m o n g archeologists). Leaving this question aside, the only thing I would like to say is that the fragmentary architectural decoration that still remains in M é r i d a (as mentioned above) belongs clearly to a late R o m a n tradition; 50 and is difficult, if not improper, to speak of "Visigothic" architecture.

4

‫ יי‬As yet there has been no full publication, but see, L. Caballero and P. Mateos, "Trabajos arqueolôgicos en la Iglesia de Santa Eulalia de Mérida", Extremadura arqueolôgica III (1992), pp. 15-51. 4fi VPE, V, 15, 1 3 : aud procul ab altario sanctissimae virginis Eutatiae honirfice tumulata quiescunt. 47 Caballero and Mateos, "Trabajos arqueolôgicos en la Iglesia de Santa Eulalia de Mérida", p. 23. 48 VPE, V, 13, 11 ff.; V, 13, 61 74. 49 P. Mateos, "El culto a Santa Eulalia y su inflencia en el urbanismo emeritense (siglos IV VI)", Extremadura arqueotôgica III, pp. 68-73, with fig. 5, p. 77). "' Against this opinion L. Caballero, "Zamora en el trânsito de la Edad Antigua

Christian life in Mérida also manifested itself in sophisticated ceremonial, which illustrates the continuity of late R o m a n forms of life, and which too is attested in the Vitas.5* It was a part of religious life on special occasions, and enhanced the prestige of the ecclesiastical hierarchy. Processions with fixed itineraries in which the people participatcd, singing and with acclamations, are well attested for particular occasions—the feast of Easter, celebration of Holy Mass, the adventus of the bishop. 2 '‫ ׳‬O t h e r characteristically christian activities complemented and reinforced the importance of the C h u r c h for the citizens: charitable attention to the sick, poor, widows, pilgrims; almsgiving; celebration of the Mass; visits to the churches . . . But this is only a partial view: Arians, Jews, even pagans were also present in Mérida. But they have 110 place in the Vitas, or they are only mentioned in passing. 53 In fact Mérida was a mixed society in which noble Romans, noble Goths, poor, slaves, ruined widows, clergymen lived next to each other. 5 4 This is a picture that we can deduce from the Vitas and that we can consider as being essentially true. C a n we deduce that this society in transition was the normal composition of the social body of cities in sixth- to seventh-century Spain? It is quite possible. It should be remembered that we know more about Mérida than any other city in Spain in this period.

Oriental merchants or "la fascination de l'Orient"? T h e origin of the first two bishops of Mérida, whose main biographical traits are related in the Vitas, require some analysis. Paul and Fidelis both came from the Eastern part of the Empire 55 — so claims the author. This detail concerning the Greek (Byzantine) origin of the bishops, has produced the theory a m o n g historians that

a la Edad Media", Historia de ^amora (Zamora, 1996), pp. 341 430; id., "Visigodo ο Asturiano?", Corsi di Cultura sull'arte revennate e bizantina, Ravenna, 39, pp. 139 190. 51 For the importance of ceremonial in Constantinople in the period studied here see now G. Dagron, Empereur et prêtre (Paris, 1996): and M. McCormick, Eternal Victoiy (Cambridge, 1986). 52 VPE, V, 5 , ' 6 0 ff.; V, 8, 42 ff.: V, 8, 59 ff.; V, 11, 11 ff. 53 VPE, V, 3, 17; V, 2, 30-32; the dux Claudius was a recently converted christian: V, 10, 33. 54 Masona was a Goth: VPE, V, 2, 2; Claudius was a Roman: V, 10, 33; Vargila was a noble Goth: V, 1 I, 87. 55 VPE, IV, 1 , 2 (Paul); IV, 3, 10 (Fidelis).

there was intense commercial activity between M é r i d a and Asia M i n o r in the sixth century, and that it was one of the causes of the wealth and richness of the Church of Mérida; and some speak frankly of "commercial oriental colonies" in the city. 56 I myself think that such an interpretation takes the text of the Vitas without considering its literary genre: that is, that the Vitas is a hagiographical text with its own rules and structure. And while it is right to accept the work's basic evidence for christianisation as f u n d a m e n t a l l y true, this is not the case for other evidence describe in the Vitas. If we follow this approach, it is possible to interpret the "reality" of the oriental presence of merchants and commercial relations in a completely different way. Let us, for a moment, follow this line of argument. First, and most important, the author begins the story of Paul's arrival in Mérida with a referunt multi, a typical narrative formula that introduces the beginning of what is usually a fantastic tale or story. And the tale goes on: Paul, a Greek physician, came to Mérida de 0úentis partibus.31 H e came as a peregrinus (what is the real meaning of this word: a foreigner in the community, as in the classical period, or a pilgrim who came to visit the t o m b of a saint?) with no fortune of his own (nihilque habens).58 This oriental m a n , a foreigner in a society totally different from his own, gradually gained the respect and admiration of his new countrymen. His medical skill surely contributed to his attaining such a preeminent position as to be elected bishop of the C h u r c h of Mérida. O n e day, the tale goes on, it happened that a group of people who also came from precisely the same country and region as Paul (again the Orient) arrived in the city. T h e y were negotiatores39 and on arrival, they immediately went to pay a visit to the bishop. T h e next day they sent a young m a n , Fidelis, to repay the bishop's hospitality with some presents. Paul was immediately struck by the boy and began to ask him about his origin, family, n a m e of his father, his city and so on, and, thus, he realized that the young m a n was none other than his own nephew! 60 H e persuaded the merchants to leave Fidelis with him, and after an appro56

So L. Garcia Moreno, "Colonias de comerciantes orientales en la peninsula ibérica. (siglos V-VII)", Habis 3 (1972), pp. 127-54. 57 VPE, IV, 1, 2; IV, 3, 3-5. 58 VPE, IV, 2, 74-76. 59 VPE, IV, 3, 4. ti0 VPE, IV, 3, 14-24.

priate education and a reasonable period of time the boy himself became bishop of Merida. Narrated like this, the story sounds like a tale from the Thousand and One Nights (the comparison is made by Collins). Set out in such a way the story of the origin of the most brilliant period of the C h u r c h of Mérida, was a guarantee of future success and prestige. W e have in this narrative all the ingredients of a fantastic story of an almost "mythological" origin. T h e r e is no need to believe in commercial relations, or in the oriental colonies in Mérida. O f course we can take the story in another way and follow the traditional interpretation of the passages of the Vitas M A man, attracted by the fame of the sanctuary of St. Eulalia in the remote extremity of the world, came from the eastern part of the Empire in the hope that he would be able to m a d e a fortune through his professional skill. He was exceptionally lucky after all, because he was not only elected to be bishop of the community, but he also found himself extremely rich, having saved the life a senator's wife by performing a caesarean operation: 6 2 he was rewarded with part of the senator's fortune, and inherited the rest when the couple died. 63 His succèssor was Fidelis, who happened to be his nephew. We have evidence for relations, travel and pilgrims between Hispania and the eastern part of the Empire from the fifth century onwards. 6 4 It is quite possible that doctors went to sanctuaries and places of pilgrimage because they offered opportunities for work, money and prestige. 65 W e know of Greeks established in the West—for instance in Gallia—who, after involvement in commercial activities, became bishops of cities. 66 Paul's case was not an exception. W h a t is, indeed, an exception is the narrative of the Vitas: the story itself. W h a t is most suspicious is that the whole story seems to be a justification for the c o m m o n r e a d e r — o r listener—to explain the origin of the wealth of the Church of Merida: that is, donations from pious people, like the Lusitanian senator. O n the other hand, the evidence for relations between Mérida

61

As for example Collins, Espana en la Alta Edad Media, pp. 121 122. VPE, IV, 2, 45-50. 63 VPE, IV, 2, 61-66. 64 Mention can here be made of the visits to Constantinople of Hydatius, Egeria, Leander, J o h n of Biclar and so on. '5‫ י‬I owe this suggestion to Peter Brown. 66 Gregory of Tours, Histories, X, 26. 62

and the East of the Empire is very slight, almost non-existent. A few inscriptions in Greek cannot be proof of contact; and the stylistic influences in the architectural decoration are better seen as North African rather than oriental. Some historians have mantained that the merchants mentioned in the Vitas were interested in commerce in marble columns. 67 This is highly improbable, especially because, as in classical period, the quarries used for building of Mérida were located outside the city, in Estremoz (prov. of Badajoz), or came from other parts of Spain (Macael). We have seen that the Vitas is a hagiographical document in which is necessary to distinguish a m o n g the variety of evidence it presents. First, it is a pamphlet with a specific and concrete purpose: to show and demostrate the essential role of the city of Mérida during the struggle against the Arians, exemplified in the figure of King Leovigild. It is also, at the same time, a document that claims for the city supremacy a m o n g other cities of sixth-century Spain, at a time when M é r i d a ' s star was declining in favour of T o l e d o . T h i s claim to supremacy is based on the well-known activities, energy and orthodoxy of the bishops, and, especially, the city's patron, Eulalia. As a consequence, the Vitas emphasises the catholic elements of the city—churches, charitable activities, processions, the cult of martyrs and saints. These elements are a palpable reality in a city that, until the fourth century, had been an undeniably pagan community. 6 8 As a hagiographical text the Vitas adorns the different episodes of the lives of the patres/bishops of the city with exotic tales a n d stories—as, for example, their origins or social activity. All the wealth and power of the C h u r c h came, not as a result of the commercial activity of foreign people or of their bishops, but as a consequence of donations from local landowners, like the Lusitanian senator. In the Vitas we must distinguish between fact and fiction. T h e churches, the christian topography, the processions, cannot be fiction; everything else could be pious invention to glorify Eulalia, the patron and defender of the city.

67

Collins, "Mérida and Toledo: 550 585", p. 203. See J. Arce, "El mito de Dionysos y Ariadna en un puteal tardorromano de Mérida", Habis 7 (1976), pp. 343-365. tili

I M A G E S AS A S U B S T I T U T E F O R W R I T I N G Peter Brown

At some time in the 680's, a dispute broke out in the Armenian churches of the south-east Caucasus: the painters (of images) became arrogant and claimed to be superior to all other ministers of the Church. "Our art", they said, "is light itself, for young and old each understand it, while only few can read the Holy Scriptures." A council was called. The artists were firmly put back in their place. They must sit below those who handled the Holy Scriptures—"the scribes, the readers, the exegetes."1 It is best to begin with this incident, recorded in the tenth-century History of the Caucasian Albanians of Movses Dasxuranci. For I suspect that, in inviting me to speak this morning on the theme of "Images as a Substitute for Writing" in the communications system of Late Antiquity, the European Science Foundation may have been guilty of the heresy of those proud painters. It is, indeed, part of the common-sense of m o d e r n times that images are invariably more accessible than writing. In the words of Ernst Curtius, "knowing pictures is easy compared with knowing books." 2 Thus, we tend to assume that a society in which images come to bear a heavy and repeated weight in the communication of messages that had previously been conveyed, by preference, through written texts must be a society groping towards the formation of a mass culture. Averil C a m e r o n reminds us, in her Christianity and the Rhetoric of Empire, that, in Christian circles, figurai and, hence, pictorial language was "only natural". It formed an essential part of the Christian creation of "a totalizing discourse", an achievement which she sees as the secret of the Christianization of the R o m a n

' The History of the Caucasian Albanians by Mouses Dasxuranci, 46, cd. C.J.F. Dowsett (Oxford, 1961), p. 171. See P.J. Alexander, "An Ascetic Sect of Iconoclasts in Seventh Century Armenia", Late Classical and Medieval Studies in Honor of Albert Matthias Friend, ed. K. Weitzmann (Princeton, 1955), pp. 151 160. 2 E.R. Curtius, European Literature and the Latin Middle Ages (Princeton, 1973), p. 15: cited by M. Camille, "Seeing and Reading: Some Visual Implications of Medieval Literary and Illiteracy", Art History 8 (1985), pp. 26-49, at p. 44.

world: for a reliance on images reflected a "striking degree of interest in reaching an audience at all levels." 3 T o others, it has seemed that not only the Christian Church but, also, the Later R o m a n State had broken with the aristocratic reticence of an earlier age. T h e emperors came to harness to their needs the rich visual culture of the lower classes. In the words of Ramsay MacMullen's classic study of 1964, "Some Pictures in Ammianus Marcellinus", the result was that "a sort of billboard brightness" came to prevail in the public life of the fourth century. Such intensely visual messages, MacMullen added, "have nothing to do with "classical" culture. T h e y reflect the upthrust of the non-Greek and nonR o m a n elements through an upper surface worn thin." 4 Altogether, anyone who faces the role of images in Late Antiquity, and their relation to written communication, is instantly led out on to perilous yet fascinating terrain—to the entire problem of the democratization of culture in the Late R o m a n period. It is a terrain on which giants have walked Riegl, Rodenwaldt, Bianchi Bandinelli, a m o n g the art historians, Rostovtzeff, Santo Mazzarino, and noster William Frend among historians.' It was on the shoulders of such giants that those of us who came to study the social role of religion and culture in Late Antiquity, in the 1960's, were happy to perch, to view, with unfeigned excitement, what seemed to be the most distinctive feature of the period—and, to a modern sensibility, alternatively its most heartening and most disquieting feature: the increased participation in the culture of the élites of hitherto excluded, nonliterate masses. But today I wish to part company from certain aspects of that potent construct. I will do so by concentrating largely on aspects of Christian Late Antiquity. I will attempt to show that the notion that 3

Averil Cameron, Christianity and the Rhetoric of Empire, Sather Classical Lectures 45 (Berkeley, 1991), p. 186. 4 Ramsay MacMullen, "Some Pictures in Ammianus Marcellinus", Art Bulletin (1964), pp. 433-455; now in Changes in the Roman Empire (Princeton, 1990), pp. 78-106, at pp. 104-105. 5 To name only a few major contributions to this debate: G. Rodenwaldt, "Eine spätantike Kunstströmung in Rom", Römische Mitteilungen 36/7 (1921/2), pp. 58 110; R. Bianchi Bandinelli. "Osservazioni sulla forma artistica in Oriente e in Öccidente", Tardo Antico e Alto Medioevo, Problemi Attuali di Scienza e Cultura 105, Accademia Nazionale dei Lincei 365 (1968), pp. 289-308; Santo Mazzarino, "La democratizzazione della cultura nel 'Basso Impero'", Antico, tardoantico e era Costantiniana (Rome, 1974), 1, pp. 74-98; H. Brandenburg, "Ars Humiiis", Jahrbuch fur Antike und Christentum 24 (1981), pp. 71-84; R. MacMullen, "The Historical Role of the Masses in Late Antiquity' 5 , Changes in the Roman Empire, pp. 71-84.

images can act as a substitute for writing—largely as a didactic substitute—reflects a precise m o m e n t in the later period of the Late Antiquity. It is a product of the spätere Spätantike. It signals, indeed, the end of Late Antiquity. For this reason, it should not be projected backwards into the late antique period proper. T o put it bluntly: essential parts of that conglomerate of notions that came to constitute the common-sense of Christian Europe on the nature of religious art were set in place in the Age of Justinian and Gregory the Great. T h e y were not there in the Age of Constantine and Augustine. Let me make my point simply, by playing the film backwards in time. W e begin in the 780's with M a r Shubhaliso', a Nestorian missionary in the wooded highlands south of the Caspian Sea. He knew what a missionary should do. H e felled the great holy tree of the mountain tribes, "the King of the Forest". H e did it in style. He made his entrance there with exceeding splendour, for barbarian nations need to see a little worldly pomp and show to make them . . . draw nigh willingly to Christianity.6 A generation earlier, a better-known missionary, St Boniface, did much the same in the woods of Germany. He wrote back to England for yet more splendid copies of the Bible, written in letters of gold, that a reverence for the Holy Scriptures may be inspired in the carnal minds of the heathen.' It was assumed both by M a r Shubhaliso' and by Boniface that minds on a low cultural and spiritual level needed the blunt visual messages which were the form of communication appropriate to that level. T h e same attitude is shown by Hypatius, bishop of Ephesus, who (if we may accept Paul Speck's redating of the fragment) was a moderate Iconoclast, and, consequently, an exact contemporary of Boniface. As he wrote to a colleague: We (ourselves) take no pleasure at all in sculpture and painting. But we permit simpler people, as they are less perfect, to learn by way of initiation, by means of the sense of sight, which is more appropriate to their natural development. 8

11

Thomas of !Marga, The Book of the Governors, tr. E.A. Wallis Budge (London, 1893), p. 480. 7 Boniface, ep. 35 (21), Die Briefe des Bonifatius, ed. R. Rau (Darmstadt, 1968), p. 114. 8 Tr. P.J. Alexander, "Hypatius of Ephesus: A Note on Image Worship in the Sixth Century", Harvard Theological Review 45 (1952), pp. 177 184, at p. 179; Paul

It is, of course, with Gregory the G r e a t that we touch the firmest g r o u n d of all. His two letters to Bishop Serenus of Marseilles (Ep. IX. 229 of J u l y 599 a n d X I . 10 of O c t o b e r 600) have come to form the basis of all subsequent Western common-sense on the religious function of art. For a picture is displayed in churches on this account, in order that those who do not know letters may at least read by seeing on the walls what they are unable to read in books . . .9 For what writing offers to those who read it, a picture offers to the ignorant who look at it, since in it the ignorant see what they ought to follow, in it they read who do not know letters: whence especially for gentes—gentiles/barbarians—a picture stands in place of reading. 10 Faced with so persuasive a statement of w h a t was to b e c o m e the obvious in Latin Christianity, it is easy to overlook the unprecedented consequentiality of Gregory's formulation. As Celia Chazelle pointed out in her study in Word and Image, of 1990, Gregory is the first Christian writer on art's relationship to words. Gregory had, indeed, attempted to effect nothing less than an alignment of church pictures with texts." In the forceful words of Guglielmo Cavallo's contribution to the last session of the Centro di Studi sull'Alto Medio Evo, devoted to Testo e Immagine, Gregory's views involved Lo schiacciare dell'immagine sulle funzioni

del testo/libro.

The crushing of the image into functions performed by texts and books.12 Given the novel sharpness of Gregory's formulation, it is worthwhile looking back to Late Antiquity as a whole to ask ourselves why this h a d not h a p p e n e d earlier, why it h a p p e n e d as it did, a n d to linger, at the end, on the significance of such a change for the religious climate of the age.

Speck, Poikila Byzanlina 4: Varia I, eds. R.-J. Lilie and P. Speck (Bonn, 1984), pp. 213-272. 9 Gregory, ep. IX 229, ed. D. Norberg, Corpus Christianorum 140A (Turnhout, 1982), p. 768; tr. Celia Chazelle, "Pictures, books, and the illiterate: Gregory I's letters to Serenus of Marseilles", Word and Image 6 (1990), pp. 138-153, at 'p. 139. 10 Gregory, ep. XI 10, p. 874; Chazelle, "Pictures", p. 139. 11 Chazelle, "Pictures", p. 144 Ū r 148. 12 G. Cavallo, "Testo e immagine: una frontiera ambigua", Testo e Immagine nell'Alto Medio Evo, Settimane di Studio del Centro Italiano di Studi sull'Alto Medio Evo 41 (Spoleto, 1994), pp. 31-64, at p. 45.

But let me turn aside, for a moment, to follow some loose threads in Gregory's otherwise compact statement. I remain puzzled by the reference, in Gregory's second letter, to the gentes: unde praecipue gentibus pro lectione pictura est. Serenus was warned not to upset "savage minds" by an abrupt destruction of images: ferocibus animis scandalum generares. A Marseilles where a bishop could be supposed by Gregory to have to take account of the feelings of refractory primitives is not the Marseilles that we have learned to know—the urban "success story" of fifth century Gaul, the repository of papyrus for Gregory of Tours, the port of entry of a thoroughly Byzantinized pretender, where Serenus' predecessor, Theodore, had been involved in a memorable face-off with the highly literate Dynamius, patncius of Provence. 13 If there was a community at Marseilles which took the issue of images seriously, I would suggest, rather, that we are dealing, not with gentes, barbarians, but with the Jews, some of w h o m had already complained to Gregory that they had been forcibly baptized by the local bishops, when travelling on business from Italy to Provence. Conversion (forced or otherwise) and the use of images tended to go hand in hand. T h e y were a lively issue for the Jews of the western Mediterranean. At the same time as Gregory wrote to Serenus, a newlybaptized J e w at Cagliari, across the sea in Sardinia, had caused a disturbance. He had celebrated his conversion by taking over the local synagogue and had placed within it a sign of the Cross, along with images of the Virgin and of Christ, besides hanging up his own baptismal robe, as an ex voto. If the need to meet the criticisms of Jews outraged by such conversions played a part in Serenus' decision to prevent, at all costs, the "adoration" of images by the Christians of Marseilles, then we have a further case in which the forcible baptism of Jews was accompanied by disputes on the "adoration" of images, similar to those which erupted, a generation later in the eastern Mediterranean, as a result of the measures taken by the emperor Heraclius. 14 This suggestion, however, does not explain why a reference to

13

S.T. Loseby, "Marseilles: A Late Antique Success Story?", Journal of Roman Studies 82 (1992), pp. 165-185; Ian Wood, The Merovingian Kingdoms: 450-751 (London, 1994), pp. 84-86. 14 Gregory, ep. I 45, p. 60; ep. IX 196—the incident at Cagliari; see V. Deroche, "La polémiqué anti-juive au vie et au viie siècle. Un memento inédit, les Kephalaia", Travaux et Mémoires 11 (1991), pp. 275-311, at pp. 290-293 and Leontios, "Apologie contre les juifs", Travaux et Mémoires 12 (1994), pp. 43 104.

gentes appears in Gregory's letter. I suspect that in both letters we are dealing with a distinctive feature of Gregory's pontificate, which Ernst Pitz has taught us to appreciate. Here as elsewhere, Gregory attempted, in writing as he did to Serenus, to put his personal stamp on a bureaucratic process, that was based on countless, unknown decisions on the same subject.' 5 Maybe the reference to the gentes and to their "savage minds" found its way into this letter from another document, now lost to us, in which Gregory had made a ruling on the use of images in a different region (one thinks of the deeply under-Christianized highlands of Sardinia) where such remarks would have been more relevant. In using that earlier letter as a model, Gregory may have wished upon the cultivated, long-Christian sea-port of Marseilles a more barbarous population than it, in fact, possessed. T h u s Gregory's letters to Serenus of Marseilles may have been the tip of an iceberg, a magisterial summing up of many similar cases. In any case, we know very little about the extent to which images played a role in bridging linguistic and cultural frontiers, through addressing gentes in the true sense, whether these were "barbarians" or pagan outsiders. Despite Gregory's reference to gentes in his letters, this does not seem to be at stake at Marseilles. In a Christian city, images could be assumed, perhaps on the basis of Gregory's Roman experience, to be used simply as a means of increasing the access to Christian preaching of non-literate Latin speakers who, as Michel Banniard has shown, remained firmly held in a net of "vertical c o m m u n i c a t i o n " , c o n d u c t e d in Latin a n d based on lectio—on the reading of Holy Scriptures that were not totally alien to them. 1 6 If we want to find images used so as to communicate between different cultures and different religious groups, we should turn to the eastern frontiers of the R o m a n world. In both Sasanian and Islamic times, statues and images of the emperors served to delineate the frontier. Reverence for these remained a touchy issue, in which the prestige of both empires was involved. Thus, imperial images at border-crossings were expected to speak of the majesty of the R o m a n Empire even to the "barbarians" on the other side of

15

E. Pitz, Papstreskripte im frühen Mittelalter (Sigmaringen, 1990), p. 205. M. Banniard, Viva Voce: Communication écrite et communication orale du rue au ixe siècle en Occident latin (Paris, 1992), pp. 108-112.

the frontier.' 7 Images were also expected to speak, to outsiders, of the majesty of one's own religion, and were partly created in order to do so. Herbert Kessler has recently suggested that it is no coincidence that the splendid frescoes of the Jewish synagogue of D u r a Europos showed evidence of foreign visitors, maybe of non-Jewish visitors, in the form of graffiti in Greek, Aramaic, and Middle Persian. Such visitors were expected to be at least impressed, if not instructed, by the monumental scenes that they saw on the walls of the synagogue.' 8 At the beginning of the seventh century, the Zoroastrian Persian, M a g u n d a t (later known as the martyred Saint Anastasius the Persian), was informed about the deeds of Christian martyrs when he saw them, for the first time, depicted on the walls of a Syrian church.' 9 Indeed, the ethnic and religious divisions of the Sasanian empire, where so many groups were "outsiders" to each other, fostered other, more bizarre forms of visual communication. W h e n the Nestorians wanted to depose their own Katholikos, they approached K h u s r o II Aparwez with petitions accompanied by graphic caricatures drawn upon the fans with which the King would fan himself. T h e y showed the Katholikos pinching a chicken to see if it was fat enough, while another showed him turning over a gold coin in his hand, while a girl sat on his knee.20 In a court where foreigners—such as the Christians were held to be, even if many of them were Pahlavi-speakers—were only allowed to communicate with the sovereign through an interpreter, images bridged the gap between Christians and the gentes. But it is a use of pictures which Gregory I (a contemporary of the maligned Katholikos) would hardly have had in mind! But the gentes could be closer to home. W e tend to forget how

17

See the evidence discussed in W. Kaegi, Byzantium and the early Islamic conquests (Cambridge, 1992), pp. 165-166. 18 K. Weitzmann and H.L. Kessler The Frescoes of Dura Europos: Synagogue and Christian Art, Dumbarton Oaks Studies 28 (Washington D.C., 1990), p. 182. 19 Vita Anastasii Persae 9, ed. B. Flusin, Saint Anatase le Perse et Γhistoire de la Palestine au début du viie siècle (Paris, 1992), p. 51. Chronique de Sçert 83, Patrologia Orientalis 13, col. 523.

full of "air pockets" the Christian culture of Late Antiquity continued to be. Persons of high education, if they came from non-Christian families, or from thinly Christianized regions, were not necessarily conversant with Christian images. T h e y could be as m u c h "outsiders" to the meaning of Christian images as were any barbarians. In the late fifth century, no less a figure than Severus of Antioch, the future founder of Monophysite theology, still needed to have the story of A d a m and Eve explained to him when he saw it, for the first time, in a church in Beirut, where he was studying as a law student well into his twenties. 21 O n e did not have to be an illiterate to draw conclusions from pictures—and, frequently, to draw the wrong conclusions. As patriarch of Antioch, Severus took exception to the m a n n e r in which representations of angels, by portraying them with the entire paraphernalia of imperial rule—that is, by including the purple robes, the diadem and the orb—implied that these beings exercised total, autonomous control over the material universe. 22 By acting in this way, as "a law to themselves", artists could hinder quite as much as they could help the process of the codification of the Christian visual imagination, in which the educated stood quite as much in need of correction as did the illiterate. Intriguing though such considerations may be, they are, in a sense, marginal. For we tend to read Gregory's letters to Serenus the wrong way round. W e read them, as it were, from wall to worshipper. W e want to know what picturae stood on those walls. W e want to know what class of persons they addressed. W e want to gauge the extent to which they relayed a message rooted in written texts—the lectio of Christian scriptures and their commentaries. But Serenus and Gregory alike looked from the worshipper to the wall. W h a t concerned them both was a charged contemporary issue— idolatry—and its heavy-handed remedy—iconoclasm: seeing people adoring images (Serenus) broke the images and threw them from the churches.23 W h a t had shocked Serenus's critics was that Christian objects in a Christian church (probably placed there, we should add, like much of the church plate, by Christian donors whose memory was main-

21 22 23

Zacharias Scholasticus, Vie de Sévère, Patrologia (Mentalis 2, col. 49. Severus of Antioch, Homiliae cathedrales 72, Patrologia Orientatis 12, cols. 83-84. Gregory, ep. IX 229, p. 768; Chazelle, "Pictures", p. 139.

taincd in Marseilles by such gifts)24 had been subjected to a form of pointed humiliation usually reserved for pagan idols. It was only by such drastic measures that Serenus thought that he could disperse the sultry atmosphere of adoratio, which had built up between worshippers and walls. All that Gregory asked of Serenus was that he should reassure his congregation that he would henceforth leave Christian walls alone: indicate that what displeased you was not the sight of the story revealed through the witness of a picture, but that adoration which had inappropriately been exhibited to the pictures.23 In effect, the bishop should change the worshipper not the wall. T h e chaste act of reading, just because it was an eminently intellectual exercise, guided by the mind, as the eyes scanned the pictures at a safe distance, to puzzle out their meaning and their spiritual relevance, was to replace the heavy body-language associated with adoratio—the bowing, the close c o n t a c t of the kiss, the worshipful "counter-presence" of a flickering oil-lamp, even the tell-tale wisps of incense. It is, perhaps, the destiny of any formulation that appears, to later ages, to carry "the power of common-sense", 2 6 that it should sueceed in shaking off its immediate context. Gregory's formulation was appealed to, frequently, in debates between Christians, throughout the Middle Ages and, with renewed vigor, in the Reformation. As a result, the sharp smell of an ancient world where idolatry had until recently been widespread drifted entirely away from it. But this was not so in 600 A.D. 27 Even if Christians took art for granted and harboured no systematic hostility to images, a lingering fear of idolatry remained real, as it did to a bishop such as Serenus of Marseilles. This was due, in large part, to the absence of a clear24 Such an action would have been the nearest equivalent to selling church ornaments given by leading families to the church: see W. Klingshirn, Caesarius of Aries: The Making of a Christian Community in Late Antique Gaul (Cambridge, 1994), pp. 114 116) and Peter Brown, Power and Persuasion in Late Antiquity (Madison, Wise., 1992), p. 96, n. 132. 25 Gregory, ep. XI 10, p. 875; tr. Chazelle, "Pictures", p. 140. 26 Herbert L. Kessler, "Pictorial Narrative and Church Mission in Sixth-Century Gaul", Pictorial Narrative in Antiquity and the Middle Ages, eds. H.L. Kessler and M.S. Simpson, Studies in the Hutoiy of Art 16 (1985), pp. 75-91, at p. 75. 27 Siri Sande, " T h e Icon and its Origin in Graeco-Roman Portraiture", Aspects of iMte Antiquity and Early Byzantium, eds. L. Rydén and J.O. Rosenqvist (Stockholm, 1993), pp. 75-84.

cut code of behaviour, for Christians, in the presence of images. W h a t Christians wished to avoid was what they observed pagans to do. Faced by any image in a solemn setting, ancient persons tended to get physical. T h e most familiar images were not messages. T h e y were presences. As presences, they were not expected to be "read"; rather, they were expected to elicit appropriate gestures. In a society ruled as much by deference as by force, this was a serious matter. M a n y images h a d r e m a i n e d necessary presences. Images of those who exercised authority were expected to trigger a carefully maintained charge of reverence, expressed through deep bowing. W e should always remember that, because expressed through a few body motions, this reverence was bound to remain disturbingly undifferentiated. In answering the challenge of a Jewish critic, a Christian pointed out that a believer had few options: a reverential bow did not have to m e a n "adoration", in the sense of idolatrous "worship", although there was no way of telling, from the physical movement alone, that it was not so. "Tell me, if you enter and worship before a king do you worship a provincial g o v e r n o r . . . or all the order of ranks with the same (position of the) head with which you have worshipped the king?" . . . The Jew said, "It is impossible for me to change the (position of my) head according to each one as they are lifted up and brought low. So I worship the king and the lower ranks with the same (position of the) head." 28 (The Syriac, more blunt, simply says "with the same head".) T o worship "with the same h e a d " could be acutely embarrassing. T h e e m p e r o r Julian the Apostate h a d placed on the imperial images, images of the gods, as they bestowed sovereignty on the emperor. Viewing the small, highly-charged tablets from a distance, Christians had fallen flat on their faces before them. 2 9 G o o d subjects of an absolute monarchy, an imperial image had released in them the same, ill-focused adrenaline of "worship" as it had done in their pagan colleagues. Bowing, of course, was never enough when it came to the gods. Pagans would blow kisses to their idols. T h e y would lift little chil-

28 The Disputation of Sergius the Stylite against a Jew XIV, 14 15, ed. A.P. Hayman, Corpus Scriptorum Christianorum Orientalium 338, Scriptores Syri 152 (Louvain, 1973), p. 41. '2‫ יי‬Gregory Nazianzen, Oratio IV, Contra Julianum 1, 80-81, Patrologia Graeca 35: 605C-608C; Sozomen, Historia Ecclesiastica 5.17.

dren up to kiss their faces. T h e y would sit and gossip with them, placing food, even letters, at their feet. T h e y wished to sleep with them. 3 0 In early medieval Constantinople, a w o m a n afflicted by soreery, dreamed constandy that she saw herself standing in the theatre of the Hippodrome, kissing the statues there and striving with lustful desire to have intercourse with them.31 W h e n confronted with objects where such behavior was deemed legitimate—the tombs of saints and holy sites in Palestine—Christians of all classes and in all regions reacted in the same, highly physical m a n n e r . T o visit the cradle of Christ at Bethlehem, was to touch it with "eyes, mouth, forehead, to gain a blessing". 32 I suspect that the pilgrims who came to St Martin at T o u r s paid less attention than we do to the images of the miracles of Martin, designated by tituli composed by none other than Venantius Fortunatus. Herbert Kessler's fundamental study of "Pictorial Narrative and Christian Mission in sixth-century G a u l " provides a fascinating glimpse of the messages that came from the walls of great basilicas that contemporaries of Gregory and Serenus could have seen. T h e subtle "splicing" of text and image that Kessler reveals provides a concrete precedent for Gregory's more abstract formulation. 3 3 Yet pilgrims came to T o u r s so as to touch and not to read. W h e n , during the Lenten season, crowds from T o u r s crossed the Loire to troop through Martin's former monastery at Marmoutier, no tituli held their minds at a distance from the immediate, fully licit physicality of a Christian holy place: forgetful, alas, of Professor Kessler, they licked and kissed and moistened with their tears each spot where the blessed man had sat or prayed or where he had eaten food or laid his body to rest after his many tasks.34 30 s.v. "Götterbild", Reallexikon ßir Antike und Christentum (Stuttgart, 1981), 11, cols. 661-775, at cols. 689, 694-695, 735-736. See also Β. Gladigow, "Präsenz der Bilder-Präsenz der Götter. Kultbilder und Bilder der Götter in griechischer Religion", Approaches to Iconology. Visible Religion 4 5 (Leiden, 1 9 8 5 6 ‫ ) ־‬, pp. 114 133; and D. Gourevitch, "Quelques fantasmes erotiques et perversions d , objet dans la littérature gréco-romaine", Mélanges de l'école française de Rome: Antiquité 94 (1982), pp. 823-840, at p. 834. 31 Vita Andreae Sali 17:132, Patrologia Graeca 111, 780C. 32 Sophronius, Anacreontica 19:45, Patrologia Graeca 87, 3813B. 33 Kessler, "Pictorial Narrative", passim. 34 Gregory of Tours, De Virtutibus Sancti Martini 1, 2, tr. Raymond van Dam, Saints and thdr Miracles in Late Antique Gaul (Princeton, 1993), p. 204.

Christians of all classes shared in this intense piety of "presence". It was only after he had gone through ancient, profoundly faceless gestures in the church of St Cassian at Imola,—stratus humi tumulo— that Prudentius lifted his eyes to see an actual image, the martyrdom of Cassianus. 35 In his fine analysis of this and of other poems in Prudentius's Peristephanon,36 Michael Roberts makes us realize the very real ambiguity of Prudentius's position: he was so near yet so far from kissing that image. W h a t he chose to emphasize were his own dramatic gestures of worship at the tomb itself clasping the tomb, shedding tears, warming the altar with my lips, the stone with my breath.37 T u r n i n g to the image, Prudentius deliberately switched to a use of ekphrasis, a literary device where, we must r e m e m b e r , the principal psychological effort consisted of evoking a living scene out of the unprepossessing (and, almost certainly, ill-lit) image of the saint and his sufferings. W h a t Prudentius did not do was sink, mindlessly, towards the image with a reverent kiss. Altogether, the poem betrays anxiety on Prudentius' part to show to his readers that he had gone so far and no farther. H e had conjured up the image; he had not adored it. Like our own President Clinton, when confronted in his youth, at Oxford, with a "joint" of m a r i j u a n a , he had, perhaps, puffed; but he had not "inhaled" the noxious substance of idolatry. Part of the answer to this ambivalence, for a m a n such as Gregory, was to replace the well-known stance of the adorator by the novel stance of a lector. In this area of Christian experience, we are dealing with an inversion of the title on which I have been asked to speak. T h e image did not act as a substitute for writing—in a m a n ner that implied a process of "democratization" of late antique culture, through the wider circulation and the greater valuing of visual images. W e are faced by the opposite. Images are pulled into the gravitational field of a sacred text, in order to save them from acting as the objects of strong and still dubious emotions. For this to occur, two changes had to take place, in the generations before and after Gregory I. And it is with these that I will end.

35

Prudentius, Peristephanon IX. Michael Roberts, Poetiy and Prudentius (Ann Arbor, 1993), pp. 37 Prudentius, Peristephanon IX. 36

5. the Cult of the Martyrs: The Liber Peristephanon of 138.148 ‫־‬ 99100 ‫ ־‬.

If paintings on a wall were to be appreciated as another form of written codex, attitudes to that particular codex itself had to change. Gregory's formulation assumed not only the previous ccntury of Christian narrative art, to which Herbert Kessler has drawn our attention. It registers the slow imaginative triumph, in late antique culture, of the notion of Sacred Scriptures. This cannot be assumed for earlier periods. This is due to no lack of loyalty to the Gospels on the part of late antique Christians. But we must r e m e m b e r that the élites of the fourth and early fifth centuries were not like those of later centuries. T h e élites of early medieval times have been aptly characterized by Michael McCormick as ces techniciens de Vécût—as technocrats of the written word. 3 8 Not so the great rhetors of the classical and late antique periods. It was as masters of the living word that they dominated their society. Even the written text struck them as no more than a dry river bed, down which the spoken word had once roared in full flood. If the written text lacked glamour, compared with the spoken word, a fortiori pictures were even less adequate to communicate the magic of speech. T h e y were far too static. T h e y lacked the flexibility, the constant element of surprise, the sheer "attack" of the grands jets d'eaux of ancient eloquence. Quintillian wrote of some lawyers, w h o m I have occasionally witnessed, bringing into court a picture of the crime painted on wood or canvas, that the judge might be stirred to fury by the horror of the sight. H e was unimpressed: the pleader who prefers a voiceless picture to speak to him in place of his own eloquence must be singularly incompetent.5'‫י‬ N o t surprisingly, in such an atmosphere, codex and pictura were allowed to drift apart in learned circles. Illustrations of major classics such as the Iliad implied that the owner of such pictures had not made that classic his own: his memory needed prompting. Artifacts such as the Tabulae Iliacae—small bas-relief panels illustrating incidents of the Trojan war—were defective as didactic tools: they showed

38 Michael McCormick, "Textes, images, et iconoclasme dans le cadre des relations entre Byzance et l'Occident carolingien", Testo e Immagine, pp. 95-162, at p. 153. 3 " Quintillian, Institutio Oratoria 6: 1: 32.

no more than a mere "flickering Gelehrsamkeit?'.*0 Products of a cultural underworld, to which the standards of accuracy d e m a n d e d of the written and r e m e m b e r e d word did not apply, paintings were known to be unreliable. Educated Christians of the fourth and fifth centuries continued this prejudice. Augustine, for instance, reports that pagans believed that Christ had been a magician and that he had passed on his lore to St Peter and St Paul (as if Paul, converted after Christ's death, could have been one of his disciples!). It was an opinion based upon well-known Christian iconographie traditions, which showed Christ as a teacher, sitting between the two aposties. Not sharing the enthusiasm for magicians now shown by Professor T h o m a s Mathews in his provocative re-interpretation of the iconography of Christ, in The Clash of Gods,4' Augustine was not amused: They thoroughly deserve to err who have sought Christ and his Aposdes not in sacred books but in pictures on walls.42 For this suspicion of images to change, attitudes to eloquence and to the sacred books themselves had to change. T h e Holy Scriptures had to take on some of the static quality of a picture. T h e "white water" rapids of the great age of Christian eloquence had to give way to stiller waters. T h e patient repetitive discipline of the lectio divina—notably, but not exclusively, practised in monastic circles— invested the Holy Scriptures with an "iconic" charge. Humble gestures of reverence for the book of the Gospels, contemplative disciplines that developed specific forms of attention to the pages of the Scriptures gave a more visual form to Christian reading. T h e sacred written words were treated as a series of signs, pointing always beyond themselves. In that sense, lectio and pictura converged. T h e pious mind played on the pages of the Gospels, seeing in each sentence, even in each word, an initial opacity that beckoned for understanding quite as much as did the silent lines and colours of a picture. Both

40

N. Horsfall, "Stesichorus at Bovillae?", Journal of Hellenic Studies 99 (1979), pp. 26-48, at p. 29. 41 Thomas F. Mathews, The Clash of Gods: A Reinterpretation of Early Christian Art (Princeton, 1993): 54-91, pp. see Peter Brown, Art Bulletin 77 (1995), pp. 499-502. 42 Augustine, De Consensu Evangelistarum 1: 10: 15-16; see G. Madec, "Le Christ des paiens d'après le De consensu Evangelistarum de Saint Augustin", Recherches augustiniennes 26 (1992), pp. 3 - 6 7 , at pp. 4 0 - 4 7 .

were ares of the same vast circle whose centre was the invisible wisdorn of God. In the words of Philoxenus of M a b b u g : It has also been said by the holy fathers when day is breaking he (the monk) may wash his hands and make some genuflection before the Cross, until his thoughts have gathered from their wanderings . . . while he prays to our Lord . . . "let your holy light shine unto my heart that my intellect may enter into the external body of ink." And when you have gathered your thoughts by means of words of prayer like these, salute the Cross and take the Gospel in your hands and put it before your eyes and your heart and go and stand on your feet before the Cross . . . and lay on a cushion every (Gospel) book, in which you read, and fall down before this Cross (with, presumably, the Gospels now arranged at its foot).43 For others, such as M a r Abbi of Amida, the page itself radiated its message with the still intensity of an icon: He would sit in the great common chapel, and cover his head, with his book on his knees . . . Thereupon he would open the book and gaze at it, and at once his tears would burst forth . . .; and he would not turn over a leaf but generally had the book open at the beatitudes, or the parable of the virgins, or that of the talents . . . or any place where the subject was that of threats and judgment. 44 In such conditions of rapt, contemplative reading, it was difficult for lectio not to become pictura. Pope Leo deeply hoped that this would be so for all the faithful. Immediately after the reading of the great Passion narratives in the G o o d Friday liturgy, he turned to the congregation and said:

ita omnium vestrum arbitror inhaesisse pectonbus ut unicuique audientium ipsa lectio quaedam facta sit visio. I think that this narrative must have come to cling to your hearts in such a way that the Gospel reading has become a vision for all of you who hear it.45 With such words in one's ears, it would be quite natural to turn, with chastened eyes, to contemplate the narrative images that stood on the walls of any R o m a n church. Such images were no longer

43

Philoxenus of Mabbug, Letter to a Friend, tr. G. Olinder, Acta Universitatis Gotoburgensis 56: 1 (1950), 14*-15*. 44 J o h n of Ephesus, Lives of the Eastern Saints 14, Patrologia (Mentalis 17, col. 215. 45 Leo, Sermo 70. 1, Patrologia iAtina 54: 380B.

seductive "presences," eliciting ill-defined sentiments and prompting suspect gestures of adoration. T h e y were picturae that could be "read", because they echoed adequately, in Leo's words, an imaginative "vision" that was firmly schooled by prolonged Christian lectio. O n e might expect such a shift to occur over time, as a result of centuries of preaching, every year, on the same dramatic themes. Biblical narratives and, apparently, histonae of saints and martyrs, were there to be seen on the walls of churches in which the same dramatic events were conjured up, at regular intervals, in the words of preachers. But behind the convergence of word and image there lies, also, a deeper change of sensibility, almost too big to be seen. I would like to touch upon it, inevitably very briefly, so as to conelude. For viewers to become the equivalent of readers, as Gregory the Great expected them to be, a highly specific kind of viewerresponse to representations had to be created. In order to work in the way that Gregory the Great and later, more radical defenders of icons expected them to work, images had to be thought of as "out there". Their heightened presence in Christian circles assumed the existence of great distances, such as only an image was thought capable of spanning. These distances were both cultural and religious. T o their illiterate viewers, images were expected to relay messages taken from written texts that were now assumed to be otherwise inaccessible to the uneducated person. T h e y had to be looked at with care, for their exact message to be understood. M o r e important still, in all cases, Christian images were held to speak of a distant world. T h e y required a heightened degree of subjective engagement on the part of the viewer, literate and illiterate alike, if that distant world was to become present and for its majestic message to sink deep into the heart. T h e sense of a widening gulf between G o d and humankind subtly ousted the spoken and written word from the eminence that they had enjoyed in previous centuries. Lectio and pictura came to be placed on a more equal footing. Both were equally distant from an invisible, utterly immaterial God. Both could function as merciful aids, through which h u m a n beings strove as best they could to regain a glimpse of heaven. 4 6 It was precisely this sharp sense of "otherworldliness" that had been gloriously absent in large areas of the late antique Christian 41

' P.A. Marivaux, "L'image selon Gregoire le Grand et la question de l'art missionnaire", Cristianesimo nella storia 14 (1993), pp. 1-12.

mentality. Like their pagan rivals, Christians thought of themselves as enveloped, for good or ill, in a mundus, a visible universe shot through with etherial (and so, usually, though not invariably, invisible) benign and hostile powers. Their problem was not to envision a so-called O t h e r World, a world "out there": a más allá, an aldilà, an au-delà, still less an outretombe. W h a t mattered was to bring into their own, perilous existence in this world touches of Paradise—a region of delight, hauntingly a d j a c e n t to themselves. Day-to-day Christian endeavour in the visual arts, as shown in the placing of distinctive Christian signs (such as the sign of the Cross or of the Christogram) on artifacts of every kind, as well as the setting up of distinctive Christian images in both public and private places, was less concerned with communicating a message as with bringing specks of protective power down from heaven into the world. Recent study of Christian motifs on clothing, the collection of household objects that formed the theme of the recent exhibition on "Art and the Holy Powers in the Early Christian House", so ably presented and commented upon by Henry M a c G u i r e and by Eunice D a u t e r m a n n MacGuire, the plethora of objects connected with pilgrimage to the Holy Places, interpreted by Gary Vikan, may cause us to revise m a n y of our previous opinions on the nature and function of Early Christian art. 4 ‫ ׳‬W e find ourselves in a world swamped with vivid Christian images. But they are not images that were intended to be " r e a d " as picturae. T h e y spoke a more muted, frankly supernatural language. T h o u g h they took the form of images, their principal function was not to s u m m o n the visual attention of their h u m a n users but to ward off demons and to attract angels. W e are dealing with an art that was, quite literally, w r a p p e d around the person to such an extent that it could not be wrenched away, so as to become a distanced object of contemplation. O n e does not take off one's flowing caftan shirt to meditate upon the 47

Henry MacGuire, "Garments Pleasing to God: The Significance of Domestic Textile Designs in the Early Byzantine Period", Dumbarton Oaks Papers 44 (1990), pp. 215-224; E.D. MacGuire, H.P. MacGuire and M.J. Duncan-Flowers, Art and Holy Powers in the Early Christian House, Exhibition Catalogue, Krannert Art Museum (Urbana, 111., 1989); Gary Vikan, "'Guided by Land and Sea': Pilgrim Art and Pilgrim Travel in Early Byzantium", Tesserae: Festschrift fiir J. Engemann, Jahrbuch ftir Antike und Christentum: Ergänzungsband 18 (Münster in Westfalen, 1991), pp. 74-92, and "'Pilgrims in Magi's Clothing': T h e Impact of Mimesis in Early Byzantine Religious Art", The Blessings of Pilgrimage, ed. R. Ousterhout (Urbana, 111. and Chicago, 1990), pp. 97-107.

images of Christ, or of the Virgin, or on the Gospel scenes embroidered upon it. For they were placed there to beat off the evil eye, not to beckon in the believing gaze. Pilgrims who carried terra cotta images of the Magi were too busy "being" those M a g i — h o p i n g for the same protection that had led the Magi, the archetypal pilgrims to the Holy Land, on their long j o u r n e y from the East to Bethlehem—to distance themselves from such precious talismans, as if they were no more than a representation of some far away histona. W e should not marginalize this evidence, as if it spoke only of a need for personal, apotropaic protection. T h e attitudes that we can see so clearly in this segment of Christian religiosity set the tone, also, for m u c h of the public Christian art of the period. T h e entire Christian congregation, from bishop downwards, wished to feel enveloped in a counter-mundus. T o enter a church was to enter a mundus transformed. T h e sudden explosion, in the fifth or sixth centuries, of exuberant decoration taken from the entire living kingdom, to which H e n r y Macguire has drawn attention in his Earth and Ocean: The Terrestial World in Early Byzantine Art,w shows the yearning of the Christian congregations to become, for a m o m e n t , participants in Paradise itself. At great shrines, such as that of St Martin at Tours, Paradise was frequently regained, as deep wafts of fragrance and the surreal "flowering" of withered limbs showed that a saint, who dwelt in Paradise, was present, also, in earthly space. T h e comforting sense of the accessibility of Paradise in such churches was communicated through an environment saturated with visual messages. 49 A redundancy of sensual "triggers," a m o n g which smell was quite as important as were sight a n d sound, engulfed the worshipper. Despite the presence, in so m a n y basilicas, of striking images and Biblical scenes, it was difficult to privilege these, to separate them out from an overall impression of Paradise come down to earth, and to project the mind upon them, as picturae, as images that carried specific messages that d e m a n d e d to be "read". T o conclude: I would suggest that we may be dealing with a watershed in the Christian imagination that falls somewhere in the late sixth century. Naturally, a change as profound and as diffuse as a shift in religious sentiment did not h a p p e n at once, in every region and, still less, did it happen irrevocably. It was rather that, 4)1

H. Maguire, Earth and Ocean: the Terrestial World in Early Byzantine Art (University Park, Penn., and London, 1987). 49 Maguire, Earth and Ocean, p. 47.

within the long-established imaginative ecology of the Christian church, one "strain" tended to come to the fore, in m a n y circles and in m a n y regions, at the expense of others. T w o close contemporaries serve to characterize the two, ultimately divergent strains that coexisted in the late sixth-century Latin West. Each displayed a different attitude to the mundus that had once engulfed the Christian. The writings of Gregory of T o u r s take us into what could still be expcrienced as a profoundly i m m a n e n t world. Palpable intimations of Paradise and Hell alike flickered, intermittently but vividly, throughout the visible universe. T h e y took the form of mysterious gusts of fragrance, of unearthly lights burning in the night in empty churches or, even, in a distant coppice, of fresh roses strewn around a tomb in the depths of winter, of the tormented shrieks of the possessed, and of the sudden "smoking" flames of the fevers that struck down those who had angered the ever-present saints. With Gregory the Great's Dialogues and the slightly later literature of the Columbanian monasteries of northern Gaul (and their equivalent in Spain) the present world tilts subtly but unmistakably towards the grave. T h e world beyond the grave was not to be seen in terms of haunting proximity. Rather, it was characterized by the vast and perilous distances which the soul must travel after death, vividly described in the novel genre of the "voyage of the soul".' 0 Encounters with angels, saints, and demons did not vanish from the mundus. But they were regarded, by a m a n such as Gregory the Great, as mere skirmishes, as incidents of a Phoney W a r that preceded the swift and massive onslaught of the O t h e r World at the end of life, when angels, demons and the welcoming saints would press in around the deathbed of each Christian. It was at that m o m e n t that their faces could, at last, be seen. T h e heavy air of the mundus had cleared, to allow a direct glimpse into the O t h e r World. This shift in sentiment should not be overlooked by historians of art. For we are dealing, in it, with the transition from a specifically late antique "period eye" to one that will become more dominant in the Early Middle Ages—a shift, perhaps, from participation to vision. W e begin with an "iconography of hints and shifting allusions", 51 in an art that did justice to the yearning of late antique

50

Claude Carozzi, Is voyage de l'âme dans l'au-delà d'après la littérature latine (ive-xiiie siècle), Collection de l'école française de Rome 189 (Rome, 1994); Peter Brown, Gloriosus Obitus: The End of the Ancient Other World (forthcoming). 51 Maguire, Earth and Ocean, p. 84.

Christians to find themselves surrounded by a mundus transformed. W e end with a wish for clear "windows" on to another world, such as could only be provided by a carefully-codified array of visual representations. And these windows look on to a Paradise where nature and the animal world, though present, have sunk into the background. Paradise, the "garden of delights", now takes second place to the precise outlines of the h u m a n face. In the 730s, J o h n of Damascus wrote, Is it not more worthy to adorn all the walls of the Lord's houses with the forms and images of saints rather than with beasts and trees?52 With such a remark, John's writings in defence of the cult of icons signalled the end not only of a distinct period in the history of Christian art, but of an entire "style" of ancient Christianity. After 700 A.D., it had become a little bit more possible for Christians to enter churches, as, perhaps, we still do today, plucked by a sharp sense of worlds held at a distance from us, of worlds wo ich nicht bin, worlds wo meine Toten auferstehen, of worlds inhabited by the holy dead, made present, on earth, by the silent faces on the icons and by the quiet, charged gestures of narrative scenes. W e should be careful not to expect this particular sensibility in earlier centuries of Christianity.

52 J o h n of Damascus, De imaginibus 1, Patrologia Graeca 94, col. 1252, cited by Maguire, Earth and Ocean, p. 83.

I M A G E S AS A S U B S T I T U T E F O R W R I T I N G : A R E P L Y Ian W o o d

Professor Peter Brown's study of "Images as a substitute for writing" is an account of a profound shift "from participation to vision", that is from a world where the recipient of the image ceased to be an adorator and became instead a lector; a world in which an image ceased to be a praesentia and became simply a pictura. A m o n g the factors bearing on this change was the lectio divina, the regular contemplation of the Bible. But the shift also had a broader context, that of the transformation of an " i m m a n e n t " world filled with prèsences, good and evil, to one in which the chief focus was on the grave and an ultimate glimpse of the O t h e r World. T h e key m o m e n t in this shift seems to have been the late sixth century, although Professor Brown adds a crucial rider: "Naturally, a change as profound and as diffuse as a shift in religious sentiment did not happen at once, in every region and, still less, did it happen irrevocably." It is the diversity in this pattern of change which is the subject of the current paper. T o investigate this I shall concern myself almost exclusively with the West, and with m o n u m e n t s rather more than texts, looking first at an example which conforms in m a n y respects to Professor Brown's model, that of Ravenna, but then turning to examples further north, notably from Anglo-Saxon England, where the chronology and extent of the change seems to be rather different. Further, these examples direct attention to additional factors which impinge on the rate of change, such as the precise audience of a collection of images, or the role of the liturgy in continuing to endow an object or a space with praesentia, even after any sense of antique adoratio had faded. First Ravenna, where the surviving m o n u m e n t s have clear contexts and are meticulously dated—although any understanding of their iconography has to be achieved by inference, since no contemporary description of them survives. Belonging mainly, but with one crucial exception, to the fifth and sixth centuries, the decoration of the R a v e n n a churches was put up before the shift from praesentia to pictura, as explored and dated by Professor Brown. T h u s the

mausoleum of Galla Placidia, constructed in the first half of the fifth century, can easily be read as a building in which the individual becomes a participant, an adorator. H e or she would have been in the presence of a relic of St Lawrence, who can himself be seen in mosaic hastening towards death—his activity, a n d the liveliness of the fire awaiting him contrasted with the solid cabinet holding the gospels. 1 T h e m o n u m e n t makes palpable the presence of Lawrence, and conjures up the reality of his sacrifice. T h e remaining iconography completes the story, showing the Paradise to which the martyr has gone, notably with the image of Christ as the Good Shepherd, surrounded by his sheep, which faces the depiction of St Lawrence. T h e onlooker is unquestionably engulfed in Paradise, while being invited to recognise the praesentia of the martyr. Yet one might also note that the gospels in their bookcase, are identifiable by the names written on them: clearly they are to be taken metonymically as representing the whole faith, but at the same time their recognition requires the act of reading, of lectio. Also in Ravenna, the two baptisteries, the Neonian (exactly contemporary with the Mausoleum of Galla Placidia) and the Arian (built nearly a century later), with their circling apostles, may illusträte a shift towards iconic representation, 2 but the buildings themselves were both equally c o n c e r n e d with praesentia by their very function: it was there that the Holy Spirit descended at every baptismal ceremony. M o r e difficult to discuss is San Apollinare Nuovo, contemporary with the Arian baptistry, because its original decorative scheme is necessarily in doubt. 3 Admittedly its remarkable New Testament narratives survive in the uppermost b a n d of decoration, but their very positioning makes them difficult to read. T h e y can scarcely be simple pictura, pictures to be read, since at the distance from which they are seen they are almost unreadable. Being so, they might usefully be thought of as encircling a sacred space in appropriately sacred images, 4 almost an architectural equivalent of the "art that was, quite literally, w r a p p e d a r o u n d the person", to quote Peter Brown again. 5 As for the mosaics added to the church in the mid 1

F.W. Deichmann, Ravenna, Hauptstadt des spätantiken Abendlandes (Wiesbaden, etc., 1969-89), I, pp. 158-64. 2 Ε. Kitzinger, Byzantine Art in the Making (London, 1977), pp. 60-1. 3 Deichmann, Ravenna, Hauptstadt des spätantiken Abendlandes, I, pp. 171-200. 4 See above, Brown, pp. 31-2. 5 See above, Brown, p. 31.

sixth century, they may seem iconic, with their formalised procèssions of martyrs and virgins, but they are best understood within the world of litanies—and more than that, litanies loaded with theological meaning, since the virgins are led by Eufemia, the patron saint of Chalcedon, and the martyrs by the confessor Martin, opponent of arians. 6 T h e saints were, therefore, a presence to be invoked. T o enter the chancel of San Vitale, built in the second quarter of the sixth century, is also to enter a world of liturgy, a world surcharged, once again, with all sorts of presences. T h e church itself was the martyrial church of St Vitalis, the supposed father of Gervasius and Protasius, whose cult was developed to enhance the status of the see of Ravenna, and to justify its independence from Milan.' T h e mass, proclaiming the greatest praesentia of all, is eternally eelebrated on the walls of San Vitale. 8 As depicted, it is celebrated by the emperor Justinian, and his dead empress T h e o d o r a — r e p r e s e n tations which remind us that liturgy itself, sacred and profane, was m a d e u p of a series of potent, living, images. T w o and a half centuries earlier the portrait of the e m p e r o r would have been the subject of adoratio. Now the emperor's image was active in an even greater liturgy. Between the imperial images of San Vitale and the central octagon of the church there are narrative scenes to be read, but chosen in order to enhance an understanding of the Eucharist—Cain and Abel, Melchisadek, A b r a h a m and Isaac are all there. So too is Moses, represented two or three times. Is it sheer coincidence that he also appears thrice in St Catherine's Sinai, a church built by Justinian—himself a lawgiver? As for St Catherine's, it too is a church in which presences are overwhelming—that of the transfigured Christ, bowling over the disciples with the force of his transfiguration, and that of his antetype, Moses, the recipient of the Law. Quite literally behind all these presences in the church of Sinai was the burning bush itself, immediately to the east of the great apse mosaic, very m u c h an i m m a n e n t power. Returning to R a v e n n a — b u t leaving aside for the m o m e n t that

6

Deichmann, Ravenna, Hauptstadt des spätantiken Abendlandes, I, pp. 199-200: II, 1, pp. 149-50. On this frieze see also Beat Brenk's contribution to this volume. 7 O. von Simson, Sacred Fortress (Chicago, 1948), pp. 5, 15. 8 Von Simson, Sacred Fortress, pp. 27-30: S. McCormack, Art and Ceremony in Late Antiquity (Berkeley, 1981), pp. 260-1.

other Transfiguration, the apse mosaic of San Apollinare in Classe— let us j u m p ahead to the same church's autocephaly mosaic of 666, which firmly postdates Gregory the Great's statement on images: this was modelled, of course, on the image of Justinian in San Vitale, but here it lacks all eucharistic connotation (other than that which visual similarity conferred on it). Instead it makes a legal point: nothing more nor less than the independence of the C h u r c h of R a v e n n a from R o m e / ' This is unquestionably an image to be read—we have crossed a divide between praesentia and pictura, though in this instance we should perhaps note that the divide is as m u c h between types of scenes represented (scenes of liturgy and law) as chronology. Yet if the autocephaly mosaic is no more than a picture, a straightforward representation of a legal event, it is set in a part of the church, the sanctuary, which most certainly had not been stripped of the divine presence. Despite its function as a legal record, the autocephaly mosaic regularly witnessed the eucharist. Moreover, the mid-sixth-century apse-mosaic of San Apollinare in Classe, which looms above the sanctuary, sits challengingly across the divide of praesentia and pictura. T h e onlooker witnesses the Transfiguration but it is a Transfiguration totally different from that of its exact contemporary at St Catherine's. This is no explosive vision, but a pastoral scene in which Peter, J a m e s and J o h n gaze up as sheep at Christ, repesented as a cosmic cross. T o mediate this vision to the bewildered congregation is St Apollinaris himself—wearing his pallium, perhaps to celebrate the mass. T o this extent a presence is conjured up. O n the other hand Apollinaris' presence is best understood in terms of theological and ecclesiastical politics—a supposed disciple of St Peter, his cult became entangled with the growing confrontation between R a v e n n a and R o m e , which began with the shift of the imperial capital to Ravenna, ran through the T h r e e Chapters controversy, and continued with the grant of autocephaly. 10 T h e apse mosaic of San Apollinare in Classe can be approached both as praesentia and pictura—praesentia because of Apollinaris, whose own relics were preserved in the church, pictura because of the theological and political meanings to be read in the mosaic. Dating to the mid-sixth century the apse mosaic of San Apollinare in Classe could mark a m o m e n t

9 10

Von Simson, Sacred Fortress, pp. 5 9 - 6 0 . Von Simson, Sacred Fortress, pp. 46-52.

before the parting of the ways—but it could also mark a functional division between religious worship and reading. It is time to move north—across the Alps to one of the few architectural m o n u m e n t s of the late seventh or early eighth century for which it is possible to reconstruct the majority of the decorative scheme: the Hypogée des Dunes of abbot Mellobaudes of Poitiers. Despite the chronology—the building is approximately dated to the year 700—the parallels to the hypogeum are R o m a n rather than medieval. Moreover the m o n u m e n t itself conjures up a world of praesentia. T h e r e are rows of angels—some of them uncanonical—as well as two thieves crucified—and we must assume an image of Christ, or the cross, to complete this crucifixion scene." These images on their own might be no more than pictura, but approached down a flight of steps, engraved with cryptic symbols and overlooked by an inscribed anathema on all defilers, and with the arcosolium itself painted with the names of saints, whose relics may have been found within the mausoleum, they belong to an ambience loaded with presences. T h e Hypogée was after all the burial place of Abbot Mellobaudes, who may have hoped that his tomb, like so many others in Merovingian Gaul, would become a locus of the Holy. At the very least the a b b o t s body awaited its afterlife, guarded by inscriptions and images, including that of the Crucifixion, which automatically carried with it overtones of Christ's burial and Resurrection, and which thus might be said to have made i m m a n e n t the Holy Sepulchre itself. T h u s the hypogeum looks to the past, and to Constantine's building of the Holy Sepulchre in Jerusalem. It is not the only building to do so, since recall or the recreation of the tomb of Christ can be paralleled elsewhere in the seventh- and eighth-century: for instance, in the marginally-earlier creation by the Anglo-Saxon Wilfrid of a Sepulchre at Ripon, 1 2 and in subsequent Carolingian work, especially that of Angilbert at Saint-Riquier. 1 3 Such buildings seem to imply more than a simple representation of a particular m o n u m e n t . It is difficult to drain these copies of the Holy Sepulchre of the notion of praesentia. This may be an indication of the cultural sluggishness

" Abbé de la Croix: work currently in progress in the ESF project is tending to vindicate the Abbé's reconstruction of the Hypogée. 12 R.N. Bailey, "St. Wilfrid, Ripon and Hexham", Studies in Insular Art mid Archaeology, eds. C. Karkov and R. Farrell (Oxford, Ohio, 1991), pp. 3-25. 13 C. Heitz, L'architecture religieuse carolingienne (Paris, 1980), pp. 51 60.

of the northern world, but perhaps, following Professor Brown's suggestion, a more important factor is the necessary relationship of funerary m o n u m e n t s with the m o m e n t of death, when Heaven and Hell burst out for a brief instant to dazzle the living. As a gateway between life and death the t o m b was necessarily a place of potential praesentia. T h e r e was no more potent funerary m o n u m e n t than that of Christ himself. If we move yet further north other factors impinge. In AngloSaxon England the iconography of architectural sculpture presents precise problems, for although many fragments of such sculpture survive, they rarely survive in situ. T h u s the iconography which is most open to study is that of space—as at Ripon. M o r e accessible to interpretation is the non-architectural sculpture. T o take the most complex and challenging piece: the Ruthwell cross: 14 here, without question, is a cycle of scenes to be read: following the standard reconstruction of the fragments, and working u p from the bottom of each of the major faces the images run as follows: the Crucifixion, the Annunciation, the curing of the blind m a n , M a r y Magdalene washing Christ's feet and M a r y and M a r t h a on one side: the Flight into Egypt, the hermits Paul and Anthony, Christ adored by the animals and Christ (or J o h n the Baptist) presenting the L a m b of God, on the other. T h e scenes are identified by captions—though not all arthistorians concur with the sculptor's n a m i n g of M a r y and M a r t h a , ' 5 and it is impossible to say whether the title Iudex Aequitatis refers to the figure holding the L a m b or to the L a m b itself—a point which graphically illustrates the possibility of misidentifying images. 16 T h e Ruthwell Cross may be read as a series of images—and the cycle as a whole may be assumed to m e a n something more than the sum of its parts. At the simplest level one may note that one side moves from recognition of the truth of the Gospel to an espousal of the active christian life—there is the Crucifixion, the premise on 14

Here I treat the Ruthwell Cross in its final form. Initially it seems to have been a pillar: see F. Orton, "Rethinking the Ruthwell Monument: Fragments and critique; tradition and history; tongues and sockets", Art History 21 (1998), pp. 65-106. 15 Confusingly the Mary in question appears to be identified by an inscription as the Virgin: see D. Howlett, "Inscriptions and design of the Ruthwell Cross", The Ruthwell Cross, ed. B. Cassidy (Princeton, 1992), pp. 73-4. Howlett offers a different interpretation, reading mr as merentes rather than mater, although a deliberate amalgamation of the Visitation and Mary and Martha, with the two Marys forming a link, may well be what was intended. 16 T h e fullest survey is to be found in Cassidy, ed., The Ruthwell Cross.

which Christianity is based; the healing of the blind, the beginning of understanding; washing Christ's feet, Christian action as penitential act; a n d M a r y a n d M a r t h a , the two forms of Christian life: the other side concerns entry into the desert, which can be understood metaphorically as the monastic state. 1 ' So far we have pictura. Yet the Ruthwell cross is more than a collection of narrative images, for the two sides of the cross are decorated with vine-scroll, and round the vine-scroll is carved a series of verse lines which are related to a p o e m known as the Dream of the Rood, in which the cross itself addresses the beholder. In N o r t h u m b r i a in the mid-eighth century a piece of stone carving has enough presence to speak. In short the Ruthwell cross is neither entirely pictura nor praesentia. Since its iconography seems to be related to the liturgy current in R o m e , the m o n u m e n t may well have had some liturgical function. 1 8 T h e cross may even have incited adoration: the wear on the crucifixion panel could have been caused as m u c h by devotion as by the actions of the seventeenth-century iconoclasts. This is not the end of the challenges offered by this remarkable piece, for while the inscriptions on the front and back faces of the cross are essentially in Latin, those on the sides are in Old English, and inscribed in runes. T h e Ruthwell Cross passes all the tests of intertextuality—and it achieves "a totalizing discourse". O n the other h a n d it is a discourse which is far f r o m being d e m o c r a t i c — t h e onlooker has to know his or her Bible, as well as the iconographie significance of the true vine: he or she has to be able to decipher two alphabets and to read Latin and Old English. Moreover these two languages throw us back upon the issue of linguistic frontiers posed by Professor Brown, though in a different way, for they seem to be addressed to a single community, a n d a community which was exclusive, in the basic sense of the word: the native population of the area round Ruthwell may well have been Celtic-speaking, a n d ill-versed in both Latin and Old English. A multilingual object is not necessarily intended to be an accessible one. T h u s , while the

17

Cf. the discussion by P. Meyvaert, "A new perspective on the Ruthwell Cross: Ecclesia and Vita Monastica", in The Ruthwell Cross, ed. Cassidy, pp. 125 38, though without accepting his view that the top of the cross should be turned round. 18 É. O'Carragain, "Liturgical innovations associated with Pope Sergius and the iconography of the Ruthwell and Bewcastle Crosses", Bede and Anglo-Saxon England, British Archaeological Reports, British Series 46, ed. R.T. Farrell (Oxford, 1978), pp. 131-7.

images, which may have been no m o r e obvious to an onlooker in the eighth century than they are now, have their captions, those captions, whether in Latin or Old English, 19 are addressed to an elite, highly literate, presumably monastic, audience. In its intertextuality the Ruthwell Cross is not unique—at almost exactly the same time on the other side of N o r t h u m b r i a a carver created a whale-bone box known as the Franks Casket. It too presents a series of scenes: on the four sides Wayland in his smithy, the Adoration of the Magi, the sack of Jerusalem, Romulus and Remus, a n d a scene which is now unintelligible, a n d which may always have been enigmatic: on the lid a perfectly labelled scene depicts an attack on Eigil's house—a story about which nothing is known. Indeed all the scenes, like those on the Ruthwell Cross, are labelled. Again Latin a n d O l d English is used. Moreover, once again the object speaks, for the front panel presents a riddle, whose answer proclaims that it was m a d e out of whale bone. 2 0 R e a d i n g an object in Anglo-Saxon England was, therefore, something which could be decidedly complicated. A simple identification of scenes was not necessarily enough. T h e reader was required to listen to the object, which in the case of the Ruthwell Cross, at least, effectively speaks to the onlooker, thus implying the active presence of the Cross itself. Although reading, lectio, is required, the text implies presence, praesentia: the cross itself may have d e m a n d e d adoratio. Such an object would have delighted the readers of Aldhelm, whose riddies suggest something more of the culture to which the Ruthwell Cross and the Franks Casket belong. 21 It would, however, be dangerous to assume that this trait was exclusive to the Anglo-Saxons. AldhelnTs riddles are in Latin, a n d they d r a w on a Latin tradition. R e t u r n i n g from the verbal to the pictorial, there is a case for seeing at least some of the carpet-pages of insular Gospels as deliberate exercises in pictorial revelation. T o take an undoubtedly Irish example, the carpet pages of the Book of D u r r o w a p p e a r to represent a puzzle whose solution—according to E m m a n u e l l e Pirotte,

l!l

Brown's comment, with regard to Late Antiquity, that " O n e did not have to be an illiterate to draw conclusions from pictures—and, frequently, to draw the wrong conclusions" (above, p. 22), is likely to have held true for seventh- and eighthcentury Northumbria. 2 " R.I. Page, An Introduction to English Runes (London, 1973), pp. 174 82. 21 See M. Lapidge, Aldhelm, The Prose Works (Woodbridge, 1985), pp. 61 94: this is an issue which is being more fully explored by Professor Joyce Hill.

building on recent work is the cross of Christ. 22 T h e reading of an image, in this instance, is something more than mere lectio: it involves contemplation and concludes in revelation. 23 Like a page of Gospel text, carpet pages required contemplation to reveal their mysteries, and the images themselves were not the simple pictures envisaged by Gregory the Great as means of teaching the illiterate. In the Gospel books, and arguably on the crosses of N o r t h u m b r i a , images were aids to contemplation for the highly literate: and contemplation may also have culminated in acts of devotion: the m u c h worn state of the crucifixion image of the D u r h a m Gospels may bear witness to some liturgical adoration. In the Book Durrow and later insular Gospels interlace was drawn into service by an artist, and its intricacies exploited to engage the thoughts of the onlooker. In some cases the symbolism is intelligible, largely because it is drawn from a c o m m o n pool of christian imagery. T h e extent to which interlace, animal and other ornament bore meaning in pre-christian, Germanic art, and therefore the extent to which the decorators of Gospels Books were exploiting an already established means of symbolic expression, is unfortunately less open to investigation. O n occasion it is possible to unravel aspects of G e r m a n i c iconography, especially where it has R o m a n origins—as for instance that of the bracteates, which is imperial, even if their meaning became something quite other. 24 Indeed, leaving aside the implications of objects as indications of status, especially when found in assemblages of grave goods, the real meanings of the G e r m a n i c iconography of the period of the transformation of the R o m a n World are only clear when they can be related to R o m a n objects, or very occasionally to texts, some of which, like the Edda, were composed long after the Migration Period. 25 22

This issue is being E. Pirotte in a later volume in this series: in general see G. Henderson, From Durrow to Kelts: the insular Gospel books (London, 1987). 23 For an additional example see also J . O'Reilly, "The Book of Kelts, folio 114r: a mystery revealed yet concealed", The Age of Migrating Ideas: early medieval art in northern Britain and Ireland, eds. R.M. Spearman and J. Higgitt (Edinburgh, 1993), pp. 106-14. 24 M. Axboe, "Danish Kings and Dendrochronology: archaeological insights into the early history of the Danish state", After Empire: towards an ethnology of Europe's barbarians, ed. G. Ausenda (Woodbridge, 1995), pp. 230 1, with refs. 25 There is clearly some relation between certain scenes and Scandinavian mythology, as set down in the twelfth century, but it is usually impossible to determine the extent to which the stories recorded by Snorre have undergone change since the migration period.

T h e use of later documents is, however, fraught with danger, for a text is a product usually of a single individual working at very precise m o m e n t s in time. Only in exceptional cases, if ever, can twelfth-century writings be a guide to seventh-century meanings. Even an eighth-century text is no certain guide to the culture of the sixth. In pin-pointing the late sixth century as a watershed in the shift from lectio to pictura Professor Brown, like others before him, has noted the significance of the evidence of Gregory the Great. T o the letters of the pope himself is sometimes added further information from Bede, w h o stated that Augustine and his followers entered Canterbury singing a Rogation a n t h e m and carrying icons. 26 Since the Rogation a n t h e m was not known in R o m e at this date it is highly unlikely that Augustine knew it—and this must cast doubt on all the detail relating to the entry into Canterbury, supplied by Bede. 27 Indeed, one may wonder whether the icons of Bede's imagination did not have more to do with those brought back by Benedict Biscop to h a n g on the walls of M o n k w e a r m o u t h and J a r r o w in the late seventh century, than with anything Augustine might have brought with him a little less than a hundred years earlier. 28 M o n k w e a r m o u t h and J a r r o w , like Kildare, 2 9 had their icons before the end of the seventh century. Certainly these had their function in spelling out the gospel story, but in such monastic contexts the gospel story was actually being represented to communities which knew their Bibles backwards. Biscop can scarcely have brought his collection of icons back from R o m e primarily to teach his monastic community. W h a t the images did was to make his churches more similar to R o m a n churches, most of which had gained their mosaics, frescoes and icons before Gregory the Great h a d pushed praesentia just that little bit further towards pictura. In filling M o n k w e a r m o u t h and J a r r o w with images of the Old a n d New Testaments, moreover, Biscop was making the Holy Land itself more present: they added to an architectural symbolism which seems already to have evoked Jerusalem. 3 0 Images in the seventh and eighth centuries had, there-

2,1 Bede, Histona Ecclesiastica gentis Anglorum, I 25, ed. C. Plummer, Venerabilis Baedae Opera Historica (Oxford, 1896).' 27 I.N. Wood, "Augustine's Journey", Canterbury Cathedral Chronicle (1998), pp. 37-8. 28 Bede, Historia Abbatum, 6, 9, ed. Plummer, Venerabilis Baedae Opera Historica. 29 Cogitosus, Vita sanctae Brigidae, 7, Acta Sanctorum, February I, (February 1st). 30 I.N. Wood, The Most Holy Abbot Ceolfiid, Jarrow I^ecture 1995 (Newcastle-uponTyne, 1996), p. 15.

fore, not entirely lost their function of making palpable a sacred space, although now, in the North at least, that space might be the Holy Land rather than Paradise. T h e intimations of Paradise which were a dominant feature of art and architecture through to the late sixth century may well have ceded their importance to the world dominated by the grave. Yet there was already a balance between the two in the decades before 600, apparent in Gregory of Tours, with his roses in J a n u a r y , implying a suspension of worldly nature, and his tombs of the holy dead. " Moreover, in certain respects this balance remained. Some Columbanian communities certainly had an eye on the grave. In Book T w o of J o n a s ' Vita Columbani the nunnery of Faremoutiers looks unblinkingly towards death in its series of holy and unholy deaths. 32 Yet in the same narrative Luxeuil is different: Agrestius accused abbot Eustasius and his brethren of being overzealous in blessing cups, or perhaps spoons, and of crossing themselves. 33 Such zeal was actually required by the Rule of Columbanus, 3 4 as also was genuflection before the cross. 35 This emphasis on genuflection in a monastic Rule may be reflected in the Ruthwell Cross, whose iconographie scheme has apparently to be read from the bottom up. Not only was there the liturgy, constantly reminding a monastic community of the heavenly presence, but there were also numerous daily acts, which indicate something of the numinous which was not concentrated at the tombs of the saints. For some the world was still immanent with presences, which elicited the performance of tiny ritual acts. Such evidence complicates the identification of a change in the reading of images towards the end of Late Antiquity. Certainly the Italian evidence—that is the writings of Gregory the G r e a t and the mosaics of R a v e n n a can be seen as implying a change a n d as suggesting that a key m o m e n t came in the late sixth century. Further north, however, the change may have been slower and, even, less 31

For Gregory's view of Paradise see G. De Nie, Views from a many-windowed tower (Amsterdam, 1987), e.g. at p. 121: for the graves, P. Brown, The Cult of the Saints (Chicago, 1981), esp. pp. 73 85. 32 Jonas, Vita Columbani, II 1 1 22, ed. B. Krusch, lonae vitae sanctorum Columbani, Vedastis, Iohannis, Monumenta Germaniae Historica, Scriptores Rerum Germanicarum in usum scholarum (Hannover, 1905). 33 Jonas, Vita Columbani, II 9. 34 Columbanus, Regula Coenobialis Fratrum, I, 3, ed. G.S.M. Walker, Sancti Columbani Opera (Dublin, 1957), p. 146. 35 Columbanus, Regula Coenobialis Fratrum, III, ed. Walker, p. 146.

thorough-going: though by the end of the seventh century there were icons at M o n k w e a r m o u t h and J a r r o w — a n d indeed at Kildare—and iconic author portraits were appearing in Gospel books. Why this change should have taken place, and why it should have taken place at different speeds in different places, arc questions we can only answer through an inspection of the fragmentary evidence that remains. T h e deployment of images for p r o p a g a n d a purposes— at a time when Christianity had placed limits on the adoratio which might reasonably be paid to an e m p e r o r — m a y have speeded up the change from praesentia to pictura in some places. At the same time the liturgy may well have been a constant force for endowing certain objects and certain places with praesentia, long after other objects had become the bearers of mere pictura. Places associated with baptism, the eucharist and with holy death are likely to have kept some sort of charge. So too Bibles will have retained a charge, even if they came to encourage a new type of contemplation—a contemplation which could exploit the riddling culture of Aldhelm, and of the art of the insular Gospel books. All of which leaves untouched the questions raised by Gregory the Great's attitudes towards the use of images to teach the illiterate—but casts up a plethora of other issues and problems, m a n y of which at least help define what was at stake in the transformation of the R o m a n World.

ADMINISTRATIVE LANGUAGE AND ITS PUBLIC D E P L O Y M E N T N. Oikonomedes

T h e Transformation of the R o m a n World is a tricky topic when seen from a Byzantine vantage point; after all Byzantium always thought that it was the direct continuation of the R o m a n Empire. But we all know that, although more " R o m a n " that any of the other contenders, it was something different from what had come before. T h e end of the R o m a n Empire and the beginning of the Byzantine one in the East has always been a contentious issue, as there was no momentous event that could serve as a landmark. T h e foundation of Constantinople and its transformation into a capital city and seat of the emperor certainly had important long term results, as the centre of gravity was displaced to an hellenophone region, but still the character of the Empire did not change much and three quarters of a century later, the principle of the two capitals prevailed, one Western, the Old Rome, and one Eastern, the New Rome. T h e fall of Old R o m e and the creation of the barbarian kingdoms in the West left the Byzantines as the only real successors of the Empire. But this happened in the fifth century, at a time when Latin was quickly disappearing from the R o m a n East. T h e problems of communication between East and West were already making their appearance. Traditionalism stood up for some time to this general trend by advocating the principle of bilingualism. Emperor Justinian, in spite of his R o m a n convictions, found himself under obligation to promulgate his new legislation in Greek. And this in spite of the fact that he was in the process of reconquering many western and latinop h o n e regions for the Empire. But in the sixth century, the theory of the two centres could not hold any water. T h e Empire was unique, centred in Constantinople, and Italy and North Africa were mere provinces. Not quite. Before the end of the sixth century, the two large latinophone regions were placed under the administration of officers with special jurisdiction, with authority over the civil servants as well as the military of their regions. T h e y were called exarchoi, using a Greek

word m e a n i n g " l e a d e r " — a n d the use of a Greek word for the leaders of latinophone provinces is of some significance, if one thinks of the decision making centre and the decision making process. T h e two exarchates would not survive for very long. T h e exarchate of Africa will fall to the Arabs before the end of the seventh century. T h e exarchate of Italy, centred in Ravenna, will fall to the Lombards in the mid-eighth. But well before that, a new administrative unit, similar to the ones of the eastern empire, headed by a strategos (another Greek title), would appear in Sicily (shortly before 695), 1 where a large grecophone population had existed since ancient times. O n e is tempted to see in all this a certain discrimination in terminology and in administrative procedure that follows a rough linguistic line of separation. In what follows, I shall try to examine this desired bilingualism in the field of the imperial administration and its results in real life. From the geographical point of view I will concentrate to the Eastern Empire, but without excluding the latinophone countries; this choice is due to the limits of my competence; also to the fact that the linguistic problem is more acute and obvious in a state that enjoyed uninterrupted continuity. From the chronological point of view, I will start with the sixth century, and will concentrate on the following ones. I will go down to the ninth century, as is required by the organizers of the meeting, but sometimes I will dare to enter the tenth and the eleventh centuries, because the general chartacteristics of the empire did not change in any radical way before the last decades of the eleventh. T h e main part of the period that I will consider, from the midseventh to the mid-ninth centuries, is the "Dark Ages" of Byzantium, so n a m e d because of the lack of narrative sources. In my research I shall use what is available, with particular stress on epigraphy, when possible, and on sigillography. Lead seals, dating mainly from the sixth to twelfth centuries, which were made to close parcels or letters and to authenticate open documents, have survived in thousands (ca. 70,000 of them are known and their n u m b e r increases every day). T h e y are now being studied and published more sys-

1

N. Oikonomides, Les listes de préséance byzantines des IX' et X' siècles (henceforth Listes) (Paris, 1972), p. 351: Catalogue of Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art, eds. J . Nesbitt and N. Oikonomides (Washington D.C., 1991), pp. I, 5.

tematically than in the past. 2 These are quantities of first hand material, especially important for administrative history, as they mainly provide us with the names and titles of public officials. T h e y are fundamental for one who wants to study administrative language, which, by their nature they deployed as publicly as possible. 5

The survival of iMtin in the. East It is certain that at the m o m e n t of the transfer of the capital in the East, high hopes prevailed that the empire would be at least functionally bilingual. This d r e a m of bilingualism is manifest during the early centuries, as is shown by the reorganization of superior teaching in Constantinople u n d e r Theodosios II (425) or by the legislation of Justinian. But in time, individuals knowing Latin become rare in Constantinople. E m p e r o r R o m a n o s III Argyros of the eleventh century prided himself for knowing Latin, 4 but what he knew he had learned because he was a jurist interested in R o m a n legislation, not in Latin culture. O n the contrary, all the great scholars of tenthand eleventh-century Constantinople, such as Arethas of Caesarea, Michael Psellos, J o h n Mauropous, Constantine Leichoudes ignored this language. It is hard to say when the eastern empire ceased to be functionnally bilingual. Already in the first half of the sixth century Justinian I declared that he would not have his Novellae written in his "ancestral tongue" (i.e. Latin), but rather in the " c o m m o n language of the Greeks", because he wanted the law to be understood by all, by "the people". 5 Latin was obviously then on its way to disappearing

2 Beyond the scholady catalogues, that have appeared recently, there is also a quasi periodical publication under the title Studies in Byzantine Sigillography, published by Dumbarton Oaks in Washington D.C., with detailed indexes, mainly containing papers presented to international colloquia on Byzantine seals. These papers address methodological questions and theoretic al problems of philology and iconography; new material is also published, especially seals found in more or less well known archaeological contexts. Also, an effort is being made to gather and index all the lead seals scattered in various publications. 3 I shall draw most of my examples from a vast collection that was published two decades ago: G. Zacos and A. Veglery, Byzantine Lead Seats (henceforth ZV) (Basel, 1972). ‫י‬ 4 Michel Psellos, Chronographie, ed. Ε. Renauld, I (Paris 1926), pp. 32.33‫־‬ 5 Justinian, Nov. 7, c. 1, 32-5: Nov. 66, c. 1, 2-3.

in the East. From the seventh century onwards, it is not mentioned in the legislation any more. Yet it seems that Latin was still spoken in the Balkans until the end of the seventh century, and this in big cities, like Thessalonica. Moreover, the text providing us with this information also shows that seventh-century Thessalonians were open to the idea of linguistic pluralism, well beyond the traditional bilingualism of the empire. O u r main texts on this subject are contained in the collections of the Miracles of St Demetrius, which provide i m p o r t a n t information concerning the Slavic attacks against Thessalonica. 6 In the first collection, written by archbishop J o h n at the beginning of the seventh century, we are informed about a soldier possessed by the devil, who, when cured by the saint, started asking unexpected questions "in the R o m a n language". W h a t is this language? If the author, being himself a Greek and addressing a Greek audience, felt it necessary to define the language in which these questions were asked, one can deduce that it was not Greek—thus it could be only Latin, called by the Byzantines "the R o m a n tongue" until the tenth century. Consequendy, we can conclude that in the army of Thessalonica Latin was still a language that could be spoken and understood in the early seventh century, but that its use was out of the ordinary. In the same collection of miracles, we find a text written in the last quarter of the seventh century (between 678 and 685), describing how a Bulgarian chieftain, an exceptional and very resourceful m a n , spoke four languages: Greek ("our language"), Latin ("the language of the Romans"), Slavic and Bulgarian. These linguistic capabilities gave this man, according to the author, the ability to infiltrate the Thessalonians and try to deliver the city to his own chief. From this text, where the languages are quite well distinguished from each other, including the Slavic and the (proto-) Bulgarian (of Turkic origin), we can deduce that Latin was still known in the west of the Balkans, even a m o n g those para-Byzantines of the countryside. It is not impossible that the survival of Latin in these regions was due to the fact that ecclesiastically Illyricum still depended on R o m e , and consequently the ecclesiastical authorities had to main-

fi

P. Lemerle, Les plus anciens recueils des miracles de Saint Demetrius I—II (Paris, 1979, 1981); see especially I, pp. 86 and 229; for the dates see II, p. 161. For the meaning of " R o m a n " = Latin, see Martha Grègoriou-Ioannidou, in Byzantiaka 3 (1983), pp. 85-90.

tain the language in order to communicate with the pontiff. Also, in the Northern Balkans, there was an important latinophone tradition. In any case, J o h n Lydos stresses that the inhabitants "of Europe" although mostly of Greek descent, spoke Latin, especially those in the public service. 7 T h e r e is not much to be added here. T h e ecclesiastical province of Illyricum was removed from the authority of R o m e by the Isaurian emperors in the first half of the eighth century. And the usage of Latin disappears in the Balkans (presumably with the exception of the Vlachs). In Byzantine milieu Latin words (but practically no phrases except for the set ones) survived in the administration and, of course, in some ceremonies and in some c o m m a n d s used by the military. By contrast, Greek was still spoken not only in Byzantine Italy but also beyond it, as this is attested for the tenth century in a miracle of Saint Dcmetrios."

Bilingual Seals I have already studied the question of bilingual seals in a paper delivered at Erice. 9 T h e question has also been studied by E. Stepanova, whose research has concentrated on seals with unilingual Latin or mixed inscriptions. O n the basis of the collection of the Hermitage, where the provenance of some of the seals is known thanks to the notes of N. Lihacev, she concludes that most of those Latin or GrecoLatin seals come from the West; thus she attributed them to Greek officials sent there after Justinian's reconquest, and dated them all between the sixth and seventh centuries. 10 It is certain that in the western part of the empire, Latin was predominant. In early eighth-century R a v e n n a , there were some orientals who were able to speak Latin but unable to write it; for this reason they signed agreements on papyri in the Latin language but written in Greek letters—they may also have done so because they

' J o h n Lydos, De magistratibus, III, 68, ed. Wunsch, pp. 158 159. Acta Sanctorum, October IV, p. 193. 9 "L'épigraphie des bulles byzantines", in Epigraßa medievale greca e latina. Icleologia e funzione, eds. G. Cavallo and C. Mango (Spoleto, 1995), pp. 153 168. 10 E. Stepanova, "Seals with Latin Inscriptions in the Hermitage Collection", Studies in Byzantine Sigillography 3 (1993), pp. 29 39. 8

may have thought that Greek was a noble alphabet." If such difficulties could exist in Ravenna, the usage of Latin in the eastern part of the Empire would certainly have been by far more problematic. Hcnce the question: what was the point in fabricating a bilingual seal in the Eastern R o m a n Empire? O n e can imagine that unilingual Latin seals were fabricated to be used in Italy or North Africa. M a n y such seals are discovered in that part of the world, seals of private individuals but also of exarchs or other imperial officials. But similar seals seem to have been used in the oriental part of the Empire. W e have, for example, the unilingual Latin lead seal of a count of Abydos, at the entrance of the Dardanelles: 11 ' was-it to be used on documents or merchandise addressed to ports of the pars occidentis? It is very uncertain, since we can say with certainty that all-Latin seals are found, albeit in small numbers, all over the territory of the Eastern R o m a n Empire. For example: Theodosio illustri, found in D o b r o u d j a ; Petru scnbonos, found in Crete; Carellu candidatu, found in Athens; Sergii illustm et comitis, found in Samos. 1 ! Consequently one has to envisage the possibility that the Latin language might also have a special significance, beyond the simple need to make oneself understood to a correspondent who happened to be latinophone. It is obvious that the necessity to be understood predominated. It is also conceivable that a certain pride might play a role when choosing the language of one's seal. W e know several individuals who possessed two seals, one in Greek, another in Latin, both featuring the same n a m e and titles. It is obvious that in this case the owner of the seals chose the language according to his correspondent. W e can quote the example of two sakellaùoi, Leontios (642647‫)־‬ a n d Maurikios (654-668) whose Latin seals have been f o u n d in Carthage. 1 4 O n the other hand, their Greek seals, with the same title of sakellarios, have been found in Constantinople. T h e r e are two possibilities. O n e would be that both have served as sakellanoi in Carthage before being promoted to Constantinople—but this would be quite

" Cf. A. Guillou, in: La Paléographie grecque et byzantine, Colloque international du CNRS, no. 559 (Paris, 1977), pp. 526 .528‫־‬ 12 ZV 2871 : Constantinou comitos Abydou. 13 Studies in Byzantine Sigillography Γ (1987), p. 77; 2 (1990), pp. 147, 242, 258. 14 Cf. Cécile Morrisson and W. Seint, "Sceaux de commerciaires byzantins du V i r siècle trouvés à Carthage", Revue Numismatique 24 (1982), pp. 236-238.

a coincidence and ultimately seems unlikely. T h e other and more likely possibility would be that while in Constantinople, they had also m a d e a second Latin seal for their correspondence with—or for certifying monetary emissions of the mint o f — C a r t h a g e . An interesting p h e n o m e n o n is that in the g r o u p of unilingual Latin seals, while the script is Latin, the language is in fact Greek, as is shown from the endings: Sergiu palatinu, Ioannu, Carellu candidatu, Constantinu scnbonos, Eutychianu illustnu11 are some examples that could easily be multiplied. O f course, we also find seals inscribed in correct Latin, such as Carello magistro militum, Constanti chartulario, Genethlio ex consule, Genethlium ex consule et patúcium,6‫ י‬but they are really very rare outside Italy a n d North Africa. T h e same can be remarked concerning the bilingual seals, the ones bearing the owner's n a m e in Greek on one side, and in Latin on the other: the Latin inscription is sometimes correct (e. g. Μαρίνου Μαήηϊ) but usually the Latin inscription presents obvious signs of Greek influence (Άθηνογένους—Atinogenus). 17 T h e same appears on seals with longer inscriptions repeated in the two languages. Usually the Greek is correct, which shows that this was the p r e d o m i n a n t language in the region. T h u s one has to conclude that in most cases the translation was m a d e from Greek to Latin. But if the owners lived in a Greek-speaking milieu, why did they bother translating the inscriptions on their seals? J u s t out of respect for bilingualism? If this was a generally accepted idea, why were there so m a n y Byzantines who chose unilingual seals? O n e could say that it was out of sheer respect for tradition and reflects a certain generation gap. For example, Ιωάννου—Mariani. i S It is clear that one of the two names is a patronym. In such a case, I would tend to think of those R o m a n families that came to Constantinople (especially the senatorial families that Constantine the Great brought there) and w h o were reaching a point of Hellcnization: the father's n a m e was written in the "ancestral" tongue, Latin, while the son's was in the prevailing language, Greek. T h u s , in m y view, the most likely translation of this inscription would be: J o h n son of Marianus.

15 16 17 18

ZV ZV ZV ZV

488A, 633, 767, 786, 806. 768, 773A, 810, 811. 2863, 2900. 2801.

But this cannot apply to bilingual seals with the n a m e in one Ianguage a n d the title in the other. T h e combination of the n a m e in Latin a n d the title in Greek is not c o m m o n , a n d usually it occurs when the n a m e is written in the form of a m o n o g r a m ; one can hypothesize that, if it was a m o n o g r a m with some tradition in the family, no one would want or dare to modify it—and so the combination of a Latin n a m e with a Greek title may have survived. Also one should stress that this is a rare combination. In the collection of Z V I found only nine such cases with the titles άπο έπαρχων, από υπάτων, βασιλικός σπαθάριος, κουβικουλάριος, πατρίκιος, ύπατος, χαρτουλάριος. T h e contrary is m u c h more c o m m o n . In the same publication I found seals with Greek names together with the following Latin titles: admissionalem, ad sabanis, apo ypaton, asecretis, basilicu spatanu, bestit0rìs, chartulariu, comitis sacrarum, discussoris, domesticu, illustriu, magistru, magistro militum, moderatoros, patrìciu, proticturns, referendanu, scholariu, scholasticu, scribonos, scriniariu, silentiariu, stratelatu, stratioticu logothetu, vicam Traciae, and m a n y others. It is interesting to note the absence of ecclesiastical tides. T h e Church, always attached to the idea of direct communication with the flock, did not partieipate in the bilingualism g a m e — t h e C h u r c h addressed the faithful in as direct a way as possible. Sometimes the Latin inscription contains Greek letters and, less often, everything is completely mixed up: Φιλιππικοΰ scholastici καί ίλλουστρίου. 19 In spite of the difficulties, some insisted on bilingual inscriptions. I think of the seal of Ευσταθίου stratioticu logothetu,2° which cannot be dated before the seventh century, but on which the owner chose to transcribe in Latin characters a Greek title that had nothing to do with any Latin tradition. It was not because the targeted audience was latinophone, because the n a m e of the officiai is written out in Greek. Obviously, one has to find a n o t h e r explanation for this inscription. Another example: in the magister militum and had a title. But he also m a d e a difference that on this seal

19 20

ZV 469. ZV 805.

sixth century, a certain Carellus b e c a m e Latin seal engraved with his n a m e and Greek seal: Καρέλλου stratelatu, with the the title of stratelates has been engraved

in Latin script. 21 This case as well as the one of the logothete of the stratiotikon can be explained only if one thinks that the Latin alphabet had the reputation at that time of being a noble alphabet, the alphabet in which it was normal to write the names of titles and offices. An alphabet noble as far as the administration was concerned. A variant of the above combination: on one side we have a monogram with the n a m e of the owner in one language, while on the other we have his n a m e and title in the other language: Εύθαλίου Euthalii chartularii‫׳‬, Ιωάννου—Ioanu scnbonos (kai) stratoros; Θεοδώρου Theodoru stratoros, or the other way round: Mariani—Μαριανοΰ σκρίβωνος; Theodon—Θεοδώρου πατρικίου. 2 ‫ ־‬Here one must suppose, I think that the owner wanted to keep unchanged a m o n o g r a m , Greek or Latin, that had long belonged to his family or that he already had on his ring or his bi-valve seal for wax.

Titles W h e n speaking of titles, one has naturally to take it for granted that old offices which existed maintained their names, maybe in a grecisized form, with no one feeling the necessity to translate them. In fact, translations of titles occurred usually in the time of bilingualism, rather than later. T h e praefectus was translated into Greek by eparchos from the very beginning, were he prefect of the praetorium or of the City. But domesticus or magister were never translated; they were just transcribed, equipped with a Greek ending—domestikos, magistros and survived like this until the e n d — b u t of course with meanings that evolved considerably in the course of the centuries. Yet one has to accept that, for example, the word domestikos had in the course of time become a Greek word, and had evolved some distance from its initial meaning. Presumably at a certain m o m e n t the domesticus of the magister ojjiciorum c o m m a n d e d , on behalf of his superior, the regiment of the scholae. T h e n the word domestikos was adopted in Greek with the meaning of " c o m m a n d e r of a regiment of the imperial guard". And thus, all such regiments that have been created subsequently, the hikanatoi in the early ninth century, the

21 22

ZV 768, 769. ZV 326, 391, 2827; 422, 526.

athanatoi in the tenth, were placed under the c o m m a n d of a domestikos. Moreover, the second in c o m m a n d to the domestikos is constantly n a m e d topoteretes, a Greek term meaning the locum tenens, although the Latin equivalent of this is never attested as far as I know. 23 In the lower ranks of officers and petty officers, one finds a good mixture of Latin and Greek terms, obviously reflecting a tradition, which in the military could not but be tenacious. Another example somehow different in nature: in R o m a n military terminology numerus meant a regiment of soldiers. In the imperial palace of Constantinople there were certainly several numeri of guards. O n e of them became the regiment of the vigla or arìthm0s, which is the Greek translation of the word numerus, and was entrusted with the guard of the closed hippodrome and, during campaigns, with the personal security of the emperor. 2 4 Another, an ex-numerus of the excubiti, kept its grecisized Latin name, noumeros, and designated the guard of one of the three m a j o r prisons of Constantinople, also situated at the great palace. 25 In this case the Greek translation of a term ended u p by meaning another entity than its Latin model. Sometimes, one has the impression that the Byzantines tried to avoid confusion that might be provoked by the usage of their own Latin terminology in a milieu where Latin was a living language. T h u s in the second half of the tenth century, the Byzantines ereated large frontier commands of heavy cavalry that they placed under a doux or katepano. T h e two terms indicate practically the same authority, but the first is of Latin etymology, dux, ducis, while the second is a very Greek one, 0 epi, 0 epano — the head of something. But we realize that in southern Italy, where a Latin speaking population has always existed as have several local duces, the term used almost without exception is katepano,26 and the c o m m a n d is called katepanikion or katepanaton—Greek terms that will be adopted in Italian as capitanata, capitano, and will end by giving birth to the m o d e r n "captain", with its various meanings. T h e survival of a terminology the initial meaning of which was not any more understood gave, as is natural, birth to all kinds of misunderstandings and false etymologies. T h e Latin term a triclinio

23 24 25 26

Listes, pp. 329 ff. Listes, p. 331. R. Janin, Constantinople byzantine (Paris, 1964), pp. 169-170. Listes, p. 344.

indicated the official in c h a r g e of the b a n q u e t hall, i.e. in c h a r g e of seating all guests in the p r o p e r o r d e r at an imperial banquet. 2 7 T h e t e r m passed into G r e e k in the grecisised f o r m atriklines—close to the original but with a G r e e k ending. As the w o r d did not have a n y obvious m e a n i n g (in spite of the fact that the w o r d triklinos = hall was still of c o m m o n use in the palace) false etymologies a p p e a r e d : artiklines, c o m p o s e d of arti = recently a n d kline = couch or b a n q u e t seat. O r , better, artoklines, c o m p o s e d of artos - b r e a d a n d kline = couch. T h e n e w w o r d s m a d e hardly better sense t h a n the original one, but at least they h a d a n obvious a n d u n d e r s t a n d a b l e , albeit a b s u r d , etymology, a n d this would suffice for a public that was not really interested in semantics, but r a t h e r w a n t e d the feeling of safety that the usage of a familiar language gives. For the rest, there is a tendency, very p r o n o u n c e d in the seventh century a n d later, to c o m p o s e n e w titles with G r e e k words. W h e n the praefectura praetorio per Onentem exploded into several "ministries" the G r e e k t e r m logothetes was m o s d y used to n a m e the chiefs; a n d the n a m e s of the services were also Greek, in spite of their Latin antecedents: genikon, stratiotikon, dromos (for cursus), agelon (for gregum), etc. At the same time, the provincial c o m m a n d e r was called strategos, a new G r e e k t e r m , i n d e p e n d e n t f r o m magistn militum (in G r e e k stratelatai) of the past. But of course this did not diminish the very i m p o r t a n t influence that G r e e k administrative terminology h a d received f r o m Latin. A very good e x a m p l e is provided by the e n d i n g -anus, of very c o m m o n usage in p r o d u c i n g n e w w o r d s related to the administration. It was a d o p t e d in G r e e k in the f o r m -anos a n d was practically accepted in c o m m o n usage: e.g. spatharius-spathams‫׳‬. so m u c h so that it was also used with foreign roots: drungus-drungarius-drounganos. A n d this productive e n d i n g did not d i s a p p e a r with the eclipse of Latin. In the tenth century it was still used in o r d e r to give a G r e e k f o r m to a d o p t e d foreign words such as the A r m e n i a n tasnak-tasinarios, or the huszar-honsarios, both indicating light c a v a l r y m e n used to raid e n e m y territory. 2 8 W h a t is even m o r e striking is the s u d d e n revival of Latin terms in the administration at a time w h e n the language was long forgotten.

27 28

Listes, pp. 27-9.

G. Dagron and H. Mihaescu, Le traité sur la guérilla de l'empereur Nicéphore Phocas (Paris, 1986), pp. 252-254.

T h e y naturally reappear in Grecicised form, but they are nonetheless surprising. T h e censor (kensor) a n d augustalis (augustalios) reappear in the tenth century, the illustris (illoustrios) in the eleventh.‫ '"־‬It is obvious that consciousness of a R o m a n past was very m u c h alive in the Byzantine empire, in spite of the fact that the Byzantines had, in the meantime, lost all practical knowledge of the Latin language. M o r e than that: in the new E u r o p e a n context, the Byzantines h a d b e c o m e antagonistic to the Latin-speakers of the West, who also claimed to be the successors of the R o m a n Empire, first with Charlemagne, then with the G e r m a n emperors. It is not impossible that this very partial "revival" of Latin in the administrative terminology of Constantinople was related to this political antagonism between the Eastern a n d the Western Empires.

Conclusion T h e two-language problem was not a novelty in the Byzantine empire; it had existed in the same milieu, with the same languages, since the time of the R o m a n conquest of the Balkans. 3 ' 1 And it has reappeared several times since, a n d is a very c o m m o n problem in m a n y societies down to our own day. Yet in Byzantium it presents several peculiarities, and is influenced by some objective and unsurmountable conditions: i) Byzantium always considered itself as the " R o m a n Empire". So m u c h so that the n a m e Romaios (Roman) was the only one used to indicate its Greek-speaking citizens, who have kept it until today. Romios means the Greek. This ideology permeated all strata of the population who identified themselves as "Byzantines". ii) This ideological attachment to R o m a n tradition was coupled, most significantly, with the political desire to monopolize this tradition. T h e clashes with the western emperors a n d the haughty attitude towards all other neighbours bear witness to this. iii) Yet, the R o m a n Empire was always conceived as a natural necessity, imposed by G o d ' s will, and consequently free from some bonds that in different conditions would have had prime importance.

29

Listes, pp. 309, 325. See for example, G. Daux, L'onomastique romaine d'expression grecque, L'onomastique latine, Colloque international du CNRS, no. 564 (Paris, 1977), p. 406. 30

O n e of the bonds that could be, and has been, neglected is the one of the language. iv) Byzantium being an absolute m o n a r c h y , political desire would normally suffice to impose its points of view, especially if this point of view conformed with a generally accepted ideology. Nevertheless, the avowed desire to maintain the R o m a n tradition and to use the "ancestral tongue", as Justinian called Latin, have not ended up by maintaining this language, even as a vehicle for the administrative terminilogy, m u c h less as a m e a n s of communication in the administration. v) It is obvious that in the christian East, with its administrative a n d monarchic tradition stretching back to Hellenistic times, with its cultural tradition that went even f u r t h e r back, a real need to use Latin in the administration was never seriously felt. Traditionalists managed only to impose some use of this language at the very beginning, a use that became all the m o r e marginal with the passage of time. This marginalization was even more p r o n o u n c e d w h e n only social (and not administrative) criteria were involved. In the seventh century, Latin was called "the language of the R o m a n s " ; in the eleventh century, "language of the Italians". In the m e a n t i m e it had become clear for all Easterners that the real languagc of the R o m a n s was the one that they were speaking themselves, Greek, ta Romaiïka, as the language is still called by some today. T h e y were ready to give up the "ancestral tongue" of Justinian, while bitterly fighting for the imperial tradition that went with it. W e all know how difficult it is to learn a foreign language.

SOME REFLECTIONS ON MANDARIN

LANGUAGE

Mayke dc J o n g

Professor Oikonomides has raised a n u m b e r of interesdng quesdons. W h a t struck me most is how quickly "the d r e a m of bilingualism"— still cherished in the fifth century—vanished in the East. Rapidly and unequivocally, Greek became the administrative language proper; although some orientals in early eighth-century R a v e n n a could speak Latin, they used Greek letters when signing documents in Latin. Apparently the reputation of Latin as a "noble alphabet" was best preserved in the domain of seals a n d honorary titles; there was even an upsurge in Latin titles at times of increased confrontation with the West. Nonetheless, the overall impression is one of a predomin a n d y archaising a n d symbolic use of Latin. T h e language of legal transactions matters deeply, for its validity depends upon the written word. Obviously Latin had some kind of snob value in Byzantium, but it did not e m b o d y power—otherwise the administration would have stuck to it. Given that the administrative language, publicly deployed, was Greek, I w o n d e r whether the label "bilingual" should be used after the sixth century. Was there no need to express the ideological attachment to R o m a n tradition by using Latin as the language of government because those in power never for once doubted their connection to a glorious past? Professor Oikonomides seems to suggest that unlike the successor states in the West, Byzantium h a d no need to prove its Romanitas. I f o u n d this a n interesting t h o u g h t , a n d was r e m i n d e d of Michel Banniard's comparison of Paul the Deacon and Alcuin—the former secure in his continuity with a R o m a n a n d Latin past, the latter stemming f r o m a country where the basically unfamiliar Latinitas c o m m a n d e d a disproportionate respect. 1 T h e Byzantine situation m a y have been exceptional, in that the written language of administration was also that of informal, oral

' M. Banniard, Viva voce. Communication écrite et communication orate du IV au IX' siècle en Occident iMtin, Collection des Etudes Augustiniennes. Série Moyen Age et Temps Modernes 25 (Paris, 1992), pp. 306-307.

communication. T h e use of M a n d a r i n languages seems to have been m o r e c o m m o n . 2 Arabic was introduced as the written and spoken language of government in a multi-lingual empire from 695 onwards, a measure which had rapid success; Old Persian was used for sacred a n d legal purposes, but stories were written down in Elamitic; M a n darin Chinese emerged around 1500 B.C. a n d continued virtually u n c h a n g e d up till this very century. W h a t are the advantages a n d disadvantages of having a M a n d a r i n language (an expression I will use here to denote a formal written a n d spoken language used for administrative purposes, which differs fundamentally from the language of informal communication)? I should like to reflect upon this briefly, drawing most of my examples f r o m the domain I a m most familiar with, the Carolingian kingdoms. For those governing a multi-lingual realm, the adoption of one administrative language has obvious practical advantages. W h y one particular language is considered suitable for this purpose is another matter; the criteria guiding this choice are variable. O n e thing is certain: it is not advisable to use a written language in its experimental stages for governmental discourse, otherwise one ends u p with an administrative disaster. W h a t e v e r its language, therefore, legal and administrative written communication tends to be extremely conservative a n d rigid. Since the Merovingians already took over a R o m a n administrative system and the Carolingians inherited that, the obvious practical solution was to go on using Latin a n d avoid misunderstanding. Accordingly, the early experiments with the written vernacular—i.e. adapting the Latin alphabet to the spoken language— predominantly stem from expert linguists trying out their skills, mainly in the religious a n d / o r missionary domain. This goes for Ottfried von Weissenburg's Evangeliendichtung, as well as for the a u t h o r of the Eulalia sequences. 3 However, language is not only used for communication, but also for hiding information in order to exclude outsiders, a n d for defining

2

I am not using the expression "Mandarin languages" in any precise or technical sense of the term, but rather as a metaphor for administrative languages which require formal and extensive training to be mastered. ! Otfrid von Weißenburg, Evangelienbuch, ed., Ο. Erdmann, 6th ed. (Tübingen, 1973). In general see W. Haubrichs, Geschichte der deutschen Literatur von den Anfangen bis zum Beginn der Neuzeit, Band 1, Von den Anfangen zum hohen Mittelalter, Teil 1: Die Anfange: Versuche volkssprachiger Schrifllichkeit im frühen Mittelalter (ca 700-1050/60) (Frankfurtam-Main, 1988).

the identity of élites. This does not preclude social climbing; on the contrary, poor bright Chinese boys could make their way up in life precisely because something like M a n d a r i n existed, just as m a n y of their counterparts in the Medieval West found their way into the corridors of power by entering the clergy and learning Latin. But even if such an o p p o r t u n i t y presents itself, the sharp distinction between "us" and " t h e m " remains. O r , to express it in the words of H r a b a n u s Maurus: "All those who are with us men of Latin" (Omnes nobiscum linguae Latinae homines) .4 This sense of exclusiveness may even get stronger if the M a n d a r i n language is also the language of the sacred texts, as was the case in Islam. Even if angels were thought to speak Persian, the language of poetry, the Arabic of the K o r a n was the ultimate yardstick for governmental discourse, for the K o r a n was (and is) "the Law" p a r excellence. T h e r e are interesting lessons to be learned here for students of the Carolingian world, where the Old T e s t a m e n t was held to be the vetus lex. T h e field of linguistic development in the early medieval West is a lively a n d fascinating one, to historians as well as linguists, for the central issue is the role of written Latin in these societies and, ultimately, the value of the written sources which historians usually rely on. T a k e the capitularies, for example. D o these only inform us of a small, mostly clerical segment of society, d r e a m ing its dreams of an ordered society in courtly or monastic seclusion, or are these the practical documents of a government at work, communicating with those in charge all over the realm? R o s a m o n d McKittcrick has been the most eloquent advocate of the second point of view. 3 She has argued against the idea of Latin as a clerical prerogative, presenting Carolingian society as one "in which the written word was central". She does not deny the growing linguistic divisions a n d their concomitant problems, but wants historians to concentrate on "what these linguistic categories meant in terms of an individual's ability to learn to read and conduct his or her daily affairs with some resort to literate modes of communication". 6 Which

4

Hrabanus Maurus, De institutione clericorum II, c. 8, ed. A. Knöpfler, Rabani Mauri de institutione clericorum linbri très (Munich, 1905). 5 R. McKitterick, The Carolingians and the Written Word (Cambridge 1989); eadem (ed.), The Uses of Literacy in Early .Medieval Europe (Cambridge 1990). 6 R. McKitterick, "Latin and Romance: an historian's perspective", Latin and the Romance Languages in the Early Middle Ages, ed. R. Wright (London/New, York 1991), p. 137.

reminded me of one of my refreshingly candid first-year students explaining to the class that " G e r m a n i c peoples are primitive a n d therefore m u c h too stupid to learn Latin." I am afraid that sentim e n t s of this sort have long i n f o r m e d historians' views on the effectiveness of Carolingian administrative language; we have certainly come a long way in the past decades. O n the other h a n d , there is the problem of the emergence of what I have called a " M a n d a r i n " language. In the course of the ninth century written Latin underwent syntactical and orthographic transformations which eventually led to the divide between Latin and R o m a n c e . N o b o d y doubts that this h a p p e n e d , but there is still disagreement about the speed of the process. Some consider the educational reforms initiated by C h a r l e m a g n e a n d Alcuin as a fast-acting catalyst, 7 while others argue for a much slower development. 8 Whatever the case, one should keep in m i n d the fact that the growing divide between "grammatical" and "rustic" Latin did not make it impossible for the lay élite to become familiar with the "high" language. If they failed to do so, they could not function within the inner circle of the court. 9 Undoubtedly, classicising tendencies served to enhance the Carolingian rulers' prestige, linking t h e m with L a t e - R o m a n traditions. Both G o v e r n m e n t a n d liturgy b e c a m e m o r e " R o m a n " by using the M a n d a r i n version of Latin. But precisely because improved Latin b e c a m e the language of both liturgy a n d secular administration, we should redefine the notion of an "administrative language, publicly

7 R. Wright, Late Latin and Early Romance in Spain and Carolingian France (Liverpool, 1982); Banniard, Vwa voce. In discussions of the emergence of Carolingian Mandarin Alcuin has been one of the leading characters. As a Northumbrian, coming from a country where Latin was not spoken, he harboured a deep veneration for the language of Rome, and even what Banniard has called an "illusion romaine et latine"—a mistaken idea of Roman continuity on the continent. In Alcuin's correspondence, a veritable dichotomy between a culture savante and a culture populaire becomes visible, with grammatica as the distinguishing characteristic. O n e may well wonder whether his irritation about "rustic Latin" would have been so great if spoken Latin had not resembled the written, grammatical variety so closely. It was not yet a different language—merely a vulgar variety of Latin. 8 M. Van Uytfanghe, "The consciousness of a linguistic dichotomy (Latin-Romance) in Carolingian Gaul: the contradictions of the sources and of their interpretation", Latin and the Romance Languages in the Early Middle Ages, ed. Wright, pp. 114‫־‬ 129; McKitterick, "Latin and Romance: an historian's perspective", ibid., pp. 130‫־‬ 145. 9 J.L. Nelson, "Literacy in Carolingian government", The Uses of literacy in Early Medieval Europe, ed. McKitterick, pp. 258-296.

deployed". T h e Admonitio generalis (789) 10 a n d the equally f a m o u s Epistola de litteris colendisu reveal to what extent the p r o g r a m m e of "correction" of public language aimed at the correct performance of liturgy, and ultimately, at the moral correction of the populus christianus. Carolingian M a n d a r i n , therefore, was not only the language of government, but also that of Christian cult. T o a large extent, the distinction is an artificial one, for the correct performance of the liturgy in royal monasteria played a crucial role within the governance of the Christian realm. In this sense, the language of official prayer could be considered as an administrative language, which goes a long way towards explaining the intense concern of Carolingian rulers with the "regularity" of liturgy a n d its language. Without wishing to detract from the importance of things R o m a n , I should like to make a case for the K o r a n and the Arabic language as a model for comparison when studying the Carolingian situation. In both cases, we are speaking of a M a n d a r i n language grounded in a sacred text; the fear voiced by the Epistola de litteris colendis is that of having to depend upon a clergy no longer able to penetrate the mysteries of the language of Scripture, and on those w h o are well-intentioned but incapable of p r o n o u n c i n g the right prayer-texts. G o o d prayer was phrased in the language of Scripture. T h e council of Frankfurt (794) took an exceptional stance when it argued that those who thought G o d could only be prayed to in three languages were wrong, for G o d listened to all m e n asking for the right thing, whatever tongue they spoke. 12 T h e m u c h more c o m m o n conviction was that G o d wished to be addressed properly in sacred language. W e are also a long way off from Byzantium, where, as Professor Oikonomides explained, "the C h u r c h , always attached to the idea of direct communication with the flock, did not participate in the bilingualism game; the C h u r c h addressed the faithful in as direct a way as possible." This no longer held true in the West from the late 10

MGH, Capitularia Regum Francorum I, ed., A. Boretius and V. Krause (Hannover, 1883), no. 22, pp. 52-62. " MGH, Capitularia I, no. 29, pp. 78 79. M. de Jong, "Carolingian monasticism: the power of prayer", The New Cambridge Medieval History II, ed. R. McKitterick (Cambridge, 1995), pp. 622 653. MGH, Capitularia I, p. 79: "Et bene novimus omnes, cjuia, quamvis periculosi sunt errores verborum, multo periculosiores sunt errores morum." 12 Council of Frankfurt (794), c. 52, MGH, Concilia. II 1, ed. A. YVerminghoff (Hannover, 1906), no. 10, p. 171. M. de Jong, In Samuel's Image. Child Oblation in the Early Medieval West (Leiden/New York/Köln, 1996), pp. 232 245.

eighth century onwards. T h e growing divide between those whose mastered "grammar 5 ' and those w h o remained stuck in "rusticity" was not a clear-cut one r u n n i n g between the clergy a n d the laity. 13 Religious communities were by no means literate as a whole. Along with M a n d a r i n Latin, a clerical elite emerged which defined itself as scholastici—at whichever stage of training they might h a p p e n to be. T h e y were often those w h o had entered monasteries at an early age, a n d had therefore received a lengthy and thorough training. T h e a n o n y m o u s report of the council of A a c h e n (816), usually called Statuta Murbacensia, explains what the scholastici were supposed to learn: first the psalms a n d hymns, then the complete text of the Rule, followed by Scripture, patristic c o m m e n t a r y a n d saint's Lives. Above all, they should speak Latin at all times. Quite a curriculum, certainly c o m p a r e d to the one expected of "those who converted to the monastic life from the lay state" (i.e. newcomers of a more advanced age), who only learned the Pater JVoster and the Creed, some penitentials Psalms, and, if they were u p to it, the rest of the Psalter. 14 T h e deep gulf between these two classes of monks, the "grammarians" a n d the "rustics", is illustrated by one of Ekkehard's stories about the Carolingian monks of St Gall. Admittedly, this is a story written in the eleventh century, but circumstances, linguistic a n d otherwise, h a d not c h a n g e d all that m u c h in St Gall since the Carolingian period. If circumstances required it, the literate élite could switch effortlessly into Latin. This was what the learned threesome Notker, Tuotilo and Ratpert did, when they were eavesdropped on during one of their nightly discussions of Scripture. W h e n the officious m o n k Sindolf listened at the window, hoping to catch any " b a d talk", Tuotilo got wind of this and immediately addressed his companions in Latin, "of which Sindolf understood nothing whatsoever". 15 Within Carolingian monasteries with their strongly textual orientation, the scholastici were the undisputed élite. Letter collections are full of references to what one could call "old boy networks": "when we grew u p together"; "when we studied together"; "when we were together at the court". It was within this group of scholastici that the 13

See now M. Innes, "Mentality, orality and literacy in an early Medieval society", Past and Present 158 (1998), pp. 1.36‫־‬ 14 Statuta Murbacensia, c. 2 and epilogue, ed. J . Semmler, Corpus Consuetudinum Monasticarum, I (Siegburg, 1963), pp. 442 and 449. Ibid., c. 2, p. 442. Ifl Ekkehard, Casus s. Gatti, c. 35, ed. H.F Haefele (Darmstadt, 1980), pp. 80-82.

early initiatives at developing M a n d a r i n were further elaborated, in the cloister as well as at the court. As for the M a n d a r i n s themselves, the scholastici, they formed a mixed group. The Aachen council of 817 indeed decreed that monasteries should only have schools for oblates, but this was by no means universally obeyed. T h e fact that the decree needed be issued at all shows how m u c h pressure there was on monastic schools to open their gates to those looking for literate training. T h e y h a r b o u r e d young oblates, clerics, and also laymen; and even oblates were not confined to the monastery walls, for they b e c a m e bishops, courtiers a n d sometimes runaways, whose literate skills must have been m u c h in demand. 1 6 I will not go into the perennial issue of the literacy of the laity. Needless to say, I do not think that G e r m a n i c people were too stupid to learn Latin. O n the other hand, on the early medieval Continent there was but one literate culture, which was both Christian and clerical, which m e a n t that all those who became literate above the level of being able to read or understand a charter or a capitulary had to immerse themselves in a M a n d a r i n of an inherently religious nature. T h e curriculum of 816 was valid for all scholastici, be they destined for the cloister or secular government. Perusing the correspondence of Alcuin's most influential student, Hrabanus Maurus, I was again struck by the need to redefine the concept of "administrative language": even exegetical writing could to a certain extent be "administrative", in the sense of belonging to the sphere of government. 1 7 H r a b a n u s was so m u c h in d e m a n d as a biblical commentator that at times he even lost track of his own writings, not being sure whether he had "covered" something or not. For kings he mostly wrote on texts suitable to kings, notably the historical books of the Old Testament. H e m a d e several references to the lectores who were part of the retinue of Louis the G e r m a n and Lothar I; they assisted the rulers in correcting the work he sent them. T h e king acted as the final judge: to him, exegetical writing was sent ad legendum et ad probandum, but he was surrounded by peútissimi lectores who might find fault as well. These "readers" seem to have read

lb

M. de Jong, In Samuel's Image, Child Oblation in the early medieval YVest (Leiden/New York/Köln,' 1996), pp. 232.45‫־‬ 17 M. de Jong, "Old law and new-found power: Hrabanus Maurus and the Old Testament 1 ', Centres of Learning. Learning and Location in Pre-Modem Europe and the Near East, eds. H.J.W. Drijvers and A.A. MacDonald (Leiden/New York/Köln, 1995), pp. 161-176.

H r a b a n u s " work to the king, discussing a n d criticising it while so doing; this is clear from one of his dedicatory letters to E m p e r o r Lothar, where he wrote: Order this to be read in your presence, and if you discover something that is not correctly explained because of the weakness of my understanding, or distorted by scribal errors, make your learned readers correct it, and thus you will be rewarded forever with your just reward in Heaven by Christ, the Lord of all, for your noble struggle and for having corrected me.18 This passage nicely shows which intermediaries intervened between the author of biblical exegesis a n d its royal recipient: not only scribes but also lectores helped to convey the message. In addition, it illuminâtes a court culture in which rulers had Latin texts read to them by professional readers, pondering a n d censuring what they h e a r d while the reading session progressed. H r a b a n u s considered textual criticism to be Lothar's bonum certamen, his "noble struggle". T h e military connotations of this expression must have pleased an e m p e r o r who c o m p a r e d himself to that other soldier of G o d , J o s h u a . T h e authority to j u d g e (probare) biblical scholarship was one of the hallmarks of a legitimate ninth-century ruler. W h a t H r a b a n u s did, in his own words, was to explain the "old law" to kings who by definition h a d to be conversant with its p r o p e r interpretation. This is what Carolingian kings needed their Mandarins for; this is why they themselves needed to be fully literate, a n d why their efforts to improve the written language centred upon the language of Scripture. T h e distinction between secular administrative language on the one h a n d a n d that of liturgy a n d exegesis on the other is therefore partly artificial. Both used scriptural models, a n d administrators were often also scholars. It is equally difficult to distinguish between "lay" a n d "clerical" literacy, for even if lay aristocrats aspired to become literate, they did so according to an ecclesiastical

18 Hrabanus Maurus, Epistolae no. 50, ed. E. Dümmler, MGH, Epistolae V (Berlin, 1899), p. 505: "Iubete illud coram vobis legi et si quid in eo propter tenuitatem sensus mei non rite prolatum vel scriptorum vitio depravatum conspexeritis, per vestros eruditos lectores facite illud corrigi, et sic vobis merces condigna pro vestro bono certamine et nostra simul correctione a Christo omnium domino perpetualiter recompensabitur in caelis." Cf. J . Fleckenstein, "Über Hrabanus Maurus. Marginalien zum Verhältnis von Gelehrsamkeit und Tradition im 9. Jahrhundert", Tradition als historische Kraft. Interdisziplinäre Forschungen zur Geschichte desfrüherenMittelalters (Berlin/New York, 1989), pp. 204 213.

model. As J a n e t L. Nelson has pointed out, in the Carolingian world Latin served as the definition of the ruling cadre, and as a "litmus test of political loyalty". 19 Its symbolic qualities centred u p o n the notion of the New Israel of the Franks; the closer to the court one aspired to be, the greater b e c a m e the need to become literate in a " M a n d a r i n " fashion. With the disappearance of the Carolingian kingd o m s the close ties between the royal courts a n d their monastic dependencies became severed; it was now u p to former royal abbeys to keep alive the M a n d a r i n tradition, and to authors such as Notker of St Gall to dream of former days of glory, when monks still roamed the corridors of power.

19

Nelson, "Literacy in Carolingian government", p. 272.

COMMUNICATING

HOLINESS

Lennart Rydén

A hagiographer is an author who writes about saints and their virtues, struggles, miracles and other achievements. But apart from that, how should a hagiographer be described? Perhaps one may say that he is a historiographer, an encomiast, a homilist and an entertainer in one person, although the proportions may differ. H e is a historiographer because he wants to save the m e m o r y of the saint for the future. H e is an encomiast because he wants to p r o m o t e his cult. H e is a homilist because he wants to show us a model way of life. H e is an entertainer because he wants to present the life of the saint in a pleasant form, and obviously enjoys telling a good story. Except for the limitations inherent in his subject, a hagiographer knows few literary restraints. H e can write as he likes and, if he has talent, give free rein to his creativity. For example, nobody expects him to tell the whole truth, i.e. the truth as we understand it. H e does not have to care about chronological coherence (although sometimes he does care), nor is he b o u n d to follow fixed rules concerning style or composition. If his own material is scanty, he can borrow from another saint's Life or from some other source. In addition he can apply any literary device that appears convenient and useful to him. This makes hagiography a highly flexible genre, which easily assimilates the forms and motifs of other literary genres, secular as well as Biblical and Christian. It further means that hagiography, which is a m o d e r n term, not a medieval one, covers a wide literary spectrum ranging from the collection of the sayings of the Fathers to the Passio, the saint's Life, the edifying short story and the miracle account. T h e most c o m m o n and most important of these is of course the saint's Life, in Greek most often called bios kai politeia, "Life and C o n d u c t " . This particular form is a literary genre in its own right, related to biography as m u c h as it belongs to hagiography, although there are characteristic differences. Unlike m o d e r n biography it takes no interest in psychological analysis and source criticism, although

it lays claim to credibility. With ancient biography it shares the distinction between biographical data on the one h a n d and deeds a n d way of life on the other. It further shares its tendency to stylize, praise a n d moralize at the expense of analysis. But whereas ancient biography is about extraordinary men of the world, saints' Lives are, in principle, about humble servants of G o d , who works through them as his chosen instruments. It is also worth observing that whereas there are thousands of saints' Lives, the n u m b e r of ancient biographies is small, and that whereas ancient biography is mainly concerned with men, hagiography also considers women. However, what distinguishes a saint's Life from secular biography more than anything else is that it constantly refers to the Bible a n d the Christian world of thought. This means that the saint and his achievements, or hers, are not seen in isolation but in the context of God's plan for the salvation of mankind. 1 T h e Gospels, especially those of M a t t h e w a n d Luke, are also a kind of biography, and the saint's Life has m u c h in c o m m o n with them. Like them its readership is not limited to an educated élite but includes all Christians, and therefore it does not have to express itself in a traditional literary style, although this also occurs. However, unlike the hagiographers, the evangelists did not stress asceticism— Christ was above asceticism—nor were they fascinated by Christ's life as such. W h a t interested them in the first place was Christ's teaching a n d the fact that he was the Son of G o d , who fulfilled the prophecies of the Old T e s t a m e n t and saved mankind through his death on the cross. N o saint's Life claims to deliver a message as revolutionary as this; on the contrary they repeat and elaborate the same Christian message, every saint being an imitator of Christ, as it were. Photios, the great patriarch a n d literary critic of the ninth century, deals with hagiography in some ten codices of his Bibliotheca (nos. 96, 198-200, 252~8). 2 With one exception they a p p e a r in two groups, which indicates that Photios regarded these writings as something special, although he does not try to make literary distinctions.

' O n these and other characteristic features of hagiography, which have more to do with content than with form, see the excellent article by M. van Uytfanghe, "L'hagiographie: un genre chrétien ou antique tardif?", Analecta Bollandiana 111 (1993), pp. 135-88, esp. pp. 171-5. 2 See Photius, Bibliothèque, ed. R. Henry, II (Paris 1960), pp. 48-63; III (Paris, 1962), pp. 95-9; VII (Paris 1974), pp. 207-28; VIII (Paris 1977), pp. 8 - 4 0 .

Evidently he did not see any literary difference between the Life of a p a g a n philosopher and that of a Christian saint. His main interest is language and style. H e appreciates good style but also accepts low style if the g r a m m a r is not too bad. In his opinion, even if the style is low a sensible and pious m a n can always benefit from the edifying contents (codex 199). T h e a p p e a r a n c e of Symeon Mctaphrastes' Menologium in the late tenth century was a turning point in the history of Byzantine hagiography. Eschewing stylistic exaggeration, he created a high, but not too high, level which not only suited the stylistic fashion of his own time but also satisfied the taste of following generations. As recent research has revealed, Symeon a n d his collaborators worked almost like scholars, tracing manuscripts, c o m p a r i n g different versions and establishing new ones on the principles of reasonable contents and moderately rhetorical s t y l e . I n this shape hagiography b e c a m e academie, as it were, m a n n e r e d and bland to our mind, yet extremely useful in the new intellectual climate that developed after the turn of the millenium. It is a great pity that Symeon did not express his aims and methods in a preface to his great work. Fortunately Michael Psellos' eleventh-century e n c o m i u m of Symeon Metaphrastes shows how a nearly c o n t e m p o r a r y literary critic j u d g e d his achievement. 4 Psellos is much less tolerant than his forerunner Photios. H e says that those w h o wrote saints' Lives before Symeon were not up to their task. Their compositions were ludicrous; m a n y people found them unbearable. T h e unattractive sentence structure, the lack of coherence, the c o m m o n vocabulary, all this m a d e t h e m unpleasant to listen to, even repellent. In these writings the miracles of the saints, their struggles and victories were ridiculed rather than honoured. Everybody criticized them but nobody did anything to remedy their shortcomings, with one exception: Symeon Metaphrastes. Symeon was in full comm a n d of the artifices of rhetoric. Yet, when he revised saints' Lives,

3

See W. Lackner, "Zu Editionsgeschichte, Textgestalt und Quellen der Passio S. Polyeucti des Symeon Metaphrastes", Byzantios. Festschrift fiir Herbert Hunger zum 70. Geburtstag, eds. W. Hörandner, J. Köder, O . Kresten, E. Trapp (Vienna, 1984), pp. 221 31, esp. 230 1: and E. Peyr, "Zur Umarbeitung rhetorischer Texte durch Symeon Metaphrastes", Jahrbuch der österreichischen Byzantinistik 42 (1992), pp. 143-55, esp. p. 155. 4 Oratio in Sanctum Symeonem Metaphrastern, in Michaelis Pselli Orationes Hagiographicae. ed. E.A. Fischer (Stuttgart, 1994), pp. 2 6 9 8 8 ‫ ־‬, esp. lines 184-99 and 257-65.

he handled them so judiciously that, on the one hand, the educated listener could enjoy the rhythm a n d sonorous vocabulary and, on the other, the c o m m o n m a n was carried away by the clarity a n d pleasant flavour, while both were impressed by the h a n d s o m e form a t and convincing presentation. In short, Symeon used a type of diction and style that appealed to his contemporaries. H e is not to be c o m p a r e d with the ancient orators a n d historians but with the evangelists. Psellos is apparently saying that just as the evangelists had used a language which fitted the audience they were addressing, so Symeon Metaphrastes adjusted the language of hagiography to suit the needs and tastes of his own time. But he does not conclude that the Gospels would need revision as well, evidently because he regarded t h e m as the unchangeable foundation, while the purpose of hagiography is to bear witness to the continued presence of the divine. Implicit in all this is that Psellos, unlike Photios, regarded the saint's Life as a genre founded on the Gospel rather than as a variant of ancient secular biography. With occasional exceptions, 5 hagiographers begin their compositions with a prologue—Symeon Metaphrastes, too, introduced his Lives with short prefaces, although he failed to provide the Menologium as a whole with an introduction. T h e prologue served the purpose of putting the biographee in a saintly context and the audience in the right m o o d to receive the message. T o achieve this end the hagiographer had at his disposal an arsenal of commonplaces, most of them inherited from the prologues of ancient secular literature or at least very like those appearing there. It is true that thanks to the large supply of commonplaces a hack could produce an acceptable prologue without m u c h effort. This has given hagiographical prologues a reputation for being high-flown, stereotyped and repetitive, and therefore boring. But it is equally true that m a n y hagiographers used c o m m o n p l a c e s in a personal a n d inventive way. O n closer inspection therefore, here as in hagiography in general, they display m o r e individuality than expected. T h u s prologues may have something to say on hagiographers as communicators.

5

For the period under consideration, see, e.g., the Lives of St Epiphanios, bishop of Salamis in Cyprus (fifth-century?), St Nicholas of Sion (sixth-seventh-century) and St Philaretos the Merciful (ninth-century). T h e authors of the first and the third Life had special reasons for not adding a prologue.

From this point of view I have perused some fifty Late Antique a n d Early Byzantine hagiographical prologues written between the middle of the fourth and the beginning of the tenth century. In order to make characteristic features stand out more clearly I have compared them with a few historical prologues ranging from Procopius of Caesarea (sixth-ccntury) to Niketas Choniates (early thirteenthcentury), which may be summarized as follows. 6 Like Herodotus, Procopius begins by introducing himself, whereupon he presents his topic and points out the value of preserving the m e m ory of important events for the future. H e says that he took part in Belisarius' military campaigns as his assessor a n d that for this reason he is better qualified to describe them than anyone else. H e contrasts historiography, which is concerned with truth, to poetry a n d rhetoric, which focus on tales a n d stylistic brilliance. In a m a n ner reminiscent of Thucydides, he underlines the superiority of his own epoch in comparison with the heroic age, not only on account of the great events that have taken place recently but also because of the progress of war technology.‫׳‬ Agathias (sixth-ccntury) begins by saying that m e n risk their lives for h o n o u r because they know that history makes h o n o u r lasting. H e further states that history is superior to political science, the latter making m e n do or avoid things only on order, while history implants virtue into the soul with exciting examples. Historians are therefore great benefactors of society. For his part, Agathias, who identifies himself carefully "according to the usual practice of historians", has always been a poet, but since such great wars have been fought in his lifetime he feels that he must write about them, preventing them from falling into oblivion, thus making himself useful. H e has been encouraged by his friend Euthychianos, according to w h o m poetry and history are brother a n d sister, divided only by metre. H e wishes

6

An important study of the relations between historiographers and hagiographers as communicators in the sixth century has been published by D.R. Reinsch: see id., "Autor und Leser in frühbyzantinischen hagiographischen und historiographisehen Werken", XVIIIth International Congress of Byzantine Studies. Major Papers (Moscow 1991), pp. 400 14. He concludes: "Was die Gesamtheit der Appell-Struktur der Texte betrifft, so scheinen wir einigen Grund zu haben, Hagiographie und Historiographie vielleicht enger zusammenzurücken als wir das gemeinhin tun." Be that as it may, a close comparison covering a longer period would certainly be useful. 7 Ed. .J. Haury, I (Leipzig, 1962)", pp. 4' 7.

that he will be u p to his undertaking a n d assures that he is giving truth pride of place, and c o n d e m n s those w h o praise contemporary rulers at the expense of their predecessors. H e appreciates Procopius, however, whose Wars he summarizes before starting his own narrative where Procopius left off. 8 T h e prologue of Theophylact Simocatta (seventh-century) is a piece of pure rhetoric, in its first half depending on sophistic ideas, in the second on the historian Diodoros, a n d it lacks immediate relevance to the following history. Like Agathias, Theophylact connects history, called teacher of all m e n and credited with spellbinding power, with poetry. 9 T h e o p h a n e s the Confessor (ninth-century) says that he is finishing the world history of George the Synkellos, which broke off with the reign of Diocletian. Aware of his ignorance and incompetence he performs his task reluctantly and only because he does not want to disobey George, who asked him on his deathbed to bring his history to completion. H e claims to have a d d e d nothing of his own but only to have put in order, particularly in chronological order, the material he has collected f r o m ancient historians. H e thinks that reading about the deeds of the ancients yields great benefit. T h e Lord should be credited with any merit that his book might have, but the responsibility for its flaws is his own. H e consoles himself with the saying "to do as best one can is dear to God". 1 0 George the M o n k (ninth-century) rebukes the Greek poets and historians for showing off their elegant diction, which most people cannot understand, rather than telling the truth and pointing out what is useful. Nonetheless he has studied both these works and those written by Christian authors a n d on this basis created an artless, concentrated and truly useful composition. Having summarized its contents and explained the disposition, he asks for his readers' intercession a n d makes a remark reminiscent of T h e o p h a n e s , namely that if his composition should prove useful, G o d alone should be given

8

Ed. Ed. übersetzt ‫ "י‬Ed. pp. 3 - 4 . 9

R. Keydell (Berlin, 1967), pp. 3 - 5 . C. de Boor (Leipzig, 1887), pp. 36-8; see Theophylaktos Simokates. Geschichte, und erläutert von P. Schreiner (Stuttgart, 1985), pp. 243-4, notes 52 and 54. C. de Boor, I (Leipzig, 1888; repr. Hildesheim and New York, 1980),

credit, but if he fails, he asks his readers to forgive him because he is an a m a t e u r both as a writer a n d as a scholar." Genesios (tenth-century) expresses the usual appreciation of history. O n the one h a n d he says that he is writing in order to benefit himself as well as future generations, on the other that Constantine Porphyrogennetos has asked him to write down the events from the reign of Leo V (813~20), since they have not yet been treated by historians. In other words he introduces himself as a continuator of T h e o p h a n e s the Confessor, although his method is different. Whereas T h e o p h a n e s collected his material from books, Genesios indicates that he is building on oral tradition. 1 2 Leo the Deacon (late tenth-century) praises history because it provides models of behaviour and saves the deeds of men from oblivion. H e says that he wants to describe the remarkable events that have taken place in his age, not only wars but also earthquakes, thunderbolts a n d other catastrophes which according to m a n y forebode the Second Coming, telling this as a lesson for future generations, unless Providence brings about C o n s u m m a t i o n before then. H e prays that he may prove himself to be up to his task, whereupon he gives his name, father's n a m e and place of birth. H e declares that he puts truth in the forefront, like Procopius quoting the saying that brilliant style becomes rhetoric, the telling of tales poetry and truth history. H e will only write down what he has seen with his own eyes or what he has learned from eyewitnesses. 13 This prologue is remarkable for its apocalyptic perspective. O n e further notices that, in the tradition of Agathias, Leo identifies himself with great precision, whereas he fails to excuse himself for his style or feign to be incompetent. J o h n Skylitzes (eleventh-century) continues where T h e o p h a n e s the Confessor finished. In his prologue he praises George the Synkellos a n d his continuator T h e o p h a n e s for their impartiality and unaffected style; he says that they almost reached the essence of the deeds they described. After them no one m a n a g e d to achieve anything similar. Later historians were superficial a n d useless; they did their readers 11 12 13

Ed. C. de Boor 1 (Leipzig, 1904), pp. 1 5 . Eds. A. Lesmüller-Werner and H. Thurn (Berlin and New York, 1973), p. 1. Ed. C.B. Hase (Bonn, 1828), pp. 3-5.

h a r m rather than edifying them. H e mentions a few of them by n a m e (among them Genesios, though not George the Monk), and accuses them of promoting their personal purposes u n d e r the guise of history. T h e y therefore differ from each other, confusing the reader. His own m e t h o d is, on the one h a n d , to remove what is partial, to smooth out discrepancies and reject absurdities, on the other to collect plausible information from his predecessors as well as from old contemporaries. His aim is to create a reliable, readable historical survey to be used by future generations. 1 4 Skylitzes does not introduce himself, although his criticism of others shows that he is self-confident. H e does not feign ignorance or literary incompetence. Michael Attaleiates (late eleventh-century) dedicates his history to the E m p e r o r Nikephoros III Botaneiates (1078-81). In his dedicatory letter he says that he has written a c o n t e m p o r a r y history in which he has not only described the events but also explained the reasons for what h a p p e n e d . T h e virtues and shortcomings of the leading characters have been pointed out and the narrative has been spiced with memorabilia. In the prologue p r o p e r he stresses the usefulness of history in traditional terms. Since history is such an important thing he has, despite his m a n y duties, decided to write down and save from oblivion a few events that he has witnessed himself. H e describes his narrative as simple and concise, for he is not delivering a speech requiring rhetorical technique but writing history, which is above deceptive and irrelevant talk. 13 John Zonaras (twelfth-century), a court official turned monk, insists that he has been persuaded to write a new world history by his friends, who complain that the existing ones are either too circumstantial or too short or decked out with speeches a n d dialogues or, the other way round, tainted with barbaric language and bad syntax. After m u c h hesitation because of the hard work needed, he accepts the task, admitting that leisure may be dangerous a n d thinking that this undertaking may bring him spiritual benefit. H e warns the reader that he will not always be as precise as one could wish, since he does not have all the necessary literature at h a n d and the

14

Ed. H. T h u m (Berlin and New York, 1973), pp. 3 - 4 . Eds. VV. Brunei de Presle and I. Bekker (Bonn, 1853), pp. 3 - 6 (letter), 7 - 8 (prologue). 15

sources available often contradict each other. His style will be inconsistent, since his sources have been written in various styles. T h e prologue ends with a summary of the contents, in which the Athenian democracy is conspicuous by its absence.' 6 Zonaras writes very good Greek and does not blame his reluetance on literary incompetence but on the conflict between unworldly leisure and scholarly work. It is a personal and fitting introduction to his narrative. A n n a K o m n e n a (twelfth-century) states that only history is able to stem the flow of time that carries everything away, things of importance and things of no account alike. She proudly introduces herself, telling of her parents and education and learning without feigning the slightest modesty, and then defines her topic. She says that she is afraid of being regarded as an encomiast and also of being misunderstood if she happens to make a critical remark on her father. With a tacit reference to Polybios she points out that one should not shy away from criticizing one's friends and praising one's enemies. She further says that her account is based on the events themselves and on those who have seen them. Lastly she explains that her history is intended to complete her husband Nikephoros Bryennios' unfinished biography of her father Alexios I. T h e latter and greater part of the prologue is devoted to a lamentation of the death of her husband and her other misfortunes. 1 7 A n n a wants to save the m e m o r y of her father for the future, but unlike most historians she does not dwell on the moral benefit of history. H e r lamentations may be typical of a widow but she does not c o m m e n t on the fact that she, a w o m a n , is performing a typically male task. J o h n Kinnamos (late twelfth-century) begins by saying that the ancients also appreciated history. W h a t time brought forth and threatened to hide again they saved f r o m oblivion in their books as if on everlasting paintings. H e regrets that he lacks both the learning and the spare time needed to write history. Nevertheless he does not want to pass over in silence the events of his lifetime. Since Alexios' reign has been adequately treated by others he concentrâtes on J o h n II and M a n u e l I, especially on the latter, in whose campaigns he took 16 17

Ed. L. Dindorf I (Leipzig, 1868), pp. 1 11. Ed. B. Leib, I (Paris, 1967), pp. 3 8.

part as a a youth. In fact, he is better equipped than anyone else to give an account of Manuel's reign. 18 Like Procopius, K i n n a m o s is convinced that he is the right m a n for the j o b , a n d like him he does not bother to speak about language and style. T h e usual author's dilemma is only hinted at in passing (lack of time and learning). Niketas Choniates says that history is a c o m m o n utility from which those w h o wish can learn m a n y good things. W h a t people older than Tithonos could have told, had they lived, those who love history can tell even before they reach m a n h o o d . History resurrects the dead. Niketas therefore wants to describe the great events that have taken place in his lifetime and a little earlier. H e continues by saying that history d e m a n d s a clear style, without artifice, and that he intends to write accordingly. History should also be brief. T h e ideal of history is truth, while that of rhetoric is impressive style and that of poetry the art of telling tales. H e will begin with the death of Alexios I, where his predecessors finished. Historical works are like links in a chain. 1 9 Surprisingly enough Niketas fails to mention Kinnamos, although they start at the same point in time. His prologue is elegant but commonplace. With the exception of A n n a K o m n e n a , all these authors say that history is of moral benefit, that it states examples and provides models, although George the M o n k accuses the Greek historians of showing off rather than edifying. M a n y also stress its ability to save the past for the future; this is the main driving force behind A n n a K o m nena, at least according to herself, a n d probably also for most of the other twelve, although they may not say so (Theophanes, Genesios and Zonaras maintain that they are writing at the request of others). A further characteristic of history is that it claims to tell the truth in explicit or implicit contrast to rhetoric and poetry—George the M o n k says that rhetoric perverts t r u t h — a l t h o u g h Agathias and Simocatta regard history as related to poetry; George the M o n k also seems to think that ancient Greek history should be put on a par with poetry. T h e claim on truth implies that the historian should

IB 19

Ed. A. Meineke (Bonn, 1836), pp. 3.5‫־‬ Ed. J.L. van Dieten (Berlin and New York, 1975), pp. 1 - 4 .

express himself clearly a n d concisely. It also urges the historian to account for his sources, some saying that they are telling what they have seen with their own eyes, others that they report what they have heard from trustworthy witnesses, still others that they have studied old books, which they have checked against each other, or they claim to have combined various kinds of sources. Characteristically, unlike m o d e r n historians they do not thank their informants. M a n y historians explain that they begin where their predecessors stopped; the predecessors do not have to be contemporary (Genesios, Skylitzes). They regard themselves as links in a historical chain (Niketas Choniates). Some praise their predecessors (Agathias, T h e o p h a n e s the Confessor), others criticize them for being contradictory, superficial or too m u c h concerned with style (Skylitzes, Zonaras). Some explain how they have arranged their material (Theophanes the Confessor) or summarize the contents (George the Monk, Zonaras). T h r e e historians identify themselves carefully "according to the usual practice of historians" (Agathias, presumably referring to Herodotus a n d Thucydides). Some also display a remarkable self-esteem, while others are more self-effacing. 20 Procopius and K i n n a m o s claim to be more qualified than anybody else to handle the topic they have chosen. In addition Procopius maintains that his own age is superior to that of H o m e r . A n n a K o m n e n a expresses shameless pride over her ancestry, education and learning. She complains of her widowhood in a way that is reminiscent of traditional female m o u r n i n g but is far from excusing herself for writing about topics that are usually treated by men. Others, on the other h a n d , assume a humble attitude. This applies particularly to T h e o p h a n e s the Confessor and George the Monk. T h e f o r m e r speaks of his ignorance and incompetence, reducing his achievement to having put a h e a p of inherited material in due chronological order, and G o d alone should be given credit for any merit of his book. His end phrase ("to do as best one can is dear to God") is typical of hagiography. George the M o n k calls himself a humble servant of G o d , a sinful m a n and an amateur, using a phrase reminiscent of the Life of St Eutychios. 21

2

" For a penetrating analysis of this phenomenon, which changed with the centuries, see J . Ljubarskij, "Writer's 'Intrusion' in Early Byzantine Literature", XVIIIth International Congress of Byzantine Studies. Major Papers (Moscow, 1991), pp. 433 56. 21 George the Monk, ed. de Boor, pp. 5, 12- 13; Life of Eutychios, ed. C. Laga (L Ū uvain, 1992), lines 2289-90 and 2788 90.

It should be borne in mind that these protestations do not always reflect the historian's real attitude. George the M o n k may give himself an air of humbleness, but his criticism of Greek literature is anything but humble. A n d w h e n he claims to be brief, few readers would agree that he is sincere. N o r is it easy to agree with Niketas Choniates when he speaks about clarity a n d says that he intends to avoid difficult phrases and daring vocabulary. Obviously, his words should be seen in the light of the opposition between history and rhetoric, a n d not as referring to what we would define as clear and straightforward style. Furthermore, all historians speak of truth but nobody of research in our sense, except that some say that they (like Symeon Metaphrastes) have c o m p a r e d their sources a n d tried to eliminate contradictions. However, all thirteen, except Theophylact Simocatta, have produced prologues that are relevant for what follows. T h e y contain plenty of topoi, but the words are not always empty. T h a t hagiography is related to historiography is evident from the mere fact that hagiographers, often in terms borrowed from historiography, declare that they want to save the deeds of great m e n from the depths of oblivion, that they intend to tell the truth and nothing but the truth, a n d that they want to edify the reader. Like the writers of history they distance themselves from poetry and rhetoric, since these either deviate from truth or blur it. For instance, the a u t h o r of the Life of St Eustratios (ninth-century) promises not to extol the saint with false ideas, for this would be impious and unworthy of the Christian way of life. 22 But since their subjects are different they also distance themselves from historiography. In his prologue to the Hist0rìa religiosa, T h e o d o r e t of Cyrrhus (fifth-century), just like another historian, compares a saint's Life with a remedy for oblivion and a support for the memory. But then he asks: Is it not absurd t h a t poets a n d historians write a b o u t great deeds d o n e on the battlefield, that tragedians exhibit on the scene misfortunes that one had better keep secret, that others waste their words on comedy a n d fun, whereas we consign to oblivion men w h o have demonstrated dispassionateness in a mortal body subject to passion, and rivalled the nature of the angels? 23 T h e a u t h o r of the Acts of Sts David, Symeon a n d George (ninth-century) approves of history insofar as it 22

Ed. A. Papadopoulos-Kerameus, Anale/cta Hierosolymitikes Stachyologias (St Petersburg, 1897), pp. 367 400, esp. pp. 368, 9 - 1 0 . 23 Eds. P. Canivet and A. Leroy-Molinghen, I (Paris, 1977), p. 126, 14-22.

is impartial and sticks to the truth, since this kind of history prevents its readers from repeating old mistakes and helps them to do what is right. But he strongly disapproves of historians w h o add myths and fiction; these, he says, paint an indelible picture of their own folly. 24 Another hagiographer of the ninth century, the author of the Life of St John Psichaites, is more negative. H e doubts that poetical or historical tales about victories a n d deeds of valour can stimulate those who listen to them to do anything similar; they only serve the purpose of saving the events from disappearing from m e m ory, whereas the deeds of pious m e n , written down, resound like trumpets, inspiring the reader to similar achievements, persuading him to spurn the visible world and preparing him for the fight against the invisible enemies. 2 5 According to this author, therefore, historiography fails to produce even worldly edification, whereas spiritual edification is powerful. As mentioned, hagiographers claim that they are telling the truth, in contrast to rhetoricians, w h o can prove that black is white, and to those who pass on myths a n d fiction. T h e y declare that they are telling nothing but what they have seen themselves or heard from reliable witnesses. O f t e n these are obviously fictitious, as. e.g. J o h n the Deacon in the Life of St Symeon the Fool by Leontios of Neapolis (seventh-century) and the Egyptian M e n a s in the same author's Lfe of St John the Merciful.2(‫ י‬Unlike the historians, however, some also ask their readers to have faith in what they read, a m o n g them Cyril of Skythopolis in the Life St Euthymios (sixth-century) and the author of the Lfe of St Stephen the Younger (ninth-century). M a n y more, beginning with Athanasius, a u t h o r of the Life of St Antony the Great, ask them not to disbelieve, although this in fact is a sign that they should do so.27 Others, like the a u t h o r of St Symeon the Stylite the Younger (c. 600) even refer to the word of Jesus, "I am the t r u t h " (John 14:6).28 In hagiography truth is subordinate to edification. This may be exemplified with the updating of Symeon the Fool from the reign 24

Ed. I. van den Gheyn, Analer ta Bollandiana 18 (1899), pp. 211-59, esp. 211 12. T h e edition is in need of correction, particularly in the first paragraph; the author's train of thought is therefore sometimes difficult to follow. 25 Ed. P. van den Ven, Le Muséon 21 (1902), pp. 103-25, esp. 103, 1-14. 2li See the Life of Symeon, ed. L. Rydén (Uppsala, 1963), pp. 125, 16-24 and the Life of John the Merciful, ed. A.J. Festugière (Paris, 1974), pp. 345-6. 27 O n this topos, see A J . Festugière, "Lieux communs littéraires et thèmes de folk-lore dans l'Hagiographie primitive", Wiener Studien 73 (1960), pp. 123 52, esp. 133-5. 28 Ed. P. van den Ven (Brussels, 1962), lines 30-31 of the prologue.

of Justinian I (52765‫ )־‬to that of Maurice (582-602), which is explained by the fact that Leontios in the two of his prologues that have been preserved expresses the opinion that the Lives of contemporary saints have a m o r e edifying effect than those of old ones. T h e attitude of the hagiographers towards the encomium, a rhetorical genre, is ambivalent. O n the one hand they associate the encomium with beautiful but empty or deceptive words, whereas hagiography is concerned with truth. T o quote the Ltfe of St Symeon the Stylite the Younger, "Naturally, we despise a style that is elaborated for the purpose of beauty; for this we do not care at all. W e are concerned with truth that overcomes everything. T h e words of truth are sweet, for they are filled with grace. T h e y do not need the artificial ornamentation of the rhetoricians or a tongue capable of making great things small and conferring greatness on that which is slight. As the saying goes, the word of truth is simple." 2 9 O r they remark that rhetoric lends lustre to the speaker rather than to the topic, as in the Miracles of St Demetrios by the Archbishop J o h n of Thessaloniki (seventh-century) 30 a n d the Ltfe of St Nicholas of Studios (early tenthcentury). 31 O n the other h a n d they quote Prov. 10:7: " T h e m e m o r y of the just is praised", and Ps. 32:1: "Praise becomes the upright" (both quotations a p p e a r in the Ltfe of St Stephen the Younger, ninthcentury). 32 T h e y further point out that a saint's Life serves as praise of the Lord (the Ltfe of St Symeon the Stylite the Younger),33 or that G o d is praised in his saints (the Miracles of St Artemios, seventh-century), 3 4 referring to Ps. 67:35, although the m e a n i n g of the Septuagint text seems to be that G o d is wonderful in his holy places. In the Ltfe of St John Psichaites praise is said to be the n o r m (kanôn) of teaching. 3 5 T h e conflict between e n c o m i u m a n d truth is therefore not a conflict between e n c o m i u m a n d hagiography but between the pagan a n d the Christian encomium. As the author of the Ltfe of St Thekla (fifth-century) observes, historians usually write on their own initiative, whereas he on his part

29

Ed. cit., lines 31-7 of the prologue. Ed. P. Lemerle, I (Paris, 1979), p. 52, 23-4. 31 Patrologia Graeca (hereafter PG) 105, col. 864Λ. 32 PG 100, col. 1069A. 33 Ed. cit., lines 17-18 of the prologue. 34 Ed. A. Papadopoulos-Kerameus, in Varia graeca sacra (St. Petersburg, 1909), pp. 2, 4. 35 Ed. cit. p. 103, 17 18. 3(1

is writing in response to a request. 3 6 Historians are p r o u d representatives of a classical literary genre, the tradition of which they are keeping unbroken, being links in a chain, as it were. T h e y often criticize their predecessors, which means that they have a high opinion of themselves. All this is alien to the hagiographers, who invariably assume a humble attitude, extolling the saints at the expense of themselves. T h e y do not criticize each other nor regard themselves as links in a chain. In their view G o d sends every generation a light from heaven in the shape of a saint to guide men and remind them of the world of eternity. Continuity is therefore not guaranteed by the hagiographers but by the saints. H a g i o g r a p h e r s do not i n t r o d u c e themselves in the m a n n e r of Procopius, Leo the Deacon and Anna K o m n e n a . Palladius, the author of the Lausiac History (fifth-century), who tells us how old he is, is unusually explicit. As a rule, they prefer to remain in the background. Some even pretend to be mere editors of texts written by others, like the authors of the Lives of St Thekla and St Hypatios (sixth-century?). Nevertheless they m a y give a vague idea of who they are by saying that they were the saint's disciple or, in passing, mentioning some other biographical detail. Most of them, however, outline a conventional self-portrait by trying to convince their readers or listeners that they are not at all u p to the task they have assumed. O n the one h a n d they complain that their way of life is far inferior to that of the saint and that they therefore are disqualified from writing his Life—the concordet-sermo-cum-vita topos. 37 T h e o p h a n e s , the hagiographer of St J o s e p h the H y m n o g r a p h e r (ninth-century), says that in his deeds J o s e p h copied the great men he honoured with his hymns, 3 8 but we may be certain that he would not have dreamt of saying anything similar of himself. O n the other hand, and this is still more frequent, they complain of their lack of education and literary competence. Using various paraphrases they confess to being amateurish, ignorant, slow of speech, almost spcechless. A conflict is building up: "I want to speak, but I c a n ' t " (the Life of St Gregory of Akragas, late eighth- to ninth-century). 3 9 Eustratios, the

36

Ed. G. Dagron (Brussels, 1978), p. 170, 29 36. O n this topos, see Festugière, "Lieux communs", pp. 140-2. 38 Ed. A. Papadopoulos-Kerameus, in Monumenta graeca et latina ad histonam Photii pair, pertinentia, II (St Petersburg, 1901), pp. 1, 19. 39 Ed. A. Berger (Berlin, 1995), p. 143, 8 9. 37

very competent a u t h o r of the Life of the Patriarch Eutychios (c. 600)— he seems to have known the writings of the Fathers by heart, espedaily those of Gregory of Nazianzos—begins by describing how he is torn between opposing feelings. His eagerness is fervent, he says, but his speech sluggish; he is aflame with zest, but his tongue is chilled, is stiff a n d hardly able to speak; his interest is broad, but his speech narrow; there is material in a b u n d a n c e for the building of the tower of David, but unfortunately he is an inexperienced builder. 40 O t h e r s say that because of their ineptitude they are divided between the opposing forces of reverence and impudence, fear and desire. T h e m o n k w h o wrote the Ltfe of St Euarestos (early tenthcentury) is a good example of this. H e says that he is about to write an encomium, unfortunately with words that are unworthy of Euarestos. H e is therefore surrounded by fear, on the one side, and love, on the other, one feeling trying to overcome the other, one calling on him to speak, the other telling him to keep silent, " a n d I do not know to w h o m I shall give precedence". 4 1 Sometimes the feeling of inadequacy develops into an author's dilemma which can only be solved through divine intervention, a famous case being the Ltfe of St Euthymios by Cyril of Skythopolis, in the ninth century imitated by the a u t h o r of the Life of St Peter of Atroa.42 Normally, however, the dilemma is solved in a less dramatic way. After much deliberation and lively protestations of unworthiness and incompetence, there occur to the hagiographer a n u m b e r of reasons why he should go ahead: his love of the saint is stronger than his scruples; he does not want to disappoint his friends or superiors, who are enjoining him to write the Vita in question; he can rely on their prayers; they will surely overlook his shortcomings; a child's attempts at talking are dear to its father; G o d appreciates that one does as best one can a n d will give the author "utterance in opening his m o u t h " (Eph. 6:19); besides, rough a n d simple style is good enough for expressing truth. Eustratios, the a u t h o r of the Ltfe of Eutychios just mentioned, points out that even if the task is lofty and

40

Ed. cit., lines 1-9. Ed. C. Van de Vorst, Analecta Rollandiana 41 (1923), pp. 295-325, esp. 296, 10 13. For Latin examples see T . Janson, Latin Prose Prefaces. Studies in Literary Conventions (Stockholm, 1964), pp. 116-17, with further references. Janson's study is of considerable help for the study of commonplaces in Greek hagiographie prologues. 42 Life of Euthymios, ed. E. Schwartz (Berlin, 1939), p. 84; Life of Peter of Atroa, ed. V. Laurent (Brussels, 1956), pp. 67-9. 11

difficult to accomplish, one should not give it up to lazy and useless silence, like a cowardly landlubber, who is frightened by the mere sight of the sea and does not dare to embark. 4 3 Similarly, the author of the Life of St Stephen the Younger states: not to deliver an e n c o m i u m that fails to meet the standard, but to abstain from speaking as best one can, brings condemnation. 4 4 T h e same author, and others as well, quote the saying that one should not hide a talent held in charge in the grounds (cf. Matt. 25:14fT): viz. one should not keep a treasure of edification to oneself. It should be borne in mind, however, that there is no strict correspondence between protestations of incompetence on the one h a n d a n d actual stylistic achievement on the other. N o r should one forget that there are hagiographers who express no feelings at all of personal inadequacy, either real of feigned. T o some the problem does not even seem to exist, while a few declare that they are writing for the uneducated, to w h o m high style is incomprehensible. Leontios of Neapolis belongs to this category. In the prologue to the Lfe of St John the Merciful he says that J o h n ' s Life has already been written by two highly cultivated m e n who expressed themselves in a learned and lofty way. H e makes a point of saying that he by no means wishes to calumniate these admirable men, nor try to emulate them. His reasons for writing about J o h n are different. First of all it would not be right to consign to silence material that could edify the listeners. Second, this material is not contained in the aforementioned composition. T h i r d , unlike his predecessors he intends to write in a prosaic, u n a d o r n e d and lowly style, so that the c o m m o n , illiterate m a n may also be edified by J o h n ' s example. 4 3 O n e notices that Leontios does not criticize his predecessors for writing in high style. H e simply seems to make a serious attempt at compensating for the limited range of a composition in high style. Also the a u t h o r of the Miracles of St Demetrì0s, the Archbishop J o h n of Thessaloniki, whose style is more disciplined than that of Leontios, says that he wants to express what is great to the mind in ordinary Greek, so that not only those who are linguistically skilled but all the brotherhood may gain a complete understanding of his narrative. 4 6

43 44 45 46

Ed. Ed. Ed. Ed.

cit., cit., cit., cit.,

lines 23-8. col. 1072B. p. 344, lines 5 0 - 6 9 . p. 52, 28-9.

As already indicated, the hagiographical prologues display amazing variety, despite all their stereotypes—the vastness of the subject, the incompetence of the author; his trust in G o d w h o will give utterance in opening his m o u t h ; his assurance that he is telling the truth; his conviction that his composition will glorify G o d , h o n o u r the saint a n d edify the reader etc. T o peruse fifty of t h e m is therefore not as boring as one might expect. It is true that rather than c o m m o n places a m o d e r n reader would prefer more precise information on the circumstances u n d e r which the hagiographie piece in question was written. O n the other h a n d , despite their vagueness one must a d m i t that the prologues usually serve the purpose of p r e p a r i n g the listener or reader for what he is going to hear or read. This becomes particularly a p p a r e n t when the hagiographer is writing about unusual or potentially controversial saints. T w o examples, one preiconoclastic, the other post-iconoclastic, m a y illustrate this. Leontios of Neapolis, w h o lived on Cyprus in the first half of the seventh century, was the first to write a full-scale Vita of a holy fool. Since the fool, St Symeon, lived in Syria in the middle of the sixth century, Leontios had never met him. In all probability he relied on written sources, as he indicates himself in his epilogue, a n d these are likely to have been anecdotal in character rather than having the form of a Vita.47 Leontios therefore invented a secret friend of Symeon, the D e a c o n J o h n , from w h o m he pretends to have received all his information, a n d gave Symeon a biographical background, according to which he had left the world during a pilgrimage to Jerusalem, become a m o n k at the J o r d a n , but soon left the monastery for the desert, where, after almost three decades of uncompromising asceticism, he discovered that he h a d m a n a g e d to attain apatheia, the highest level of ascetic perfection. After this achievement he was i m m u n e to sin a n d could enter the city of Emesa to play the fool for the sake of Christ, without running the risk of losing his moral purity. All these preliminaries constitute a kind of preface to the main part of the Vita. It is preceded by a substantial prologue in the usual sense. Here Leontios first exploits two well-tried commonplaces, namely the concordet-sermo-cum-vita topos a n d the advantage of having

4

' C. Mango, "A Byzantine Hagiographer at Work: Leontios of Neapolis", Byzanz und der Westen, ed. I. Hutter (Vienna, 1984), pp. 25-41, esp. p. 30, claims that it was "a paterikon, i.e. a collection of disconnected anecdotes, a type of composition that was very popular of the time".

an example in the shape of a contemporary saint, w h e r e u p o n he focuses on the kind of edification one can reap from the case of Symeon, unusual though it is. H e then asks his listeners and readers to receive his narrative with fear of Christ and faith and trust. K n o w i n g that it may make some people laugh, he reminds them of St Paul, who says that he who wants to be wise in this age must become a fool that he might become wise (1 Cor. 3:18), and so forth. H e further maintains that Symeon's folly was not the result of a sudden impulse but the ripe fruit of long years of preparation; Symeon was like a soldier w h o does not step forth to fight in single c o m b a t until he has received p r o p e r training. At last he points out that Symeon, by leaving the desert and entering a city in order to save his fellow creatures, was a true imitator of Christ. This prologuc, especially in its second half, a n d in combination with the introductory biographical part, is extremely well suited to prepare the reader for the topic. In the ninth century a Jewish boy living at Synada in Phrygia witnessed a series of wonderful signs which eventually m a d e him leave his bride at the wedding, like another Alexios, m a n of G o d , and flee to M o u n t Olympos, where he converted to Christianity and b e c a m e a m o n k u n d e r the n a m e of Constantine. Later he almost suffered m a r t y r d o m when he tried to convert the J e w s of nearby Nicaea, who of course bitterly resented his attempts. T h e prologue of the Vita, which was written by a younger contemporary of Constantine, begins with a traditional comparison between a statue and a saint's Life. Not surprisingly, a Lfe appears to be much more informative than a statue. T h e a u t h o r further declares that he intends to tell the truth rather than glorify the saint according to the rules of rhetoric, although this does not m e a n that his style is particularly straightforward. O n the contrary it makes an overambitious impression, a n d it is not always clear. However, after these preliminaries the author makes the most of the fact that Constantine was a J e w who had become a convert to Christianity of his own free will. H e expresses wonder at the consistency with which the Saviour continues to save the souls of men by transforming persecutors into evangelists, biasphemers into theologians, Sauls into Pauls. Constantine, he continues, was brought up according to the old, corrupt law but b e c a m e a believer in Christ, the perfecter of law. H e reached such a degree of virtue that he was even granted the charisma of prophecy. H e was an equal to both Elijah and J o h n the Forerunner. R e b o r n by

baptism he put off the old nature of m a n a n d put on the new one. And so forth. 4 8 This is not a great piece of introduction but, like the prologue of the Life of Symeon the Fool, it illustrates the way in which hagiographers combine topoi of no more than general relevance with others that refer directly to their topic. Despite its flexibility, hagiography is a conservative genre. Nevertheless there are differences between fourth- to seventh-century and ninth-century Byzantine hagiography, not only with regard to the geographical and social spread of the authors a n d their heroes, but also concerning forms a n d attitudes. T h e most striking one is that, with some exceptions, the Lfe of St Philaretos the Merciful (early ninthcentury) being an obvious example, in the ninth century the style has become higher a n d the references to classical literature and culture more frequent, although less justified. This of course reflects the renaissance of classical studies that began in the late eighth century—why else refer to statues, athletes, Olympic games and the like in the ninth century? This new development must have narrowed down the n u m b e r of people who could profit from contemporary hagiography, although it raised its literary status—certainly very few were able to appreciate the Lives written by the Patriarch Methodios (d. 847), to take an extreme example. It also resulted in a more frequent appearance of the author's dilemma. Instead, narrative devices, such as imaginary sources in the form of alleged eyewitnesses or old books retrieved, have become less c o m m o n . T h e borderline between hagiography a n d historiography has become more fluid. As might be expected, there is now m u c h m o r e talk of icons a n d their relation to the saints and their Lives. T h e contrast to Leontios of Neapolis, who defended the veneration of icons against Jewish criticism, but never mentioned icons in his hagiographical writings, is remarkable. 4 9 Moreover, if I am not mistaken, ninth-century hagiographers have become m o r e prone to saying that they are writing at the end of time. This foreshadows the pessimistic remarks of Leo the Deacon, mentioned above, as well as the apocalypses in hagiographie form that b e c a m e popular in the tenth century. An interesting new fea-

48

T h e life of St Constantine the Jew (c. 900), ed. in the Acta Sanctorum, Nov. IV, pp. 628-56, esp. 628B-629A. 49 Leontios' defence of the icons has recently been edited with introduction, translation and commentary, by V. Déroche, Travaux et Mémoires 12 (1994), pp. 45-104.

ture is also the intolerant attitude demonstrated by the prominent ninth-century hagiographer Ignatios the Deacon. In his prologue to the Life of the Patriarch Nikephoros I he bluntly declares that he does not want those who fail to share Nikephoros' view of the icons, i.e. those who are still iconoclasts at heart, to read his Vita.50 But with this we have come to the antithesis of communication.

50

Ed. C. de Boor, in his Niceph0rì opuscuta histonea (Leipzig, 1880), pp. 139 217, esp. 141, 5-12.

T H E USE AND ABUSE O F LATIN H A G I O G R A P H Y IN T H E EARLY M E D I E V A L W E S T Ian W o o d

Hagiography is not one genre, but a multiplicity. This is true even at the most basic formal level. T h e r e are short texts and there are long texts: there are works, usually short, written for litugical use, whether in a small monastic c o m m u n i t y or in a large u r b a n cathedral: a n d there are works, sometimes long, written to be read and contemplated in private, whether by c h u r c h m e n or laymen, and, a m o n g c h u r c h m e n , by monks as well as secular clergy. Function and audience vary, a n d they affect the form of the hagiographical text. As Alcuin pointed out, his two versions of the Ltfe of Willibrord, similar as they are in content, were m e a n t to be treated in different ways: the prose version was intended to be read to the monastic c o m m u nity of Echternach at large, publice fratnbus in ecclesia, while the poetic version was to be contemplated in secreto cubili inter scholasticism Sometimes form, function and audience may be indicated in a preface, even if prefaces are not always the most revealing point of entry into a work of hagiography, replete as they usually are with topoi—though true mastery or otherwise of something as c o m m o n as the modesty topos can shed a deal of light on the skills of a writer or the expectation he or she has of his or her audience. 2 N o t that all prefaces in Western hagiography are so limited, even if detailed information relating to the author or the circumstances of writing are rare: J o n a s of Bobbio is exceptional in providing an account of his life immediately prior to the composition of the Ltfe of Columbanus in the preface to that work, 3 and he inserted similarly autobiographical

1

Alcuin, Vila Willibrordi, praef., ed. H.-J. Reischmann, Willibrord, Apostel der Friesen (Darmstadt, 1989). 2 T. Janson, Latin Prose Prefaces: Studies in Literary Conventions (Stockholm, 1964), pp. 162-8, discusses the preface to Gregory of Tours' Liber in Gloria Confessorum, though he has little to say about other hagiographical prefaces. E. Curtius, European Culture and the Latin Middle Ages, trans W.R. Trask (London, 1953), pp. 159-62, is a little fuller. 3 Jonas of Bobbio, Vita Columbani, ep. to Waldebert and Bobolenus, ed., Β. Krusch, Ionae Vitae sanctorum Columbani, Vedastis, Iohannis, Monumenta Germaniae Historica (henceforth MGH), Scriptores Rerum Germanicarum (henceforth SRG) (Hannover, 1905).

material in the incipit to his Lfe of John of Réomé.4 M o r e often to be found are spiritual or theological statements, the majority of them admittedly topoi or second h a n d , but which provide at least something of a key to the author's attitude to his or her subject matter. T o take a single famous, albeit a b n o r m a l , example: in the preface to his Liber Vitae Patrum Gregory of T o u r s argued that "it is better to speak of the 'Life of the Fathers' rather than the 'Lives of the Fathers', the m o r e so since there is a diversity of merits a n d virtues a m o n g them, but the one life of the body sustains them all in this world." 5 T h e notion of a single Christian Life was crucial to Gregory of T o u r s , a n d indeed to m a n y hagiographers w h o drew without qualms on earlier texts for their models and to fill gaps. Yet, however strong the theological case for seeing saints as participating in a single life, and one largely built on Biblical models, 6 research over the last few decades has tended, rightly, to stress the variety within the Vitae written about those saints during the period from the end of the R o m a n Empire through to the tenth century and beyond. Quite apart from form, there is the issue of subject matter. H a g iography deals with m e n a n d w o m e n ; clerics and ascetics; martyrs a n d confessors—some of them the focus of cult throughout Christend o m , and some of them culted in one or two places; each deserving a n d receiving different treatment. T h e r e are episcopal Lives; though instantly one might contrast two of the earliest: Sulpicius Severus' Lfe of Martin,' which stands at the head of m u c h western hagiography, a n d which was cited regularly by later hagiographers, was written before the saint's death, m o r e to justify Martin's ascetic lifestyle, than to promote the image him of as a bishop: by contrast Constantius' Lfe of Germanus of Auxerres is a fifth-century text on a fifth-century bishop, written for a second bishop, Patiens of Lyons, a n d then addressed to a third bishop, Censurius of Auxerre. 9 Presumably Ger-

4

Jonas, Vita Iohannis, incipit, ed. Krusch, ioc. cit. Gregory, Liber Vitae Patrum, pref., ed. B. Krusch, MGH, Smptores Rerum Merovingicarum (henceforth SRM) I, 2 (Hannover, 1885): trans. E. James, Gregory of Tours, Life of the Fathers (Liverpool, 1985), p. 28. 6 M. van Uytfanghe, Stylisation biblique et condition humaine dans l'hagiographie mérovingienne [600-750] (Brussels, 1987). 7 Ed. J. Fontaine, Sulpice Sévère, Vie de saint Martin, Sources Chrétiennes 133-5 (Paris, 1967-9).' 8 Ed. R. Borius, Constance de Lyon, Vie de saint Germain d'Auxerre, Sources Chrétiennes 112 (Paris, 1965). 9 Constantius, ep. to Censurius, ed. Borius, pp. 114-7. 5

m a n u s ' dealings with heretics a n d with barbarians were intended as a model for a subsequent generation of bishops faced with yet greater barbarian and heretical pressure, as Euric established the Visigothic kingdom in the remnants of R o m a n Gaul. 1 0 Unlike a host of later works of episcopal hagiography, neither of these vitae has m u c h interest in the cult of their subjects: in the case of the Life of Martin this was inevitable, since the bishop was alive at the time of writing. T h e r e are monastic Lives, like the sixth-century Vita Patrum Jurensium, whose coverage of the monastic day is so thorough as to suggest that the Life itself may even have been intended to be the monastic Rule which the author promises to include in his work, but which some scholars have thought has been detatched from the text and lost." T r u e to his monastic concerns, the a u t h o r spends no time on posthumous miracles, which might have brought pilgrims and disturbed the peace of the community. 1 2 There are even monastic Lives, like Braulio of Saragossa's Ltfe of Aemilian, which include more than a note of warning against would-be imitators of a saint whose bchaviour had attracted a certain a m o u n t of criticism. 13 T h e r e are texts concerned with holy m e m b e r s of the laity, a m o n g them women, like the extremely complex Vita Genovefae, again a sixth-century Ltfe.4‫י‬ H e r e the saint performs m a n y of the duties of a bishop, perhaps not so m u c h because she really did so, as because episcopal sanctity was the only model available when it c a m e to writing u p a saint whose life was lived outside the monastic c o m p o u n d . ‫י י‬

10

See the comments of Borius, Constance de Lyon, Vie de saint Germain d'Auxerre, pp. 13-15, on relations between Constantius and Sidonius Apollinaris, bishop of Clermont: P. Brown, The Rise of Western Christendom (Oxford, 1996), p. 67. " Vita Patrum Iurensium, 4 (I, 1) ed. F. Martine, Vie des Pères du Jura, Sources Chrétiennes 142 (Paris, 1968). See I.N. Wood, "A Prelude to Columbanus: the monastic achievement in the Burgundian territories", Columbanus and Merovingian Monasticism, British Archaeological Reports, International Series 113, eds. H.B. Clarke and M. Brennan (Oxford, 1981), p. 4. 12 Vita Patrum Iurensium, 61 (I 19), 117 (II 16): compare Gregory of Tours, Liber Vitae Patrum, I, 6. 13 Braulio, Vita sancti Emitiani, praef (6), xxiii (30), ed. L. Váquez de Parga (Madrid, 1943): see A. Fear, Lives of the Visigothic Fathers (Liverpool, 1997), pp. xxviii xxix. See also Gregory of Tours, Liber Vitae Patrum, X (Friardus), X V (Senoch). 14 Ed. B. Krusch, MGH, SRM III (Hannover, 1896): see M. Heinzelmann and J.-C. Poulin, Les Vies anciennes de sainte Geneviève de Paris, Bibliothèque de l'Ecole des Hautes Études, IVe section, vol. 359 (Paris, 1986). 15 See the comments of Heinzelmann and Poulin, Les Vies anciennes de sainte Geneviève de Paris, pp. 80-103, and the critique of I.N. Wood, "Forgery in Merovingian

Some texts m a y have had a p r o g r a m m a t i c purpose. T h e r e is, for instance, Alcuin's Vita Willibrordi, low on fact, but carefully constructed a r o u n d an idea which stems from Gregory the Great: "although the ministry of evangelical preaching is to be prefered to the working of all miracles a n d demonstration of signs, nevertheless I do not think that those things which have been related should be kept secret, to the greater glory of G o d " — a sentence which is physically in the centre of the text. 16 W h a t follows is a series of tales about divine rather than h u m a n exaction of vengeance, intended perhaps as a critique of Charlemagne's forcible conversion of the Saxons. Alcuin's hagiographical exploration of mission was directed at a relatively small g r o u p — i n the first i n s t a n c e t h e c o m m u n i t y of Echternach, although doubtless it was intended to have a larger, secondary circulation—and it was expressed in theological terms. O t h e r hagiography could be m o r e overtly political, at least in part: there is R i m b e r t ' s ninth-century Life of Anskar, which at one level is a defence of the combined diocese of Hamburg-Bremen, levelled against the onslaught of the archbishop of Cologne, and which must therefore have been directed at a larger audience than the monks of Corbie mentioned in the opening chapter. Yet at quite another level the Vita Anskani seems to be a personal meditation on the perils of mission. 17 This quality of personal meditation recurs in other texts written by hagiographers w h o were missionaries, and suggests that some hagiography is displaced autobiography—implying a depth of individuality ignored by those who see only in the twelfth century the discovery of the individual. 18 Rimbert's Lfe of Anskar is by no means the only text to have a multiplicity of functions. T h e Vita Columbani by J o n a s of Bobbio is for the most part a c o m m e n t a r y on the spirituality to be found in seventh-century Luxeuil, Bobbio and Faremoutiers, especially in the

Hagiography", Monumenta Germaniae Historica: Schriften 33, Fälschungen im Mittelaller V (1988), pp. 376-8. 16 Alcuin, Vita Willibrordi 14. 17 Rimbert, Vita Anskani, ed., W. Trillmich, Quellen des 9. und 11. Jahrhunderts zur Geschichte der hamburgischen Kirche und des Rdches (Darmstadt, 1978). See the comments of I.N. Wood, "Christians and Pagans in ninth-century Scandinavia", The Christianisation of Scandinavia, eds. B. Sawyer, P.H. Sawyer and I.N. Wood (Alingsàs, 1987), pp. 36-67. 18 See in particular Bruno of Querfurt, Vita quinque fratrum Poloniae, ed. R. Kade, MGH, Scriptores (henceforth) SS XV, 2, pp. 716-38). O n this see I.N. Wood, The Missionary Life (forthcoming).

generation following C o l u m b a n u s ' death.' 9 But it is also a text concerned with monastic rights established after the death of Brunhild, whose m e m o r y is blackened, in accordance with the damnatio memoriae encouraged by the successful C h l o t h a r II, whose final succession to all parts of the Frankish kingdom is prophetically heralded on two occasions during the Vita.20 In terms of the damnatio memoriae of Brunhild, the Vita Columbani is only one of a g r o u p of texts with a shared political propaganda: other related works include the Visigothic king Sisebut's account of the Passion of Desiderius of Vienne, a n d the marginally earlier a n o n y m o u s Passio Desiderii: the first a work a p p a r ently intended at least in part to support better diplomatic relations between Visigothic Spain a n d Frankish Gaul: 2 ' the second a text illustrative of the overlap of court support a n d local interests, particularly over the matter of building u p a saint-cult. 22 Even m o r e directly associated with court politics are the Passiones of Leodegar of Autun a n d Praeiectus of C l e r m o n t , two martyrs overwhelmed by the political crisis of the 670s in Francia. 2 3 T h e s e texts emerge from a very precise political situation, and reveal a hagiography closely tailored to the peculiar political circumstances of the second half of the seventh century in Francia—although it is as well to r e m e m b e r that there had been earlier episcopal m u r d e r s in the Merovingian kingdoms, 24 that the Passiones of Leodegar and Praeiectus were not the first records of c o n t e m p o r a r y martyrs to be written in Francia, 2 3 a n d that the second Passion of Leodegar, at least in its current form, is not c o n t e m p o r a r y with the martyrdom. 2 6 T h e first

19

I.N. Wood, "The Vila Columbani and Merovingian hagiography", Peritia I (1982), pp. 63-80. 20 Jonas, Vila Columbani, I 20, 24: see also 29. I.N. Wood, 'Jonas, the Merovingians and Pope Honorius: Diplomala and the Vita Columbani" After Rome's Fall, ed. A.C. Murray (Toronto, 1998), pp. 99-120. 21 J . Fontaine, "King Sisebut's Vita Dedderii and the political function of Visigothic hagiography", Visigothic Spain, ed. E . J a m e s (Oxford, 1980), pp. 93 129. 22 Vita Desiderii, ed. B. Krusch, MGH, SRM III: see Wood, "Forgery in Merovingian Hagiography", pp. 373-6. 23 Passiones Leudegarii, ed. B. Krusch, MGH, SRM V (Hannover, 1910): Passio Praeiecti, ed. Krusch, 10c. cit.: for translations and commentary see P. Fouracre and R.A. Gerberding, l/ite Merovingian France: History and Hagiography 640 -720 (Manchester, 1996): see also P. Fouracre, "Merovingian History and Merovingian Hagiography", Past and Present 127 (1990), pp. 3 - 3 8 . ‫י‬ 24 Gregory of Tours, Decern Libri Historiarum VI 37 (Abbot Lupentius of Javols), VIII 31 (Bishop Praetextatus of Rouen), ed. B. Krusch and YV. Levison, MGH, SRM I 1 (Hannover, 1951). 25 See the anonymous Passio Desiderii. 26 See Fouracre and Gerberding, Late Merovingian France, pp. 206 8.

Passio of Leodegar seems to have been directed against the maior Ebroin, or rather his memory, in an attempt to exculpate Herminarius, the bishop of A u t u n , from any involvement in his predecessor's martyrdom; it also seems to have m a r k e d an unsuccessful bid for possession of the martyr's body. 2 7 T h e Passio of Praeiectus seems at once to m a r k the reconciliation of the two parties involved in the dispute underlying the m a r t y r d o m , a n d at the same time to attempt to prom o t e the shrine of the m a r t y r at Volvic. 28 Theological, spiritual and moral statements, political statements, p r o p a g a n d a for a cult-site, works for the liturgy, even history as understood by classical authors; 29 hagiography could encompass them all and a single text could embrace m o r e than one function at once. T h e various functions of hagiography in the Early Middle Ages gave rise to a variety of forms. W e even tend to underestimate the variety because we usually forget that a substantial proportion of the acts of the christian martyrs of the R o m a n period were compiled in the fifth century a n d later. T h e destructive review by Delehaye of D u f o u r c q ' s Etudes sur les Gesta Martyrum may have been a factor in the subsequent comparative neglect of these texts. 30 T h i s has unfortunately served to strengthen an understandable emphasis on those texts published within the various series of the Monumenta Germaniae Hist0úca, most notably the Scnptores Rerum Merovingicarum, leaving only the most devoted researcher to wrestle with sometimes inadequate editions to be found in the Acta Sanctorum,3' a n d the non-Frankish specialist to work with texts which are usually published in isolation. 32 Even the Merovingianist m a y sometimes find him- or herself

27

See Fouracre and Gerberding, Late Merovingian France, pp. 201-6. See Fouracre and Gerberding, Late Merovingian France, pp. 266-7. 29 See Fear, Lives of the Visigothic Fathers, pp. xvii-xviii. 30 A. Dufourcq, Etudes sur les Gesta Martyrum romains, 4 vols. (Rome, 1900-10): reviewed by H. Delehaye in Analecta Bollandiana 27 (1908), pp. 215-8: some exceptional texts (the Lives of Lucius of Chur, Quirinus of Tegernsee, Afra and the martyrs of Agaune) were edited by B. Krusch in MGH, SRM III. 31 T h e r e are, of course, plenty of well edited texts in Acta Sanctorum and in Analecta Bollandiana. The texts to be found in the most recent volumes of the Acta Sanctorum to have been published (vitae for the month of November), especially those edited by such scholars as A. Poncelet, compare in quality with anything in the MGH. 32 For the Anglo-Saxon material see, however, the useful editions by B. Colgrave, The Life of Bishop Wilfrid by Eddius Stephanus (Cambridge, 1927); Two Lives of Saint Cuthbert (Cambridge, 1940); Felix's Life of Saint Guthlac (Cambridge, 1956); The earliest Life of Gregory the Great (Kansas City, 1968). For the Visigothic material see the translations, with bibliography of editions, in Fear, Lives of the Visigothic Fathers, 28

wrong-footed by the failure of B r u n o Krusch a n d Wilhelm Levison to include a m a j o r work in the Smptores Rerum Merovingicarum.33 In the face of the variety present within the sphere of hagiography the most immediate p r o b l e m facing the historian w h o has to work regularly on saint's Lives is the search for a set of organising principles. M u c h has been d o n e by students of hagiography on m a n uscript tradition a n d survival 3 4 ^ though one should add that, despite this, individual hagiographical works are still too rarely considered within their p r o p e r manuscript c o n t e x t — a n d m u c h research has also been carried out on literary form. 3 5 Yet the results of such studies are registered more often than not in the d o m a i n of the textual or literary historian or the theologian, rather than in that of the student of political or social history. S o m e very basic categorisation has, however, yielded clear, albeit circumscribed, dividends, even in the world of the historian of politics or society, as, for instance, has the study of the vitae of female saints. 36 O t h e r apparently basic categorisation has arguably been deleterious. T h u s Krusch's decision to place the Lives of Gibichung, Merovingian a n d Pippinid saints in a single volume, volume two of the Smptores Rerum Merovingicarum, gives a spurious uniformity to a g r o u p of totally unrelated texts, separated chronologically as well as by intention and value. Similarly, Krusch's decision to gather together, in volume three of the Scnptores Rerum Merovingicarum, the Lives of fifth- a n d sixth-century saints, regardless of their place or date of composition, has tended to emphasise the division between Merovingian sanctity pre- a n d a n d p o s t - C o l u m b a n u s , whose Vita, by J o n a s of Bobbio, opens the fourth volume of the Scnptores. T h i s division has

although the introduction is not entirely in tune with much current thinking on hagiography: for Italy the crucial work is Gregory the Great's Dialogi, ed. A. de Vogiié, Sources Chrétiennes 251, 260, 265 (Paris, 1978 80): for early Irish hagiography one now has as a guide, R. Sharpe, Medieval Irish Saint's Lives: an introduction to the Vitae Sanctorum Hibemiae (Oxford, 1991), though crucial texts still lack modern editions. 33 See especially the Life of St Samson, now usefully reedited by P. Flobert, IM Vie ancienne de saint Samson de Dot (Paris, 1997). 34 See the regular comments on hagiographical work in Francia, and the related work of M. Heinzelmann, ed., Manuscrits hagiographiques et travail des hagiographes (Sigmaringen, 1992). 35 Van Uytfanghe, Stylisation biblique et condition humaine dans l'hagiographie mérovingienne [600 750], is seminal. 3 '‫ י‬J.-A. M c N a m a r r a and J.Ε. Halborg, Sainted Women of the Dark Ages (Durham, North Carolina, 1992).

h a d a p r o f o u n d a n d unhelpful impact on the reading of Merovingian religious history, which all too easily falls a p a r t a r o u n d the year 600 because of the deaths of Gregory of T o u r s a n d Venantius Fortunatus, a n d of the absence of any c o m p a r a b l e writer for the later period. Although J o n a s of Bobbio famously argued that C o l u m b a n u s reintroduced to Francia after a long interval of neglect the paenitentiae medicamenta, by which he seems on occasion to have m e a n t monasticism rather than penance, 3 7 his total œ u v r e suggests a m u c h greater sense of religious continuity between the early Merovingian period a n d the seventh century than this might a p p e a r to imply. Apart f r o m writing the Vita Columbani for abbots W a l d e b e r t of Luxeuil and Bobolenus of Bobbio, J o n a s wrote the Life of the early sixth-century Gallo-Roman, J o h n of R é o m é , for that saint's distant successor, abbot C h u n n a , w h o was closely associated with Luxeuil. 3 8 H e m a y also have written an account of a n o t h e r sixth-century saint, Clovis' cont e m p o r a r y Vedast, for Autbert, bishop of C a m b r a i , a n d previously m o n k of Luxeuil. 3 9 J o n a s thus f o u n d no difficulty in recording the Life of one, a n d probably the Lives of two saints of the early sixth century. However m u c h m o d e r n historians m a y wish to see a distinction between the religious history of the sixth a n d seventh centuries, with the break being caused by C o l u m b a n u s himself, the latter's own hagiographer, while seeing the Irish saint as a restorer of standards, did not see him as the instigator of a novel spiritual revolution. 4 0 Yet, at the same time, J o n a s ' three hagiographical works belong together as evidence for the religious history of the mid seventh century, a n d indeed Krusch's editions of all three texts are to be found appropriately reissued in the Scnptores Rerum Germanicarum in usum scholarum. R e a d as a g r o u p they illuminate the religious a n d cultural history of Neustro-Burgundy in the days of Clovis II a n d Balthild. 41 Like all hagiography, the works of J o n a s of Bobbio need to be read in the precise context of their composition, regardless of what else they say about their subject-matter. 37

Jonas, Vita Columbani, I 5 (perhaps penance), 10 (monasticism); see also II 2 (penance), 8 (perhaps monasticism), 15 (penance), 19 (penance), 25 (penance). 38 Jonas, Vita Iohannis, incipit: see Krusch, Ionae Vitae sanctorum Columbani, Vedastis, Iohannis, p. 326, n. 2. 39 Krusch, Ionae Vitae sanctorum Columbani, Vedastis, Iohannis, pp. 295-6: I.N. Wood, The Merovingian Kingdoms, 450-751 (London, 1994), p. 188. 40 Wood, "A prelude to Columbanus", p. 3. 41 Wood, " T h e Vita Columbani and Merovingian hagiography": Wood, "Jonas, the Merovingians and Pope Honorius: Diplomata and the Vita Columbani".

T h e study of a single author, like J o n a s , or like Arbeo of Freising a century later, 42 or indeed of A b b o of Fleury in the tenth century, 4 3 or the study of a single region, like that between the S a m b r e a n d Meuse in the Carolingian period a n d beyond, 4 4 can clearly yield significant rewards. Indeed, most hagiographical texts can only be fully understood when their author or at least the context of their composition is known. Unfortunately it has to be recognised that this a p p r o a c h , which privileges those texts written by known authors, assignable to known monasteries, or at least texts which are firmly dated, tends once again towards the exclusion of anonymous, dateless texts, such as the Gesta martyrum- arguably it also leads to an overemphasis on works which are of importance to the political, social or ecclesiastical historian, to the detriment of those texts— probably the majority of hagiographical works—which are of most use to the theologian, liturgist or the student of spirituality. T h e possibility, however, of establishing models from clearly placeable texts, and subsequendy applying them to those without provenance is worth consideration. T h e integration of hagiographical material into historical interpretation has unquestionably h a d its problems—not least over the question of the reliability of texts belonging to a genre which often, though not always, has recourse to the miraculous. 4 5 T h e problems posed by this aspect of h a g i o g r a p h y have most effectively been addressed by m o d e r n approaches drawing on social anthropology, the most eye-catching examples of which for early medievalists are those of Peter Brown. 4 6 For the most part this work has drawn on

42

Ed. B. Krusch, Arbeonis episcopi Frisingensis vitae sanctorum Haimhrammi et Corbiniani, MGH, SRG (Hannover, 1920). See also H. Löwe, "Arbeo von Freising: eine Studie zu Religiosität und Bildung im 8. Jahrhundert", in id., Von Cassiodor zu Dante (Berlin, New York, 1973), pp. 75-110. 43 M. Mostert, The political theolog)! of Abbo of Fleury: a study of ideas about society and law of the tenth-century monastic reform movement (Hilversum, 1987). 44 A. Dierkens, Abbayes et Chapitres entre Sambre et Meuse (VII'-ΧΓ siècles) (Sigmaringen, 1985). 45 For a fallacious statement of the role of the miraculous in hagiography, albeit one that reflects common assumptions, Fear, Lives of the Visigothic Fathers, pp. xix-xx. T h e miraculous is more or less absent from Willibald's Life of Boniface and from the Vita Lebuini, to name only two eighth-century Lives. 40 P. Brown, Relics and Social Status in the Age of Gregory of Tours, Slenton Lecture 1976 (Reading, 1977): id., The Cult of the Saints. Its Rise and Function in Latin Christianity (Chicago, 1981): see also the critique by J . Fontaine, "Le culte des saints et ses implications sociologiques. Réflexions sur un récent essai de Peter Brown", Analecta

clearly dated texts, and its revelations about the mentalités of the fifth a n d sixth centuries have been dramatic. T h e r e has, however, been a growing realisation that it is mentalités rather than realités which are most certainly exposed. Parallel to, and in m a n y ways compatible with, this development is a growing appreciation of the painstaking work of the Bollandists, with its emphasis on literary models and structures. Nevertheless the cult of the saints has, at least until recendy, stubbornly refused to enter the mainstream of historical reconstruction of the early medieval period. 4 7 It has largely remained a subject on its own. O n e other approach, which may seem to have placed hagiography m o r e clearly at the centre of our understanding of a period, may be more flawed, not least because it has used a vast range of hagiographie texts without p r o p e r emphasis on their date of composition. Frantisek Graus' Volk, Heiliger und Herrsher, which was for m a n y the seminal text in opening u p early medieval hagiography, is a very problematic work. 48 T h e words Volk and Herrscher in the title of a work written in a communist country by a dissident ought to have signalled to its readers that m o r e than a straight reading of M e r o v i n g i a n h a g i o g r a p h y was at stake. Equally the almost total absence of references to unquestionably Merovingian texts ought to have alerted readers to the chronological problems involved in Graus' interpretation of the development of the importance of aristocratic saints, in G e r m a n Adelsheiliger.49 G r a u s ' discussion of Merovingian sanctity was, quite simply, advanced without due attention being paid to the date of the sources. It is worth pausing over two families boasting Adelsheiliger, the Etichonen and the Pippinids. T h e Etichonid saint Odila was certainly culted at St Odile in southern Alsace, a n d her t o m b gives some indication that she was venerated at least within the c o m m u n i t y of Mont-Ste-Odile during the Merovingian period, 3 0 but the liturgical

Bollandiana 100 (1982), pp. 17 41. For another, more spiritual, approach see B. Ward, Miracles and the Medieval Mind: theolog)), record and event 1000-1215 (London, 1982). 4 ' Brown, The Rise of Western Christendom, is a major exception. 48 F. Graus, Volk, Henscher und Heiliger im Räch der Merowinger: Studien zur Hagiographie der Merowinger zeit (Prague, 1965). See, however, F. Prinz, "Heiligenkult und Adelsherrschaft im Spiegel merowingischer Hagiographie", Historische Zeitschrift 204 (1967), pp. 529-44. 49 For a more generous assessment of this tradition of scholarship, see J . M . Wallace-Hadrill, The Prankish Church (Oxford, 1983), p. 89. 50 R. Will, "Le monastère de Hohenbourg et ses abords", Le Mont Sainte-Odile, eds. F. Pétry and R. Will (Paris, 1988), pp. '101-2.

and hagiographie development of her cult is harder to date—according to Levison the Life itself was written between the late eighth and the early tenth century. 3 1 With the Pippinids there is at first sight a better case for a Merovingian tradition of Adelsheiliger. T h e saints in their family, Gertrude and Arnulf, are held to be the subjects of early vitae. T h e Vita Geretrudis is unquestionably a seventh-century Lfe of the daughter of Pippin I. 32 She, a n d the monastery of Nivelles, can reasonably be seen as providing the Pippinid family with a nucleus of sanctity. 53 Arnulf, Metz and the Vita Amulfi, however, are rather different. T h e r e is no clear evidence to place the Lfe of Arnulf in the seventh century. It is closest in style and content, despite Lcllia C r a c c o Ruggini's recent observations on its unusual emphasis on poverty, 3 4 to the Vita Amati and the Vita R0marìci, the Lives of Amatus and Romaricus, both abbots of Remiremont, the monastery to which Arnulf retired. Krusch dated these texts to the Carolingian period. 3 3 T h e r e is, however, little to distinguish t h e m from the Lfe of Arnulf. In narrative and linguistic terms they unquestionably form a very close-knit group. T h e y could all be placed in the late seventh or early eighth century. 5 6 But even that is not the end of the problems of using the Vita Arnulfi as evidence for a notion of Adelsheiliger a m o n g the early Pippinids, because there is no Merovingian evidence to support the notion that Arnulf had a son called Ansegisel who married Begga, and therefore there is no compelling evidence to prove that the Pippinids were related to the bishop of Metz. T h a t information is not to be found before the late eighth century and Paul the Deacon's Gesta of the bishops of Metz, and it could be a political fabrication of that time. 57 Despite the significance of G e r t r u d e for the Pippinids, it cannot be said that their family cult proves the existence of a wide-spread notion of Adelsheiliger in the Merovingian period.

51

Vita Oditae abbatissae Hohenburgensis, ed. YV. Levison, MGH, SRM VI (Hannover, 1913), pp. 24-29. 52 B. Krusch, in MGH, SRM II (Hannover, 1888), p. 448. 53 Even if the point is rather overstated by M c N a m a r a and Halborg, Sainted Women of the Dark Ages, p. 222. 54 L.C. Ruggini, "The crisis of the noble saint: the Vita Amulfi", Le septième siècle: changements et continuités/The seventh century: change and continuity, eds. J . Fontaine and J . N . Hillgarth (London, 1992), pp. 116-49. Ruggini does not examine the problem of Krusch's dating of texts, nor the varying functions and audiences of texts. 55 Ed. Krusch, MGH, SRM IV (Hannover, 1902), pp. 208-25. 56 YVood, "Forgery in Merovingian Hagiography", pp. 370-1. 57 Paul the Deacon, Liber de episcopis Metlensibus, ed. G.H. Pertz, MGH, SS II (Hannover, 1829), p. 264: Donatus, Vita Trudonis, ed. YV. Levison, MGH, SRM VI,

In fact the case for a broadly disseminated G e r m a n i c notion of Adelsheiliger before the late eighth century is very weak. By contrast there is clear evidence for G a l l o - R o m a n aristocratic family saints in the sixth century. Gregory of T o u r s wrote the Vitae of three of his relatives, Gallus of Clermont, Gregory of Langres and Nicetius of Lyons. 38 H e also makes an indirect claim to be related to the Lyons martyr of 177, Vettius Epagathus. 5 9 H e r e was a family that was making m u c h of its sanctified members, even if Gregory himself is somewhat indirect about asserting his own membership of that family. 60 It was most certainly not a G e r m a n i c family seeking to find in holy relatives compensation for lost pagan sacrality. An even better exampie of Late R o m a n concern to p r o m o t e a recognition of Christian sanctity within the aristocratic kin group is to be found in Avitus of Vienne's epic, the De Consolatoria Castitatis Laude, concerned with the female saints of his family. 61 T h e notion of family saints began in the Late R o m a n period: it should not be endowed with a G e r m a n title, which is entirely misleading with regard to the notion's origin. 62 Despite the fact that saint's Lives often resemble, and indeed borrow f r o m , one another, a n d that the majority of t h e m pedal an image of the Christian Life which is derived in one way or another from the Bible, close concern with the date at which a Vita was produced is a sine qua non of historical research. With dated texts the possibilities for interpretation are considerable. In recent years this has been recognised most dramatically by Walter Goffart in his discussion of "Bede a n d the ghost of Bishop Wilfrid". 6 3 T h e juxtaposition of the Lindisfarne Life of Cuthbert (699-705), 6 4 the Whitby Life of

23: on the context of the composition of the Vita Trudonis, B. Rosenwein, "Association through exemption: Paderborn, Salonnes and Metz", Von Kloster zu Klosterband, eds. H. Keller and F. Nieske (Munich, 1977), pp. 68-87. 58 Gregory, Liber Vitae Patrum, VI (Gallus), VII (Gregory), VIII (Nicetius). 59 Gregory, Liber Vitae Patrum, VI, 1, reveals that Gallus was related to Vettius Epagathus, while the Gregory's own relationship to Gallus is revealed in Liber Vitae Patrum, II, 2, and Liber de Virtutibus sancti Juliani, 23, ed. Krusch, MGH, SRM, I 1. 60 Gregory's family connections have to be reconstructed from a large number of asides, rather than any direct description of his genealogy. 61 Avitus, carm. VI, ed. R. Peiper, MGH, Auetores Antiquissimi VI 2 (Berlin, 1883). 62 An analogous point should be made with regard to Eigenkirchen and Eigenkloster: these were Late Roman in origin, and the use of a modern German noun should not be allowed to imply that the origins of both were in any sense Germanic. 63 W. Goffart, The Narrators of Barbarian History (Princeton, 1988), pp. 235-328. '4‫ י‬Ed. Colgrave, see above, n. 32.

Gregory the Great (after 704), 65 Bede's verse •Life of Cuthbert,(>6 the W e a r m o u t h Lfe of Ceofid (after 716), 67 S t e p h a n u s ' Lfe of Wilfiid (before 720), 68 Bede's prose Lfe of Cuthbert (before 721), 69 his History of the Abbots (before 725) 70 a n d Histona Ecclesiastica (after 731), 71 is a clear indication that hagiography h a d suddenly a n d dramatically become a f o r u m of debate in Anglo-Saxon England, a n d that a n u m b e r of writers writing close to each other in time a n d place were using hagiography to put forward alternative ecclesiastical a n d religious views. T h e sheer range of opinion is in itself an indication of how precise one should ideally be in locating a hagiographical text, its a u t h o r a n d context: exact contemporaries used hagiography to different ends. T h e r e is no need here to go into the question of quite what was being debated between hagiographers in N o r t h u m b r i a d u r i n g the second a n d third decades of the eighth century. W h a t is certain is the significance of clusters of hagiographie texts. T h e y can be found elsewhere in the British Isles, notably in the mid-to-late seventhcentury Irish Lives, beginning with the two Lives of Β rigid, 72 the Lives of Patrick by Tírechán and Muirchú 5 ‫ ׳‬and Adomnân's Life of Columba,74 all texts apparently involved in a competition for religious status in Ireland. 7 3 Clusters can be identified elsewhere in the seventh century. T h e Passiones of Leodegar, Praeiectus a n d G e r m a n u s of Grandval 7 6 all belong to a single period of political crisis. 77 A m o r e sustained cluster

65

Ed. Colgrave, see above, n. 32. Ed. W. Jaeger, Bedas metrische Vita sancli Culhberti (Leipzig, 1935). 67 Ed. C. Plummer, Baedae opera historica (Oxford, 1896). 68 Ed. Colgrave, see above, n. 32. 69 Ed. Colgrave, see above, n. 32. 70 Ed. Plummer, Baedae opera historica. 71 Ed. Plummer, Baedae opera historica: on the date, Goffart, The Narrators of Barbarian History, p. 242, n. 36. n Vita Prima sanctae Brigidae, Cogitous, Vita sanctae Brigidae, ed. Acta Sanctorum, February 1st, vol. I, pp. 1 1 9 . 4 1 ‫־‬ 73 Ed. L. Bieler, The Patrician Texts in the Book of Armagh (Dublin, 1979). 74 Ed. A.O. and M . O . Anderson, Adomnân's Life of Columba (revised edn. Oxford, 1991). 75 See J.-M. Picard, " T h e Purpose of Adomnân's Vita Columbae", Peritia 1 (1982), p. 172, with the reservations of M. Herbert, Iona, K'ells and Deny (Oxford. 1988), p. 146. 7r> Vita Germani Grandivallenh, ed. B. Krusch, MGH, SRM V (Hannover, 1910). 77 See Fouracre and Gerberding, Late Merovingian France, pp. 193-215, 254-70. B(i

or g r o u p of clusters, a n d one with which I w a n t to end this discussion, begins with Willibald's Life of Boniface, written between 754 a n d 768. 78 T h e richness of the hagiographie material relating to the Anglo-Saxons on the continent, and to their Frankish successors, has provided generations of historians with a r e m a r k a b l e seam f r o m which to write the history of the christianisation of Germany. 7 9 T h a t very richness might also suggest that the hagiographers were involved in a debate, or rather a series of debates, similar to those identified by Goffart in early eighth-century England. 8 0 S o m e specific debates have indeed been recognised: thus EigiPs Ltfe of Sturm clearly springs from conflicts relating to Fulda, t h o u g h not necessarily f r o m the conflict between Lull a n d S t u r m described, or perhaps created by, the hagiographer. 8 1 O t h e r conflicts can also be observed in this pool of early Carolingian hagiography. For instance Willibald's Life of Boniface was written between 754 a n d 768. 82 Arbeo's Ltfe of Corbinian was written c. 769 a n d his Ltfe of Emmeram was written in 772. 83 This tight chronological cluster is particularly worthy of note in the light of Willibald's claim that Boniface divided Bavaria into four dioceses, appointing J o h n to Salzburg, Erembert to Freising a n d Garibald to Regensburg. 8 4 T h e fourth diocese, Passau, was left u n n a m e d , not surprisingly, since Vivilo, its bishop, h a d already been appointed by pope Gregory III. 85 Arbeo's decision to write accounts of saints associated with two of these dioceses immediately after Willibald's composition of the first Vita Bontfatii seems particularly significant. 86 In the person of E m m e r a m he h a d a subject w h o h a d been bishop of Regensburg, and whose

78

Ed. W. Levison, Vitae sancti Bonifatii archiepiscopi Moguntini, MGH, SRG 57 (Hannover, 1905). 79 For example S.J. Crawford, Anglo-Saxon Influence on Western Christendom 600-800 (Oxford, 1933): and more recently Brown, The Rise of Western Christendom, and R. Fletcher, The Conversion of Europe (London, 1997). 8,1 I.N. Wood, "Missionary Hagiography in the eighth and ninth centuries", Ethnogenese und Überlieferung, eds. Κ. Brunner and Β. Merta (Vienna, 1994), pp. 189-199. 81 Ed. P. Engelbert, Die Vita Sturmi des Eigil von Fulda (Marburg, 1968). 82 Wood, The Merovingian Kingdoms, 450-751, p. 306. ‫ "י‬Wood, The Merovingian Kingdoms, 450 751, p. 307. 84 Willibald, Vita Bonifatii, 7. 85 Boniface, epp. 44, 45, ed. M. Tangl, Die Briefe des heiligen Bonifatius und Lullus, MGH, Epistolae selectae I (Berlin, 1916). 8 '‫ י‬Wood, The Merovingian Kingdoms, 450 751, pp. 307-8: Wood, "Missionary hagiography in the eighth and ninth centuries", p. 192.

body had been translated by Boniface's own appointee, Garibald. 8 7 In the person of Corbinian he had as subject another bishop, who h a p p e n e d to be the b r o t h e r of Boniface's appointee to Freising, Erembert. 8 8 Further the Vita Corbiniani was addressed to Virgil, bishop of Salzburg, another of the sees supposedly created by Boniface, a see whose history actually went back into the seventh century with bishop Rupert. 8 9 In short, Arbeo seems to have been using hagiography to reply to Willibald, a n d he seems to have been justified in so doing. Further, that Garibald translated the body of E m m e r a m , and that Boniface appointed Corbinian's brother to Freising seems to suggest that Willibald, or his patron Lull, was more concerned to deny the christian past of Bavaria than was Boniface and other of his disciples. N o r was the disagreement between Arbeo and Willibald the only one to surface in the hagiography of the closing decades of the eighth century. Shortly after Alcuin completed the Life of Willibrord in 796 a response appeared in the shape of the Ltfe of Wulfram of Sens, a text which is known to have been copied out, if not composed, by the St Wandrille m o n k H a r d u i n before 807. 90 T h e Vita Vulframni attempted to downgrade Willibrord's mission to R a d b o d , attributing to W u l f r a m near-success in the conversion of the Frisian m o n a r c h a near-success which may well have had some basis in fact, though that success would have occured two decades earlier than the hagiogr a p h e r implied. 91 Unfortunately the hagiographer, w h o claimed to be a near-contemporaiy of his subject, made Wulfram, who is known to have died before 696, outlive R a d b o d , w h o died in 719. For Krusch, historical inaccuracy was one of the chief methods for exposing the fraudulence of claims to contemporaneity. 9 2 In Krusch's terms the Vita Vulframni is unquestionably a forgery: it is, however, a very interesting one, apparently intent on u n d e r m i n i n g Alcuin's views on the importance of Willibrord's work in Frisia.

87

Arbeo, Vila Haimhrammi, 35. Arbeo, Vita Corbiniani, 30: Willibald, Vita Bonifatii, 7. 89 See in particular, Conversio Bagoaùorum et Carantanorum, 1, ed. F. Losek, Die Conversio Bagoaiiorum et Carantanorum und der Brief des Erzbischofs Theotmar von Salzburg, MGH, Studien und Texte 15 (Hannover, 1997). 90 Ed. W. Le vison, MGH, SRM 5 (Hannover, 1910): I.N. Wood, "Saint Wandrille and its hagiography", Church and Chronicle in the Middle Ages: essays presented to John Taylor, eds.' I.N. Wood and G.A. Loud (London, 1991), pp. 3, 13-14. 91 S. Lebecq, "Le baptême manqué du roi Radbod", Les Assises du pouvoir, Mélanges Jean Dévisse (Vincennes, 1994), pp. 141 50. 92 Wood, "Forgery in Merovingian hagiography", pp. 369 84. 88

H a d Krusch edited the Lives of saints active in the second half of the eighth century his criteria for identifying forgeries would have had problems with Liudger's Life of Gregory of Utrecht. 9 3 Liudger was a pupil of Gregory and ought to have known about his master's life, but m u c h of what he included in the Vita Gregorii is either unlikely or impossible. In particular he extended the three years that Boniface spent in Frisia between 719 a n d 722 to thirteen, 9 4 and this must call into question much else in the Lfe, which cannot be supported by other evidence. Even the supposed close association of Gregory and Boniface must be regarded as unlikely, given the lack of reference to Gregory in Boniface's own writings. While the Vita Gregorii is apparently no forgery, it seems to have no authority as an account of Gregory's life. 95 For apparenly reliable information on Gregory's career we must turn to the Lfe of Liudger by Altfrid, which describes Gregory's position, at least during the part of his career spent at Utrecht. 9 6 Altfrid also demonstrates that the very concerns which p r o m p t the most obviously debateable sections of the Vita Gregorii, the problems of establishing a C h u r c h in Frisia, a n d of coping with pagan incursions, were central to Liudger's own career. 9 7 It m a y well be that the Vita Gregorii was Liudger's attempt to set down a personal theory of mission—much as Alcuin, albeit in a more theoretical m a n n e r , did at roughly the same time, and m u c h as R i m b e r t would later use the Lfe of Anskar to explore his personal responses to the missionary life. By contrast with Liudger a n d his Lfe of Gregory, Altfrid in his Lfe of Liudger may well have written a rather more historical text, but that may have been because his concerns, with the sanctity of a relative, a n d one w h o established the diocese of which he was bishop, could be dealt with by what appears to be an accurate narrative. Hagiography, then could be history 98 as much as it could be liturgy, theology, edification a n d p r o p a g a n d a , w h e t h e r spiritual, cultic or

9:!

Liudger, Vita Gregorii, ed. O. Holder-Egger, MGH SS 15, 1 (Hannover, 1887), pp. 66-79. 94 Liudger, Vita Gregorii, 2. 95 O n the conflicting notions of authority and authenticity see Wood, "Forgery in Merovingian hagiography", pp. 369-84. % Altfrid, Vita Liudgeti, ed. W. Diekamp, Die Vitae Sancti Liudgeri (Münster, 1881). 97 Altfrid, Vita Liudgeri, esp. I 14-16, 21 .22‫־‬ 98 See Rydén, above: see also the comments in Fear, Lives of the Visigothic Fathers, pp. xvii-xviii.

political. It was an infinitely flexible m e d i u m , with a vast pool of Biblical or other models to d r a w on. O n e might argue that early medieval hagiographers, in adapting their models for a variety of puiposes, abused what ought to have been a spiritual genre. Such an argument, yet again, underestimates the fact that the majority of hagiographical texts, especially those written for liturgical reading, are first and foremost examples of a spiritual genre, dependent largely on Biblical models. Nevertheless, for the majority of historians, those texts which can be attributed to known authors or placed in known contexts, and which help to elucidate social a n d political issues, hold most interest: but that does not exempt those historians from recognising the complexity of the g e n r e . "

w This paper was not delivered at the Mérida congress. T h e original respondent to Professor Rydén was Professor Alain Dierkens, who decided not to offer his response for publication. In order to keep a balance between eastern and western contributions I have substituted the present paper, which was originally given at the 1993 International Congress on Medieval Studies held at Kalamazoo. A few additions and alterations have been made to pick up on some of Professor Rydén's points and to take account of work done since 1993.

SOCIAL LANGUAGE AND ITS PRIVATE

DEPLOYMENT

Averil C a m e r o n

T h e challenge of this project is to explain change. O u r very title, " T r a n s f o r m a t i o n " , d e m a n d s it. At the same time, the title of this particular contribution (suggested by the organizers) suggests a consideration of the relation of public and private, centralized a n d m o r e diffused, incentives to change. T h e interaction between public and private is a central theme of current historical inquiry, and we must attempt to extend it into the b r o a d e r issues of change with which we are ourselves concerned. O n e of the most obvious manifestations of this process of change is the formation of new states. Even as this project was being planned, and as it went through its early stages, the configuration of Europe and the definition of what it means to be a European nation was subject to p r o f o u n d change. At the same time, some of the older E u r o p e a n countries were themselves undergoing rapid and uncertain social change. O n e can hardly avoid a comparison between the processes of change of this kind in the m o d e r n world and similar ones in the period covered by our project. Such a comparison involves a consideration of the respective types of society and social formation, m o d e r n or otherwise. T h a t is where "language" comes in the other part of my title. I shall therefore consider how language—taking of course a broad definition—may have been instrumental in this process of social change, and how studying it can help us to understand both the nature of the change and the societies themselves. For the latter question—the nature of the societies which in some sense replaced the R o m a n order —is still as open as that of the Late R o m a n order itself. In one of his last published contributions, the anthropologist Ernest Gellncr drew a contrast familiar in other recent works between hierarchical societies (he had in mind eastern Europe before 1989) and "civil societies", borrowing a long-established term that has been enjoying a revival a m o n g political theorists. 1 T h e debate

1

E. Gellner, Conditions of Liberty. Civil Society and its Rivals (London, 1994).

is widespread, not least in East E u r o p e a n countries themselves. Its salient point is that while hierarchical societies are essentially totalitarian, controlling all or most of the social institutions and cultural apparatus, "civil society" has a relatively weak or at least limited government, a n d m a n y other ways in which the citizens can combine in social a n d cultural bonds. Only in limited ways, by comparison with hierarchical societies, does the government dictate matters. C h a n g e , therefore, comes in all sorts of ways, not just top-down through the will of those at the top, but laterally, and from the bottorn. It follows that individuals in civil societies are susceptible to cultural influences from many different directions, whereas individuals in hierarchical societies are "atomized"—exposed and powerless. T h e social organization of civil society is (in Gellner's eyes) characterized by "modularity". It is m o r e like a network than a vertical structure, a n d people typically belong to a variety of different sub-groups. Any one of these sub-groups may change independently of the rest, in which case the social a n d cultural condition of each one of its m e m bers changes too. O f course, changes in individual sub-groups are not usually completely i n d e p e n d e n t — b u t that is the very question of how change happens. In a completely " m o d e r n " society, few people belong to single clearly defined social groups, let alone classes, but instead, are possessed of a high degree of choice a n d freedom of movement between them, and relate to the rest of society in a multiplicity of ways—the social structure is highly permeable; in contrast, in a hierarchical or totalitarian society very little choice is possible and there is little permeability of the social structure. Such a contrast is not the preserve only of academic debate. It represents a m a j o r set of issues which are being debated in the wider political sphere, b o t h in relation to eastern E u r o p e in the postC o m m u n i s t era a n d to western societies generally. "Big government" has been discredited, if only temporarily, and in contemporary Britain we have been witnessing a reconfiguration of politics as a result. Already obvious comparisons suggest themselves with the period covered by our project. T h e Later R o m a n Empire is still sometimes represented in the guise in which it has frequently featured in the older literature, as a totalitarian régime, in which the individual was controlled by terror; 2

2

See notably R. MacMullen, Corruption and the Decline of Rome (New Haven, 1988).

this is a picture that seems to find confirmation in the tone adopted in the Theodosian C o d e and from some of the colourful a n d violent language used, for example, by A m m i a n u s Marcellinus. 3 Punishments laid down in law became steadily m o r e ferocious and bloodthirsty, surely the m a r k of greater authoritarianism. 4 Again on the older view, Diocletian had instituted a system of extreme centralization and autocracy; it is natural then to see its continuation in the later part of the century. T h e treason trials of the 370s and, more broadly, an increasingly hectoring tone in imperial legislation seem to provide the corroboration. It has further been suggested that the very process of Christianization, far f r o m softening the effects of this tendency, actually itself contributed to uniformity, intolerance and "closing-in". 5 T h e argument is still more familiar in its application to the eastern empire and to Byzantium; indeed, Alexander K a z h d a n actually characterised "Byzantine m a n " as "atomized", an individual yet without freedom. 6 T w o responses suggest themselves. First, if a society is as rigid as these depictions suggest, it becomes extremely difficult even in principle for historians to account for change. Yet change certainly did come about. T h e second response is that these descriptions are simply not true. In the case of the Later R o m a n Empire this might seem a banal observation, for the younger generation of scholars have already become accustomed to think of the Late R o m a n State as a creaking and leaky boat rather than a well-controlled ship of state, and of the outpourings of emperors as rhetorical bombast rather than indications of what really h a p p e n e d . Even so, however, the actual structure of Late R o m a n social networks is still very m u c h open to discussion—and essential to understand in any consideration of change. As for Byzantium, the eastern Empire, with its centuries of centralized government, might seem to provide an obvious foil to the fragmented and changing medieval West. But here too, recent scholarship is progressively removing the monolithic aspect of Byzantium, a n d 3

Ammianus , s language: R. Seager, Ammianus Marcellinus. Seven Studies in his Language and Thought (Manchester, 1986). 4 Id., "Judicial savagery in the Roman empire", Chiron 16 (1986), pp. 43 62. 5 Peter Brown, Power and Persuasion in Is!te Antiquity (Madison, 1992); R.A. Markus, The End of Ancient Christianity (Cambridge, 1990); on this idea see Averil Cameron "Ascetic closure and the end of antiquity", Asceticism, eds. V. Wimbush and R. Valantasis (New York, 1995), pp. 147-61. 0 A. Kazhdan and G. Constable, People and Power in Byzantium (Washington, D.C., 1982).

replacing it with the image of a society no less open to change than the West, a n d one that was marked, despite appearances, by a high degree of political instability. 7 I a m suggesting therefore, that there was far m o r e "modularity" about society in our period than might have been at first supposed. This fact m a y help as part of the explanation of the social a n d cultural changes that were taking place. Individuals—the "private" elem e n t of my title—could a n d did bring about change without its having been h a n d e d down from the top; change was indeed lateral and even bottom-up as well as top-down. W h a t then about language? W h e r e does that fit? In the first place, culture itself has often been defined as c o m m u nication, as language. Secondly, language, of varying kinds, is the means whereby social organization is expressed. T h e language/culture m e t a p h o r may be a helpful one. According to this m e t a p h o r , sub-cultures are like dialects of the same language—except that in our case, nothing is static—the dialects are in the process of evolving into languages of their own. T h e process is not, at least after the first phase, one of acculturation, which could more properly be likened to translation from one language to another. Rather, both the main language a n d its dialects are undergoing change. I would like us to look now at some of the ways in which this can be seen to be happening in our period, a n d how the social use of language can be seen to mirror a changing historical situation. O n e such usage which has received some attention within the project is that of n a m i n g — i n this case the n a m i n g of groups or peoples. At at early stage in the project, Working G r o u p O n e considered the process of ethnogenesis—the emergence of social identities, through genealogy, which involves the appropriation of names and the creation of social or collective memory. 8 N a m i n g , in this context, is not something arbitrary and fixed; rather, it requires choice a n d deliberation. But relevant also are the ways in which the "old" ' See notably Kazhdan and Constable, People and Power in Byzantium, and the works of G. Dagron, most recently Empereur et prêtre. Étude sur le "césaropapisme" byzantin (Paris, 1996): and for the Byzantine church, A. Ducellier, L'Église byzantine. Entre pouvoir et esprit (3131204 ‫) ׳‬ (Paris, 1990); political instability: J.-C. Cheynet, Pouvoir contestation à Byzance (963-1210) (Paris, 1990). Generally on perceptions of Byzantium, Averil Cameron, The Use and Abuse of Byzantium, Inaugural Lecture, King's College London, 1990 (London, 1992). 8 See now W. Pohl and H. Reimitz, ed., Strategies of Distinction: the Construction of Ethnic Communities, 300 -800 (Leiden, 1998).

inhabitants refer to the new. Far f r o m being a mere matter of pointless archaizing, the continued use in our period of traditional terms such as "Scythians" for the new barbarian peoples, or the resort when describing them to the familiar vocabulary and categories of ethnography, not only affected contemporary ways of thought but also colours the history of the " b a r b a r i a n " invaders to this day." As Working G r o u p O n e has found in its attempts to define ethnicity, the self-definition of new groupings, and eventually of new states, cannot be studied in isolation either from the names adopted by themselves and their contemporaries or from the use of these names in m o d e r n historiography. T h e issues involved are lasting ones: they recur centuries later, for example, in the first uses of the term Rus, and the names by which the Byzantines of the eleventh century and later referred to the Scandinavian founders of Kievan Russia; 1 " in the same way they have recurred yet again and are still of critical importance in recent Balkan history. N a m i n g is important in other ways too. O n e of the problems inherent in a project such as this one is precisely that posed by the practical necessity of dividing up the subject into discrete segments for group work. In practice, several groups found that they were pursuing the same themes, notably those of power a n d representation. W e have had proposed as topics of study rituals and power, images and power, and group representations, all of which concern every scholar working in the period, of whatever discipline; I suspect that it will become increasingly difficult to separate the work of the individual groups. This is not simply because certain ways of approach are either m o r e or less in vogue, but rather because in studying change we are not (by definition) studying fixed entities. Similar problems occur in the choice of subject-divisions for encyclopaedic or multi-author works on this period, and for the same reasons. Staying with the conventional divisions will only lead us back into the existing dead-ends; they will not enable us to escape them or to do justice to a situation where m u c h was fluid, let alone to grasp the very nature of that fluidity. For this reason alone, the very modes of communication, the ways in which overlapping social groups, circles or 3

W. Pohl, "Tradition, Ethnogenese und literarische Gestaltung: eine ZwischenBilanz", Ethnogenese und Überlieferung. Angewandte. Methoden der Frühtnillelalterforschung, eds. K. Brunner and B. Merta (Munich, 1994), pp. 9.26‫־‬ 10 See E. Patlagean, "Nommer les russes en grec, 1081 1204", in Le origini e 10 sviluppo délia cristianita slavo-bizantina, .Nuom Studi Storni 17 (1992), pp. 123-41.

networks were in fact permeable, must indeed be central to our investigation. I would like to suggest here that the notion of permeability, if not quite "modularity" in Gellncr's sense, can be a useful way of approaching our topic. W e are of course attuned as historians to the diachronic view—to measuring change over time. All the same, it sometimes helps to think laterally—to ask ourselves how ideas, habits, rituals are transmitted between groups a n d individuals, what are the actual mechanics by which social m e m o r y is created." In our period, not only new states, but also new states of consciousness were created— all the more reason, then, to look at the how of the process. It m a y be felt that it is beyond us to track the moral a n d subjective history of a period so lacking in the necessary source materials. Yet others have tried it in the past when seeking to define the essence of the "medieval", even without the benefit of our wealth of archaeological material. R e t u r n i n g to our own day, it is already a c o m m o n p l a c e that private behaviour, a n d even private thought, can be read as enactments and internalizations of social language. Even if we watch little television, rarely go to the cinema and never read the popular press, each one of us is both influenced and constrained by the messages they deliver. In turn those messages are influenced a n d constrained by material a n d economic conditions, and by the institutional framework of society. T h e material evidence from our period of study can surely provide some of the same tools for understanding, as indeed is explored in m a n y existing studies, not least the articles in volume I of the enterprise edited by Philippe Ariès and Georges Duby, the History of Private Life}"1 It will be distincdy easier to understand the "private", however, if we adopt a diffused or distributed rather than a centralized model of cultural communication. T h e r e are, I suggest, indications of uncertainty in our period about the status of knowledge itself. This is hardly surprising, in a situation of rapid change. T a k e first the dissemination of learning, of which as medievalists we are highly conscious. It is a consciousness shared by o u r sources. Late antique Christianity was a bookish religion; not only did leading c h u r c h m e n write energetically themselves—they also explicitly debated the p r o p e r nature of education, the texts to

" James Fentress and Chris YVickham, Social Memory (Oxford, 1992). 12 See especially P. Ariès, G. Duby, eds., Histoire de la vie privée I, ed. P. Veyne (Paris, 1986).

be read and the status of secular versus divine learning. Cassiodorus's Institutes followed naturally from Augustine's De Doctrina Christiana and City of God. O n e is struck at once by its scholastic approach. Divine learning, according to Cassiodorus, fell quite obviously into thirtythree headings, since that was the age of Christ when by dying he gave eternal life; secular learning was divided into seven heads, just as David praised G o d seven times in a day, Wisdom built seven columns a n d G o d said to Moses, " M a k e seven lamps". 1 3 T h e reorganization and reordering of the sum of knowledge into a proper Christian whole was a m a j o r preoccupation of m a n y early medieval writers. T h u s the writings of Isidore of Seville combine history, exegesis a n d encyclopaedism; his hand-list of the works of fifthand sixth-century ecclesiastics 14 provided a key to Christian writing, his Etymologies and Divisions of Mature a c o m p e n d i u m of h u m a n and natural knowledge. Different in kind, in that the secular is excluded altogether, is J o h n of Damascus' Fount of Knowledge, but this too is effectively an encyclopaedia of orthodox faith. T h e m a n y catalogues of heresy and arguments against Jews known and surviving from our period (both types are represented in the works of Isidore and J o h n ) are far more a subset of this zeal for organization and prescription of knowledge than they are a guide to the actual religious situation. Some of these works were m a d e easier to put together by the compilation of collections of useful extracts—florilegia. T h e y might be of m a n y types—collections of quotations from Scripture or from the Fathers chosen to prove a doctrinal position, perhaps pro- or antiChalcedon, or later, pro- or anti-Monothelite. T h e r e were also antiJewish florilegia ready to h a n d . Any religious dispute, such as the debate about images, called forth new collections. Indeed, elaborate precautions had to be taken by the seventh- and eighth-century councils to verify their sources, especially after leading eastern clerics had been found to be introducing bogus citations. N o self-respecting ecclesiastical author would d r e a m of failing to add his own florilegium or list of heresies to his latest treatise or prise de position. In the early seventh century Sophronius of J e r u s a l e m a p p e n d e d just such a list to his synodal letter. It did not perhaps matter very m u c h if the material had been used before—witness the extent of J o h n of Damascus'

13 14

Cassiodorus, Inst. II, praef. 1-2. De viris illustnbus.

borrowing f r o m the fourth-century heresiologist Epiphanius—for the object was not (as it might be nowadays) originality. R a t h e r , the purpose was to reclaim the past and mould it to one's own image. This was the creative m e a n i n g given to "tradition", a concept greatly valued in this period, a n d one that was claimed, like the label "orthodox", by all parties. T h e more inventive one's range of citations from past writers, the stronger one's claim. Alongside florilegia c a m e another range of literature, less formal in character—teaching-aids, catecheses, questions a n d answers. W h o was reading them? T h e seventh-century Palestinian writer Anastasius of Sinai called his long a n d technical treatise against Monophysites a Hodegos—a "Guide to Life". It is a pretty dry guide by m o d e r n standards. All the same, while little is known of the audience of this or of another c o m m o n genre, the Adversus Iudaeos texts which continued to be written in our period by both well-known and obscure authors, their level of inter-dependence shows that they a n d the florilegia on which they drew must have circulated widely. W e know equally little of the processes by which popular apocryphal texts, saints' lives and monastic collections or edifying stories came to be translated into so m a n y different languages, east a n d west. Yet used, read a n d copied they undoubtedly were. T h e tradition of the single learned a u t h o r continued, but is set against a background of anonymous writers and readers. Not surprisingly, this gives us m a j o r problems of attribution. But the p h e n o m e n o n of anonymity is interesting in itself. W h o were the writers, a n d who the readers? Merely to describe them as "monastic" does not seem to be adequate; somehow we have to a c c o m m o d a t e the notion of a reading public interested in improving and edifying reading materials. T h e concept of the " D a r k Ages" is one which needs as m u c h discussion as that of "fall" or "collapse". For some, the D a r k Ages are dark because they are u n d o c u m e n t e d in the way we would like— that is, by secular, and to some extent analytical, histories. 15 But for how m a n y periods in history is this expectation actually fulfilled? It is in any case for us an extremely self-serving expectation. I suggest rather that we should think in terms of an energetic striving after knowledge. In the East, at least, the sheer volume of surviving textual

15

T h e lack of traditional secular histories in Greek after Procopius is discussed by Michael Whitby, "Greek historical writing after Procopius: variety and vitality", The Byzantine and Early Islamic Near East I: Problems in the Literary Source Material, eds. Averil Cameron and Lawrence I. Conrad (Princeton, 1992), pp. 25-80.

material in Greek and in languages such as Syriac, Georgian and Coptic, even without considering Arabic, is quite at odds with the view of a descent into ignorance. Recent work shows also the important impact of Palestinian Christian culture in seventh-century Italy and Sicily, a n d its continued influence in Constantinople after the A r a b conquests and through to the iconoclast period. 1 6 If on the whole as historians we are still looking for "classical" works a n d for continuities, and have not yet learned to deal with what is in fact available, that is our own problem: the fact that our eyes are often still resolutely closed does not m e a n that the period itself must be "dark". This is particularly true of the East. O p e n i n g our eyes, however, entails the re-evaluation of well-known works—the Histories of Gregory of T o u r s is a good example, as is the Chronicle of J o h n Malalas in the sixth-century East. In each case, what used to seem like naïve credulity is now read as original and indeed creative. 1 ‫׳‬ T h e demonstration of spiritual truth, whether in chronicle or history d e m a n d s its own subtle methodologies, very different from the analytical m a n n e r of a Thucydides. So if in our scholarly language "transformation" is to be preferred to "fall", I suggest that the search for secularity on our own part should be another casualty of this revisionist mood. It might seem at first sight that there was greater control of knowledge in the medieval than in the classical period. Certainly elements of the C h u r c h tried hard from time to time to c o n d e m n what they did not approve themselves. Heresy took over from paganism as the arch-enemy. Further, the authority of ecclesiastical pronouncements, and the enforcement of "officiai" views b e c a m e on the face of it easier as bishops acquired m o r e power. In these changing societies, C h u r c h and C o u r t b e c a m e increasingly the centres of patronage, and of intellectual activity. Sometimes they succeeded in acting in unison to purvey a powerful political message. Justinian, for instance, clearly wanted this to happen. 1 " O n e has the impression, perhaps

16

Recently Cyril Mango, "Greek culture in Palestine after the Arab conquest", Scritture, Librì e Testi nette Aree Provinciali di Bisanzio, eds. G. Cavallo, G. de Gregorio, M. Maniaci, Ath del seminario di Erice, 15 18 sett., 1988 (Spoleto, 1992), pp. 149-60. 17 In the case of Malalas, understanding has been transformed by the work of a group of Australian scholars, for which see Elizabeth Jeffreys et al., John Malalas. A Translation (Melbourne, 1986); Elizabeth Jeffreys, with Brian Croke and Roger Scott, eds., Studies in John Malalas (Sydney, 1990). 18 And has accordingly been seen as "Caesaropapist"; for discussion, see C. Capizzi, Giustiniano I tra politico e religione (Messina, 1994), pp. 151 64.

just because of the relative absence of a secular alternative, of a long series of strident proclamations of ecclesiastical a n d political unity. Justinian, still ruling over the apparatus of R o m a n culture, declared himself the restorer of the past. But emperors and kings alike of necessity allied themselves with the C h u r c h and sought to use it as they struggled to control their realms. Justinian was also a legislator, against pagans, heretics and Jews. Indeed, our period is characterised by increasingly severe legislation against J e w s and J u d a i s m . Yet the level of control actually attained by C h u r c h or State was in reality low. A strident rhetoric of condemnation stood for respectability, and for a power that was often enough lacking in fact. Knowledge was, in actuality, if not in theory, distributed. Some centres of learning, like the Lérinian or the J u d a e a n monasteries, h a d little to do with political power structures: their authority was differently derived. 19 T h e very level of official p r o n o u n c e m e n t s against so-called heresy, the stubbornness with which "old" heresies keep on recurring—sometimes u n d e r different names, a n d the struggles with which governments tried to suppress them, all point to actual plurality. Just as "the C h u r c h " was represented in each instance by its most forceful or prestigious representatives—it does not yet act as a united institution e m b e d d e d in the social structure—so "knowledge" was a matter of what one could get a n d where one could get it. T h e movement of peoples was of course in itself one of the main factors in bringing this situation about. A n o t h e r was, particularly in the East in the seventh century, the crumbling of the Late R o m a n administrative structure which h a d required a steady supply of m e n trained in the classical style, a n d for which the cities, most of which now shrank into defensive enclaves, provided the necessary educational facilities. Aphrodisias in Caria, a city without any particular links with the capital or the administration, provides a striking example: here, in the sixth century, well-off families sent their sons to Alexandria or elsewhere for training in rhetoric and philosophy, whereas by the late seventh century this city, that had been one of the most important centres of the sculptor's art in Late Antiquity, had shrunk into insignificance and adopted a new name, Stauropolis, the city of the Cross. 20

19

For the J u d a e a n monasteries in the period see J . Patrich, Sabas, Leader of Palestinian Monasticism (Washington, D.C., 1995); Y. Hirschfeld, 7he Judaean Desert Monasteries in the Byzantine Period (New Haven, 1992). 2(1 Charlotte Roueché, Aphrodisias in Late Antiquity (London, 1989).

H e r e there is an i m p o r t a n t difference between East a n d West, for the fragmentation of the R o m a n imperial structure at an early date in the West m e a n t , paradoxically, that some towns were able to develop as intellectual centres in new ways, a n d that the traditions of secular learning could be continued or absorbed into newer structures without the same degree of d e p e n d e n c e on central government a n d administration. T h e Latin letter collections a n d other literary works of western ecclesiastics in the late sixth c e n t u r y — I a n W o o d has recently d r a w n attention to Merovingian examples 2 1 —are not so easy to parallel in the East, where letters were indeed written in plenty, but were mainly of a m o r e theologically utilitarian kind. T h u s an epistolary culture already well established in the late fourth a n d early fifth centuries continued, even if with suitable variation, irrespective of political changes, a n d great episcopal figures continued to see themselves as intellectuals a n d cultural patrons as well as pastoral teachers. A n o t h e r way in which traditions of R o m a n writing a n d record-keeping persisted in a different environment was in the bureaucratic records kept by the papacy. 2 2 In contrast, we know little about the patriarchate of Constantinople in this regard, except that it possessed a library a n d kept copies of d o c u m e n t s pertaining to C h u r c h councils; m a n y works necessary for citation in the debates of the iconoclastic period were however missing from the patriarchal library, a n d the iconophile side engaged in a feverish search for copies elsewhere. 2 3 Until the "revival" of learning in Constantinople of a r o u n d the year 800, one has the impression that there was m o r e continuity of learning in the West, for all its political decentralization. After it, despite a genuine impulse towards literary revival, forgery, interpolation a n d fabrication in religious texts were, if anything, m o r e rife in Byzantium than before, as recent history was rewritten or texts doctored to prove a point. T h e location and nature of schools, the possibilities of travel, the existence of patronage—all these determined the dissemination, a n d

21

Ian Wood, The Merovingian Kingdoms, 450 751 (London, 1994), p. '242. See Thomas F.X. Noble, "Literacy and papal government in late antiquity a n d the early middle ages", The Uses of Literacy in Early Medieval Europe, ed. R. McKitterick (Cambridge, 1990), pp. 82 108. 23 See P. Van den Ven, "La patristique et l'hagiographie au Concile de Nicée de 787", Byzantion 25-27 (1955-57), pp. 325-62; Cyril Mango, " T h e availability of books in the Byzantine empire, A.D. 750 850", Byzantine Books and Bookmen (Washington, D.C., 1975), pp. 2 9 - 4 5 . 22

often enough the n a t u r e of knowledge. Again in the West m u c h m o r e than in the East, law-codes a n d legal d o c u m e n t s were equally important; it is m u c h less clear whether, despite the thousands of seals which survive from the seventh century onwards, early medieval Byzantium can be described, as the Merovingians have been, as a "bureaucratic society". 24 O f course knowledge, in the form of intellectual training, was available only to the very few. But it was not the only, or even the main, m e a n s of acquiring social language. W e m a y regard the level of debate in both East a n d West about the respective merits of pictures a n d writing as a n o t h e r manifestation of this uncertainty a b o u t the status of knowledge. 2 3 I have suggested already that religious images, viewed as ways of attaining access to the truth, functioned as a sign system c o m p l e m e n t a r y to written or verbal knowledge. 2 6 T h a t such a system should have itself developed is a n o t h e r indication of lateral c o m m u n i c a t i o n a n d a further m a r k e r of change. But that so intense a dialogue should have developed between the supporters of words a n d the advocates of pictures is m u c h more interesting: it shows us a society in change, where f u n d a m e n t a l s are being renegotiated. P e r h a p s we should read the emphasis on right belief, a n d the attempt to enforce it, in a similar way. Anthropologists lay stress on the language of distinction, the laying d o w n of categories whereby it is hoped that society will be regulated. 2 ' Heresiologies, polemics a n d religious rulings all fall into this class, in o u r period. O f course there is ample evidence of the ruling powers seeking to impose their authority in these ways. But it would be a mistake to take t h e m too 24

Ian Wood, "Administration, law and culture in Merovingian Gaul", The Uses of Literacy in Early Medieval Europe, ed. R. McKitterick, p. 63. 25 See J.-M. Sansterre, "La parole, le texte et l'image selon les auteurs byzantins des époques iconoclaste et posticonoclaste", Testo e Imaggine nell' alto medioevo, 1521 aprile 1993, Settimane di Studi del Centro suit alto medioevo 41 (Spoleto, 1994), pp. 197-243; H. Kessler, "'Pictures fertile with truth': how Christians managed to make images of God without violating the Second C o m m a n d m e n t " , Journal of the Walters Art Gallery 4 9 / 5 0 (1991/92), pp. 5365‫־‬, especially pp. 59 ff.; L. Brubaker, "Perception and conception: art, theory and culture in ninth-century Byzantium", Word and Image 5 (1989), pp. 19-32, especially p. 27 on the function of images and pp. 28 f. on images and text. 26 Averil Cameron, " T h e language of images: the rise of icons and Christian representation", in The Church and the Arts, ed. D. Wood, Studies in Church History 28 (Oxford, 1992), pp. 1 - 4 2 . 27 P. Bourdieu, Distinction, Eng. trans. (London, 1986), pp. 466 ff.; in relation to material culture: M. Douglas and B. Isherwood, The World of Goods: Towards an Anthropology of Consumption (London, 1979).

m u c h at face value as indications of a genuinely hierarchical system, for they can be found e m a n a t i n g from m a n y different sources, not just from the top down, a n d they too are accompanied by an intense level of competitive disagreement. A functional explanation sees such manifestations as indicative of a society u n d e r strain. 28 But they can equally be read as indicating a high level of pluralism and interaction between that society's members, in m u c h the same way as Gilbert D a g r o n reads the m u s h r o o m i n g during this period of imaginative and exotic tales and explanations as indicative of a whole territory of inventiveness and experiment. 2 9 T h e "questions" and the sometimes bizarre answers found in collections of the period, like the catalogues of miracles and "true-life stories" represent the other side of the coin of the heresiologies and ecclesiastical pronouncements; without them, the latter have no substance. Miracles, too, have a role not just as signs—which indeed, as tekmeúa or semeia, they were f r o m the b e g i n n i n g — b u t also as social indicators, points of intersection in the communications network. Moreover, we see most clearly of all in the contemporary miracle collections, if not the exact process of communication (though often enough that is in fact indicated), then at least and clearly the fact itself. Packed with circumstantial detail, the collections bring before us people—individuals—of all classes, whose consciousness has been reached by awareness of a shrine, a holy m a n , or a possible cure. Lateral communication is patent for all to see. T h e particular process of change with which we are concerned is only partly a matter of acculturation, for acculturation is asymmetrical—it implies the adaptation of the weaker to the stronger, whereas we have to do with adaptations, translations and creativity on multiple levels. Some scholars have seen "the birth of the Middle Ages" in terms of a breaking down of classical order and a confusion of styles. T h a t too is one-dimensional: it imposes value-judgements and presupposes decline. Yet it is a more useful model than that of acculturation, just because it highlights change and fluidity—and from these new things can be born. In a context in which there was, as we have seen, a visible striving for authority, I suspect that it is going

2

" J. Halcion. Byzantium in the Seventh Century•. The Transfonruition of a Culture (Cambridge, 1990). 29 See especially "Le saint, le savant, l'astrologue", in id., IM romanité chrétienne en Orient (London, 1984), IV; Constantinople imaginaire (Paris, 1984).

too far to talk in terms of a democratization or simplification of culture. 30 Nevertheless, in all changing societies the existing balance is a n d must be upset. Old mechanisms of social a n d cultural reproduction are no longer present or no longer work in the same way. T h e y may be replaced by a degree of cultural opportunism. I would suggest that in such circumstances the individual, and private life, cannot help but be affected, a n d that the relation between public a n d private is necessarily disturbed. Michel R o u c h e has written of our period in these terms: The new governments established in fifth-century-Gaul, be they Visigoth, Burgundian or Frank, had little success imitating either the political institutions or the social arrangements of the Roman empire. From the court down to the lowest official, in city, and countryside, in religious and professional organizations, private persons took center stage.31 W e might not want to go as far as that. But this first plenary conference was concerned with communication. H o w then in this period could individuals be influenced to learn and to adopt new social languages? T h a t I think is the question which has united all the scholars a n d all the Working G r o u p s in the project. I began by suggesting a model of permeability. How, then, to conelude, was that permeability effected? Some very obvious factors are: travel, or rather, m o v e m e n t or displacement (for the travel m a y not be voluntary); the visual environment (exposure to different kinds of buildings, different artefacts, unfamiliar settlement patterns or urban topography); linguistic variety a n d the need for translation—these are only some of the possibilities. 32 I would like to see us in this project looking not for continuity, but for mixture and for juxtaposition; emphasising openness instead of searching for derivations; thinking in terms of multiple new self-identities and imagined communities as well as the continuation of old ones. I take "social language", or perhaps better, "social grammar", 3 3 to be the way in which the organization of society is expressed a n d established. Every society's identity is expressed through a system of

30

For which see Peter Brown, The World of Late Antiquity (London, 1971), p. 174. History of Private Life I, Eng trans. (Cambridge, Mass., 1987), p. 419. 32 D. Lowenthal, The Past is a Foreign Country (Cambridge, 1985), explores the effect of such factors in the recreation of a sense of the past; they are equally potent in the creation of new self-identities. 33 See C.J. Berry, The Idea of Luxury (Cambridge, 1994), p. 37. 31

information and communication (not necessarily verbal) of which this social g r a m m a r is a part. 3 4 It embraces all types of representation, whether verbal, visual or matters of praxis. As culture embraces the whole of society and its practices, so social language cannot but be expressed in private life as well as public. T h e t h e m e of c o m m u n i c a t i o n , it seems to me, comes near to including the whole of what this project has been about. Above all, it implies movement, and therefore it accommodates change.

34

Douglas and Isherwood, The World of Goods (see above n. 27).

SOCIAL LANGUAGE, IDENTITIES AND THE CONTROL OF DISCOURSE Walter Pohl

W h a t do the private d e p l o y m e n t of social language, knowledge, m o d e s of b e h a v i o u r , l e a r n i n g a n d subjectivity have to do with "Imperium, gentes et régna" (the topic of the working group [1] of the T R W project), a n d with ethnogenesis and the development of barbarian kingdoms (my own field of studies)? I think that this is the kind of "cross-over" and of experimenting with new concepts that large-scale projects like the " T r a n s f o r m a t i o n of the R o m a n W o r l d " can a n d should stimulate. In this paper, I can only offer a few miniatures in which language, communication, learning, and subjective identities become relevant within a confrontation between the Empire a n d the gentes: small incidents that culminate in personal tragedy for individuals subjected to tensions between powers beyond their comprehension. As a starting point, I can take the double question Averil C a m e r o n proposed in her paper: W h y do people act as they do— and why do they think as they do? This also implies a relationship between thinking and acting, between concepts and decisions, between the objective and the subjective, between language and power, that in m o d e r n scholarship has sometimes been obscured by specialization. Political and social history on the one h a n d and the history of culture and literature on the other hand, while retaining their specific sets of questions a n d methods, should therefore be studied as aspects of a whole—as was admirably achieved in most of the contributions to the Mérida conference. T h e same applies to archaeology, the most dynamic discipline in early medieval studies; the experience of the group on "Power a n d Society", for instance, has certainly shown that a symbolic reading of archaeological evidence can offer some very attractive hypotheses. T h e term "social language", proposed by the organisers of this conference (one might also use "social grammar"), can provide a useful label to cover both verbal a n d non-verbal communication, discourse a n d representation. Averil C a m e r o n has crossed this wide field with a broad survey of various kinds of text, taking us as near

to the individual and the private as we can get in Late Antiquity. 1 She has reminded us that even a large-scale process like the transformation of the R o m a n World can be seen as an aggregate of modules, of micro-histories and singularities. O u r evidence does not allow us to reconstitute the individuals of a distant past as sovereign subjects of a historical process, and indeed recent philosophies may warn us that we should not try to do so. 2 But instead, traces of specific practices a n d perceptions can be a key to the understanding of structures; structures that are not prior to the event, but constituted by it. Structural history, even w h e r e it reaches out for processes of "longue durée", consequently cannot brush aside events and the narratives that relate them. 3 A closer reading of events as singularities can show a wide range of modes of behaviour that result in suecesses and failures to adapt to changing circumstances. Let me, therefore, a t t e m p t a brief glimpse on four m i n o r events; the first as described by an a u t h o r who again reminds us of Averil C a m e r o n ' s work, Procopius. 4 In the course of his Gothic War, Procopius's narrative briefly touches u p o n the tragic fate of Gilakios, an A r m e n i a n by birth, the comm a n d e r of a small force of Armenians: This Gilakios did not know how to speak either Greek or Latin or Gothic or any other language except Armenian alone. When some of the Goths happened upon this man, they enquired who he might be. For they were quite averse to killing every man who came in their way, lest they might be compelled to destroy each other in fighting at night, as might easily happen. But he was able to make no answer except indeed that he was Gilakios, a strategos; for this title which he had received from the emperor he had heard many times and so had been able to learn it by heart. The barbarians, accordingly, perceiving by this that he was an enemy, made him a prisoner for the moment, but not long afterwards put the man to death. 3

1

See, for instance, A. Cameron, Christianity and the Rhetoric of Empire (Berkeley, 1991): eadem, History as Text (London, 1989): and her contribution to this volume. 2 See, for instance, J . Derrida, L'écriture et la différence (Paris, 1967): M. Foucault, L'ordre du discours (Paris, 1971). 3 For a classic statement against "l'histoire événementielle" see the introduction to F. Braudel, La Méditerranée et le monde méditerranéen à l'époque de Philippe II, vol. 1 (Paris, 1949). Proposing a "diachronic structure" that can relate events and structures: R. Koselleck, Vergangene Zukunft (Frankfurt, 1989), pp. 144-57. 4 A. Cameron, Procopius and the Sixth Century (London, 1985). 5 Procopius, De Bello Gothico (BG) 7, 26, 24-7, ed. H.B. Dewing (Cambridge, Mass., 1914-28).

I have chosen this example for a n u m b e r of reasons. As far as the title of this conference is concerned, this story symbolises an attempt to communicate between East and West that completely failed, with fatal consequences. Moreover, it is one of not too m a n y references to languages and language problems in historiographie sources of the period. The context is multilingual, as was usual in the R o m a n army, and in Ostrogothic Italy with its trifariae linguae,6 just as it was, for instance, at Attila's court, in P a n n o n i a u n d e r Avar rule, in L o m b a r d Friuli (or, of course, in Palestine, where Procopius came from).' Quite exceptionally, a problem to make oneself understood arises here. It is, moreover, quite extraordinary that in this specific case, language is shown to function as a sign of identity. Theoretically, this is what m a n y authors assume (for instance Isidore of Seville). 8 But we rarely hear that language actually served for distinction, as in this case, between friend and foe. O f course, we are left to wonder if in an army as heterogeneous as Totila's that would have been a practical solution. Was his inability to c o m m u n i c a t e the reason that Gilakios was put to death? In any case, a soldier's fate obviously could depend on the power of words, on his ability to persuade or negotiate. W e know f r o m m a n y examples given by Procopius that the treatment of defeated enemies could be negotiated, and that it varied according to specific identities; Totila once excluded Sicilians f r o m his clemency because they h a d been the first to betray the Goths, 9 whereas Isaurians enjoyed special favour because some of them had h a n d e d over Rome. 1 0 As a rule soldiers were given the choice to join the winning army or leave," quite unlike civilians in captured towns w h o ran a much greater risk of being massacred. 1 2 W h a t , at first glance, seems a straightforward account of the fate of an individual, could, however, also be deciphered as a parable or 6

Cf. Β. Meyer-Flügel, Das Bild der ostgotisch-römischen Gesellschaft bei Cassiodor (Bern, 1992), p. 396: P. Amory, People and Identity in Ostrogothic Italy, 489 554‫( ׳‬Cambridge, 1997). 7 W. Pohl, "Strategie und Sprache‫ ־‬zu den Ethnogenesen des Mittelalters", in Entstehung von Sprachen und Völkern, ed. S. Ureland (Tübingen, 1985), pp. 93 102: id., "Telling the difference: signs of ethnic identity", in Strategies of Distinction, ed. W. Pohl (Leiden, 1998), pp. 17-69. R Isidore, Etymologiae 9, 1, 14, ed. VV.M. Lindsay (Oxford, 1910). 9 Procopius, BG 7, 16. 10 Procopius, BG 7, 21. " Urbino: Procopius, BG 6, 19: Faesulae, Auximum: BG 6, 27; Cottian Alps: BG 6, 28. 12

Milan: BG 6, 21: Tibur: BG 7, 10.

a piece of fiction to add depth a n d colour to the narrative. T h e r e is a series of rather bizarre ignoble deaths in Procopius which usually point towards some moral. 1 3 This problem of source criticism frequently occurs in historiography and has also been extensively discussed in our group in connection with references to treaties and their specific rhetoric. In this case we may ask: Would there really have been generals in Justinian's armies who did not even know the most rudimentary military language? This is certainly what Procopius wants us to conclude, not without a note of implicit criticism reminiscent of an a r g u m e n t in the Anecdotal T h e way in which Gilakios is depicted, stubbornly echoing the word strategos like a parrot, is ironic to the point of caricature. Procopius is more outspoken in the speech he puts into Totila's m o u t h before the battle of T a d i n a e (552), in which the king of the Goths says: The vast number of the enemy (i.e. the army led by Narses) is worthy only to be despised, seeing that they present a collection of men from the greatest number of peoples (έθνή). For an alliance (συμμαχία) which is patched together from many sources gives no firm assurance of either loyalty or power, but being split up in origin (τοις γένησι) it is likewise divided in purpose. 15 This statement, however, is not without irony either, being spoken by someone about to lose a battle a n d his life against the a r m y he has thus decried; being himself, of course, in charge of an army as mixed in origin. O n e might add that the Ostrogoths themselves had originally become what they were in such a συμμαχία with the R o m a n army. T h e p a t c h e d - u p alliance at T a d i n a e that Totila referred to included over 5000 L o m b a r d s , for w h o m the συμμαχία with the Empire was to pave the way to general recognition and to dominion over Italy as well. 16 T h e implicit criticism of Justinian's b a r b a r ian policy in the Wars (and its explicit counterpart in the Anecdota) belongs to a series of similar, a n d often very rhetorical, statements

13

For instance in the case of Rikilas, BG 7, 11. Procopius, Anecdota 8, 5 - 6 ; 11, 3-11; 19, 13-16; 24. 15 Procopius, BG 8, 30, 17. IG H. Wolfram, Die Goten (3rd ed., Munich, 1990): W. Pohl, "The F.mpire and the Lombards: Treaties and Negotiations in the Sixth Century", Kingdoms of the Empire: The Integration of Barbarians in Late Antiquity, ed. W. Pohl (Leiden, 1997), pp. 89-98. 14

throughout Late Antiquity, a tendency that gained force towards the end of the emperor's reign, as Evangelos Chrysos has shown. 17 T h u s , we might expect that for the "model reader" (according to U m b e r t o Eco's terminology)"' of the Wars, the Gilakios episode would ring a bell. If we look at this kind of context, we need not worry too m u c h whether Procopius's story is a factual account or a piece of pure fiction, or whatever mixture of the two. It has rich potential to exemplify the ways in which the account of an incidental casualty is explicitly or implicitly presented as meaningful. In this context, the contradiction between event and narrative, between reality a n d the text, which has inspired so m u c h radical rhetoric in recent years can, at least partly, be bridged. T h e call for caution in taking "fiction of fact" for factual history has been an important challenge to historical research (although traditional source criticism should to some extent be equipped to deal with this problem). 1 9 But if we not only ask whether something has actually h a p p e n e d or not, but also look for what could have h a p p e n e d , what was j u d g e d possible or impossible, right or wrong, we can regain some confidence in the text, and the discourse in which it is e m b e d d e d . In the context of our little story, Gilakios might also be understood as an emblematic figure for a process of learning and a construction of identity brought about by the R o m a n army. Averil C a m e r o n has put a lot of emphasis on learning; and the army certainly had some part in its diffusion, at least up to the end of the sixth century. T h e e m p e r o r has appointed Gilakios strategos, a n d he is frequently addressed by that title; thus, in extreme danger, and under pressure to disclose his identity, he is able to repeat his Byzantine title. H e does not try to say that he is A r m e n i a n , or Greek, or R o m a n (all of these would have been appropriate labels in his situation); he defines himself as a strategos. Identity, and also ethnic identity is a "situational construct", an anthropological concept introduced into medieval studies by Patrick Geary.‫ "'־‬Conflict and confrontation

'‫ י‬Unpublished lecture in Vienna, 1989. 18 U. Eco, Lector in fabula. IM cooperazione interprelativa nei lesti nanativi (Milan, 1979). 19 Cf. Hayden White, The Content of Form (Baltimore, 1987): A. Fletcher, ed., The Literature of Fact (New York, 1976): W. Goffart, The Narrators of Barbarian History (Princeton, 1988): for exemplary discussions and a reappraisal of historical source criticism, see A. Scharer and G. Scheibelreiter, eds., Historiographie im frühen Mittelalter (Vienna, 1994). 20 P. Geary, "Ethnic Identity as a Situational Construct", Mitteilungen der anthropologischen Gesellschaft in Wien 113 (1983), pp. 15 26.

can become the keys to testing it. As usual, we are left to guess what the reality behind the ethnic a n d social labels that Procopius uses so freely really is. Within the R o m a n army, they make a lot of sense; they describe different categories of troups, and often also denote some special skill; if Belisarius wanted to lay an a m b u s h , he needed Slavs, 21 whereas fighting on difficult ground was the speciality of the Antes. 22 But outside its ranks, this system of denominations could easily lose its significance. From a n u m b e r of examples of professional or class solidarity cutting across ethnic or political bonds in Procopius, we m a y assume that any captured officer might have deliberately played on his status rather than mentioned his nationality; at least this could show that he was worth a ransom, the most c o m m o n fate of prisoners in early medieval wars. But then again, such an opportunity would have to be negotiated, a n d that is where Gilakios failed. Let us look at a similar situation with a different outcome, in an episode from the Balkan wars of the 580s as related by Theophylact Simocatta about half a century later. 23 A Byzantine soldier, n a m e d Busas, is caught by a b a n d of Avars who threaten to kill him. He, unlike Gilakios, manages to persuade them to lead him to the walls of his native town of Appiareia a n d accept ransom (we do not hear about any language problems in this case). But there his persuasive power fails: in spite of a long speech in which Busas enumerates all of his military achievements, "displaying the marks of his wounds on his body as if they were signs of his labour", the townspeople refuse to ransom him "at the instigation of a m a n reputed to have secret intercourse with the wife of Busas". Again, the Avars are more receptive to the rhetoric of their prisoner, for he, u n d e r the imminent threat of death, promises to make t h e m masters of the town. And indeed, Busas teaches t h e m the art of the siege and the construction of machines, which they had not known previously, a n d soon they storm Appiareia a n d m a n y other towns. T h e broader narrative, however, strangely counteracts this conelusion a n d the significance of the whole episode. It is situated in the context of an Avar invasion after a n u m b e r of fortresses have

21

BG 6, 26. BG 7, 22. 23 Theophylaktos Simokates, Historiae 2, 16, ed. and trans. M. and M. Whitby, The History of Theophylact Simocatta (Oxford, 1986). 22

been overrun, so it is difficult to believe that the Avars learnt something from Busas which they h a d not known before; moreover, immediately after the event, we see the invaders engaged unsuccessfully in two m o r e sieges. 24 A n o t h e r intratextual relation should be noted: after this war, there are protests in Constantinople, a n d some citizens openly reproach the e m p e r o r a n d his generals that the Avar successes were due "not to coincidence but to incapacity". Fatal coincidence is exactly what the Busas story attributes the b a r b a r i a n vietories to; a n d a l t h o u g h — o r even b e c a u s e — T h e o p h y l a c t goes out of his way to claim that nobody believed the critics, the political implications of the episode s o m e h o w r e m a i n a m b i g u o u s . T h i s is not unusual; barbarians always figured prominently in the discourse of Empire; and emperors represented themselves—and were perceived in triumph over barbarians, which opened a wide field for ironic a n d apologetic discourse. I have recently tried to argue that the ethnie discourse that was f u n d a m e n t a l for R o m a n state ideology paradoxically prepared the ground for the identification and later integration of the barbarians. 2 5 T o a certain extent, the Empire needed the prèsence of the " O t h e r " just as the b a r b a r i a n s needed the Empire, as the wars u n d e r M a u r i c e demonstrate. Apart f r o m that, the Busas episode is a parable of barbarian learning. H o w can the Avars, perceived as the most savage a n d uncivilised of barbarians, so successfully threaten the Empire? Stories like this d e m o n s t r a t e the need for specific explanations w h e n the tension between prejudice a n d perception is felt too strongly. In general, this corresponds to a stereotyped preoccupation by sedentaries about the way peoples of the steppe appropriate their skills a n d knowledge, often with fatal consequences. 2 6 A c c o r d i n g to Priscus, the H u n s employed a R o m a n to build a magnificent bath, where he then had to serve as the lowest of slaves. 27 A n d J o h n of Ephesos mentions a R o m a n architect the e m p e r o r sent to the Avars, w h o was then forced to build a bridge to cut off the city of Sirmium from its Byzantine hinterland. 2 8 T h e s e stories deal with the control of knowledge, a n d with its dangerous potential. O n e person, one coincidence was enough

24

For the context, see W . Pohl, Die Awaren (Munich, 1988). Pohl, "Telling the Difference: Signs of Ethnic Identity". 26 J . Deleuze, F. Guattari, Milk Plateaux (Paris, 1980), p. 490. 27 Priscus, Frag. 8, ed. R.C. Blockley, The Fragmentary Classicizing Historians of the Later Roman Empire, vol. 2 (Liverpool, 1983). 211 J o h n of Ephesos 7, 24, ed. E.W. Brooks (Leuven, 1935-6). 25

to let it leak out to powers that would use it against the res publica which was its natural h o m e ; a situation that also implied tragedy for those w h o were forced to betray it. A n o t h e r element in the Busas story is the role of rhetoric that constitutes the basic sequence of its narrative. Busas twice persuades the Avars to let him live, but fails to persuade his fellow citizens to ransom him. T h e o p h y l a c t is certainly one of the writers w h o lays great stress on "the sure-working, ancient magic of Greek words", as Peter Brown put it. 29 In a n u m b e r of other instances, he elaborately shows how the Avar khagan or rebellious soldiers are placated by the power of words. T o quote Peter Brown again, in "a political system where change h a p p e n e d , if at all, not through the effect of persuasive oratory", there was a strong element of mystification in this view. It m a y be noted that in the case of Busas, his rhetoric only worked w h e n he h a d some bargain to back it u p with. In spite of his insistence, in the middle of the seventh century, when T h e o p h y l a c t wrote, the p o w e r of words was on the decline, a n d the power of the W o r d h a d taken its role as a c o u n t e r p a r t to worldly violence. Western writers resorted less to classicising speeches a n d m o r e to miracles when individuals were to escape imprisonment a n d imminent danger. T h e r e is a third story about a prisoner, again f r o m Theophylact. In the 590s, a y o u n g Gepid in the imperial a r m y h a d m u r d e r e d one of the e m p e r o r ' s guards a n d stripped him of his golden belt; when he tried to have it melted down in Constantinople, he was arrested. At the praetor's office, the Gepid was asked (with the help of an interpreter, as we hear this time) where that belt c a m e from. H e responded with an elaborate narrative a b o u t Alboin, king of the L o m b a r d s , w h o fell in love with the d a u g h t e r of the Gepid king C u n i m u n d and abducted her. This led to war in which the Byzantines were also involved on the side of the Gepids (this passage echoes the account of these affairs in Procopius and M e n a n d e r ) . T h e Gepid prisoner claimed that he had personally killed an illegitimate son of Alboin in the ensuing battle won by the R o m a n s , a n d that the belt was his booty. H e had already been released when one of the magistrates realised that this battle had been fought thirty years ago, when the Gepid could not even have been born, a n d the m u r d e r e r was finally executed. 3 0 29 30

P. Brown, Power and Persuasion in Late Antiquity (Madison, 1992), p. 30. Theophylaktos Simokates, Historiae 6, 10.

This is an interesting case of, as we might call it, private deployment of political a n d ethnic discourse. Some of the story clearly represents Theophylact's historical reading. But we may also assume distant echoes of ethnic memory. The murderer is a Gepid, for w h o m the confrontation with the Lombards, culminating in the final defeat and death of C u n i m u n d in 567, was still so present that he could easily project himself into events that had h a p p e n e d before he was born. T h e Gepid kingdom had fallen, so that its history and its myths were open for personal reinterpretation. It is remarkable that this y o u n g Gepid pictures himself as fighting in the R o m a n army, a n d not with the Gepids, although still against the Lombards. O n the other h a n d , the story revolves a r o u n d king Alboin who was a figure that almost immediately inspired a halo of legends, in which his adventures against the Gepids and his death at the instigation of R o s a m u n d , C u n i m u n d ' s daughter, stood out. T h e m u r d e r story in Theophylact is the only version in which R o s a m u n d ' s abduction by Alboin is presented as the cause, not the consequence of the war, quite consistent with folk tale logic. At the same time, the legends connected with king Alboin demonstrate how useless it would be to ask what was "authentically" G e r m a n i c in them. A courtroom d r a m a around a young Gepid who was a Byzantine soldier, written down by a classicizing historiographer about half a century later; various shorter versions by western chronicles; a detailed account by Paul the Deacon, the L o m b a r d monk of Montecassino; and YVidsith's king Aelfwine—the variety of these traces shows that an analysis of intertextual relations can hardly assume exclusively written diffusion. But they do suggest that the production of myths and legends was not restricted to G e r m a n i c bards, and that the stories followed similar patterns everywhere. 3 1 An analogous conclusion can be m a d e about the belt, which, of course, is a part of social language too. It is obvious that a belt worn by the emperor's guard can as easily be ascribed to a L o m b a r d prince. It does not serve as a sign of ethnic identity, and that would hardly surprise any archaeologist working with central European prèstige objects of the period. But it is a sign of social status; a simple Gepid soldier in the Byzantine army is automatically u n d e r suspicion when it is found in his possession. This does not m e a n that precious 31

For an analysis of the legends of Alboin's death see O. Gschwantler, "Die Heldensage von Alboin und Rosimund", Festgabe Otto Höfler, ed. H. Birkhan (Vienna, 1976), pp. 214-47.

objects and parts of dress could not suggest ethnic identifications as well, something that social anthropology has always assumed; this is being discussed in several contributions to the second volume of group 1, Strategies of Distinction—The Construction of Ethnic Communities, 300-800.3‫צ‬ But on the whole, it is surprising how little evidence for such identifications can be found. Isidore of Seville, to prove his point that peoples were distinguished by outward signs, has to go to somewhat remote times and places for his evidence. 3 3 Even the long beards of the Lombards, a sign that is at once bodily, semantic, narrative a n d symbolic, are surrounded by an area of ambiguity. In the Origo gentis Langobardorum it is really the women whose fake long beards earn the L o m b a r d s their n a m e . M e n wearing long beards do not figure prominently in L o m b a r d iconography, but they are c o m m o n on Byzantine coins of the period. 3 4 Narratives like these involving captivity a n d uncertain identities a p p e a r in c o n t e m p o r a r y Western texts; for instance in Gregory of T o u r s ' story of the hermit Hospitius who predicts a L o m b a r d raid and then stays on alone in his tower while all the others flee. Because of his hermit's clothes and chains, they mistake him for a punished murderer. "And after they h a d called an interpreter, they asked him what evil he h a d done so that he had to undergo such severe punishment. But he said he was a m u r d e r e r and guilty of all crimes. T h e n one pulled out a sword to sever his head; but when he struck his right h a n d was arrested in motion so that he could not pull it back." W h e n the holy m a n cures the arm with the sign of the cross, the m a n w h o wanted to kill him becomes his most faithful pupil, and some of the others a b a n d o n their raid. 35 Here, the situation is resolved in a paradoxical way: the spiritual leader of a community styles himself as an outsider, awaits those w h o capture him, in selfinflicted captivity, a n d professes a wrong identity; but his double powerlessness, in chains and in the middle of his enemies, results in a display of supreme power through a miracle. Again, although this

32

Ed. Pohl (above, n. 7). Isidore, Etymologiae 19, 23: cf. Pohl, "Telling the Difference. Signs of Ethnic Identity". 34 Pohl, "Telling the Difference. Signs of Ethnic Identity". 35 Gregory of Tours, Historiae 6, 6, ed. B. Krusch and VV. Levison, MGH, SRM I 1 (Hannover, 1951): used also in Paul the Deacon, Historia Langobardorum 3, 2, ed. G. VVaitz, MGH, SRI. (Hannover, 1878): cf. Sulpicius Severus, Vita Martini, 15, ed. J . Fontaine (Paris, 1967-9). 33

need not have been essential for the plot of the story, communicadon is given a central role. Before attempting to kill him, the Lombards call an interpreter and ask who he is and what he has done (although they have already formed their own ideas about his identity). T h r o u g h the power of the holy m a n , mistaken identities lead to a change in identity: the fierce L o m b a r d , in eodem loco conversus, tonsorato capite, becomes a faithful monk. Captivity a n d the power to release captives is one of the key motifs in Merovingian hagiography, whether this is achieved through ransoming, persuasion, trickery or miracles. 36 But in most cases, unlike in the Hospitius story, the identity of these captives does not matter, they are mere objects of the holy m a n ' s powers. So far, I have tried to present examples of how political discourse, uncertain identities and problems of communication could affect individual lives. In all four cases, people react to mortal dangers with statements of their identity, which include a reference, however rudimentary, to their community, their social role and status within it, a n d their achievements. These statements, if recognised by the powers they are aimed at, might gain pardon; 3 , for these powers, even the Avars or the Lombards, are not portrayed as indiscriminately violent. But each of these professions of identity is somehow flawed: incomplete in the case of Gilakios who only utters his title; inappropriate with Busas who defines himself by military exploits that do not earn him any credit; too obviously fake in the case of the 36

Cf. for instance, Gregory of Tours, De virtutibus sancti Martini 1, 23; 2, 35; 4, 26, 39, 41, ed. B. Krusch, MGH SRM I 2 (Hannover, 1885): Vita Eparchii, 9: Vita Richarii primigenia, 7: Vita Sigirarnni, 16-18: Vita Amandi, 9, 13: Vila Filiberti, 23-6: Constantius, Vita Germani, 7, 36, ed. R. Borius (Paris, 1965). O n these and other examples and for a thorough discussion see František Graus, "Die Gewalt bei den Anfangen des Feudalismus und die Gefangenenbefreiungen der merowingischen Hagiographie", Jahrbuch für Wirtschaftsgeschichte 21 (1961), pp. 61-156. Graus coneludes that the innocence or guilt of the captives is not the point of these stories; it is the power of the holy man to liberate them. Cf. also R. Van Dam, Saints and their Miracles in Late Antique Gaul (Princeton, 1993), p. 103: W. Klingshirn, "Charity and Power: Caesarius of Aries and the ransoming of captives in sub-Roman Gaul", Journal of Roman Studies 75 (1985), pp. 183-203. Gregory of Tours (De virtutibus sancti Martini 1, 23) offers a metaphorical reading at the end of a story in which a whole series of chains are broken, freeing Wilichar: "If only the blessed confessor might deign to show such power to me, so that he might release the bonds of my sins just as he crumbled those massive heavy chains on top of Wilichar!" 37 For the ordinary functioning of the rhetoric and ritual that could gain pardon from Byzantine and Frankish authorities, see Brown, Power and Persuasion: G. Koziol, Begging Pardon and Favour: Ritual and Political Order in Early Medieval France (Ithaca, 1992)'.

young Gepid who has reconstructed Gepid history as his own; a n d deliberately wrong with regard to Hospitius, who pretends to be a criminal instead of trying to be spared. In the first three cases, those who have the power over life and death refuse these self-definitions, with the result of actual or social death (betrayal a n d loss of the community); only the profession of the holy m a n is taken at face value, which gives him the opportunity to demonstrate his own superior power. W e may conclude that our static conception of identity is rather inappropriate to grasp all these attempts of identification, set in a precarious communication that is h a m p e r e d by problems of translation, the humiliating context of supplication, and the incalculable threat of resort to speechless violence. A n d yet, c o m m u nication plays a central part in these situations between life and death. T h e r e is a need for recognition, for orientation. At the root of each situation, there is a displacement—one of the conditions for the model of permeability presented in Averil C a m e r o n ' s paper: a reshuffling of populations; a confrontation of strangers; a changing of allegiances a n d identities: a n A r m e n i a n in Italy; a G e p i d in Constantinople; a T h r a c i a n townsman chancing upon hostile Avars in the n e a r b y woods; a holy m a n living on a diet of i m p o r t e d Egyptian herbs, confronted by L o m b a r d raiders. T h e first three are Byzantine soldiers; Hospitius is a "Byzantine" holy m a n in Gaul; all of t h e m live in the twilight between romanitas a n d barbarianism. But their fates do not make sense as particles of any overwhelming a n d all-inclusive confrontation or contradiction, be it between the Empire a n d the barbarians or between classes or élites. 38 Conflicts in Late Antiquity could be very specific; to make them manageable a n d meaningful, they had to be defined as part of a confrontation on some higher level. W h a t Procopius and Theophylact, implicitly or explicitly, make of their narratives is already part of such a work of interpretation. This shaping of events was m a d e possible by some matrix of m e a n i n g that we could call a "great code" (Northrop Frye), a grand récit (Lyotard) or a "discourse" (Foucault). In some way or other m a n y c o n t e m p o r a r y sources are traces of this continuous effort to define groups and communities: impeúum, gentes et régna. This ethnic and political discourse has often been interpreted

38

In this context Gilles Deleuze's idea of understanding conflicts in societies not in terms of overwhelming contradictions but of individual "escape routes" would be appropriate—see G. Deleuze, Foucault (Paris, 1986).

as a reflection of a world in which rather homogeneous peoples b e c a m e the true subjects of history. Certainly, claims for power could now be based on ethnicity, quite contrary to the R o m a n system. This explains some of the strong ethnic statements in royal titles, law-codes or histories. 39 But it does not imply that such a strong sense of identity had to be wide-spread. T h e "notion of difference", or "language of distinction", 40 could also be a rhetorical device. Ethnic discourse in the regna responded to a double challenge: one was an environment of complex and shifting identities of which we have just had a few glimpses. A n d secondly, the barbarian kingdoms were confronted with the overwhelming reality of two of the most suecessful social languages (or discourse systems) ever: the R o m a n and the Christian. This created a tension that distinguishes medieval ethnogenesis, migration a n d statehood from ethnographic analogies, from Polynesian migrations to the Aztec state. It is a tension that is at the heart of early medieval ethnic and political discourse, a n d accounts for m u c h of its ambiguity. O n c e again, this takes us back to the work of Averil C a m e r o n . Societies, she has written, have characteristic discourses or "plots" whose control is a key to social power. 41 She has admirably described the power of Christian discourse and its potential to construct reality through text. T h e term discourse, in this sense, does not only m e a n the conscious rhetoric that Busas, or, on a m u c h more sophisticated level, professional rhetors, historiographers or preachers employed. In a semiotic sense, it r a t h e r implies the notions of the possible or impossible, rational or insane, true or false that lie at the basis of every statement, speech or narrative in a given society. 42 In this sense, the enormous wealth of texts that Averil C a m e r o n mentions in her p a p e r constitutes the neccssary context for a political history. By contrast with what is implied by the traditional notions of p r o p a g a n d a or ideology, discourse in this sense could not simply be created by a ruling class, whose potential to manipulate or control it was limited. O f course, the hold on discourse was firmer in the Christian sphere where to say the wrong thing might be fatal;

39

See, for instance, H. Wolfram, Intitulatio I. Lateinische Königs- und Fürstentitel bis zum Ende des 8. Jahrhunderts (Vienna, 1967). 40 Cameron, Christianity and the Rhetoric of Empire, p. 7. 41 Cameron, Christianity and the Rhetoric of Empire, p. 1. 42 M. Foucault, L'ordre du discours: R. Barthes, L'aventure semiologique (Paris, 1985).

but in spite of the enormous efforts that went into codification and control of the Christian discourse, even there success was limited. Still, a change in discourse and strategies to control it also contributed to the transformation of the Empire and the development of gentes and regna. W e may ask u p to which point the success of the barbarian kingdoms was also due to their ability to control and channel ethnic a n d political discourse. Ostrogothic Italy is an obvious example (Herwig W o l f r a m ' s analysis of Amal ideology is still fundamental), 4 3 but also the Merovingians whose legitimacy remained undisputed in spite of all criticisms of single rulers (here, Ian W o o d has contributed a lot to our understanding). 4 4 These analyses may be extended beyond ideologies, to the assumptions that make them possible. W h y do people cease to think in terms of imperium and start to think in terms οΐ gentes et regna? A n d why do they act accordingly? Military decisions and displays of force are ultimately powerless against this type of change, otherwise Justinian's achievements would not have come to so little so soon. It is remarkable that Procopius, the historian of Justinian's conquests, ended u p c o n d e m n i n g these very exploits in the Anecdota, even before their ultimate failure became obvious. Looking at changing identities through changes in discourse has an advantage: the sources that have come down to us, so fragmentary in other respects, move to the heart of our problem themselves. T h e r e is a drawback, of course: we can say m u c h less about communities that are not being perceived a n d constructed by social language. Therefore, other methods, for instance settlement archaeology or history of economic relations, must add to our knowledge. Studying social language should not be understood in the limited context of Geistesgeschichte, of the history of ideas. 45 T h e change of discourse has to be seen in the same matrix as social reproduction a n d power relationships. This is a process of transformation that escapes totalizing reconstructions, and has to be traced in the variety of its manifestations. For historians w h o deal with such powerful abstractions as "Imperium, gentes et regna" it is useful to be confronted with the elu-

" Wolfram, Die Goten, pp. 25 -8: id., "Origo et religio", Early Medieval Europe 3 (1994), pp. 19 38: cf. also P. Heather, "Theodoric, King of the Goths", Early Medieval Europe 4 (1995), pp. 145-74: Amory, People and Identity in Ostrogothic Italy. 44 I.N. Wood, The Merovingian Kingdoms 450-751 (London, 1994). 45 Cf., for instance, M. de Certeau, The Writing of History (New York, 1988).

sive singularities, the "particles of change" (Joni Mitchell) 46 that added u p to the transformation of a macrocosm like the R o m a n World. Anthony Burgess, in a recent novel, goes one step further: " T h e r e was a philosopher who spoke of the cat that mews to be let out and then mews to be let in again. In the interim, does it exist?" 4 ' At times, we might be even m o r e u n s u r e about the identity of the R o m a n Empire or the gens Gothorum than about the SchrödingerBurgess cat. For all we ever hear as historians is the cat's mew. But we m a y also be reassured: there are m a n y ways in which cats make themselves understood.

46 47

Joni Mitchell, "Hejira", Hgira (LP, 1976). A. Burgess, A Dead Man in Deptford (London, 1994), p. 3.

M I T WAS FÜR M I T T E L N KANN EINEM PHYSISCH ANONYMEN AUCTORITAS VERLIEHEN WERDEN? Beat Brenk

προς ζωγράφον σύ μοι άλλότριον σχήμα πώς έδίδους, έταΐρε; οίόν με είδες, τοιούτον και γφάθον. (Julianus imp.) Transit laetitia in serenitatem, membrorum vigor animi vigor mutatur. Virtus corporis in virtutem spiritus migrât. Pallescit jejunio speciosa faciès, et prius succi plena, fit plena gravitatis. (Hilarius, Vita

Honorati)

Praeliminarien Die Entwicklung der spätantiken Porträtkunst zwischen dem 4. und dem 6. Jh. wird gekennzeichnet durch ein N e b e n e i n a n d e r von individuell-physiognomischcn und a n o n y m e n , stereotypen, "attributiven" Porträtdarstellungen. Allmählich verschwindet das M a r m o r - und Bronzeporträt griechisch-römischer Prägung. V o m späten 6. J h . an tritt an die Stelle der individuellen Charakterisierung das, was m a n etwas unreflektiert als Typus, Formel, Idealbildnis bzw. als geometrisiertes, stilisiertes oder idealtypisiertes Bildnis bezeichnet hat. Mit was für Kriterien sollen formelhafte Bildnisse beurteilt werden? Soll bloß das Formelhafte, der Stil oder der verbleibende Rest physiognomischer Anspielungen analysiert werden? J e d e s Porträt hat bekanntlich Stil u n d propagiert infolgedessen m e h r oder weniger ein Ideal, wie dieses auch i m m e r umschrieben sei, das über seine physische Beschaffenheit hinaus u. U. auch einen Inhalt anvisiert. "Inhalt" ist somit nicht an eine individuell-physiognomische Darstellung gebunden. Auch physisch a n o n y m e Bildnisse vermögen inhaltliche Signale auszusenden. D a r ü b e r nachzudenken, ist Ziel der folgenden Zeilen. Das M a r m o r - und Bronzebildnis wird vom 5 . / 6 . J h . an vom gcmalten Bildnis in Fresko- und Mosaiktechnik und dem enkaustischen Porträt verdrängt. Die V e r ä n d e r u n g des Porträts betrifft nicht n u r

den Darstellungsmodus sondern auch den Bildträger und die damit implizierten technischen und bildnerischen Aspekte. Diese notorisehen, a b e r im einzelnen noch nie mikrohistorisch u n t e r s u c h t e n Prozesse w u r d e n häufig mit d e m historiographischen Problem der Dekadenz in der Spätantike in V e r b i n d u n g gebracht, wofür i m m e r wieder der Konstantinsbogen als Beweis herhalten mußte. Zu den klassischen Äusserungen gehört eine Passage im Konstantinbuch von J a k o b Burckhardt: Der Constantinsbogen beim Kolosseum ist allerdings ein Werk der Hast und Eile, und dies erklärt und entschuldigt hinlänglich die große Roheit der plastischen Ausführung, nicht aber die Häßlichkeit der Gestalten und die Verkümmerung der Züge. Wohl gibt es Zeiten, in welchen die Kunst sich darauf etwas einbildet, ihr Ziel einseitig im Charakteristischen statt im Schönen zu suchen, um jenes sogar bis ins Häßliche zu steigern, ohne daß die den Künstler umgebende Welt daran schuld wäre. Allein hier ist von einer solchen Vorliebe für den Charakter nicht die Rede, sondern ganz einfach von der Unfähigkeit, an den klassischen Schönheitsidealen auch nur oberflächlich festzuhalten, während die Außenwelt keine Beziehung mehr zu denselben hat. W ä h r e n d Burckhardt Vorwürfe erhebt, indem er von der Unfähigkeit der Künstler spricht, hält Evelyn W a u g h in der Novelle Helena die Beurteilung des Reliefs a m Konstantinsbogen schlicht für Geschmackssache, freilich im depreziativen Sinn. Konstantin w ä h n t seinen T r i u m p h b o g e n noch 11 J a h r e nach Baubeginn für unfertig und findet die Skulpturen "atrocious". "A child could do better". D e n Entwerfer kritisiert Konstantin: "Your figures are lifeless a n d expressionless as dummies. Your horses look like children's toys". D a sich kein besserer Künstler finden läßt, befiehlt Konstantin dem Entwerfer: "pull the carvings off Trajan's arch and stick them on mine". Die Diskussion u m das Problem der Qualität ist heikel und führt i m m e r wieder zu Verlegenheiten. D a die klassischen Archäologen vor allem anhand der Kaiserbildnisse in M a r m o r , Bronze, auf M ü n z e n und Medaillen eine Chronologie zu ermitteln suchten, w ä r e n sie dazu prädestiniert gewesen, eine Antwort auf das Problem des verblassenden Interesses a m Individualporträt im 4 . - 6 . J h . beizusteuern. W a r es ein n i c h t - m e h r - K ö n n e n oder ein nicht-mehr-Wollen? O d e r haben sich die Interessen verlagert? Diese Fragen konnten von den Archäologen so leicht nicht beantwortet werden, weil das Ziel ihrer hauptsächlich auf Chronologie fokusierenden Betätigung nicht feststeht. G e h t es u m Chronologie

um der O r d n u n g und Logik willen oder um Stilchronologie oder um eine Chronologie imperialer M o d e n und Attitüden? O d e r verä n d e r n sich die Konzepte der Menschendarstellungen? Eine noch viel grössere Verlegenheit macht sich bei den Kunsthistorikern bemerkbar, die sich nicht in erster Linie mit Kaiserbildnissen, sondern mit den Bildnissen Christi und solchen von Propheten, Aposteln, Engeln, Bischöfen, M ä r t y r e r n und Heiligen auseinanderzusetzen haben. Diesem höchst disparaten Material aus Holz, Metall, Glas, Mosaik, Fresko und M a r m o r konnte bisher kein verbindliches begriffliches Instrumentarium abgewonnen werden. 1 Angesichts dieser vagen historiographischen Situation empfiehlt es sich, eine Fallstudie v o r z u n e h m e n , welche ein Licht auf die Problematik der veränderten Mentalität in der Spätantike zu werfen vermag. Wie im Titel angezeigt, möchte ich das P h ä n o m e n der physischen Anonymität unter die Lupe nehmen. 2 Physische Anonymität ist paradoxerweise erstmals und seit der tetrarchischen Zeit bis ins 6. J h . immer wieder von den Kaiserbildnisscn propagiert worden. Zwischen 290 und 300 wurden, wie es scheint, erstmals in Kyzikos u n d N i k o m e d i a im N o r d w e s t e n Kleinasiens Kaisermünzen mit einem kantig stereotypisierten Bildnistypus geprägt.'

1

G. Ladner, Die Papstbildnisse des Altertums und des Mittelalters, Bd. I (Città del Vaticano, 1941), pp. 2-7; A. Grabar, Le portrait en iconographie paléochrétienne. L'art de la fin de l'antiquité et du moyen Age, Bd. I (Paris, 1968), pp. 591-605; E. Kitzinger, "Some reflections on portraiture in byzantine art", The Art of Byzantium and the Medieval West. Selected Studies, hrsg. von E. Kleinbauer (Bloomington, 1976), pp. 256-264; J.D. Breckenridge, IJkeness. A conceptual History of Ancient Portraiture (Evanston, 1968); H.P. L'Orange, Ukeness and Icon (Odense, 1973); C. Belting-Ihm, "Heiligenbild", Reallexikon fur Antike und Christentum XIV (1988), pp. 66-96. 2 Das Thema ist bisher fast ausschließlich von den Historikern und Literaturwissenschaftern behandelt worden. J . Fürst, "Untersuchungen zur Ephemeris des Diktys von Kreta", Philologus 61 (1903), pp. 374 440; Ε. Evans, "Descriptions of Personal Appearance in Roman History and Biography", Harvard HSCPh 46 (1935), pp. 43-84; E. Evans, "Physiognomies in the Ancient World", Transactions of the American Philosophical Society 59 (1969), pp. 1 101; L'invention de l'autobiographie d'Hésiode à S. Augustin. Actes du 2ème colloque d'équipe de recherche sur l'héllénisme postclassique, hrsg. von M.-F. Basiez, P. Hoffmann, L. Pernot, Paris, Ecole Normale supérieure (1993); G. Dagron, "Holy Images and Likeness", Dumbarton Oats Papers 45 (1991), pp. 2333‫ ;־‬bes. pp. 25-28 mit weiterführender Literatur. 3 H.P. !,'Orange, Studien zur Geschichte des spätantiken Porträts (Oslo, 1933), pp. 15-46; C. Vermeule, "Maximianus Herculius and the Cubist Style in the Late Roman Empire 295 to 310", Museum of Fine Arts, Boston Bulletin 6 0 / 3 1 9 (1962), pp. 9 20; C. Vermeule, "A Graeco-Roman Portrait of the Third Century A.D. and the Graeco-Asiatic Tradition in Imperial Portraiture from Gallienus to Diocletian",

In der Folge der diokletianischen M ü n z r e f o r m von 294 wurden die Bildnisse der Tetrarchen auf Münzen und auf einigen politisch relevanten Porträts einander so weit angeglichen, dass es kaum m e h r möglich ist, ein Bildnis Diokletians von einem solchen des Maximianus Herkulius zu unterscheiden. Dieser kaiserliche Porträtstil wurde wie ein Gesetz dekretiert und in den hintersten Winkel des römischen Reiches transportiert. Das bezeugen die in Ägypten entstandenen Porphyrtetrarchen an der Fassade von S. Marco in Venedig (Abb. 1). Die vier Tetrarchen sehen einander zum Verwechseln ähnlich. Ziel ihrer physiognomisehen Similitudo war die politische Concordia.4 Die gerunzelte Stirn und die aufgerissenen Augen verbildlichen als ikonographische Attribute das angestrengte Herrschen. Die Tränensäcke und die Bärte evozieren die Auctoùtas und Gravitas des Alters, insbesondere aber herrscherliches Imponiergehabe. Die politische Concordia k o m m t selbstverständlich n u r d a n n zum T r a g e n , wenn alle vier T e t r a r c h e n gleichförmig dargestellt sind wie das in S. M a r c o in Venedig und an den T e t r a r c h e n im Vatikan der Fall ist. Die Einzelporträts der Tetrarchie verabscheuen physiognomische Individualität keineswegs. Häufig aber sind gewisse Attribute wie der kurze und eckige Haarschnitt, die aufgesperrten Augen, die doppelten Vertikalfalten an der Nasenwurzel und die gerunzelte Stirn als dynastische "Markenzeichen" beibehalten, und der Porträtist konnte es sich erlauben, innerhalb dieser vorgegebenen Koordinaten persönlich-individuelle Züge "einzutragen". Dies zeigt ein Vergleich des 1991 gefundenen Porphyrkopfes des Kaisers Galerius aus Gamzigrad-

Dumbarton Oaks Papers 15 (1961), pp. 1 22‫ ;־‬H.A. Cahn, "Kunstgeschichtliche Bernerkungen zu Diocletians Münzreform", Festschrift Karl Schefold (Bern, 1967), pp. 91-95; R. Calza, Iconografia Romana Imperiale da Carausio a Giuliano. 287-363 d.C. (Roma, 1972); R. MacMullen, Roman Government's Response to Crisis A.D. 235-337 (New Haven, 1976), pp. 16-23 (freundlicher Hinweis von Peter Brown); J . Meischner, "Die Porträtkunst der ersten und zweiten Tetrarchie bis zur Alleinherrschaft Konstantins 293-324", Archäologischer Anzeiger (1986/2), pp. 223-250; H.P. L'Orange, Das spätantike Herrscherbild von Diokletian bis zu den Konstantin-Söhnen 284 361 n.Chr., Das römische Herrscherbild Abt. III, Bd. 4 (1985). 4 H.P. L'Orange, "Die Bildnisse der Tetrarchen", Acta Archaeologica 2 (1931), pp. 29 52; H.P. L'Orange, Studien zur Geschichte des spätantiken Porträts (Oslo, 1933), pp. 16-18; J . Strzygowski, "Orient oder Rom. Stichprobe: Die Porphyrgruppe von S. Marco in Venedig", Bdträge zur alten Geschichte II (1902), pp. 105-124; M. Cagiano de Azevedo, "Note sui Tetrarchi di Venezia", Rendiconti delta Pontificia Accademia romana di archaeologia (1966/7), pp. 153 159; Zum Problem der Concordia‫׳‬. W. Seston, Dioctétien et la Têtrarchie (Paris, 1946), pp. 232, 256; F. Baratte, "Observations sur le portrait romain à l'époque tétrarchique", Antiquité tardive 3 (1995), p. 70; Eusebius, de laude. Constantini, III, 4.

Romuliana mit dem Galeriusbildnis am kleinen Ehrenbogen aus dem Oktogon im Palast zu Saloniki (Abb. 2-3). Zu den erwähnten "Markenzeichen" treten individuelle Züge wie die verfettete Kinnpartie, die starken Nasen-Mundfalten sowie die weit hinaus gezogenen und dann j ä h abfallenden Brauen. Diese Züge und die F u n d u m s t ä n d e der beiden Bildnisse e r l a u b e n keinerlei Zweifel an der A u t o r s c h a f t des Galerius hinsichtlich des Palasts in Saloniki und der Residenz in Gamzigrad-Romuliana. 5 V o r allem aber geht aus dem Vergleich der vier T e t r a r c h e n in Venedig mit den beiden Galeriusbildnissen hervor, daß Stil in diesem Fall etwas mit W a h l zu tun hat, und nicht mit K ö n n e n oder N i c h t - K ö n n e n . In der theodosianischen Zeit ist erstmals das Ideal unvergänglicher J u g e n d in V e r b i n d u n g mit einer High-Brow-Effeminiertheit systematisch auf Kaiser- und Privatbildnissc übertragen worden. Jugendlichkeit bzw. unverbindliche Alterslosigkeit galt als neues Herrscherideal. Von den theodosianischen Kaiserbildnissen‫ '׳‬ging eine weit r e i c h e n d e Signalwirkung aus: Jugendlichkeit war unter gleichzeitigem Verzicht auf eine individuelle Physiognomie zum Herrscherideal und damit zu einer allgemein gültigen Aussage geworden. H.P. L ' O r a n g e hat einige Privatbildnisse, die sich die theodosianische M o d e anverwandelten, mit der Etikette "subtiler Stil" versehen.‫ ׳‬Er nennt in diesem kurzen Aufsatz eine Togastatue des 2 . / 3 . J h . im T h e r m e n m u s e u m , die (Abb. 4) um 400 mit einem neuen Kopf ausgerüstet wurde. Die Schöneinigkeit der Nase und Brauen und die kokette Schrägstellung der Augen verleihen dem Bildnis weibliche Züge. Besonders subtil wirkt die Blickführung, weil die Iris des linken Auges ganz leicht nach außen verschoben ist, sodaß ein irisierender Blick entsteht. Diese modischen, nicht-individuellen Züge können auch an M e n s c h e n d a r S t e l l u n g e n auftreten, die a priori keinen Anspruch auf Individualität erheben wollen und können, so z.B. an dem Lebewesen, welches den Menschen (nach Apok. 4. 7) im Baptisterium in Neapel verbildlicht (Abb. 5). An diesem 39 cm h o h e n Mosaik-Porträt überrascht die effeminierte lcichte D r e h u n g des Gesichts nach rechts, der abermals

5 D. Srejovic (Hg.), Roman Imperial Towns and Palaces in Serbia (Beograd, 1993), p. 50; D. Srejovic, C. Vasic, Imperial Mausolea and consecration memorials in Felix Romuliana (Garnzigrad, East Serbia) (Beograd, 1994). 6 J . Kalkvitz, Oströmische Plastik der theodosianischen Zeit (Berlin, 1941), pp. 81-131. 7 H.P. L'Orange, "Der subtile Stil. Eine Kunstströmung aus der Zeit um 400 n.Chr.", Antike Kunst 4 (1961), pp. 6874‫־‬.

etwas schillernde Blick, welcher durch die verschieden geformten Augen zustande kommt. 8 N e u ist somit die attributive Behandlung der Porträtzüge, d.h. es können Porträtzüge, die für das Privatbildnis der kulturell f ü h r e n d e n Oberschicht ausgedacht worden sind, aus ihrem Kontext herausgelöst und wie eine Vokabel in einen gänzlich verschiedenartigen Kontext transplantiert werden. Das Lebewesen des Menschen im Baptisterium von Neapel wirkt d a h e r erstaunlich porträthaft und modisch, und eben die modischen Elemente bekommen den Anschein des "Aufgesetzten". Dass vom 4 . / 5 . J h . an Porträtzüge attributiv eingesetzt werden können, m a c h e n auch Darstellungen von Personifikationen (Flußgottheiten etc.) deutlich, die erstaunlich lebendig und präsent wirken können, wie z.B. der b e r ü h m t e Neilo-Stoff aus Achmim-Panopolis im Puschkin-Museum in Moskau. Kopftypus, Physiognomie und Physis waren zum Gegenstand attributiver Wahl geworden. D e r Verzicht auf Individualität hatte notwendigerweise eine stärkere Konzentration auf die Attribute des sozialen Stands und der Mode zur Folge. Attribute konnten zuweilen ohne besonderen Realitätsbezug ü b e r h a n d n e h m e n , sich beinahe verselbständigen.

1. Christus und der Bart der Philosophen Mit was für Mitteln kann einem physisch A n o n y m e n Auctoritas verliehen w e r d e n ? Die P r o b l e m a t i k ist so alt wie die G ö t t e r u n d Heroenbildnisse der Griechen und R ö m e r . Die Künstler der frühchristlichen Zeit m u ß t e n eigene Lösungen finden. Das vollplastische m a r m o r n e Götter- und Heroenbildnis griechisch-römischer Tradition wurde von den Christen vehement abgelehnt, bekämpft und gelegentlieh zerstört, weil es den Götter- und D ä m o n e n g l a u b e n der Heiden repräsentierte. D e n n o c h ist vereinzelt im Bereich der römischen Sarkophagplastik für das Signalement Christi auf heidnische Götterdarstellungen zurückgegriffen w o r d e n . Auf den sog. polychromen Fragmenten im T h e r m e n m u s e u m , welche zu den frühsten christlichen Sarkophagrelicfs gehören, ist für das Signalement Christi eine

8 J.L. Maier, Le baptistère de Naples et ses mosaïques (paradosis 19) (Fribourg, 1964); P. Pariset, "I mosaici del battistero di S. Giovanni in Fonte nello sviluppo della pittura paleocristiana a Napoli", Cahiers Archéologiques 20 (1970), pp. I 13‫־‬.

dem Asklepius ähnliche Gestalt (Abb. 6) gewählt worden. 9 Auf d e m Relief der Bergpredigt ist Christus mit nacktem O b e r k ö r p e r wiedergegeben und seine Haartracht gleicht der des Asklepius. Dieser Asklepiustypus wird aber in der Brotvermehrung und in der Heilung der beiden Blinden (Abb. 7) in ein unspezifisches Dutzendgesicht verwandelt. In der Heilung des L a h m e n und in der Heilung des Wassersüchtigen ist Jesus regulär mit T u n i k a und Pallium bekleidet, und bärtig ist er alle Male, und sein H a a r ist ungepflegt, wie es die kynisehen Philosophen zu tragen pflegten. Der unbekannten Physis Jesu ist gleichsam attributiv Auctoritas verliehen w o r d e n . Die kynische Philosophenfrisur und die Philosophentracht sollten die Auctoritas des Philosophen Jesu evozieren. W a s hatte sich der C o n c e p t e u r dieser p o l y c h r o m e n G r a b v e r schlußplatten wirklich gedacht? Die im G r u n d e doch sehr verschiedenartigen Antlitze Christi erwecken den Eindruck eines Zögerns u n d Suchens. Zweifellos w a r der A u f t r a g g e b e r ein Privater. Die Kirche als Institution hätte wohl kaum ein derartiges Signalement für den Kopftypus Christi propagiert. Die W a h l dieses asklepiusförmigen Christusbildnisses, dessen Ideologie nicht wirklich geklärt werden kann, blieb innerhalb der frühchristlichen Kunst eine Episode. Die Menschen der frühchristlichen Zeit m u ß t e n sich sehr bald damit abfinden, dass Christus in der K u n s t sehr verschieden dargestellt wurde. Für Augustinus war dies ein fait accompli: " D e n n auch das leibliche Antlitz des H e r r n wird in vielfacher Auffassung verschieden dargestellt, wiewohl er doch nur ein einziges hatte" (Aug., De Trinitate VIII 4. 7).10 Noch im 6. J h . hatte sich kein Konsens in der Christusdarstellung herausgebildet. Erstaunlicherweise ließ die Kirche den Auftraggebern und den Künstlern in dieser Sache freie H a n d . C h a rakteristisch ist die Situation z. B. in Ravenna, wo die Apsis von S. Vitale (Abb. 8) den jugendlich-bartlosen Christustypus zeigt, in S. Apollinare in Classe (Abb. 9) dagegen ist ein bärtiger Christustypus mit langen auf die Schultern fallenden H a a r e n in vorgerücktem Alter gewählt worden. Dies ist u m so auflallender, als es sich bei beiden Darstellungen u m himmlische T h r o n b i l d e r mit eschatologischem

!l

E. Dinkier, "Christus und Asklepios. Z u m Christustypus der polychromen Platten im Museo Nazionale Romano", Sitzungsberichte der Heidelberger Akademie der Wissenschaften. Philos. -Histor. Klasse 1980 (2). 10 G. Dagron, "Holy Images and Likeness", Dumbarton Oak Papers 45 (1992), p. 24.

Gehalt h a n d e l t . " Wieder anders wird Christus im neutestamentlichen Zyklus von S. Apollinare N u o v o verbildlicht. 12 D e m jugendlich-bartlosen Jesus begegnen wir in den Szenen des Wunder-Zyklus, dem bärtigen Jesus in den Szenen des Passionszyklus. Zweifellos hat sich der Entwerfer dabei etwas gedacht. Offensichtlich sollten zwei Abschnitte des Lebens Jesu unterschieden werden. Vor dem Abendmahl wird Jesus jugendlich konterfeit, nach d e m A b e n d m a h l dagegen bärtig. D a das A b e n d m a h l die Passion Christi bzw. seine zweite Lebenshälftc einleitet, war der Bart ein anerkanntes Verständigungsmittel, mit dem etwas über "vorgerücktes Alter" u n d / o d e r über "Leiden" mitgeteilt werden konnte. D a m i t ist aber der Aussagewert des Bartes in der Spätantike keineswegs erschöpft. Das lehrt u.a. ein Blick auf die Darstellungen von heidnischen Philosophen. 1 3 Plotin verweigerte einem Porträtisten Modell zu stehen. Plotinus, der Philosoph, der zu meiner Zeit lebte, war die Art von Mann, die sich dessen schämt im Leibe zu sein; aus solcher Gemütsverfassung wollte er sich nicht herbeilassen, etwas über seine Herkunft, seine Eltern oder seine Heimat zu erzählen. Einen Maler aber oder Bildhauer zu dulden, wies er weit von sich, ja er erklärte dem Amelius, der ihn um seine Einwilligung bat, daß ein Bild von ihm verfertigt werde: es soll also nicht genug daran sein, das Abbild zu tragen, mit dem die Natur uns umkleidet hat, nein, du forderst, ich soll freiwillig zugeben, dass ein Abbild des Abbildes von mir nachbleibe, ein dauerhafteres, als sei dies Abbild etwas Sehenswertes! So sagte er ab und weigerte sich, einem Künstier zu sitzen; aber Amelius hatte Carterius zum Freunde, den derzeit besten Maler, und den brachte er mit in die Vorlesungen und ließ ihn so mit dem Meister zusammentreffen (denn es stand jedem frei, der Lust hatte, in die Vorlesungen zu kommen); durch langanhaltendes Aufmerken gewöhnte er ihn daran, den Seheindruck immer eindringlicher zu erfassen; so konnte er dann nach der in seinem Gedächtnis niedergelegten Vorstellung ein Bild malen, und indem dann Amelius half, die Ähnlichkeit der Züge zu verbessern, entstand dank der Begabung des Carterius ein Porträt des Plotinus, ohne dessen Wissen, das sehr ähnlich war. (Porphyrios, Vita Plotini, c. 1-4).14

" F.W. Deichmann, Frühchristliche Bauten und Mosaiken von Ravenna (Baden-Baden, 1950), Abb. 351, 388. 12 Deichmann, Frühchristliche Bauten und Mosaiken von Ravenna, Abb. 155, 159, 171, 189, 191, 194, 204, 213. 13 P. Zanker, Die Maske des Sokrates. Das Bild des Intellektuellen in der antiken Kunst (München, 1995). 14 H. von Heintze, "Vir sanetus et gravis. Bildniskopfeines spätantiken Philosophen", Jahrbuch für Antike und Christentum 6 (1963), pp. 35-53, bes. p. 47.

Helga von Heintze folgerte aus diesem Passus: "In dieser Einstellung Plotinus und seines Kreises haben wir mit den G r u n d dafür zu suchen, dass wir erstens im Verhältnis zur großen Anzahl von Philosophen dieser Zeit so wenige Bildnisse von ihnen besitzen und d a ß zweitens diese wenigen kaum porträthaft sind".. . . Weiter: "Die Überw i n d u n g des Körpers durch den Geist, j a seine Negierung, das war das Ziel dieser späten Philosophen und dieses Ziel haben die Künstler j e n e r Zeit erkannt und in den Porträts festgehalten." 1 5 Helga von Heintze n a n n t e den in den Philosophenbildnissen des späten 3. J h . (Abb. 10) neu zutage tretenden T y p u s den Vir sanctus et gravis. Es ist der T y p u s des weltabgekehrten ernsten Mannes, für den der Körper n u r ein Abbild ist, nicht wert, im Bilde festgehalten zu werden. "Deswegen wird dem Kopf als solchem, sowie dem K ö r p e r keinerlei Bedeutung m e h r beigemessen." 1 6 " D e r Kopf mit dekorativem langem H a u p t - oder Barthaar konnte j e d e n beliebigen alten M a n n darstellen, gleichgültig ob einen sogenannten Philosophen oder einen Flussgott." 17 D e r von Helga von Heintze kommentierte Passus der Plotin-Biographie sagt zunächst n u r aus, dass trotz der Weigerung Plotins, einem Porträtisten Modell zu sitzen, ein ihm ähnliches Porträt hergestellt wurde. Zutreffender interpretierte Patricia Cox diesen Text: The artist Carterius is described as working with "striking fantasies" (phantasias plektikoteras)\ what he drew is characterized as a "likeness of his mental impression" (indalmatos to eikasma; indalma can also mean "hallucination"!); what Amelius helps him improve "to a greater likeness" (eis homoioteta) is his "sketch" (to ichnos: literally, "footprint", "track"). All of these terms suggest that for Porphyry the portrait was not an opaque image of the man, nor was it merely an imitation of his physical features. It might better be described as a creative mirroring, an act of the imagination more true to reality than direct perception. This kind of portrait could be described as a visual text full of signs, footprints of the man's deep self.1"

15 Von Heintze, " Vir sanctus et gravis. Bildniskopf eines spätantiken Philosophen", p. 48. "‫ י‬Von Heintze, "Vir sanctus et gravis. Bildniskopfeines spätantiken Philosophen", p. 50. 17 Von Heintze, "Vir sanctus et gravis. Bildniskopfeines spätantiken Philosophen", p. 52. 18 P. Cox, Biography in iMe Antiquity. A Quest for the Holy Man (Berkeley, 1983), pp. 108 109. H. Pekary, "Plotin und die Ablehnung des Bildnisses in der Antike", Festschrift H. Brandenburg, Boreas 17 (1994), pp. 177 186) hat eine eindrückliche Reihe von Beispielen für die Ablehnung des Bildnisses in der Antike zusammengestellt.

Es ist nun allerdings fraglich, ob diese Aussagen für die Philosophenbildnisse der Spätantike f r u c h t b a r g e m a c h t w e r d e n k ö n n e n . Die H a a r t r a c h t der Philosophen auf Sarkophagen ist recht variabel, aber es sind stets bärtige ernste Köpfe ohne individuelle Züge. Schon G. Rodenwaldt meinte, "dass das T h e m a dieser Sarkophage zum ersten Male d e m Christentum die Möglichkeit einer A n k n ü p f u n g bot." 1 9 Aber es wäre verfehlt, in den Philosophensarkophagen die direkten Voraussetzungen für die Darstellung Christi, von Heiligen, Bischöfen und Asketen zu erblicken. Die Voraussetzungen für den allmählichen Verzicht auf individuelle Menschendarstellung in der Spätantike waren vielfältig und multifaktoriell bedingt. Die attributive Verwendung von Gesichtszügen und Haartrachten war zweifellos keine christliche Erfindung. Dennoch ist der Beitrag der Christen nicht zu übersehen. Sind nämlich die große M e n g e der heidnisch-römischen Bildnisse in M a r m o r skulptiert, so entschieden sich die Christen gegen die Skulptur für das Mosaik u n d die Malerei. Das Bildnisrecht der Bischöfe konkretisierte sich stets in Tafelbildern, nicht in Marmorstatuen. Im frühen Mittelalter kam den frühchristlichen Bischofsbildnissen so viel Auctontas zu, daß sich Agnellus in seinem u m 840 verfassten Liber pontißcalis ecclesiae Ravennatis auf alte bildliche Darstellungen beruft, wenn er etwas über das Äußere seiner H e r o e n berichtet: quia pictura insinuât mihi illorum vultus.20 W u r d e in der frühen und späten Kaiserzeit (bis ins 5. Jh.!) ein Pater einer Stadt, ein Eparch, ein Dichter, R h e t o r oder sonst ein bedeutender Euergetes mit einer M a r m o r s t a t u e geehrt, so sind Christus, Bischöfe, Heilige und Aszeten ausschliesslich in Mosaik und Malerei dargestellt worden. Hinter dem Entscheid der Christen gegen die figürliche Monumentalskulptur, der übrigens kein absoluter war, standen die negativen Erfahrungen der Gläubigen mit dem Kaiserkult vor d e m J a h r e 312 sowie die A b l e h n u n g der a n t h r o p o m o r p h e n Got-

19

G. Rodenwald, "Zur Kunstgeschichte der Jahre 220 bis 270",Jahrbuch des deutsehen archäologischen Instituts 51 (1936), pp. 82-113, bes. p. 105; F. Gerke, Die christliehen Sarkophage der vorkonstantinischen ζρ1 (Berlin, 1940), pp. 244-291, 312-315; T h . Klauser, "Studien zur Entstehungsgeschichte der christlichen Kunst III". Jahrbuch für Antike und Christentum 3 (1960), pp. 112-133, und 6 (1963), pp. 96-100; P.'Brown, Augustinus von Hippo (Frankfurt a.M, 1973), pp. 35-36, 70, 126; Zanker, Die Maske des Sokrates. Das Bild des Intellektuellen in der antiken Kunst. 20 VV. Berschin, "Personenbeschreibung in der Biographie des frühen Mittelalters", in Historiographie im frühen Mittelalter, hrsg. von A. Scharer und G. Scheibelreiter (Wien, 1994), pp. 186-193, bes. p. 190,'

tesdarstellung ganz allgemein. Es führt also kein direkter W e g vom heidnisch-römischen Philosophen- und Kaiserporträt zum Christusbild. D e r christliche C h a r a k t e r der Statuette im T h e r m e n m u s e u m ist aus den oben angeführten G r ü n d e n fraglich. Das einzige Attribut der Philosophen, das die Christen rezipierten, war der Bart. 21 Als einflußreichster Christustypus erweist sich der bärtige, dessen H a u p t h a a r schräg hinter der rechten Schulter verschwindet. Erstmals begegnet dieser eindrückliche Christustypus auf der b e r ü h m t e n Ikone auf d e m Sinai (Abb. 11), die d e m 6. oder 7. J h . angehört. Es ist die älteste erhaltene Christusikone. D a sie auf Nahsicht gearbeitet ist, bekannte der Künstler explizit Farbe in Bezug auf das Aussehen Christi. Meines Erachtens ist der Bildnischarakter dieser Ikone bisher nicht richtig eingeschätzt worden, denn es genügt nicht, von einer Starre des Ausdrucks oder von strengen asketischen Zügen zu sprechen. Die meisten Autoren, die sich mit dieser Ikone beschäftigt haben, nahmen zu spekulativen Deutungen Zuflucht. Ein Autor meinte: "Die Ikone steht einer spekulativen Schau des inkarnierten Gottessohnes offen." 2 2 Zunächst erstaunt der Kontrast zwischen dem ausgesprochen längliehen schmalen Gesicht und der Fülle der H a a r e , die sich beinahe perückenartig auftürmen. Die Haare wirken, wie wenn sie vom Friseur sorgfältig gekämmt und zurecht geföhnt worden wären. Diese bewusst schönheitliche K o m p o n e n t e ist nicht zu übersehen. Es ist also keineswegs das H a a r eines Asketen, sondern eines Menschen, der es darauf abgesehen hat, zu wirken und zu repräsentieren. Die H a a r e sind genau so repräsentativ behandelt wie der Nimbus, das Buch und die p u r p u r n e G e w a n d u n g , die den K ö r p e r großzügig verhüllt. Ich weise vor allem auf die haarnadelförmige Schlaufe des Palliums hin, aus welcher die rechte segnende H a n d auftaucht. Diese Schlaufe evoziert eine schönlinige Eleganz. A m erstaunlichsten ist aber doch das Gesicht mit seinen übergroßen und n a h e zusammenstehenden Augen. Es ist ein absolut faltenloses Gesicht mit einem hellen, fast é c r u f a r b e n e n Inkarnat, o h n e

21

O. Thulin, "Die Christus-Statuette im Museo Nazionale Romano", Römische Mitteilungen 44 (1929), pp. 201 259; Die Philosophenkleidung, Tunica und Pallium, ist seit den Anfängen der christlichen Kunst für biblische Gestalten gewählt worden. Zum Bart: B‫׳‬. Kötting, RAC XIII (1986), pp. 177-203, sub voce "Haar". 22 R. Warland, Das Brustbild Christi, 41. Suppl. lieft der Rom. Quartalschrift fur christliehe Altertumskunde (1986), pp. 200-202, Abb. 75; Κ. Weitzmann, The Monastery of Saint Catherine at Mount Sinai. The Icons, vol. 1 (Princeton, 1976), pp. 13-15.

einen Anflug von grünlichen Schatten. Das Gesicht erscheint wie wenn es frontal von künstlichem Licht angestrahlt würde. Die nicht ganz symmetrisch gezeichneten Augen sind zwar groß aufgesperrt, und doch erscheinen sie wie eingefallen. Sie sitzen in tiefen Augenhöhlen, denn um die Augen h e r u m sind in raffinierter Weise Schatten angeordnet. A m meisten Individualität tritt an den Brauen zu Tage: es sind typische M ä n n e r b r a u e n : assymetrisch und affektgeladen und eine gewisse mässige intellektuelle Tätigkeit verratend. Die Stirn zeigt in ihrer Glätte und mit der gänzlich symmetrischen Stirntour eher wieder weibliche Züge. Die Nase ist im Vergleich zur Breite des Gesichtes und zur G r ö ß e der Augen extrem schmal und recht lange. Die W a n g e n wie gesagt faltenlos. Der M u n d wirkt ausgesprochen jugendlich; die fleischige Mundpartie steht in einem deutlichen Kontrast zu den leicht eingefallenen W a n g e n . D e r Betrachter wird m. a. W . i m m e r wieder verunsichert durch intrigierende Gegensätze, die er einem Christusbildnis, das 84 χ 45, 5 cm mißt, nicht einräumen möchte. U n d doch sind diese Gegensätze da: 1. Die schönheitliche Haartracht kontrastiert mit dem ernsten Ausdruck der Augen. 2. Der weibliche Haaransatz kontrastiert mit den männlichen Brauen. 3. Die eingefallenen Augen kontrastieren mit d e m sinnlich rosafarbenen M u n d . 4. Die überaus schmale Nase kontrastiert mit d e m geradezu athletischen Hals. Am meisten sprechen den Betrachter die Augen an. Sie sind nicht eigentlich individuell geformt, wohl aber die Brauen. Es sind auch die einzigen Gesichtsteile, an denen K o n t u r e n v o r k o m m e n . Sonst fehlen K o n t u r e n gänzlich. Sie bewirken den etwas starren Blick. Die K o n t u r wird bewußt eingesetzt, u m die Augen auf den Betrachter hin zu fixieren. W e r den S c h n u r r b a r t isoliert betrachtet, muss zugeben, dass es sich u m eine eher dandyhafte Form von S c h n u r r b a r t handelt: er scheint mit H a a r ö l behandelt und ausgekämmt worden zu sein, wirkt aber letztlich doch eher wie ein Pinselwischer denn als Schnurrbart. Der Künstier dieser Ikone war sicher nicht der erste Künstler der frühchristlichen Zeit, der diesen T y p u s eines Christusbildes schuf. Er interpretierte diesen T y p u s in eigenwilliger Weise. Der Vergleich mit

der viel späteren Mosaikikone in den Staatlichen Museen 2 3 in Berlin m a c h t den Unterschied deutlich: d e r frühbyzantinische K ü n s t l e r erlaubte es sich, Individualität, Schönheitlichkeit und Repräsentation zu vereinen. Die Gegenüberstellung der beiden Ikonen aus dem 6. und dem 11./12. J h . bringt in erster Linie das lange d a u e r n d e Wirken eines Bildnistypus zur Evidenz, bei dem die H a a r - und Barttracht, die Gestik und die K ö r p e r h a l t u n g attributiv eingesetzt sind. Modische, künstlerisch-stilistische und inhaltlich-mentale V e r ä n d e r u n g e n konnten wie in gleich bleibende K o o r d i n a t e n des Bildnistypus eingetragen werden. Diese Gesetzmäßigkeit erweist ihre Gültigkeit nicht n u r a m Christusbild, sondern auch am Heiligenbild, dem ich mich jetzt zuwenden möchte.

2. Bischofs- und Heiligenbildnisse Ich stelle an den Beginn ein b e r ü h m t e s Zeugnis, welches wie eine christliche Version der oben zitierten Stelle aus der Plotinbiographie wirkt: Brief 30 von Paulinus von N o l a gibt uns Einblick in die V e r w e n d u n g von Bildnissen von 'Glaubenshelden' durch Bischöfe. Sein Freund Sulpicius Severus bat Paulinus kurz nach dem J a h r e 400, ihm ein Bildnis zu senden, damit er es im neu erbauten Baptisterium in Primuliacum z u s a m m e n mit d e m Bild des M a r t i n von T o u r s anbringen könne. Paulinus kleidete seine Absage an Severus in einen gewundenen Traktat, den er mit Zitaten aus den Paulusbriefen spickte. N a c h d e m Severus nochmals nachfragte, sandte Paulinus seinem Freund statt des gewünschten Bildnisses einfach Verse, die nicht ganz frei sind von unterwürfiger Eitelkeit. Was soll ich dir auf eine solche Bitte antworten, ich möge dir mein Bild malen und schicken lassen? Bei unserer innigen Verbundenheit rufe ich dich an: welche Befriedigung wahrer Liebe suchst du bei den leeren Formen? Was für ein Bild begehrst du, dass ich schicke: des irdischen Menschen oder des himmlischen?—Ich weiss, du begehrst jene unverwesliche Gestalt, die der Himmelskönig in dir selber liebgewonnen hat; es kann dir ja von uns keine andere Form wertvoll sein als die, der du selber nachgeformt bist. . . . Aber, ich Armer und

23

Warland, Das Brustbild Christi, p. 146, Abb. 76.

Betrübter! Noch bin ich mit einem qualmig-irdischen Bilde verwachsen und spiegle in meinem fleischlichen Fühlen und irdischen Handeln mehr von dem ersten als von dem zweiten Adam. Wie darf ich mich für dich malen lassen, wenn ich durch meine irdische Verderbnis das Bild des himmlischen Menschen offenkundig verleugne? Beiderseits bedrängt mich die Scham: ich erröte zu malen, was ich bin, und wage nicht zu malen, was ich nicht bin . . . Dies ist Verwandlung durch die Hand des Höchsten, wenn wir dereinst aus uns in den Menschen verwandelt werden, der Gott gemäß geschaffen ist und dessen Bild himmlisch ist, während wir den ablegen, der gemäß den Begierden des Irrtums verwest. Dieses letzteren Bild möge doch Gott in mir zerstören—aufrichten aber und vollenden wolle er in uns Sein Bild, mit dem wir ohne Scham gemalt werden können. 24 Der zentrale Satz lautet: qualem cupis ut mittamus imaginem tibi: terreni hominis an coelestis? Die Unterscheidung zwischen einem irdischen und einem himmlischen Menschen fußt auf d e m 1. Korintherbrief 15. Paulus sagt: " u n d es gibt himmlische Leiber (σώματα επουράνια) und irdische Leiber (σώματα έπίγεια)." Diese Unterscheidung trifft Paulus im Hinblick auf den Auferstehungsleib: "es wird gesät ein natürlieher Leib, es wird auferweckt ein geistiger Leib", und: "der erste Mensch ist von der Erde, irdisch, der zweite ist vom H i m m e l " , " u n d wie wir das Bild des irdischen getragen haben, werden wir auch das Bild des himmlischen tragen" (1. Kor. 15. 49). Paulinus sendet seinem Freund nun nicht—wie m a n vielleicht erwarten würde—ein Bild des "himmlischen Paulinus", sozusagen o h n e Porträtähnlichkeit, sondern er wünscht sich im Sinne von Genesis 1. 26 eine gottebenbildliehe Darstellung seiner selbst. Aber das war bloß Wunschdenken.

24

Paulinus von Nola, Epistulae X X X . 2 5 ‫ ־‬. Hsg. von G. de Härtel, CSEL 29 (Wien, 1894), pp. 263-266: 2. Quid enim tibi de Uta petitione respondeam, qua imagines nostras pingi tibi mittique iussisti? obsecro itaque te per uiscera caritatis, quae amoris um solatia de inanibus formis petis? qualem cupis ut mittamus imaginem tibi? terreni hominis an caelestis? scio quia tu illam incorruptibilem speciem concupiscis, quam in te rex caelestis adamauit. neque enim alia potest tibi a nobis necessaria esse quam ilia forma, ad quam ipse formatus es, . . . sed pauper ego et dolens, quia adhuc terrenae imaginis squalore concretus sum et plus de pnmo quam de secundo Adam cameis sensibus et terrenis actibus refero, quomodo tibi audebo me pingere, cum caelestis imaginem infitiari prober corruptione terrena? utnmque me concludit pudor: erubesco pingere quod sum, non audeo pingere quod non sum;. . . haec est enim mutatio dexterae excelsi, cum inmutabimur a nobis in ilium hominem, qui secundum deum creatus est, cuius imago caelestis est, deponentes eum qui corrumpitur secundum desideria erroris. huius imaginem in me, quaeso, deus conterat et ad nihilum redigat imaginem nostram id est terrenam in ciuitate circumstantiae et instauret in nobis atque perfiaat imaginem suam, in qua nos pingi non pudet. A. Cizek, "Das Bild von der idealen Schönheit in der lateinischen Dichtung des Frühmittelalters", Mittellaldnisches Jahrbuch 26 (1991), pp. 3-35, bes. p. 13.

In T a t und Wahrheit ließ Paulinus kein Bildnis von sich anfertigen. So bekannt diese Äusserungen von Paulinus sind und so häufig sie zitiert wurden, ihre geschichtliche Aussagekraft im Hinblick auf das spätantike Bildnis ist bisher wenig ausgeleuchtet worden. Die Gottebenbildlichkeit des Menschen ist in der frühchristlichen Kunst kein T h e m a geworden. Dieses Problem erfüllte die Köpfe der Theologen, nicht die der Künstler! 2 3 An diesem Text sind a n d e r e Aspekte bedeutungsvoll. Erstens handelt es sich u m ein recht frühes Zeugnis für ein Bildnis eines lebenden Bischofs, das mit dem Bildnis des prominenten Mönchsbischofs Martin von Tours, der soeben erst verstorben war (a. 397), assoziiert wird. Das Heil, das dem Mönchsbischof zugeschrieben wurde, sollte auf den noch lebenden Paulinus übergehen. Das Memorialbildnis des T o t e n sollte eine gewisse Heilswirkung evozieren. Zweitens handelt es sich um ein Bischofsbild in einem Baptisterium. In Baptisterien sind häufig Märtyrer bestattet worden. Ennodius besingt die Darstellung von Märtyrern, deren Leiber in einem Baptisterium rekondiert waren: In baptisterio Ugello ubi picti sunt martyres quorum reliquiae conditae sunt ibi.26 Drittens wird hier deutlich, daß für Paulinus von Nola sein eigenes Bildnis Gegenstand f r o m m e r Überlegungen ist. Das Bischofsbild soll im Sinne eines moralisch hochstehenden Exempels andächtig betrachtet werden. W e n n m a n die Äußerungen des Paulinus nicht als eitles rhetorisches Feuerwerk abtun will, dann ist immerhin der Gesichtspunkt beachtenswert, d a ß die Betrachtung eines Bischofs als potens et vere apostolicus episcopus27 sowie als Patron eine gewisse Andacht zu evozieren pflegte. D e m Bischof als einem "holy m a n " konnte sehr bald einmal V e r e h r u n g z u k o m m e n , d e n n Martin und Paulinus waren für Sulpicius Severus als Mönchsbischöfe besonders verehrungswürdig. Viertens wird deutlich, d a ß Paulinus nicht einmal bei seinem eigenen Porträt Wert auf eine individuelle Darstellung legte. Zweifellos war er in dieser Hinsicht nicht der erste oder der einzige, der so dachte (s. oben Plotin). Das Faszinierende an diesem T e x t besteht

25

G. Ladner, Ad imaginent Dd. The image of man in medieval art, Wimmer Lecture 1962 (Latrobe, Pennsylvania, 1965); man vergleiche dazu die Rezension von E. Kitzinger, Speculum 43 (1968), pp. 355~359. 26 F.W. Deichmann, Ravenna. Hauptstadt des spätantiken Abendlandes. Kommentar 3. Teil (Stuttgart, 1989), p. 183. 27 P. Brown, "The Rise and Function of the Holy Man in Late Antiquity", in: Society and the Holy in Late Antiquity (Berkeley, 1989), p. 137.

in der Duplizierung des Interesses für einen himmlischen und einen irdischen Leib auf der G r u n d l a g e von l . K o r . 15. Paulinus von Nola war ein ausgesprochen kunstsinniger Patron mit weit reichendem Einfluß. W e n n er im Z u s a m m e n h a n g mit seinem eigenen Bischofsbildnis von einem himmlischen und einem irdischen Leib sprechen konnte, d a n n dürfte diese Sehweise auch für die Künstler seiner Zeit, an die er sich wandte, belangreich geworden sein. N u r wenige J a h r e nach Paulinus von Nola hat Hilarius, Bischof von Arles, in seiner Lebensbeschreibung des H o n o r a t u s ein ähnliches Bild des neuen Menschen der christlichen Spätantikc entworfen: Rediguntur ad breves capillos luxuriantes comae. Transfertur ad nitorem menüs vestium splendor. Cervicis lacteae decus palliis rigentibus occupatur. Transit laetitia in serenitatem, membrorum vigor animi vigor mutatur. Virtus corporis in virtutem spiritus migrât. Pallescit jejunio speciosa facies, et prius succi plena, fit plena gravitatis (Migne PL. 50. 1253B).

Bischofsbildnisse verkörperten ein eigenes Genus. Sie waren die christliehen Nachfolger der Kaiser- und Beamtenporträts aus M a r m o r . Wie den Kaisern und den h o h e n Beamten stand den Bischöfen das Bildnisrecht zu. Aber für Bischofsbildnisse wählte m a n wie gesagt nie M a r m o r , sondern Holz, Fresko und Mosaik, weil dem M a r m o r und der Bronze allzu sehr ein heidnisches Imaginaire anhaftete. An dieser Stelle erscheint es mir angezeigt, auf das früheste erhaltene Bischofsbildnis R o m s einzugehen: das Bildnis von Papst Liberius (Abb. 12) in der Praetextatkatakombe. 2 8 An der rechten Stirnwand des Arkosolgrabes der Celerina ist eine männliche, bartlose Gestalt in Tunika, Pallium und Sandalen mit einer geschlossenen Rolle unter Girlanden stehend dargestellt. Die Inschrift LIB(. . .)S ist ergänzbar als L I B E R I U S . Die untere Front des Grabes zeigt j e n e b e r ü h m t e Szene mit Susanna als L a m m (Abb. 13), die von den beiden Wölfen (beschriftet mit S E N I O R I S ) bedroht wird. Im Tonnengewölbe des Arkosolgrabes ist Christus im Medaillon erkennbar, den Petrus, Paulus, Sixtus und Laurentius begleiten. Der N a m e der verstorbenen Celerina ist an der Front des Grabes zu lesen: "Le fresquiste a voulu évoquer la délivrance du danger: en se plaçant sous le patronage de Suzanne,

28

C. Dagens, "Autour du pape Libère. L'iconographie de Suzanne et des martyrs romains sur l'arcosolium de Celerina", Mélanges d'Archéologie et d'Histoire de l'Ecole française de Rome 78 (1966), pp. 327-381.

Celerina proclame son espérance d'être sauvée." 2 9 Die Wölfe stellen höchstwahrscheinlich nicht n u r die beiden Alten, sondern ganz spezifisch Christenverfolger dar. Dies konnte C. Dagens wahrscheinlich m a c h e n . Als G a r a n t e n ihrer O r t h o d o x i e ließ Celerina die beiden Apostelfürsten mit Sixtus und Laurentius sowie den Papst Liberius darstellen, mit dem sie so eng v e r b u n d e n war, dass ihr N a m e im Martyrologium Hieronymianum z u s a m m e n mit Liberius a m 24. September notiert w o r d e n ist. Liberius (352-366) war der persönliche Protector und Intercessor der Celerina, 3 0 n a c h d e m er selbst vom arianischen Kaiser Constantius in die V e r b a n n u n g geschickt worden war. Entsprechend den Gewohnheiten der Katakombenmaler ist für das Porträt des Liberius ein Kopftypus gewählt w o r d e n , d e m m a n auch sonst in der zweiten Hälfte des 4. J h . in den K a t a k o m b e n begegnet: starke s c h w a r z e K o n t u r e n , grosse s c h w a r z e A u g e n ; die A u g e n blicken auffallend lebendig z u m G r a b . Das Inkarnat ist das eines j u n g e n M a n n e s . Die H a a r e sind im Sinne des Zeitstils kalottenartig auf das längliche Gesicht aufgesetzt. Bemerkenswert ist die Ganzfigurigkeit des Papstes: er ist mit seinem ganzen irdischen K ö r p e r vergegenwärtigt. Celerina versteckte sich in einer vielsagenden alttestamcntlichen M e t a p h e r und identifizierte sich mit d e m L a m m , sprich mit Susanna, aber Liberius ist in voller Gestalt dabei, fast lebensgroß u n d unübersehbar. M a n m u ß sich vorstellen, dass Liberius bereits verstorben w a r , als C e l e r i n a n o c h zu ihren Lebzeiten das P r o g r a m m ihres Arkosolgrabes ausdachte. Offensichtlich war Celerina nicht im Besitz eines Bischofsbildnisses von Liberius; u n d falls sie d e n n o c h ein solches besaß, d a n n hat sie keinen W e r t d a r a u f gelegt, d a ß dieses in der K a t a k o m b e bei ihrem G r a b getreulich wiedergegeben würde. Sie begnügte sich mit der Ganzfigur des Liberius, der übrigens nicht einmal betend dargestellt ist. Ü b e r das Formelhafte hinaus verlieh der Künstler d e m Antlitz des Papstes einen lebendig-anteilnahmsvollen Blick. O f f e n b a r war Celerina mit d e m religiös-politischen Schicksal des Liberius derart v e r b u n d e n , d a ß sie nicht p r i m ä r seiner Fürbitte bedurfte, sondern die M e m o r i a des römischen Bischofs sollte mit Hilfe seiner Ganzfigur und d e m Blickkontakt z u m G r a b vergegenwärtigt werden.

29

Dagens, "Autour du pape Libère. romains sur l'arcosolium de Celerina", 30 Dagens, "Autour du pape Libère. romains sur l'arcosolium de Celerina",

L'iconographie de Suzanne et des martyrs p. 345. L'iconographie de Suzanne et des martyrs pp. 355-6.

N u n kann m a n sich d a r ü b e r streiten, ob mit der Ganzfigur der irdische oder der himmlische Leib des Liberius gemeint war. D a er verstorben war u n d demzufolge im Refrigerium interim die Auferstehung erwartete, dürfte die G a n z f i g u r eher seinen himmlischen Leib, den Auferstehungsleib beinhaltet h a b e n . D a ß die Diskussion des Paulinus mit Sulpicius Severus ü b e r sein Porträt keine akademische war, beweisen die Mosaiken in S. Vittore in Ciel d'oro in Mailand. 31 Diese Mosaiken illustrieren den Briefwechsel auf erstaunliche Weise. Bevor Ambrosius, der große Erzbischof von Mailand, die Basilica Martyrum auf dem Coemetmum ad Martyres Mailands errichtete, s t a n d e n dort verschiedene kleine K a p e l l e n , sog. Cellae Memoriae; eine von ihnen war d e m Hl. Viktor geweiht. Im J a h r e 375 bestattete Ambrosius seinen Bruder Satyrus in S. Vittore in Ciel d'oro. D e r kleine Bau misst 8 x 5 m . Vermutlich im 5. J h . erhöhte ein E h e p a a r n a m e n s Paneciria u n d Faustinus die Kapelle mittels einer K u p p e l und stiftete W a n d m o s a i k e n . Es ist in h o h e m Masse wahrscheinlich, dass sich Paneciria u n d Faustinus in dieser Kapelle neben d e m Bruder des Ambrosius, Satyrus, bestatten Hessen. Das von i h n e n gewählte M o s a i k e n p r o g r a m m ist d e m P r o g r a m m der Celerina in der Praetextatkatakombe durchaus vergleichbar. Dargestellt sind in der K u p p e l der Hl. Viktor, an den W ä n d e n die mailändi-

31

F. Reggiori, IM Basilica Ambrosiana. Ricerche e Restauri 1929-1940 (Milano, 1941), pp. 169-273; A. Calderini, Storia di Milano I (1953), pp. 677-678, Farbtafeln ad pp. 392, 432 und 448; G. Bovini, "Gli edifici di culto milanese d'età pre-ambrosiana", Corsi di cultura sull'arte ravennate e bizantina 8 (1961), pp. 47-72: über S. Vittore in caelo aureo: pp. 64 70; G. Bovini, "I mosaici di S. Vittore in ciel d'oro di Milano", Corsi di cultura sull'arte ravennate e bizantina 16 (1969), pp. 71-80; C. Berteiii, "Mosaici a Milano", in: Atti del X. Congresso Internazionale di Studi sull'Alto Medioevo (1986), pp. 333-354; C. Berteiii, "Milano, una capitale da Ambrogio ai Carolingi", II millenio ambrosiano (Mailand, 1987), pp. 8-31; M. Raspe, "Un naturale ritratto di Santo Ambrogio: Carlo Borromeo und das Mosaikporträt in S. Vittore in Ciel d ' O r o zu Mailand", in: Festschrift H. Brandenburg, Boreas 17 (1994),pp. 203-215. Raspe hält die ganze Gestalt des Ambrosius für eine Arbeit des letzten Viertels des 16. Jh., die Carlo Borromeo darstellen soll. Allein, diese Hypothese steht auf schwachen Füssen. Zunächst sind die neuzeitlichen Restaurierungen in S. Vittore von bloßem Auge zu erkennen. Zu diesen gehört freilich die Gestalt des Ambrosius nicht. Den Restauratoren, welche zwischen 1980 und 1989 die Mosaiken von S. Vittore mit grosser Sorgfalt konsolidierten, wäre selbstverständlich eine Restaurierung der Gestalt des Ambrosius nicht entgangen. Darüber geben die Berichte von A. Surace, C. Ferrari da Passano und J . Kosinka Aufschluss, welche in den Atti del convegno nazionale sul restaure in situ di mosaici parietali, Ravenna 1~3 Ott. 1990 mit dem Titel Mosaici a S. Vitale e altri restauri (Ravenna, 1992) publiziert worden sind; G. Mackie, "Symbolism and Purpose in an Early Christian Martyr Chapel: T h e Case of San Vittore in Ciel d'Oro, Milan", Gesta 34.2 (1995), pp. 91-101.

sehen Bischöfe M a t e r n u s (gest. 304) und Ambrosius (Abb. 14) (gest. 397) sowie die M ä r t y r e r p a a r e Gervasius (Abb. 15) u n d Protasius (Abb. 16) und N a b o r und Felix. Die Botschaft des P r o g r a m m s ist darin zu erblicken, dass auf der einen W a n d der Bischof Ambrosius von den Heiligen Gervasius (Abb. 15) u n d Protasius, auf der gegenüberliegenden W a n d der Bischof M a t e r n u s von den Heiligen N a b o r und Felix für den H i m m e l empfohlen werden. Die Assoziation von Ambrosius mit Gervasius und Protasius ist vielsagend. Hier finden wir eine ähnliche K o m b i n a t i o n von Heiligen mit Bischöfen wie in Primuliacum und in der Praetextatkatakombe. Felix, N a b o r und Viktor waren Soldaten aus M a u r e t a n i e n , welche in Lodi und Mailand den M ä r t y r e r t o d erlitten hatten. Die Leiber der beiden M ä r t y r e r Gervasius und Protasius (Abb. 1 5 1 6 ‫ ) ־‬hatte Ambrosius (Abb. 14) aufgrund einer Vision in Mailand entdeckt und in der Basilica Apostolorum beigesetzt. Sie galten alle als echte mailändische Märtyrer. Die nicht nimbierten Gestalten sind inschriftlich bezeichnet und wie am Celerinagrab in Ganzfigur dargestellt. D a m a n nicht wusste, wie Gervasius und Protasius einst aussahen, griff m a n zu etablierten Vokabeln. Z u r Verbildlichung von Anonymi boten sich grundsätzlich zwei Möglichkeiten an: Jugendlichkeit bzw. Bartlosigkeit sowie Alter und Weißhaarigkeit mit Bart. Im Vergleich zu den M ä r t y r e r n ist j e d o c h Ambrosius mit individuellen Gesichtszügen versehen worden. Er hält sein H a u p t etwas geneigt, worin m a n einen Hinweis auf sein vorgerücktes Alter erblicken kann—dies im Gegensatz zu den Märtyrern, die gänzlich aufrecht und frontal stehen. Er nennt auffällig große O h r e n sein eigen—die Märtyrer h a b e n dagegen kleine Mäuseohren. Der M u n d des Ambrosius ist voll-lippig, vielleicht sogar leicht geöffnet (redend?), der Blick reflektierend und wach. Der Ausdruck der Märtyrer wirkt d e m g e g e n ü b e r stereotyp und undifferenziert. Zweifellos hatten die beiden Auftraggeber des M o s a i k e n p r o g r a m m s von S. Vittore in Ciel d ' o r o Z u g a n g zu einem Porträt des Ambrosius. D a Ambrosius Erzbischof war, kam ihm das Bildnisrecht zu. "Baldigst, n a c h d e m der neue Bischof auf die K a t h e d r a erhoben war, wurde sein Bildnis in der Kathedrale aufgehängt". 32 Falls der Mosaizist den 397 verstorbenen

32

F.W. Deichmann, Ravenna. Geschichte und Monumente (Wiesbaden, 1969), p. 123; allgemein zu diesem T h e m a ebd., pp. 123 125; Kommentar 1. Teil (1974), p. 120; Kommentar 2. Teil (1976), p. 180; vgl. auch die in Masse hergestellten Bildnisse des Bischofs Meletius in Antiochien, über die J o h . Chrysostomos berichtet: PG. 50, p. 516; dazu Grabar, Le portrait en iconographie paléochrétienne, p. 602.

Metropoliten Mailands gekannt h a b e n sollte, d a n n kann das Mosaik nicht viel nach 400 entstanden sein. Eine so frühe Entstehung der Mosaiken ist bisher nicht erwogen worden. Das D a t u m dieser Mosaiken ist nicht genau bestimmbar. Sie gehören vielleicht eher in die zweite Hälfte des 5. J h . als in dessen erste Hälfte. Mein Fazit lautet: Die Kultpropaganda für die mailändischen Märtyrer Protasius, Gervasius, Felix, N a b o r und Viktor wird mit Hilfe p r o m i n e n t e r mailändischer Bischöfe verstärkt. Die Aufgabe von Ambrosius war es, mit seinem individuellen Porträt mailändische Authentizität zu evozieren. Die K o m m e m o r a t i o n der M ä r t y r e r und Bischöfe erfolgt mit Hilfe nicht betender Ganzfiguren wie am Celerinagrab. Das M a n k o ihrer physisehen Anonymität wurde durch die Vokabeln des Alters u n d der J u g e n d gleichsam geheilt. Die individuelle Physiognomie bei Bischöfen begegnet in Italien häufiger und in viel ausgeprägterer Form als im griechischen Osten. Das erweisen die Mosaiken von S. Vitale in R a v e n n a , wo Maximian (Abb. 17) als noch lebender Bischof höchst individuell inmitten des kaiserlichen Gefolges dargestellt ist. Erstaunlich ist dagegen der prachtvoll individualisierte Kopf des Hl. Vitalis (Abb. 18) in der Apsis, den in Ravenna sicher niemand kannte. Der Neubau einer Memorialkirche für den Hl. Vitalis hatte die wachsende Beliebtheit des Märtyrers im 5. J h . zur Ursache. 3 3 Zu Beginn des 6. J h . hatte sich in R a v e n n a eine neue Vitalis-Legende etabliert, nach welcher dieser in R a v e n n a und nicht in Bologna den Märtyrertod erlitten h a b e n soll. Die neue Legende schrieb ihm a u ß e r d e m die beiden mailändischen Heiligen Gervasius und Protasius als Söhne zu. In T a t und Wahrheit hatte Ambrosius die Agricola- und Vitalisreliqien in Bologna höchst persönlich ausgegraben u n d rekondiert. Die Vitalislegende m u ß t e im 6. J h . in R a v e n n a neu erfunden werden, weil es dort schlechterdings keine ortsansäßigen Märtyrer gegeben hatte. Das größte Defizit der Kaiserresidenz R a v e n n a beruhte im Mangel an autochthonen M ä r tyrern. D a h e r verschafften sich die R a v e n n a t e n im 6. J h . eigene Reliquien. Im Wetteifer mit Mailand und auf Kosten Mailands haben die R a v e n n a t e n ihre Stadt als kirchliche Metropole durchgesetzt. 3 4

33

Deichmann, Ravenna. Geschichte und Monumente. Kommentar 2. Teil, p. 7. Deichmann, Ravenna. Geschichte und Monumente. Kommentar 2. Teil, p. 8; R.A. Markus, "Carthage—Prima Justiniana—Ravenna: An Aspect of Justinian's Kirchenpolitik", Byzantion 49 (1979), pp. 277-302; für den Hinweis auf diese Publikation danke ich Evangelos Chrysos. 34

D e r Luxusbau von S. Vitale war vielleicht darauf angelegt, den Zentralbau von S. Lorenzo in Mailand zu überbieten, indem im G r u n d und Aufriß auf die kaiserliche Kirche von Hagios Sergios und Bacchos in Konstantinopcl verwiesen wurde. Die original mailändischcn M ä r tyrer Gervasius und Protasius wurden sozusagen ravennatisiert und bekamen das topische Antlitz von glatt rasierten Adolescentcn. V o r allem aber wurden sie in die Serie der Apostel a m C h o r b o g e n von S. Vitale aufgenommen. Sie haben beide weiße Gesichter, sind jugendlich-bartlos und genau so dargestellt wie die Apostel über ihnen. In S. Vittore in Ciel d'oro in Mailand waren sie bärtig. O f f e n b a r hatte sich keinerlei V e r a b r e d u n g über ihre Physis ausgebildet. Die beiden mailändischcn Heiligen w a r e n n u n m e h r selbst Apostel und zwar ravennatische geworden! Der bolognesische Märtyrer Vitalis 35 dagegen, den weder Ambrosius noch die Ravennaten de visu kennen konnten, erhielt in S. Vitale in R a v e n n a das selbstbewußte Gesicht eines Offiziers sowie gepflegt coiffierte grau melierte H a a r e . Der Vergleich mit dem Kirchengründer Ecclesius im selben Apsismosaik, der auch schon längst tot war aber möglicherweise existierte von ihm ein offizielles Bischofsbild—ergibt, daß sich die Augenpartie mit den stark abgewinkelten Brauen bei Ecclesius und bei Vitalis wieder findet. U n d doch unterscheidet sich die Physiognomie der beiden beträchtlieh. Fazit: Die vermeintliche Individualität des Vitalis war keine Verlegenheitslösung des Mosaizisten sondern sie sollte ravcnnatischc Auctontas evozieren. Endlich wußten die Ravennaten, wie ihr Märtyrer ausgesehen hatte! M a n versteht n u n besser, weshalb sich Agnellus im 9. J h . in seinen Beschreibungen der ravennatischen Bischöfe auf die bildlichen Darstellungen beruft (vgl. A n m . 20). Wie kam es überhaupt dazu, d a ß die R a v e n n a t e n den Heiligen Vitalis, Protasius und Gervasius einen derart aufwendigen Sakralbau innerhalb der Stadtmauern errichten wollten? W a r hier die Konkurrenz zur Apostelkirche in Konstantinopel oder zu San Lorenzo in Mailand mit im Spiel? Diese beiden kaiserlichen Residenzstädte konnten bedeutende Märtyrerkirchen aus d e m 4. J h . ihr Eigen nennen. R a v e n n a hatte in dieser Hinsicht ein schmerzliches Defizit wettzum a c h e n . S. Apollinare in Classe wurde gleichzeitig mit S. Vitale, j e d o c h in traditioneller Weise extra muros über dem M ä r t y r e r g r a b

3

•' Deichmann, Ravenna. Geschichte und Monumente, p. 255; zum Bildnischarakter des Vitalis und des Ecclesius äusserte sich Deichmann eher heiläufig: Kommentar 2. Teil, pp. 194-195, wo vom Naturalismus des Porträts die Rede ist.

erbaut. D a ß m a n die Heiligen Vitalis, Protasius und Gervasius in der Stadt mittels eines luxuriösen Zentralbaus, der offenkundige Konstantinopolitaner Innovationen protzig zur Schau stellte, ehren wollte, war m. E. vielsagend und wirkt wie eine rhetorische Geste, u m die Gläubigen zu überzeugen, d a ß Vitalis, Protasius und Gervasius nunm e h r ravennatische Heilige waren. Für diese ravennatisierten Heiligen war nur die anspruchsvollste Architektur gut genug, eine Architektur, deren kaiserliche Vorbilder offenkundig waren. D a m a n offenbar für die Mosaikdarstellung der ravennatischen Bischöfe Ecclesius u n d M a x i m i a n echte Bischofsbildnisse verwenden konnte, war es notwendig geworden, auch für den im G r u n d e physisch gänzlich unbekannten Vitalis eine rhetorisch überzeugende Physiognomie zu kreieren. Der Entscheid fiel zu Gunsten eines selbstbewußten und modisch coiffierten Offiziers. M a n kann nicht über Heiligenbildnisse in R a v e n n a sprechen, ohne sich mit den Mosaiken von S. Apollinare N u o v o aus der Agnelluszeit zu beschäftigen. Zu diesen Heiligenprozessionen des 6. J h . bilden niederländische Gruppenbildnisse des 17. J h . , wie sie R e m b r a n d t , Frans Hals und V a n Hoogstraten malten, 3 6 den d e n k b a r größten Kontrast hinsichtlich der individuell-physiognomischen Charakterisierungsmöglichkeiten. Ziel der niederländischen Künstler war die möglichst individuelle Persönlichkeitsdarstellung mit Hilfe der Physiognomie und der Körperhaltung. Die niederländischen Gruppenbildnisse sind mit den Heiligenprozessionen in R a v e n n a nur deswegen vergleichbar, weil hier wie dort die Kleidung im Sinne einer Uniform i e r u n g den H a u p t e i n d r u c k bestimmt. Erst bei längerem Verweilen erkennt der Betrachter in S. Apollinare N u o v o individuelle Unterschiede. Die künstlerische u n d darstellerische Problematik dieser Mosaiken ist von den meisten Autoren gänzlich verkannt worden. 3 7 Es fällt sofort auf, daß die J u n g f r a u e n (Abb. 19) untereinander viel ähnlicher charakterisiert sind als die männlichen Heiligen (Abb. 20), welche mittels verschiedener Bart- und H a a r t r a c h t e n und verschiedener Altersstufen differenziert sind. Die J u n g f r a u e n sind im wörtlichen Sinn durchwegs als j u n g e Frauen charakterisiert. Variiert wird vor allem ihre Kopfhaltung, ihre Augengröße und ihre Blickrichtung,

30

C. Brusati, Artifice and Illusion. The Art and Writing of Samuel van Hoogstraten (Chicago, 1995), pp. 83, 128-130, 247-249. 3 ‫ י‬Deichmann, Ravenna. Hauptstadt des spätantiken Abendlandes, Kommentar 1. Teil, pp. 149-150 mit weiteren Literaturangahen.

nota bene bei völlig gleich bleibendem rosa Inkarnat. Variiert sind auch die S c h m u c k d i a d e m e , aber die Frisuren sind u n t e r e i n a n d e r weitgehend gleich. Gänzlich falsch wäre es, von einer einfallslosen Repetition zu sprechen. Keine der J u n g f r a u e n stimmt mit einer anderen genau überein. Das Auge des Betrachters wird im Gegenteil aufgefordert, die Unterschiede im Ausdruck und in der Motivik abzutasten und vergleichend über die riesige Fläche zu eilen. O b w o h l die ganze Prozession still steht und nicht vom Fleck kommt, evozieren die verschiedenartigen Blickrichtungen und Kopfhaltungen der J u n g f r a u e n beim Betrachter Phasen geringerer und stärkerer Aufmerksamkeit. Auf Agnes wird m a n nicht nur wegen des Attributs des Lammes, sondern auch wegen ihres intensiven Blickes aufmerksam. Einen deutliehen Akzent setzt die schönheitlich aufgefasste Valeria (Abb. 19) mit ihrer selbstbewußt-aufrechten H a l t u n g und ihrem erwartungsvollen Blick. 58 Der Gesichtsausdruck der beiden weiter links stehenden J u n g f r a u e n Perpetua und Vincentia wirkt im Vergleich zu Valeria zurückgenommen. Sehr viel m e h r A u f w a n d hat der Künstler mit der variierten Charakterisierung der Gewänder getrieben. Die unterschiedliehen Designs der G e w ä n d e r und der modischen Accessoires fallen m e h r ins Gewicht als die verschiedenen Physiognomien, die letztlich alle aus ein- und demselben G r u n d m u s t e r hervorgehen. M a n m u ß a n n e h m e n , d a ß der C o n c e p t e u r dieser Mosaiken die Variationen des Designs, der Kopftypen, des Ausdrucks, der K ö r p e r h a l t u n g und der G e w ä n d e r sorgfältig plante, denn das Auge des Betrachters wird aufgefordert, dieselben auszuloten. Eine Repetition des Designs war nicht beabsichtigt; auch wäre es verkehrt, von Stilisierung, Typisierung und Stereotypie zu sprechen. Die Physiognomien, H a a r t r a c h t e n , K ö r p e r haltungen und die Elemente der M o d e sind zwar wie Attribute behandelt, aber dennoch variiert. Dieses attributive Denken tritt auch an den männlichen Heiligen zutage. Die Optionen der Porträtdarstellung lauten wie folgt: bartlos, bärtig, krause oder glatte H a a r e , T o n s u r oder keine T o n s u r , b r a u n e oder weiße H a a r e , Alter oder J u g e n d . Alter wird mit dem Attribut der weissen H a a r e , gelegentlich auch mit d e m Attribut der Runzeln auf Stirne und W a n g e n evoziert. 59 Selbstverständlich sind Alter und J u g e n d wichtige Attribute, dies ganz

38

F.W. Deichmann, Frühchristliche Bauten und Mosaiken in Ravenna (Baden-Baden, 1958), Tf. 134. 39 Deichmann, Frühchristliche Bauten und Mosaiken in Ravenna, Tf. 127.

im Gegensatz zu den J u n g f r a u e n . U n t e r den männlichen Heiligen sind zahlreiche Greise (Abb. 20) zu erkennen. Die Propagierung des "guten Alters" bzw. des καλόγηρος erfolgte erstmals durch die Christen seit der Zeit u m 400. 40 Possidius konnte von Augustinus sagen, dieser sei enutritus in bona smectute.n Heiligmässige Greise und alte Asketen erhalten in der 4 1 9 - 4 2 0 entstandenen Historia Lausiaca das Attribut des καλόγηρος. 12 Das gute Alter 1st "bewußt und intensiv aul die Zukunft, auf das ewige Leben ausgerichtet". 4 3 Z u r Charakterisierung des zum ravennatischen Heiligen avancierten Vitalis genügte das Attribut der krausen Haare. 4 4 Gervasius führt die Attribute der dunklcn kurzen H a a r e mit T o n s u r und der Bartlosigkeit mit sich herum; sein Gesicht ist ausgesprochen jugendlich. 4 5 Im G r o ß e n und G a n z e n sind die männlichen Heiligen physiognomisch stärker differenziert als die weiblichen Heiligen, die ganz einseitig Jugendlichkeit propagieren. Recht deutlich sind auch die Goldkränze und die G e w ä n d e r der Heiligen variiert. Abermals: von Stereotypie zu sprechen, ginge an der Sache vorbei. Das Problem muss in größerem R a h m e n der Darstellung physischer Anonymität gesehen werden, die nicht zu verwechseln ist mit der idealisierten Darstellung antiker G ö t t e r und Personifikationen. Weiter oben wurde darauf hingewiesen, dass die vier Porphyrtetrarchen von S. M a r c o in Venedig (Abb. 1) einander physiognomisch angeglichen w u r d e n , weil ihre Similitudo politische Concordia bezeugen sollte. Die Heiligen in S. Apollinare N u o v o sind einander physiognomisch keineswegs angeglichen, obgleich ihre Physis de facto gänzlich unbekannt war. Die physiognomische Similitudo der Porphyrtetrarchen von S. M a r c o in Venedig konnte nur deshalb ideologisch-politisch wirken, weil ihre w a h r e Physiognomie d u r c h a u s bekannt war und auf verschiedenen Marmorbildnissen dokumentiert ist. In S. Apollinare N u o v o geht es keinesfalls u m die Verbildlichung von Similitudo und Concordia, sondern um die Verbildlichung des Ideals der J u g e n d bei den J u n g f r a u e n und die Ideale der J u g e n d und des Alters bei den männlichen Heiligen. Die Auctoritas der Heiligen wurde mithilfe ihres Alters (bzw. J u g e n d ) und mithilfe von Attributen evo-

40

Ch. Gnilka, "καλόγηρος. Die Idee des 'guten Alters' bei den Christen ",Jahrbuch für Antike und Christentum 23 (1980), pp. 5 21. 41 Gnilka, "καλόγηρος. Die Idee des 'guten Alters' bei den Christen", p. 9. 42 Gnilka, "καλόγηρος. Die Idee des 'guten Alters' bei den Christen", p. 12. 4:1 Gnilka, "καλόγηρος. Die Idee des 'guten Alters' bei den Christen", p. 17. 44 Deichmann, Frühchristliche Bauten und Mosaiken in Ravenna, IT. 125 und 352. 4s Deichmann, Frühchristliche Bauten und Mosaiken in Ravenna, Tf. 125 und 338.

ziert. Ihre künstlerische Mission hingegen war es, Physis und Physiognomie innerhalb eines vorgegebenen Figurendesigns möglichst subtil zu variieren, so dass der Betrachter aufgefordert wird, von Figur zu Figur zu schreiten und die künsderische Varietas des rein Darstellerischen aufzuspüren. N a c h d e m das Ideal unvergänglicher J u g e n d bereits von den theodosianischen Kaiserbildnissen propagiert worden ist, kann es nicht weiter erstaunen, wenn sich später auch das Heiligenbildnis dieses Ideals bemächtigt. W i r h a b e n soeben bemerkt, dass Gervasius (Abb. 15) u n d Protasius (Abb. 16) in S.Vittore in Ciel d ' o r o in M a i l a n d in vorgerücktem Alter, in S. Vitale in R a v e n n a dagegen als weißhäutige u n d glattrasierte Adolescenten verbildlicht worden sind. Eine ähnliche V e r j ü n g u n g s k u r zeichnet sich an den Bildnissen des Heiligen Demetrios in Thessaloniki ab. 46 Das frühste datierbare Mosaik mit seiner Darstellung aus d e m nördlichen Seitenschiff von Hagios Demetrios in Thessaloniki (heute ausgestellt im weissen T u r m ) zeigt den Heiligen, dessen Physiognomie übrigens von seinen Biographen nie beschrieben w o r d e n ist, mit matt bräunlicher H a u t und einigen Falten unter den Augen u n d u m den M u n d (Abb. 21). D e r Heilige ist nicht m e h r ganz j u n g , sondern als reifer M a n n wiedergegeben. An den etwas späteren Pfeilermosaiken im C h o r von Hagios Demetrios (Abb. 22) ist d a n n alle Male ein eindeutig jugendlicher Heiliger mit sehr heller, fast weißer H a u t und großen aufgesperrten Augen verbildlicht w o r d e n . Dieses bleiche 'Milchgesicht' ist nicht zufällig von den Mosaizisten gewählt w o r d e n , sondern es war Ausdruck einer f r o m m e n Ideologie, w o n a c h d e m Heiligen zwar ein j u g e n d l i c h e s (sprich: ästhetisch angenehmes) Äußeres eingeräumt wird, wogegen seinem nicht sichtbaren, aber d e n n o c h präsenten Inneren die Lebense r f a h r u n g eines Greises z u k ö m m t . Dieses seit d e m 5. J h . in griechischen, seit dem 6. J h . in lateinischen Heiligenviten propagierte Ideal des Paidariogeron, das als solches weit in die griechische Zeit zurückreicht, m a c h t e sich auch in den künstlerischen Darstellungen bemerkbar. 4 7 Zu der E n t s t e h u n g und Entfaltung dieses Ideals hat C. Gnilka eine differenzierte Darstellung beigesteuert. 48 In den Texten

4,

‫ י‬Β. Brenk, "Von der Rhetorik der Wallfahrtsorte und Heiligenbilder", Architectural Studies in Memory of Richard Krauthämer, hrsg. von C.L. Striker (Mainz, 1996), pp. 33-36. 47 Brenk, "Von der Rhetorik der Wallfahrtsorte und Heiligenbilder". 48 Ch. Gnilka, "Aetas spmtalis. Die Überwindung der natürlichen Altersstufen als Ideal frühchristlichen Lebens", Theophaneia 24 (Köln, 1972).

der Spätantike ist allerdings nie vom ä u ß e r e n Aussehen des Paidariogeron die R e d e , und es wäre d a h e r verfehlt, von einer direkten künsderischen Umsetzung des in den Heiligenviten propagierten Ideals des Paidariogeron zu sprechen. Die g r u n d l e g e n d e n Textzeugnisse lauten wie folgt: 1. Antonius, Vita Symeons des Styliten c. 2: "In bezug auf dein Alter sehe ich dich j u n g , a b e r mit den G e d a n k e n eines Greises." 2. Historia Lausiaca c. 17 (Makarius der Ägypter): " E r wurde gewürdigt, eine so große G a b e der Unterscheidung zu besitzen, dass m a n ihn den 'jungen Greis' (Paidariogeron) nannte, c. 61 (Melania die Jüngere): "Es wäre nämlich nicht gerecht, sie wegen ihres d e m Fleische nach noch jugendlichen Alters gering zu schätzen u n d deswegen ihre so reiche T u g e n d zu verachten, übertrifft diese doch wahrhaftig auch die älterer und vortrefflicher F r a u e n . " 3. Kyrill von Skythopolis, Vita des Hl. Sabas c. 11: " I n d e m er die Lebensart des Sabas zur K e n n t n i s g e n o m m e n hatte, n a n n t e ihn der verehrungswürdige Euthymios in vollem V e r t r a u e n Paidariogeron." 4. G r e g o r der Grosse, Dialogi II, V o r r e d e (über Benedikt): fuit vir vitae venerabilis . . . ab ipso suae pueritiae tempore cor gerens senile. ("Schon als K i n d trug er das H e r z eines Greises in sich.") 5. Beda, Leben der Abte von Wearmouth and Jarrow ( Vita sanctorum abbatum monasterii in Wiramutha et Girvum c. 1, PL. 94, 713 C). Es ist schwierig zu sagen, welche dieser Textstellen a m meisten die künstlerische Phantasie anregten und in welcher Kunstiandschaft erste Paradigmen geschaffen w u r d e n , zumal sich außerordentlich wenige frühchristliche Heiligenbildnisse erhalten h a b e n . Deutlich f a ß b a r wird das Ideal des Paidariogeron in den Mosaiken von Hagios Demetrios in Thessaloniki, wo nicht n u r der Heilige Demetrios, sondern auch der Heilige Sergios mit einem Adolescentengesicht ausgestattet wurden. D e r Heilige Sergios ist a u f g r u n d einer von weitem erkennbaren Inschrift sogleich identifizierbar. D e r Heilige Demetrios ist dagegen nicht i m m e r beschriftet, u n d da seine Physiognomie u n d H a a r t r a c h t von Mal zu Mal variiert, kann seine Identität nach heutigem Ermessen nicht h u n d e r t p r o z e n t i g ausgemacht werden (Abb. 22). A m südöstlichen Pfeilermosaik wird D e m e t r i o s zwischen den beiden Stiftern gezeigt. H i e r ist sein Gesicht länglich, schmal und oval, und die H a a r e sitzen wie eine halbkreisförmige K a p p e auf d e m Schädel. Auf d e m nordöstlichen Pfeiler ist Demetrios (?) mit zwei K i n d e r n abge-

bildet, denen er offenbar Hilfe geleistet hat. Sein Antlitz ist m e h r breit als hoch, und die H a a r e ähneln denen des Heiligen Sergios: sie sind lockig-aufgebauscht und bedecken beidseits die O h r e n . W e n n es sich hier beide Male u m den Heiligen Demetrios handelt, d a n n muss beim Betrachter eine gewisse Toleranz hinsichtlich der Porträtattribute vorausgesetzt werden, d. h. es genügte d e m Betrachter, in der Demetrioskirche einen jugendlichen Heiligen mit wechselndem Gesichts- und Frisurentypus w a h r z u n e h m e n , u m zu wissen, d a ß damit stets der Patron der Kirche gemeint war. Das einzige für das Signalement ausschlaggebende Attribut war das faltenlose und bleiche Adolescentengesicht. An dieser Stelle sei auf die bekannte Ikone mit der Darstellung der beiden Heiligen Sergios u n d Bacchos vom Sinai in Kiev (Abb. 23) hingewiesen, 49 weil die beiden Heiligen einander zum Verwechseln ähnlich sind: vor allem aber sind sie jugendlich. Die Versuchung, einen Bogen zu den T e t r a r c h e n von S. M a r c o in Venedig zu schlagen, bringt keinen Erkenntnisgewinn. Die Similitudo der Heiligen Sergios und Bacchos ist zwar u n ü b e r s e h b a r , wirkt aber d e n n o c h repetitiv, denn im G r u n d e wird zweimal dasselbe über zwei verschiedene Personen ausgesagt: beide w e r b e n u m A n e r k e n n u n g u n d Auctontas mit ihrer Jugendlichkeit, die das Wissen eines Greisen miteinschließt. I m lateinischen Westen scheint sich das Ideal des Paidariogeron trotz der beträchtlichen V e r b r e i t u n g der Dialogi Gregors des G r o ß e n nicht durchgesetzt zu h a b e n . S p ü r b a r ist es aber in den ravennatisehen Mosaiken von San Apollinare N u o v o u n d San Vitale sowie in der erzbischöflichen Kapelle. Mit d e m kaiserzeitlichen Individualporträt wurden Physiognomie und C h a r a k t e r im R a h m e n des römischen Tugendsystems angeboten. Die Künstler u n d Biographen christlicher Heiliger w a r e n an der Individualität ihrer H e r o e n uninteressiert, weil sie den Akzent auf das übernatürliche legten. Dies beschreibt Sulpicius Severus, den wir als K o r r e s p o n d e n z f r e u n d des Paulinus von Nola kennen lernten, in der Vita des Hl. Martin von Tours: Nemo umquam illum vidit iratum, nemo commotum, nemo moerentem, nemo ndentem: unus idemque semper coelestem quodammodo laetitiam vultu praeferens, extra naturam hominis videbatur.

49 Weitzmann, 7tie Monastery of Saint Catherine at Mount Sinai: The Icons, vol. I., pp. 28 30 (B9), Tf. 12, 5253‫־‬.

Nicht N a t u r und Persönlichkeit w a r e n gefragt, sondern die Sichtb a r m a c h u n g der himmlischen Freude extra naturam.M Die attributive Verbildlichung physisch a n o n y m e r Personen ist nicht z u m ersten M a l im 6. J h . erfolgt. Das Problem hatte sich längst vorher in der nicht-christlichen Kunst gestellt. Dies zeigt uns in exemplarischer Weise die Hermengalerie von Welschbillig, die H . W r e d e "in die letzten J a h r e des T r i e r e r Aufenthalts Gratians vor 380" datierte. 51 In der Achse vor der Villa rustica von Welschbillig lag ein längsrechteckiger Weiher, den allseits H e r m e n s t ü t z e n in motivgleichen P a a r e n u m g a b e n . Das P r o g r a m m kann grosso modo als griechisch-römische "uomini famosi" umschrieben werden, denn es umfaßt Philosophen (Sokrates), F e l d h e r r e n (Philipp II. von Makedonien), Kaiser (Vespasian) (Abb. 24), G ö t t e r (Zeus) (Abb. 25), aber auch sog. Porträtfiktionen des 4. J h . n . C h r . und ethnische T y p e n (Barbarendarstellungen) (Abb. 26). W r e d e vermutete, daß der kaiserliche Auftraggeber—wir dürfen wohl sagen Gratian— die Hermenbalustrade von Welschbillig zu einem Abbild der antiken Welt und mit ihm zu einem Ausdruck seines kulturellen Erbes und Bildungsideal ausgestalten Hess, ein Memento, dieses Kulturgut mit seiner kaiserlichen Macht zu schützen und zu tradieren. 52 Diese D e u t u n g greift m. E. zu hoch. Die zahlreichen Porträtfiktionen des 4. J h . und die Barbarendarstellungen passen schwerlich zu einem Bildungsideal; dieses wird zwar d u r c h die b e r ü h m t e n Griechen und R ö m e r evoziert und zwar in eigentümlicher Weise: der Betrachter wird aufgefordert, die sattsam bekannten Figuren der griechisch-römischen Geschichte wieder zu erkennen. Dabei werden ihm in Welschbillig gelegentlich " u n g e n a u e Zitate" a n die H a n d gegeben, mit denen er sich selbst zurechtfinden m u ß t e . Es war nicht m e h r so klar, welche Persönlichkeiten n u n eigentlich die maßgeblichen Autoritäten der Philosophie, Dichtung, Rhetorik, der Politik u n d der Kriegsführung etc. waren. Z u m m e h r oder weniger intellektuellen Spiel des Wiedererkennens b e r ü h m t e r Griechen und R ö m e r gehörte es, d a ß der Entwerfer des P r o g r a m m s R ö m e r k ö p f e der caesarischen, spättrajanisehen u n d spätseverischen Zeit 3 3 einstreute. Die eigentliche C r u x w a r e n u n d sind noch heute die Porträtfiktionen des 4. J h . Sie f ü h r e n

50 51 52 53

Sulpicius Severus, Vila S. Martini, c. 27. H. Wrede, Die spätantike Hennengalerie von Welschbillig (Berlin, 1972). Wrede, Die spätantike Hennengalerie von Welschbillig, p. 89. Wrede, Die spätantike Hermengalerie von Welschbillig, pp. 55, 57, 60.

den Betrachter aufs Glatteis, denn er erkennt nur gerade den "Rekrutenhaarschnitt", rundliche K o p f f o r m und jugendliche Gesichtszüge.' 4 W r e d e hat sie mit Recht mit theodosianischen Köpfen (Abb. 27) verglichen, denn sie propagieren in der T a t das neue kaiserliche Ideal ewiger und unverbindlicher J u g e n d . W e n n dieses Ideal auf Anonymi übertragbar ist, d a n n entlarvt es sich als aufgesetztes Attribut, nicht als Inhalt. In diese Richtung weisen auch die Barbarendarstellungen. 5 5 W r e d e möchte sie als " B e w a h r e r des römischen Staates und des R ö m e r t u m s " verstehen. Diese D e u t u n g ist problematisch, weil es sich in Welschbillig um a n o n y m e G e r m a n e n handelt. Es empfiehlt sich daher, auf den optischen Befund zu rekurrieren, auf das, was m a n sieht: a n o n y m e Männerbildnisse mit verschiedenen g e r m a n i s c h e n H a a r - und Barttrachten. "Uomini famosi" sind sie gewiß nicht, sondern eben "Uomini anonimi". Das T h e m a der Darstellung b e r ü h m ter Griechen und R ö m e r ist ausgeweitet worden auf die Darstellung unbekannter Personen des 4. J h . Hier wird das Spiel des Wiedererkennens tricky. Die Anonymität ist zum salonfähigen T h e m a geworden, und sie wird mit Attributen, insbesondere der H a a r - und Barttracht evoziert. T h e m a der H e r m c n g a l e r i e von Welschbillig sind somit bekannte und unbekannte, a n o n y m e Griechen und R ö m e r , die d e m spielerischen Intellekt des spätantiken Betrachters zum Wiedererkennen überlassen werden. Diese Idee war neu und folgenreich auch für christliche Bischofs- und Heiligendarstellungen. Sie ist Ausdruck eines neuen Verständnisses des Porträts seit dem 4. J h . : das Porträt ist jetzt nicht m e h r n u r ein Abbild bekannter Persönlichkeiten, sondern auch Form und Erscheinung a n o n y m e r Personen, denen verschiedene Attribute, H a a r - und Barttrachtcn, aber auch Physiognomien appliziert werden konnten. Gemeinhin erkannte m a n im Individualporträt der römischen Zeit Inhalt, das Individuum per se war der Inhalt, und Individualität war Privatbesitz. Im 4. J h . betrachtete m a n das Individualporträt nicht m e h r als das einzig mögliche Ziel des Porträtisten. Neben das Indidividualporträt trat n u n m e h r das anonyme Bildnis, das mit Attributen, H a a r - und Barttrachten, Kleidungsstücken und sogar mit Physiognomien angereichert werden konnte. Ziel der Hermengalerie von Welschbillig war das Wiedererkennen des griechisch-römischen Bildungshorizonts, der durch die Einfügung von

54 55

Wrede, Die spätantike Hermengalerie von Welschbillig, pp. 61 .3‫־‬ Wrede, Die spälantike Hermengalme von Welschbillig, pp. 67 8.

Anonymi spielerisch verrätselt worden ist. Auch in der christlichen Kunst gehörten die Propheten, Apostel, Heiligen, Märtyrer und die Bischöfe z u m Bildungshorizont, aber ihr Nachleben oder, wie m a n heute sagen würde, ihre Memoria wurde nicht durch das Individualporträt gewährleistet, sondern durch das anonyme Bildnis mit Attributen und Inschrift. J e d e r Gebildete der Spätantike wusste, wie Plato und Sokrates aussahen. Die gebildeten Christen legten theoretisch keinen Wert auf das Aussehen ihrer Geistesgrössen, aber das " P h ä n o m e n Kunst bzw. Porträt" brachte sie in Zugzwang, und es m u ß t e n in diesem Bereich Entscheidungen getroffen werden. Die Erfindung des attributiven Porträts, in welchem nicht n u r Kleidung, H a a r - und Barttracht und Schmuck, sondern auch eine m e h r oder weniger individuelle Physiognomie attributiv eingesetzt werden konnte, war für die chrisdichen Entwerfer von Propheten-, Apostel- und Heiligenzyklen ein willkommenes M e d i u m , den christlichen Glaubenshelden Auctoritas und Dignitas zu verleihen.

H O W DID RISING CHRISTIANITY COPE W I T H PAGAN SCULPTURE? Niels H a n n e s t a d

Abstract The Second Commandment forbids the making of any likeness of a living being, but how literally was this prohibition taken by early Christianity? The pattern of iconography in the transitional phase between Antiquity and the Middle Ages is rather ambiguous. Acculturation rather than confrontation is the tendency. The Late Antique World, based as it was upon Hellenistic tradition, was a world of images: thus the likeness of Christ was created from pagan prototypes. Some Christians, though, were fundamentalists, and they have made a disproportionate impression on posterity. Most Christians, however—and this goes particularly for the upper class—seem to have had a rather relaxed attitude to the pagan past. Apparently many were even proud of their cultural heritage: they commissioned traditionally decorated silverware; their homes were embellished with mosaics with pagan motifs; and mythological sculptures were displayed alongside portraits of men of letters, philosophers and famous men of the past. In diis respect diey behaved no differendy than their pagan fellow citizens—or Jews, when these belonged to the urban Hellenized society. This paper will focus upon sculpture of the crucial fourth century. It can be established that the tradition of production (copying) of the Early and High Empire continued well into the fifth century. We see, however, a new selection of motifs. There is a tendency to avoid those that could be offensive to Christian belief, and many motifs are re-interpreted to fit into a new mental framework. Sculpture of the period mirrors very well the elements of tradition, as against innovation, in the ongoing process of formation. Though shalt not make to thyself a graven thing, nor the likeness of anything that is in the heaven above, or in the earth beneath, nor of those things that are in the waters under the earth. The Second Commandment

The image of Christ T h e risk of idolatrous worship is a matter of great concern in the Old T e s t a m e n t . Hence, the prohibition against making any likeness

of a living being has been one of the basic tenets of the three m o n o theistic w o r l d religions: J u d a i s m , C h r i s t i a n i t y a n d Islam. H o w ever, this ban has not always been taken too literally, especially in Christianity, which is so firmly rooted in the Hellenistic tradition. Christianity could even employ the three-dimensional form, thereby c o m i n g d a n g e r o u s l y close to i m i t a t i n g the s h a p i n g h a n d of the Almighty. U n d e r the High Empire the trend towards monotheism, which culminated in the victory of Christianity, b e c a m e increasingly strong. Being an offspring of J u d a i s m , Christianity h a d no iconography of its own, but to appeal to a Hellenized population it h a d to a d a p t to their culture, i.e. to establish an iconography, in the same way as its sacred literature b e c a m e formulated in Greek. N o other option existed than to d r a w on existing sources, in other words the stock of motifs of the G r a e c o - R o m a n world. Late Antiquity, in particular the third a n d fourth centuries, b e c a m e the formative period of Christian iconography. T h e m a i n figure, Christ himself, appears in the early phase in several forms, with two basic types: a y o u n g beardless m a n , mosdy with long curly hair, a n d an older m a t u r e one with long hair a n d a long or short beard. T h i s can hardly have been bewildering to the period since several gods, still favoured, could be rendered either as juniores or as seniores. T h e earliest known rendering of Christ in sculpture in the r o u n d m a y be the e n t h r o n e d small-scale statue in the T e r m e (figs. 1~2), dated to the mid-fourth century or a little later. 1 T h e identification, however disputed, is strongly supported by the a p p e a r a n c e of the exactly similar figure of Christ on early Christian sarcophagi, such as the one m a d e for the city prefect J u n i u s Bassus, w h o died in 359. 2 T h e technique, style, a n d size (0.72 m. high) of the e n t h r o n e d Christ

1

Museo Nazionale Romane. Le Sculture 1,1 (ed. A. Giuliano) (Rome, 1979) cat. 192 (not Christ): YV.F. Volbach, Early Christian Art (London, 1961) no. 36 (Christ): B. Brenk, "Zwei Reliefs des späten 4. Jahrhunderts", ActaAArtHist 4 (1969), pp. 547‫( ־‬not Christ): Age of Spirituality. Late Antique and Early Christian Art. Third to Seventh Century. The Metropolitan Museum of Art (ed. K. YVeitzmann) (New York, 1979) cat. 469 (Christ(?)): T.F. Mathews, The Clash of Gods. A Reinterpretation of Early Christian Art (Princeton, 1993), p. 128 (Christ). T h e fundamental survey is still F. Gerke, Christus in der Spätantiken Plastik, 2nd ed. (Berlin, 1941), on the T e r m e statuette, pp. 4 2 3 ‫( ־‬Christ!). 2 E. Strunther Malbom, The Iconography of the Sarcophagus ofJunius Bassus (Princeton 1990). T h e father was also a Christian, see The Prosopography of the Later Roman Empire I (eds. A.H.M. Jones et al.) (Cambridge 1971), BASSVS 15 (and 14).

in the T e r m e are characteristic of the period. T h e piece is unique and the context unknown, but its identification has to be accepted, owing to the parallel representation on the sarcophagi, and to the fact that as a type the figure is unparalleled in a pagan context. It would, if intended to represent somebody else, hardly have been comprehensible to any observer of the time. However dubious a source the Smptores Historiae Augustae m a y be, the statement that the E m p e r o r Severus Alexander in his laranum also had statues of holy men, a m o n g them one of Christ, must have been plausible to the society of the early fourth c e n t u r y . ' A b r a h a m , O r p h e u s a n d the philosopher Apollonios of T y a n a are also mentioned; the latter two pagan figures arc frequently represented in Late Antique sculpture, as well as in other art forms. T h e size of the T e r m e Christ would have been suitable for a private context. As a type, this piece is most illustrative of the formation process in progress. T h e rendering of the facial features constitutes no portrait, but a mask. In this way the rendering is in h a r m o n y with long established practice: gods or mythological figures were identified by their general a p p e a r a n c e or attributes, imperial "portraits" by inscription. This transformation from ruler portrait to ruler mask took place during the third ccntury, to be codified under the Tetrarchs, a n d finding its final form u n d e r Constantine the Great. 4 Moreover private portraits of the same period are also mask-like. T h e portrait of the young Christ is, therefore, no true portrait (and could hardly have been so) but an icon. Its various elements can be sorted out. T h e head is very m u c h related to that of the Sun: the softly modelled face and the long curly hair are to be found in a piece of similar size a n d date (fig. 3), a Sol from a fourth-century suburban villa of Constantinople. 5 T h e cult of the Sun was longlived as one of the strong monotheistic competitors to Christianity. Furthermore, it was closely related to the imperial cult. From literary sources we learn that pagan gods were also present in the new

3

Alexander Severus 29.2. O n this development, see N. Hannestad, Roman Art and Imperial Policy (Aarhus 1986), chaps. VI VII. ‫ '־‬N. de Chaisemartin and E. Orge n, Les documents sculptés de Silahtaraga (Paris 1984), cat. 1 ("Apollo"); B. Kiilerich and H. Torp, "Mythological sculpture in the fourth century A.D.: the Esquiline group and the Silahtaraga statues", Istanbuler Mitteilungen 44 (1994), pp. 314 16; Ν. Hannestad, Tradition in Late Antique Sculpture. Conservation—Modernization !*reduction (Aarhus 1994), p. 122. 4

capital, Constantinople, a n d that Sol was a m o n g the favourites. 6 Sol is represented in all art media until the very end of Antiquity, 7 even in large-scale sculpture. In a group from the Esquiline Hill in R o m e , originally consisting of perhaps ten pieces, of which five are now exhibited in the Ny Carlsberg Glyptotek in C o p e n h a g e n , Sol is present a m o n g other gods a n d mythological figures. By its inscriptions the ensemble is dated to the late Constantinian period. 8 A head (0.30 m. high), made for insertion, from the M i t h r e u m of the Animals in Ostia might be late, a n d another also from Ostia, from the Mithr e u m of the Imperial Palace, definitely is.9 This head of the so-called Helios/Alexander type measures 0.40 m. including neck (chin to top of head 25.5 cm.). It is m a d e of a very irregular piece of marble, revealing the absence of good quality stone in the late period. T h e neck is only roughly finished in the lower part, which is presumably intended for inserting into a statue. T h e back is flat and uncut, a n d the sloppy modelling and lack of sculptural depth combined with the rough edge makes it a p p e a r as if it were broken off from a relief. T h e eyes have no plastic rendering of the iris, but irregularly drilled holes for the pupils. T h e lips of the badly defined m o u t h are opened by a row of uneven holes. Finally the skin of the face has been polished. All these traits together point to the end of the fourth or the early fifth century. This makes it perhaps the latest representative of the type on this scale. In the iconography of Sol the likeness of Alexander the G r e a t is often influential, sometimes to a degree that only the presence of rays or information of provenance makes it possible to decide which one is represented. This is just another example of fluctuation between

6

G. Dagron, Naissance d'une capitale. Constantinople et ses institutions de 330 à 451 (Paris 1974), pp. 368-9 with references. O n pagan sculpture in the new capital, see C. Mango, "Antique Statuary and the Byzantine Beholder'', Dumbarton Oaks Papers (henceforth DOP) 17 (1963), pp. 55-75. Most relevant to the questions raised in this paper is H. Saradi-Mendelovici, "Christian Attitudes toward Pagan Monuments in Late Antiquity and their Legacy in Later Byzantine Centuries", DOP 44 (1990), pp. 47-61. 7 LIMC (— Lexicon iconographicum mythologiae classicae I ff., (Zürich 1981 seq.)), Helios/Sol IV, 1 cols. 592-625 and V,1 cols. 1005-047. 8 M. Moltesen, "The Aphrodisian Sculpture in the Ny Carlsberg Glyptotek", in Aphrodisias Papers (London 1987) (= Journal of Roman Archaeology Suppl. 1) (Ann Arbor, 1990), pp. 133-46; Hannestad, Tradition in iMte Antique Sculpture, pp. 110-13; Kiilerich and Torp, "Mythological sculpture in the fourth century A.D.", pp. 307-16. 9 Scavi di Ostia Ii 1 mitrei, ed. G. Becatti (Rome, 1954), p. 56, pl. X X X I I , 1 2; LIMC IV no. 12.

types of gods; a p h e n o m e n o n which is so characterisdc of the Ancient World and still d o m i n a n t in the transitional phase to the Middle Ages. A very curious example of re-cycling is a head found in the M i t h r e u m beneath San Clemente in Rome. 1 0 It is—presumably—a second-ccntury portrait of Alexander later modified to become a representation of Sol. T h e close relationship between these two figures m a d e such a transformation no big deal, and more examples of such modifications are k n o w n . " T h e motif of the e n t h r o n e d Christ is taken over from imperial iconography. T h e T e r m e Christ sits with a scroll in his left h a n d , as could the e m p e r o r when acting as a magistrate. T h e scroll Christ holds contains the law, which he is about to h a n d over to Peter. In an exactly similar form, he is rendered in other art media of the period. O n sarcophagi Christ can be represented e n t h r o n e d above the veil of heaven, carried either by a young Caelus or the more traditional old one, the latter version being the supporter chosen for the sarcophagus of J u n i u s Bassus. Placed together as a pair, the tradidonal Caelus with a young counterpart are carriers of the all-victorious emperors, the twin augusti in the frieze on the Arch of Galerius in Salonica. 12 Christian emblems gradually became distinct within the traditional pagan framework, in the same way as Jewish symbols did. This tendency is apparent, especially in sepulchral art, and the result is often a strange mixture: all m e m b e r s of this social class bought their sarcophagi from the same range of suppliers. In a Jewish sarcophagus of a r o u n d 300, now in the T e r m e , a tondo with a representation of the seven-branched M e n o r a h is held by a pair of Victories above a grape-trading scene. T h e Victories are flanked by the four seasons. 15

10

O n this head, see Hannestad, Tradition in Late Antique Sculpture, pp. 1 0 0 1 , with references. O n representations of Alexander and the influence of his portrait, see N. Hannestad, "Immitatio Alexandri in Roman Art" in Alexander the Great. Myth and Reality (= Analecta Romanica, Suppl. XXI) (Rome, 1993). Further on the importance of Alexander in late Antiquity, see below. 11 I.e. LI MC V,1 (Helios pp. 1005-047), nr. 176: Bologna; c.f. also a head formerly in the Collection Schott in Jena, E/l 1465 6. 12 M.S. Pond Rothmann, " T h e Panel of the Emperors Enthroned on the Arch of Galerius", Byzantine Studies 2 (1975), pp. 1 9 4 0 ‫ ־‬: H.P. Laubscher, Der Reliefschmuck des Galenusbogens in Thessaloniki (Berlin, 1975), pp. 69 78, and for the interpretation of the twin Caeli, see Hannestad, Roman Art, pp. 3 1 6 . 1 7 ‫־‬ 13 Only the central part, together with the right part of the front, has been preserved. T h e exact location of the find is unkown, Museo Nazionale Romano. Le seuliure 1,8 I, ed. A. Giuliano (Rome, 1985), cat. 111,10; A. Konikoff, Sarcophagi from the

In a sarcophagus purchased in R o m e , now in the Ny Carlsberg Glyptotek, a portrait tondo is carried by a pair of putti. 14 T h e tondo depicts a bust of the y o u n g Christ, his head encircled by a nimbus. Most unusually, Christ is handling a metal vessel. Beneath, nearly as a support of the tondo, is the imperial mark: the eagle of J u p i t e r in the imperial crown (later this eagle was replaced by the Cross). Further, the scene is mixed with a blend of pagan and Christian motifs: two cocks flank the representation and below are the symbols of water a n d earth, to the left a dolphin and a ketos (like the one swallowing J o n a h ) a n d to the right sheep. It makes a lot of sense: to establish the iconography of the new ruler of the world, you borrow iconographie elements from R o m a n ruler representation a n d the Sun, mixed with a little of Alexander, the youthful conqueror. T o stress the saviour element, some traits from Asclepios, m u c h favoured in the period, are also added. This last element emphasizes the growing importance of the sage or holy m a n : Christ was such a person. 1 3 (Cf. fig. 8). In particular, the figure of the m a t u r e Christ is heavily dependent on the iconography of Asclepios. In an early representation of the Sermon on the M o u n t , Christ is dressed in the Greek himation and is very similar to this god in appearance.' 6 A full-scale head representing this same type of Asclepios has been found in an early Christian

Jewish Catacombs of Ancient Rome (Stuttgart, 1986), cat. 14; Gardens and Ghettos. The Art of Jewish Life in Italy, ed. V.B. Mann (Berkeley, 1989), cat. 16. A very important contribution to the problems involved is L.V. Rutgers, "Archaeological Evidence for the Interaction of Jews and Non-Jews in Late Antiquity", American Journal of Archaeology 96 (1992), pp. 101-18. It has recently been suggested that the upper part of the front relief has been recarved. I owe this information to J . Stubbe Ostergaard. 14 I.N. 1552. F. Poulsen, Catalogue of Ancient Sculpture in the Ny Carlsberg Glyptotek (Copenhagen, 1951), cat. 830. T h e sarcophagus is discussed in the forthcoming catalogue Imperial Rome (Romersk kejsertid) I (Copenhagen, 1995) by J.S. 0 s t e r g a a r d (et al.), under the entry cat. 54. It has recently been suggested that the upper part of the front relief has been recarved. I owe this information to J . Stubbe 0stergaard. 15 In his recent book, The Clash of Gods, T.F. Mathews stresses the association of Christ with the holy man. No doubt Christ was a holy man. As pointed out, not least by Peter Brown in several works, the times favoured the holy man. T o subdue the connection between the imperial iconography and the victorious Christ, Mathews has to twist the material excessively. O n the problem, see further below and note 72. 16 Museo Nazionale Romano, inv. 67607. Gerke, Christus in der Spätantiken Plastik, pp. 11 14: Age of Spirituality, cat. 373; J . Engemann, Untersuchungen zur Sepulkralsymbolik der späteren römischen Kaiserzeit (Münster, 1973): Mathews, The Clash of Gods, pp. 68-72. On small-scale statues of Asclepios in late Antique villas, see below n. 75.

Basilica in Gcrasa in Jordan. 1 7 T h e head is apparently rather late. It may represent Asclepios, hut an early representation of Christ may also be suggested. If this is to be accepted, it could be explained as an Asclepios transformed into a Christ in the same way as Alexander could be turned into a Sol. 111 Constantinople there was a tradition that Constantine the Great erected a statue of Christ in his new capital. 1 " Considering the spirit of the time, the story is not at all unlikely. Even G o d the Father could be depicted creating the world or A d a m and Eve (sometimes attended by an improving Christ). T h e latter scene, when appearing on early Christian sarcophagi, can in some versions be traced back to that of Prometheus creating man. 1 9 T h e Sun and Christ are often fused in the early phase of Christian iconography. As mentioned, iconographie driftings between various deities were a familiar p h e n o m e n o n , and the difference between these particular divinities may not have been too distinct in this early period. A well-known piece of mixed iconography is represented in a vault mosaic from a mausoleum beneath St. Peter's in the Vatican, rendering Sol and Christ in a single figure. T h e god is standing in his chariot, and around his head shines a radiant nimbus, from which some of the rays form a crux gemina.2° Representations of the Sun were all-embracing, even forming a part of mosaic pavements in late Antique synagogues in Palestine— often combined with the most revered Jewish symbol, the Menorah. 2 ' 17 K. Stemmer, "Ein Asklepios-Kopf in Amman" Annual of the Department of Antiquities, Amman 21 (1976), pp. 33-9 (pis. 1 VI pp. 155-60). Stemmeris main concern is the question of the (Greek) original, but he does not reject the idea that the head and its statue could have been used in a Christian context. Stemmer compares the Amman head with another Asclepios head, now in Foligno, Museo Civico (= LIMC II no. 352). This head appears to belong to the high Empire, but it has been reworked in Late Antiquity—to represent a Christ? 18 Dagron, Naissance d'une capitale, p. 390. 19 See Spätantike und frühes Christentum. Ausstellung im Liebighaus Museum alter Plastik hianlfurt am Main (Frankfurt, 1983), pp. 3 2 0 - 4 . O n the Prometheus motif, see G. Koch and H. Sichtermann, Römische Sarkophage (München, 1982), pp. 183-4. 20 Also termed the Patriarchal Cross or Cross of Lorraine. Age of Spirituality, cat. 467: c.f. also Hannestad, Roman Art, p. 301 with references, and G.B. Ladner, Handbuch der frühchristlichen Symbolik (Stuttgart, 1992), pp. 36 8. 21 T h e fourth-century synagogue in Hammath-Tiberias displays Sol in his chariot encircled by a Zodiac combined with a pair of Menorahs: see M. Dothan in Ancient Synagogues Revealed, ed. L.I. Levine (Jerusalem, 1981), pp. 63 69. Better llustrations are found in M. Ben Dov and Y. Rappel, Mosaics of the Holy IAnd (New York, 1987), pp. 56 60‫־‬. Very abstractly this same motif is rendered in a synagogue in Beth Alpha dated by inscription to the period of Justin I (518 527): see H.-P. Stähli, Antike Synagogenkunst (Stuttgart, 1988), pp. 55 68. Another scene of the door mosaic

Apparently such representations were considered as purely decorative and therefore not offensive to the community. T h e Late Antique World was so deeply rooted in Hellenism that picture-making continued to be an integrated part of everyday life, and the multi-ethnic flavour of society is mirrored in the fact that inscriptions in Aramaic, H e b r e w a n d Greek can a p p e a r in one a n d the same synagogue. It is a puzzling fact that in a period when the world was tending to b e c o m e illiterate, writing in art, especially mosaics (and painting) b e c a m e widespread in all contexts: explanatory inscriptions for individual figures a n d long dedicatory inscriptions, often bilingual, at the entrance of a room. T h e m a n y well-preserved mosaic floors of churches (and synagogues), particularly in Palestine and Syria, testify to a basically pagan tradition unbroken to the A r a b conquest. Against the abstract and floral design of H e r o d i a n and early Imperial Jewish mosaics, late antique synagogues can be amazingly lavishly decorated (some wall paintings are also preserved) with pictures, some of which relate to the old T e s t a m e n t , while others, as mentioned, are purely pagan. As expounded in the treatise Abodah %arah ("idolatry") of the Mishnah, m u c h could be accepted in times of peace! Even sculpture could in certain circumstances be owned by Jews, but of course never adored. 2 2 T h e strict rules that were, after all, not too strict b e c a m e further softened in the Christian realm. Mentally, iconoclasm was a long way in the future.

The setting As demonstrated, the image of Christ, y o u n g or mature, is established on pagan prototypes. Building on this recognition, I would like to focus on some of the iconographie models that continued far

of this synagogue depicts the sacrifice of Isaac, including the hand of God coming down in a cloud. In general on the subject, see B. Narkiss, "Pagan, Christian, and Jewish Elements in the Art of Ancient Synagogues 1 ', The Synagogue in Late Antiquity, ed. L.I. Levine (Philadelphia, 1987), chap. 13. 22 Translated by H. Danby, The Mishna, corrected ed. (Oxford, 1938) with several reprints. O n picture-making in synagogues, see also Stähli, Antike Synagogenkunst, pp. 917‫ ־‬and on sepulchral art, Rutgers, "Archaeological Evidence for the Interaction of Jews and Non-Jews in Late Antiquity". In general on the situation in Rome, see L.V. Rutgers, The Jews in I Ale Ancient Rome. Evidence of Cultural Interaction in the Roman Diaspora (Leiden, 1995).

into the Christian period, to enter the melting-pot of the Late Antique/ Early Medieval World. I will illustrate my point mainly by sculpture. In particular, the fourth century is crucial as the formative period of Christian iconography. T h e pattern of tradition contra innovation is indeed very complex. 2 3 Mentally, the Late Antique World is a world of reflection. T h e m e s in all media were taken u p again. Literary pastiches were c o m m o n , with works like the vast and learned p o e m Dionysiaca from the fifth century by the author Nonnos, who also wrote a paraphrase of St. J o h n ' s Gospel. 24 In the same way, sculptural pastiches (copies, adaptations etc.) in the Graeco-Roman tradition were continuously produced. Even 'portraits' were often sculptural citations of previous periods. T h e last pagan emperor, Julian the Apostate, had himself rendered in the T r a j a n i c - H a d r i a n i c tradition. 2 3 In R o m e , old m o n u m e n t s , in particular those of the early Empire, were conserved with great care, a n d the forest of statues in the City was looked after by the curator statuarum.26 Iconographie themes in the arts were revived; in some art forms, for instance ivory carving, one could even argue that Antiquity never ended. As regards sculpture (and this goes to some degree for other art forms too) there is in scholarship a tradition that a dividing line exists between pagan art and Christian art. I feel that this is at best an oversimplification, if not totally wrong. T h e classical tradition in

23

For some recent attempts to focus on the aspect of acculturation rather than confrontation, see G.YV. Bowersock, Hellenism in Laie Antiquity (Cambridge, 1990): Saradi-Mendelovici, "Christian Attitudes toward Pagan Monuments in I •ate Antiquity": and Hannestad, Tradition in Late Antique Sculpture·, c.f. also Rutgers, The Jews in Late Ancient Rome. 24 T h e poem, as long as the Iliad and the Odyssey combined, describes the god's triumphal progress to India, presenting him as a predecessor of Alexander the Great. A fine survey of the problems involved is given by G.VV. Bowersock in a review of recent works by Pierre Chuvin, "Nonnos Rising", Topoi 4 (1994), pp. 385-99. 25 H.P. L'Orange and M. VVegner, Das spätantike Henscherbild von Diokletian bis zu den Konstantin-Söhnen (= Das röm. Herrscherbild III.4) (Berlin, 1984), pp. 1623‫־‬, pis. 76c-d, 77. It is worth noticing that on a pair of ivory plaques with scenes of the infancy and miracles of Christ, the evil ruler, Herod the Great, who is supervising the soldiers butchering the new-born of Bethlehem, is rendered exactly as Julian, Age of Spirituality, cat. 406. 26 O n conservation of old monuments, see Hannestad, Tradition in iMte Antique Sculpture, part I. Along the same line is A. Giuliano, "Un palinsesto su marmo", Bolletino dei Musei Comunali di Roma 7 (1993), pp. 5 1 4 ‫־‬ and "Augustus-Constantinus", Bollettino d'Arte 68- 69 (1991), pp. 3 1 0 . A curator statuarum named Fl. Magnus Ianuarius appears in an inscription dated c. 337 (CIL VI 31906).

sculpture—often in correlation with new trends—continued far beyond the limits conventionally set by the handbooks. Let us take a closer look at the problem: how did the Christians actually respond to sculpture—this long-established N e a r Eastern and then G r a e c o - R o m a n art medium? In the mid-fifth century, the Syrian bishop T h e o d o r e t complains about pagan sculpture still being produced and displayed in public. 27 H e indignantly enumerates what shameless types were on show: "Dionysos that limb-loosener a n d effeminate creature" being his main dislike (cf. fig. 4). T w o generations earlier the Christian poet Prudentius praised the sculptural glory of Rome. 2 8 T h e new capital of Constantinople was adorned with traditional sculpture, a n d collections were established there. 2 9 General opinion must lie in between the two extreme views expressed by the bishop and the poet. T o some it was a matter of concern whether the sacred house should be decorated with pictures of living beings. A contemporary of Prudentius, the wealthy Paulinus of Nola, a former imperial official, thus raises the issue when commissioning churches. 3 0 In the case of each individual one can assume that attitudes were closely related to social standing; the u p p e r class in general being m u c h concerned about its cultural heritage. T h e lower classes, the humiliores, hardly cared! Some Christians were indeed fundamentalist. T h e y might smash sculpture. T h a t is what has h a p p e n e d , for example, to the Odysseusgroups in the Tiberian sculptural embellishment of the Sperlonga grotto, which all ended u p in the circular pool. In this case the excited believers m a d e a p r o p e r j o b of it—some sculptures are now a jigsaw puzzle of thousands of pieces. T h e central group of the mighty Scylla attacking the ship of Odysseus may also have been regarded as exceptionally threatening: representing a m a n - e a t i n g 27 C.f. E.K. Gazda, "A Marble Group of Ganymede and the Eagle from the Age of Augustine 5 ', in Excavations in Carthage 1977, vol. 6 (ed. J . Humphrey) (Ann Arbor, 1981), pp. 125-78, esp. 167, with a translation of the relevant part of the Hellénikon 'Therapeutiké Pathématon 3.79 84. 28 M.A. Malamud, Poetics of Transformation. Prudentius and Classical Mythology (Ithaca and London, 1989), esp. the verse quoted on pp. 22—3. 29 Dagron, Naissance d'une capitale, index: "sculpture". See further Mango, "Antique Statuary and the Byzantine Beholder": Saradi-Mendelovi, "Christian Attitudes toward Pagan Monuments in Late Antiquity": and B. Kiilerich, "Sculpture in the Round in the Early Byzantine Period: Constantinople and the East", Aspects of Late Antiquity and and Early Byzantium, Swedish Research Institute in Istanbul, Transactions 4, eds. L. Rydén and J . O . Rosenqvist (Stockholm, 1993) pp. 9 9 . 1 1 2 ‫־‬ 3 " C. Davis-Weyer, Early Medieval Art 300-1150 (Toronto, 1986), pp. 17-23.

dragon. T h e cave as such may have been regarded as an evil place of worship. 31 At some point in Late Antiquity the sculptural and architectural setting in the C a n o p u s area of the Villa Adriana was demolished, when a small village grew up in the area. 32 All the sculpture, including two centrally placed Scylla groups, was pushed into the basin, while most of the columns with capitals were brought to neighbouring Tivoli for church-building. In this late period, when temples and other public buildings went out of use and the open space of the forum or agora b e c a m e clustered with churches, m u c h sculpture became transferred for re-use, mainly in the baths. 3 ' But through aversion to the naked body such public sculpture could become sexually mutilated. T h e males were castrated (this was before the invention of the fig leaf). Such a violation could even be inflicted on the three Graces (fig. 5).34 Another way of rendering a spirit, was to mark it with a mainly in portraiture. In most scratched in, as seen in a pair of

31

figure harmless, to get rid of the evil cross on the forehead. This is met cases it is a simple cross carved or enthroned colossal statues of Augustus

One may wonder if there could be some influence from Jewish belief. According to the treatise Abodah Zahrah of the Mishna (3.3 c.f. above n. 22), sculpture representing dragons should be particularly avoided. By this prescription, such sculpture should be smashed and thrown into the sea—preferably the Dead Sea. For much help and valuable discussions concerning this matter, I wish to thank my collegue Per Bilde. 32 O n the C a n o p u s valley, see S. A u r i g e m m a , Villa Adriana ( R o m e 1961), p. 100-33; N. Hannestad, "Über das Grabmal des Antinoos. Topographische und thematische Studien im Canopus-Gebiet der Villa Adriana", Analecta Romanica 11 (1983), pp. 69-108: and M. de Franceschini, Villa Adriana. Mosaici- Pavimenti—Edifici (Rome, 1991), pp. 294 315 and pp. 563-76. 33 For this development, see B. Ward-Perkins, From Classical Antiquity to the Middle Ages. Urban Public Building in Northern and Central Italy A.D. 300-850 (Oxford, 1984), passim (index: "statues"). In Caesarea (Cherchel) the civic authorities removed statues from the temples and placed them on new bases in the bath of the town, c.f. CIL VIII, 20963 and 20965: see also C. Lepelley, "The survival and fall of the Classical city in late Roman Africa", The City in Iuite Antiquity, ed. J . Rich (London, 1992), esp. p. 59—I owe this information to Peter Pentz, working on a dissertation entitled From Proconsulaiis to Ifraqiyah. 34 This goes for a group in natural size found in a bath in Perge, see M.E. Özgür, Skulpturen des Museums von Antalya 1 (Istanbul, 1987), cat. 30—a better illustration is to be found in Antalya Museum, eds. E. Ozgen and I. Ozgen (Ankara, 1988), cat. 101. An over life-size Apollo from a similar context in Salamis, Cyprus, has been castrated, but his nose has also been chiselled away, V. Karageorghis, Sculptures from Salamis (Nicosia, 1964), no. 3 (c.f. also no. 8—a late Antique headless statuette of Meleager). Sculpture in a private context was, of course, never mudlated in this way.

a n d Livia found in Ephesos. 3 3 T h e founder couple of the imperial system, hardly divine any longer, were still venerated in the late period. In other cases the cross could take the shape of the letter x, reminiscent of the salvation mark in oil of Jewish tradition. 3 6 Most sculpture, of course, simply disappeared for obvious economic reasons: bronze was costly and so melted down, and marble sculpture was very h a n d y for the lime kilns. O u r main source of knowledge of m a n u f a c t u r e of sculpture in Late Antiquity is provided by the current excavations in Aphrodisias. This city was one of the Empire's important sculptural productioncentres: busy in creating sculpture in any style, subject, and size until the very end of Antiquity. In the late period, m u c h sculpture was m a d e for long distance transport, but still the city was the h o m e base of travelling sculptors in the old tradition. 37 Aphrodisias seems to have been the m a j o r supplier of sculptural settings of the late period as regards the embellishment of upper-class dwellings. From Britain to Syria, a series of houses and villas from this period is known. In several cases parts of the sculptural content from these structures have been preserved; normally only a few pieces have been excavated, but occasionally an a b u n d a n c e . As far away as Britain, traditional sculptural settings of the late period can be found, as at a villa in Woodchester in Gloucestershire; in this fourth-century villa, famous for its O r p h e u s mosaic, battered pieces of small-size marble 35

J . Inan and E. Alföldi-Rosenbaum, Römische und frühbyzantinische Porträtplastik aus der Türkei (Mainz, 1979), nos. 3 and 5. Found together with this couple is a more than life-size head of Augustus, also marked with a cross on the forehead (no. 2). For the meaning of this mark, see 0 . Hjort, "Augustus Christianus—Livia Christiana: Sphragis and Roman Portrait Sculpture", Aspects of Late Antiquity and and Early Byzantium, pp. 99-112. 36 This applies to the general on the Ludovisi sarcophagus in the Terme (Museo Nazionale Romano. Le sculture 1,5, ed. A. Giuliano (Rome, 1983) cat. 25); Hannestad, Roman Art; by some identified as Decius' son Hostilian who died in Rome in 251. This type of mark is sometimes considered to be the sphragis, the mark of Mithras, see also above, Hjort, "Augustus Christianus—Livia Christiana". 37 T h e latest excavation programme was inaugurated in 1961, and, as a project, mainly fulfilled through the persistent energy of the late Kenan T. Erim. Despite its rather bizarre character, his book Aphrodisias. City of Venus Aphrodite (London, 1986) gives a glimpse of the wealth of material about to be published, especially in the series of papers of the Aphrodisias seminars. Concerning problems of production in the late period, see in particular P. Rockwell, "Unfinished statuary associated with a sculptor's studio", in Aphrodisias Papers 2 (= J R A Suppl. 2), eds. R.R.R. Smith and K T . Erim (Ann Arbor, 1991), pp. 127 43—the studio was destroyed at some time in the fourth century by an earthquake, and the material is highly informative about sculptural technique.

sculpture have been excavated, undoubtedly belonging to the same period as the mosaic. 3 8 T h e late antique house of the u p p e r class—what could be termed a stately home—flourished d u r i n g the fourth a n d earlier part of the fifth century, to decline some time after 500 A.D. In the West, except for R o m e , sites are mainly rural, while in the East we are faced mostly with s u b u r b a n villas. T h e r e m a y have been a difference d e p e n d i n g u p o n economic structures, degree of urbanization, etc., but in some areas archaeological tradition combined with the destruction of the landscape may have distorted the picture. T o my knowledge, only two such wealthy rural villas are known in Asia M i n o r , both located on the shore of the west coast a n d only one published. 3 9 M a n y m o r e m a y have existed, as we would expect f r o m literary sources. 40 However, T u r k e y has no tradition for allowing aerial photography for research, and not very m a n y surveys have been done in this area. For the R o m a n period, the focus has been on the splendid marble cities of the High Empire. Further, the destruction of the landscape could be somewhat drastic. Apparently this process gathered speed in Late Antiquity when the large harbours along the west coast were silted up with debris f r o m the deforested mountains. T h e imperial shrine in Pisidian Bubon, p l u n d e r e d by clandestine excavation in spring 1967, was thus covered with about 10 m. of soil.41

38

See G. Clarke, "The Roman villa at Woodchester", Britannia 13 (1982), pp. 197-228, esp. pp. 207 ff. and pis. X V f. A delicately carved female statuette without head (h.: 0.44m.) of Diana Luna is the only well-preserved piece. 39 A badly preserved villa presumably from the second century is located on a small island close to Elaia, the harbour of Pergamon: see A. Hoffmann, "Ein römisehe Meervilla an der kleinasiatischen Westküste", Istanbuler Mitteilungen 43 (1993), pp. 437-43. Unpublished is a major group of substantial buildings east of the small harbour of Torba, on the north side of the Halicarnassos peninsula, which I noticed in 1970, when working in the Danish Halikarnassos project. In December 1993 the complex was severely threatened by touristic urbanization. Prominent among the various structures is a barrel-vaulted building with an apse, by the locals termed kilisi (church). T h e front, excecuted in marble, faces the seashore. T h e rest of the masonry consists of irregular stones laid in mortar. T h e floor, only the first part preserved and visible, is black and white mosaic in abstract patterns (garland and meander). Based on the style of masonry, marble façade and the mosaic floor, a date of the third or fourth century seems likely. T h e design of this building is rather close to the so-called IVasserkultbau (mid-fourth century), a dominant part of the late antique villa at Milreu in southern Portugal: see Hispania Antiqua. Denkmäler der Römerzeit, ed. Α. Nünnerich-Asmus (Mainz, 1993), pp. 2 3 3 - 4 with references. 40 c.f. J.J. Rossiter, " R o m a n villas of the Greek east and the villa in Gregory of Nyssa £/>.20" Journal of Roman Archaeology 2 (1989), pp. 101-10. 41 A rescue excavation was carried out in J u n e 1967: on the structure, see J . Inan,

T h e favourite position of the R o m a n m a n o r house will be on a slope, raised above the plain, with a fresh breeze a n d an attractive vista, a very vulnerable position w h e n the entire landscape is on the move. T h e Aphrodisias material is not very well published, but this m a y come. This, however, also applies to m a n y villas. Further, several of the late villas containing m u c h sculpture were either excavated very early or by clandestine digging in recent times. 42 O f t e n the sculpture was already removed f r o m its original place in Antiquity, before the e n d came: m u c h has been f o u n d in d u m p s . It is very often stated of villas of the fourth a n d fifth centuries— w h e t h e r they were only established in the late period or h a d already been occupied for a long period by that t i m e — t h a t the sculptural content in general should be dated to the High Empire. This is not correct! It is beyond the limits of this p a p e r to give a full account of the sculptural p r o g r a m m e of the late villa. As a t h e m e the concept of the late villa calls for f u r t h e r investigation: studies about architectural a n d decorative schemes will extend o u r knowledge about the role of the late villa in general, its cultural heritage a n d its part in the process of transformation. In remote areas, the villa could most persistently reflect traditional R o m a n culture in the same way as structures in f o r m e r colonies can maintain the past of a colonial power. This could raise the issue: did the villa in remote areas function as a part of the country or was it isolated f r o m it?43

Some collections of sculpture Despite the fact that only a r a n d o m choice of sculpture has survived a n d then only in marble (or other types of stone), the finds allow us to get an idea of the pattern of types, subject preferences, etc.,

"Der Bronzetorso im Burdur-Museum aus Bubon", Istanbuler Mitteilungen 2 7 / 2 8 (1977/78), pp. 267-96. For a survey of the potentialities of landscape archaeology in this respect, see J . Percival, " T h e fifth-century villa: new life or death postponed?", Fifth-Century Gaul: A Crisis of Identity?, eds. J . Drinkwater and H. Elton (Cambridge, 1992), chap. 14. 42 This goes, for example, for a suburban villa in Pisidian Cremna, looted in the late 1960s: see Hannestad, Tradition in Late Antique Sculpture, pp. 122 f. 43 Some relevant questions concerning Gaul are raised by J . Percival, " T h e fifthcentury villa: new life or death postponed?".

just as they also provide us with important information on conditions of production and distribution. By far the most exuberant sculptural setting in a private context is the embellishment of a villa in Chiragan (Martres Tolosanes) in southern France. 44 T h e dating of this vast and luxurious villa complex is highly debated. Pieces of sculpture c a m e to light in the seventeenth century, a n d organized diggings were carried out before the mid-nineteenth century, finishing with two excavation campaigns just before the turn of the century, when architecture and finds were finally published in a rather s u m m a r y form. Further investigation of the sculptural finds, with few exceptions all in the museum in nearby Toulouse, are now in progress, but until this project is finished, we must basically rely on the corpus of Espérandieu. 4 5 H e enumerates 115 entries, some consisting of several pieces. A simple count gives 144 pieces of sculpture. M o r e sculpture is on display in the m u s e u m as well as in store. Approximately 150 pieces can be related with certainty to this villa (but some give m u c h larger figures). 46 Beyond dispute, the material is enormous, competing in n u m b e r with that of imperial estates. T h e dating of the complex mainly depends on the sculptural content: the villa is argued to have been established in early Imperial

44 T h e basic publication is L. Joulin, Les établissements gallo-romains de plaine de Martres-Tolosanes (Paris, 1901). T h e sculpture is collected in É. Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine (Paris, 1908), pp. 2 9 - 9 3 (nos. 891 1006). In general, see also Princeton Encyclopedia of Classical Sites (Princeton, 1976), p. 554 (Martres-Tolosane M. Labrousse) with bibliography. On the sculpture most recently, Hannestad, Tradition in Late Antique Sculpture, pp. 127 41; L.M. Stirling, Mythological Statuary in Late Antiquity: A Case Study of Villa Decoration in Southwest Gaul, Diss. University of Michigan (Ann Arbor, 1994), esp. pp. 3 4 8 ‫ ־‬and pp. 24252‫ ;־‬and Le regard de Rome. Portraits romains des musées de Mérida, Toulouse et Tarragona (Barcelona, 1995) passim. O n the marble and the relation to Aphrodisias, see M. Bergmann, "Un ensemble de sculptures de la villa romaine de Chiragan, œuvre de sculpteurs d'Asie Mineure, en marbre de Saint-Beat", Les marbres blancs des Pyrénées. Entretiens d'archéologie et d'histoire (1995), pp. 197 205. 43 Supra n. 44. T h e material is now being organized for a coming publication. Additional portraits are listed 111 Le regard de Rome, cats. 8 (head of Trajan); 17 (fragment of Antonine or Severan male bust); 152 (inscribed base of bust of a certain C. Aconius Taurus); 172 (head of a boy, Tetrachic or later). A fragmented head of Demosthenes is treated in Joulin, Les établissements gallo-romains de plaine de MartresTolosanes, p. 321, pl. XIV 197E. 46 See the discussion in Stirling, Mythological Statuary in Late Antiquity, n. 63. When J . Boube, "La nécropole palôchretienne de Martres-Tolosane", Pallas 3 (1955), pp. 80-115, gives the figures 272 pieces, including 73 imperial busts, he furnishes no arguments.

times, to have been greatly expanded in the second century, and still a place of great importance in its final period in the fourth century. T h e villa c a m e to a sudden end in the fifth century when it was b u r n e d down. However, most of the sculpture had at that time been d u m p e d in various pits. Because of a wide selection of imperial busts, a connection with the imperial house has been suggested. All the portraits found seem to have been busts: there are 61, of which 25 are imperial, ranging from Augustus to the early Severans. Only one is female, namely Sabina. T h e Julio-Claudians are scarce, the Flavians absent, but T r a j a n is e m b o d i e d in five copies a n d Septimius Severus and his son Caracalla in four each. Apparently the owner intended to a acquire a representative series of emperors. A few significant emperors were lacking, so they had to be specially m a d e for the collection. This applies to the likenesses of Hadrian, 4 7 Marcus Aurelius, 48 and the adult Caracalla. 4 9 T h e y were produced in a marble-like local limestone by the same workshop; the characteristic eye-rendering of the pupil as an incised circle surrounded by an iris as a semicircle indicates a dating to the earlier part of the fourth century. T h e imperial busts have all been reworked, from a slight retouching to heavy tooling a n d joining of pieces not originally belonging together. Some have been cut down from statues. M u c h of the work was obviously done by the same restorer. 30 Most of these imperial busts (or original statues) were, when acquired, in a very b a d state of preservation. T h e d a m a g e suffered could certainly not have taken place in the villa itself. T h e final scheme of decoration must have been created by acquiring a stock of scrapped sculpture. T h e next question would be to enquire about the source—an Augusteum would be an educated guess. T i m e had run out for the institution of the imperial cult, so the likenesses from

4/

Not mentioned in Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine. Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine, no. 966. This emperor is also represented in a well-preserved bust (only the right shoulder has been attached) executed in high quality municipal style, Espérandieu, no. 961. 4!l Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine, no. 968; Le cirque romain, Musée Saint-Raymond (Toulouse, 1990), cat. 13. 50 Unfortunately, several of the early excavated pieces (mainly from the period 1826-30) have been restored with new noses, and some have been treated with acid. In general, however, these post-Antique restorations are modest. On these restorations, see Hannestad, Tradition in Late Antique Sculpture, pp. 129-33 (compare figs. 8 3 - 4 with cat. 123 in Le regard de Rome (supra n. 44). 48

an imperial shrine must have been outdated a n d accordingly cheap. However, we learn from legal sources of the fourth century that collecting statues (and other material) from existing buildings had become a m a j o r problem all over the Empire, a n d Christians a p p e a r to have c o m p e t e d with pagans w h e n it c a m e to acquiring material for "decorating b a n q u e t i n g halls or porticoes". 5 1 Except for a few late pieces, the private busts range from the early Imperial period to the mid-third century. Generally, they are in a better state of preservation, but m a n y have been retouched a n d some heavily reworked. T h e better state of the private busts makes it possible that some were commissioned for the villa. From the final period of the villa is a large-scale (0.33 m.) portrait of a boy12‫ ׳‬a n d a few female portraits executed in a provincial late Constantinian style. Most are sculpted in the local limestone, but a single specimen is m a d e in the coarse-grained Pyrenean m a r b l e of Saint-Béat. 5:i In the case of this m o r e than life-size head (0.33 m.), the rather deeply cut eye rendered in the f o r m of two concentric semicircles could indicate an even later dating. T h e latest portrait of all appears to be a superb female m a r b l e h e a d of the same size. It is very close to Constantinople-court tradition, a n d a dating of the early fifth century m a y be proposed. Unfortunately, the connection with C h i r a g a n is not absolutely certain. 5 4 Only one portrait stands apart: a head of the orator Demosthenes. 5 3 T h e portraits make u p a b o u t a third of the sculptural content. T h e rest is mythological sculpture or reliefs. O f traditional full-size copies, only seven have been found. T h e y can be dated to the early a n d High Empire a n d are just as battered (and restored) as the imperial busts. T h e s e were apparently w h a t was available to a customer in this area w h o was not short of m o n e y , a n d they must have been costly collector's items. T h e rest of the sculpture in the r o u n d , about

51 See Saradi-Mendelovici, "Christian Attitudes toward Pagan Monuments in Late Antiquity", esp. pp. 50-1 with references. 52 Le regard de Rome (supra n. 44), cat. 172, identified as a Tetrarchic prince (J.Ch. Baity), but the soft modeling of the face and the curls of the back hair could point to an even later dating. T h e material is stated to be white marble (Pyrenean from Saint-Béat?). 53 Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine, no. 1004 (neg. M a r b u r g 32.646). 54 C.f. Le regard du Rome (supra n. 44), cat. 7 (Aphrodisian marble?). 55 See above, η. 45. A portrait of Socrates (Le regard de Rome (supra n. 44), cat. 63) has been attributed to the villa by some. It appears to be late.

37 pieces, is late a n d must have been part of the final decoration scheme. Nearly all are small-scale mythological statuary. Presumably m a d e for the reception rooms of the villa are twelve t o n d o busts a n d twelve corresponding reliefs depicting the Labours of Hercules, both series executed in the local m a r b l e of Saint-Béat. T h e tondos, with a diameter of between 0.74 m. a n d 0.90 m., feature various gods a n d personifications. O n display in the m u s e u m is the couple of Asclepios a n d Hygeia. 3 6 In terms of style, eye execution, etc., the C h i r a g a n tondos bear a close resemblance to a series of portraits m a d e for a philosophical school in Aphrodisias a r o u n d 400. 7 ‫ י‬Such tondos, all depicting pagan figures, seem to have been an Aphrodisian speciality of the late period, a n d they were often hoilowed out from the back to reduce their weight, thereby fitting t h e m for export. However, these are not hollowed out, having been carved locally, as revealed by the marble. U n d e r n o r m a l circumstances, an Aphrodisian sculptor would have preferred the marble familiar to him (or a similar high-quality type), which explains the above-mentioned problems of transport in the late period. 5 8 Such difficulties m a y also be an explanation for the absence of c o n t e m p o r a r y full-scale sculpture in the round. T h e Aphrodisian sculptors m a y have considered the local Pyrenean marble too inferior (e.g. too fragile or too flawed), or simply felt that is was too unfamiliar for such an enterprise. C a r v e d in the same type of m a r b l e but closer to the tradition of free-standing sculpture are the reliefs (1.50 m. high) depicting the Labours of Hercules. Again several stylistic elements relate to the socalled School of Aphrodisias. 5 9 T h e Aphrodisian touch is also clearly felt in the mythological sculpture, all pieces being p r o d u c e d in white marble except for a few in black. It is of course difficult to say for sure if every piece has been carved by Aphrodisians. T h e portable

56

Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine, no. 892, 1.12‫־‬ ' R.R.R. Smith, "Late Roman philosopher portraits from Aphrodisias", Journal of Roman Studies 80 (1990), pp. 127- 55. Compare, for example, Pindar (pis. VI-VII) with Espérandieu, Recueil général des bas-reliefs de la Gaule Romaine, no. 892, 8. O n the style and motifs, see Hannestad, Tradition in Late Antique Sculpture, pp. 135-9, against L· regard de Rome (supra n. 44), cat. 171. O n these tondos, see further below. 58 O n this problem as evidenced in sculpture, see Hannestad, Tradition in Late Antique Sculpture, part II. 59 Hannestad, Tradition in Ijite Antique Sculpture, pp. 139-41. O n the discussion of the artistic circle termed the School of Aphrodisias (a term invented by M. Floriani Squarciapino), see ib. pp. 117 ff.. 5

size and shape of the sculpture is characteristic of the period, a n d stylistically the setting has a very homogeneous character. It may be concluded that the reliefs a n d mythological sculpture in the round, i.e. approximately half of the a m o u n t of sculpture, have been produced by Aphrodisians for the final setting. C o n c e r n i n g the themes (in so far as they can be identified), the range is rather narrow a n d very conventional—Dionysos and his circle: Dionysos himself represented in four replicas (fig. 4), four satyrs (one of them part of a group with a Hermaphrodite), two putti a n d one Ariadne. Following the same long tradition is Venus in four copies, a n d Hercules in two (besides the series of reliefs). Also rendered are Mars, Isis, Diana and Athena, with one representation of each, and two barbarians. T h e saviour aspect is represented by statues of Asclepios and Harpocrates. These two, together with Hygeia, were also present in the tondi. Finally there is one genuine copy: a statue of an old fisherman in black marble. W h e n the setting is c o m p a r e d with similar collections from stately homes of Late Antiquity one misses, for example, the T h r e e Graces or the Graecian T y c h e (the city goddess mainly in the shape of busts), both well known f r o m two-dimensional art such as mosaics and silverware. T h e Old Fisherman is present but there is no herdsm e n . Also the traditional gods are absent; but more puzzling is the nearly total lack of m e n of letters and their muses. As mentioned, a head of Demosthenes has been found, and a relief fragment represents a seated Socrates in front of a H e r m e . Perhaps the owner of the estate was no m a n of literary taste or the library pit has not been found. O n e should not forget, however, that m u c h could have been m a d e in bronze melted down when the marble sculpture was d u m p e d . Further it should be noted that the sculptural content of the villa provides no clue regarding the beliefs of the owners. Were they Christians or pagans? T h e graveyard would have told us, but it has not been located. T h e lack of access to quality m a r b l e — o r rather the problems of transport of large blocks of marble—is strongly felt in the C h i r a g a n material. It appears to have become a general problem: most late villas, when sculpture is found, contain only a few small-scale mythological pieces. T h e d e m a n d for the g r a n d scale was, however, still present. A striking example of how m u c h effort could be invested in the remoter parts of the Empire for acquiring grand marble sculpture

in the Classical tradition is exemplified in a group featuring the myth of Daidalos, found in an otherwise unknown villa in A m m a n (ancient Philadelphia). 60 T h e 1.90-m.-tall main figure, Daidalos himself, is now on display in the National M u s e u m of A m m a n . It is an extraordinarily wellexecuted piece of sculpture, accurate in its carving, highly polished, but with some odd traits distinguishing it f r o m the style of the High Empire. T h e sculpture features stylistic characteristics that point to a late dating, presumably mid-fourth century, and indicate Aphrodisian workmanship. T h e most amazing thing about this group, however, is the fact that it was m a d e piecemeal a n d then transported, to be re-assembled at the point of destination. The figure of Daidalos consists of five different pieces: torso with head, arms, bottom part of torso with right leg a n d left leg. T o this should be a d d e d the now missing body of Ikaros. T h e right hip and leg carried the entire weight through an internal pole probably of metal. T h e marble is of different colouring, mostly bluish-grey, but the load-bearing right hip with leg is carved in a creamy white also quarried in Aphrodisias. T h e sculptor has m a d e m u c h effort to hide the joins by letting them follow the lines of the drapery. Unfortunately we know nothing of the villa or its sculptural p r o g r a m m e besides this statue group, but it may be presumed that it was the only m a j o r marble sculpture belonging to the villa. O t h e r parts of a sculptural setting may have been in bronze. Some centuries earlier, during the High Empire, it would have been no great matter to furnish an entire villa in A m m a n with full-scale groups of marble. In northern Gaul, in the m o d e r n town of Welshbillig not far from Trier, an imperial (?) landowner of the late fourth century had his 60

It came to light in the confused circumstances of 1947 and was taken to Jerusalem, soon to be returned. During this process the fragments of Daidalos' arms and the body of Ikaros disappeared. It has often been suggested that these pieces have ended up in Jerusalem, but this can be rejected: oral communication with chief curator Ruth Peled, T h e Rockefeller Museum. T h e group has been published very carefully by J . H . Iliffe, "A Heroic Statue from Philadelphia-Amman", Studies ftesented to D.M. Robinson I (Saint Louis, 1951) pp. 705-12, pis. 75-80. T h e photographs used for this publication are still in Jerusalem: information from Gideon Foerster. O n the interpretation, see Möbius, "Ein hellenistischer Daidalos" Jahrbuch des deutschen archäologischen Instituts I 68 (1953), pp. 96-101: c.f. also LIMC III 320, no. 52; and A. Claridge, "Ancient Techniques of Making Joins in Marble Statuary", Marble,. Art Historical and Scientific Perspectives on Ancient Sculpture (Malibu, 1990), pp. 150-3. Also Hannestad, Tradition in Late Antique Sculpture, pp. 144-9—the group will be discussed in a forthcoming paper by the author.

park adorned with a gallery of herms lining an oblong water basin nearly 60 m. long. 61 About 112 herms cut in local limestone were originally erected, 71 of which have been found more or less well preserved. T h e herms render a mixed gallery of characters: the Olympian gods, Hercules, Satyrs and Putti, barbarians of various kinds, and portraits ranging from m e n of letters such as Socrates, Demosthenes, and M e n a n d e r , to Hellenistic rulers and R o m a n emperors. Obviously this looks rather confusing, but if you need 112 herms with different heads, you will have to make a clean sweep of types available. Despite obviously provincial craftmanship, the style is pretty close to the ideals of the Thcodosian Renaissance. Most of the heads imitate marble sculpture, but some clearly reflect bronze prototypes. T h e habitation part of the villa is covered by the Medieval town, so very litde can be said about architecture, decoradve schemes, etc. However, it can by its garden design a n d choice of types for the herms be considered a supplement to the above-discussed villa of Chiragan, thereby providing us with a general idea of the setting of an upperclass rural villa of the late period. T h e s e two grandiose villas belong to the Christian era, but as mentioned, Christianity is not reflected in the sculptural schemes. This appears to be the pattern for stately homes all over the Empire, and this same pattern is also to be found in the much better preserved elements of architectural decorations, particularly in mosaics. D u r i n g the fourth, fifth, and sixth centuries, no indisputable Christian mosaic seems to have been laid in a private house (and the sacred buildings show a rather mixed pattern). Does this m e a n that the u p p e r class as such of that period was not Christian?—hardly! Sculpture could in the late period be considered as a pagan medium that continued into the Christian era. J u n i u s Bassus, the Christian city prefect of R o m e , in all likelihood had his house furnished with traditional sculpture. In principle, nearly every motif of the early and High Empire could be represented, except for the big complicated groups or sculpture carved in coloured marble. However, reduction in scale a n d types appears to be characteristic of the period. In this respect a parallel development takes place in relief sculpture. T h e

61

H. Wrede, Die spätantike Hermengallerie von Welshbillig (Berlin, 1972); Trier. Kaiserresidenz und Bischofssitz• Die Stadt in spätantiker und frühchristlicher ^eit (Mainz, 1984), cats. 151-2. T h e estate was lined by a 72 km. long wall (with no defensive purpose), enclosing an area of 220 km‫־‬. T h e estate continued into the Frankish period.

"grand tradition" in state reliefs, the pride of the Principate, also disappears, leaving the political messages of the regime to be expressed in two-dimensional art, mosaics and paintings, or in luxury crafts such as ivory carvings (diptychs) and silver, or in coins and medallions.

Favoured motifs As for previous periods, m u c h sculpture must have been chosen for aesthetic or decorative reasons. This goes, for example, for the innumerable statuettes of Venus equally popular in Hellenistic houses as in Pompeii and late antique villas. T h e ancient gods were still produced in the sculptural workshops, but with less appeal. G o n e are the traditional copies of the g r a n d pieces by M y r o n , Polycleites and other old masters, but their influence is present in the sculptural tradition, and old pieces were collected a n d treated with great care. 62 Examining the material it appears, however, that certain motifs become particularly popular in the late period, and fashion alone can hardly explain why. As regards the choice of subjects, it seems to be very m u c h the same as in the other art forms of this same period. In this "Age of Spirituality" revival played a m a j o r role. Old motifs and old styles were taken up again in all media including the fashionable new miniature forms: ivory diptychs and contorniates. 6 3 Pastiches (also literary) are frequent, a n d in the realm of art, various elements can be mixed in an astonishing way. T h e above-mentioned portrait of J u l i a n the Apostate is an example of this trend; another is a late fourth-century marble portrait of a prince provided with a reminiscence of the "crab-claw" motif of the forelocks invented by

62

A full-scale head of the Ares Borghese type is among the pieces from a late antique house in Antioch. T h e head must be dated as late as the mid-fifth century, but it is doubtful whether the body was also reproduced, see Hannestad, Tradition in Late Antique Sculpture, p. 119, figs. 78-79. T h e basic publication is D.M. Brinkerhoff, A Collection of Sculpture in Classical and Early Christian Antioch (New York, 1970), pp. 38-9. 63 For some reason, the Nachleben is particularily manifest in these two art forms, supporting the idea that some sort of crypto-paganism should be expressed only in small-scale art. T h e exact function of the contorniates is still disputed, see A. Alfbldi and E. Alfoldi, Die Kontomiat-Medaillons 12‫ =( ־‬A M U G S VI 1 2‫( )־‬Berlin, 1976/1990). T h e main work of reference on late antique ivory carving is still W.F. Volbach, Elfenbeinarbeiten der Spätantike und des frühen Mittelalters, 2nd ed. (Mainz, 1976).

Augustus, and used in various forms by some of his heirs. 64 Sometimes a view f r o m one angle suggests a different dating than another: a m o r e than life-size togate statue in Aphrodisias (figs. 6~7) wears the Augustan type of toga; the eyes have no plastic rendering of iris or pupil, as was the sculptural tradition before the late Hadrianic period, so f r o m the front one could take it as a piece of sculpture from the High Empire. 6 5 But seen in profile the bald skull defines it as a piece of the early fifth century. In this period the bald skull becomes for some reason a c o m m o n trait of male portraits, anticipating the medieval tonsure. A most spectacular clash of style a n d type is illustrated in a large silver plate, the Theodosius Missorium in M a d r i d : the c o m i n g era is manifested by the Christian emperors enthroned above in hierachic order, but they are supported by the G r a e c o - R o m a n personification of the earth, Tellus, a key motif since the Augustan period. 6 6 As the glory of the past played an important role, famous m e n could be represented, often in series, in the same way as they always had been by the R o m a n s . T h e h c r m continued to be a popular form, as did the bust: the latter with growing Aphrodisian influence often encircled by a tondo. Related to the above tondos of C h i r a g a n is the series of eleven portrait tondos found in Aphrodisias. 6 7 T h e y a d o r n e d a building of the late fourth or early fifth century, presumably a philosophical school. Some arc n a m e d by an inscription, while others can be identified by the portrait. Classical literature is represented by the poet Pindar; founding fathers of philosophy by Socrates and Aristotle; and their most important pupils Alcibiades a n d Alexander the Great are present. As noted, the iconography of Alexander became influential for that of Christ. Portraits of the great heroic king a p p e a r in all art media throughout Antiquity a n d the Alexander Romance, a work of fiction of the third century, was widely read in Late Antiquity

64

Age of Sprituality, cat. 18; Spätantike und frühes Christentum, cat. 56; Gesichter. Griechische und römische Bildnisse aus Schweizer Besitz, eds. H. Jucker and D. Willers (Bern, 1983), cat. 125. 65 T h e excavator accepted this dating, so the statue is intentionally rendered in three frontal views: see Erim, Aphrodisias, p. 129; also Hannestad, Tradition in iMte Antique Sculpture, p. 160 and figs. 103-5. 61 ' Age of Spirituality, cat. 64 and Spätantike und frühes Christentum, cat. 228. A similar juxtaposition of two realms is met in Renaissance icons of Cyprus, where Italianlooking worshippers are attending a Byzantine-style ceremony: the two parts of the icons were presumably painted by the same artist. e7 Smith, "Late Roman philosopher portraits from Aphrodisias".

a n d the Middle Ages. Related to the current Neoplatonic school are Pythagoras and Apollonios of T y a n a . Further there are some unidentified learned m e n (one a m o n g t h e m a boy), some of w h o m could be "professors" of the period, as suggested by the scholar who published them. 6 " Plotinus, e m b o d i e d in three fourth-century heads f r o m a philosophical school in Ostia, might have been present. 6 9 A n o t h e r character w h o might also have been included in this distinguished c o m p a n y is M e n a n d e r , p r o d u c e d in great quantities, because he was considered by literary critics to be second only to H o m e r , and certainly m o r e accessible to the reader of the time. 70 M e n of letters were often a c c o m p a n i e d by Muses, of which we have a n a b u n d a n c e in the late period. Illuminating is the scene on an ivory diptych from about 500, representing the seated poet a n d his muse. 71 T h e wise m a n or sage, in particular, was a favoured figure offering a bridge to Christianity, a bust from Constantinople itself of T h e o d o s i a n date being p e r h a p s the most impressive piece of the type (fig. 8).72 As noted, Christ was a wise m a n , but some could be claimed as his predecessors: Socrates a n d Apollonios of T y a n a , both present in the Aphrodisias tondos a n d frequently portrayed in the period. 7 3 O n e thing that society could indeed lay on Christianity (and did!) was that it was a new religion—something really suspicious 68

Smith, "Late Roman philosopher portraits from Aphrodisias", p. 151. No. 11 named the "sophist" surely looks like a middle aged university teacher of the 1968generation. 69 T h e identification is conjectural. T w o more portraits of this type exist, both deriving from Rome: see VV. Heibig, Führer durch die öffentlichen Sammlungen klassischer Altertümer in Rom IV (4th ed., Tübingen, 1972), nos. 3135-7, with further references. 70 This was the opinion of the respected literary critic, Aristophanes of Byzantium, as evidenced by a (late) epigram written on a herm of Menander, and further attested in other written sources: see Real Encyclopädie (Pauly-Wissowa), X X I X (1931) col. 715. O n some late busts of Menander "flattened" for long distance transport, see Hannestad, Tradition in Late Antique Sculpture, p. 120, n. 196. 71 Milano capitale dell'impero romano 286-402 d.c. (Exhibition 24 J a n . 22 April 1990) (Milan, 1990), cat. 5b.Id—colour plate p. 311: c.f. also Hannestad, Tradition in Late Antique Sculpture, p. 122. T h e muses often combined with wise men became popular motifs in sarcophagi of the 2nd and 3rd century: see Koch and Sichtermann, Römische Sarkophage, pp. 197-206. 72 Inan and Alföldi-Rosenbaum, Porträtplastik, no. 274. O n the subject, see G. Fowden, "The Pagan Holy Man in Late Antique Society", Journal of Hellenic Studies 102 (1982), pp. 33-59: and for their predecessors, G. Anderson, Sage, Saint and Sophist. Holy Men and their Associates in the Early Roman Empire (London, and N.Y. 1994). C.f. also above, note 15. 73 Statues of Apollonios of Tyana are often mentioned in Byzantine sources, see Saradi-Mendelovici, "Christian Attitudes toward Pagan Monuments in Late Antiquity", p. 57 with references: on Socrates, see below n. 91.

to the R o m a n mind. Religious legitimacy could be obtained by relating to wise men of the past; a n d Socrates a n d Apollonios of T y a n a were particularly fitted for that purpose owing to their learning, way of life a n d persecution by the authorities. A passage in the tenth chapter of Justin's Second Apology is eloquent of the way rising Christianity usurped Socrates, so to speak. 74 " O u r doctrine, then, appears to be greater than all h u m a n teaching. T h o s e who by h u m a n birth were m o r e ancient than Christ, w h e n they attempted to consider and prove things by Logos (Reason), were brought before the tribunal . . . A n d Socrates, w h o was m o r e zealous in this direction than all of them, was accused of the very same crimes as ourselves." A m o n g the old gods some gained strength in the late period. The Sun, often mixed with Christ, has been mentioned and so has Asclepios. O n e might expect to find a statuette of Asclepios, often accompanied by a Hygeia, in a Late R o m a n villa. T h e couple is c o m b i n e d in an ivory diptych of the late fifth century. 7 n As noted, this old saviour god influenced the iconography of Christ, but in his own right he still h a d a great i m p a c t — m e n are always looking for the miraculous. 76 A n o t h e r aspect of this tendency is the various mystery cults of the East: the strange cult of Atds and Cybele together with Egyptian Isis. A long-established oriental cult was that of Dionysos, the main aversion of the above-mentioned bishop T h e o d o r e t . Dionysos a n d his circle (but also Apollo) come to be frequently represented. At first it is bewildering to meet so m a n y d r u n k e n satyrs in the late antique context, but T h e o d o r e t ' s dislike and N o n n o s ' poems show

74 Adapted from G.M.A. Hanfmann, "Socrates and Christ", Harvard Studies in Classical Philology 60 (1951), pp. 205-33. 75 Age of Spirituality, cat. 133; Spätantike und frühes Christentum, cat. 168. O n smallscale statues of Asclepios, see Hannestad, Tradition in Late Antique Sculpture, index: "Asclepios". T h e Asclepios from Chiragan is discussed in Stirling, Mythological Statuary (supra n. 44), pp. 133-4; an Asclepios of similar size and date has been found in the villa (?) of Valdetorres de J a m a r a near Madrid: C. Puerta, Μ.Λ. Elvira, and T . Artigas, "La colecciôn de esculturas hallada en Valdetorres de J a m a r a " , Archivo Espanol de Arqueologia 67 (1994), pp. 179-200, no. 9. I wish to thank Prof. Elvira for showing me all the material and for valuable discussions (even if we do not agree) and for providing me with a draft of the final publication. I am also indebted to Prof. Javier Arce for our discussions of this material. Further examples of late antique pieces can be found in LIMC 2 Asklepios—no. 23, a statuette in Athens from Eleusis dated by inscription to A.D. 309. 76 See H. Clark Kee, Miracle in the Early Christian World (New Haven and London, 1983), chap. 3 (Asklepios).

that such representations were m o r e than m e r e fashion. T h e popularity of Dionysos a n d his circle was at its height in this period. Apparently the old Eastern mystery cult was updated to fit into the spiritual f r a m e w o r k of the times. A striking illustration of acculturation is found in a mosaic from a fourth-century house in Paphos, Cyprus, 7 7 featuring a series of Dionysiac motifs, each individual figure identified by inscription as had now b e c o m e c u s t o m a r y — o n e wonders if sculpture was also labelled by inscription. A m o n g the scenes, one shows the god H e r m e s presenting the infant Dionysos to the Muses at M o u n t Nysa; little Dionysos is portrayed sitting in the lap of H e r m e s , in accordance with the same iconographie scheme as that in which the infant J e s u s traditionally sits in the lap of the Virgin M a r y . H e r m e s covers his h a n d s with his g a r m e n t a n d all wear the nimbus. Another scene represents the punishment of Marsyas: the seated Apollo lets the p r e s u m p t u o u s satyr be dragged away by two Scythian slaves, one ready with his knife. T h e scene has been extended with a female personification w h o stands back in horror. T h e inscription names her as Plane (Error), a Christian concept introduced in the pagan setting that could still be renewed—guilt a n d p u n i s h m e n t had c o m e to stay! Miniature sculptures of the hanging Marsysas are c o m m o n , a n d the scene is a stock motif of mosaics. Various offshoots of this mythological complex also gain popularity: Dionysiac sarcophagi continued to be produced, with the motif of Theseus and A r i a d n e — t h e latter also being frequent in sculpture in the r o u n d — a n d the fight between Theseus a n d the minotaur, mainly reproduced in mosaics. F r o m this t h e m e springs the labyrinth, to become the important Christian symbol of the h a r d path to paradise. Eschatological elements also catch the spirit of the time. G a n y m e d e a n d the eagle are extremely frequent in all media, but Daidalos surprisingly not; the A m m a n g r o u p being unique. Hercules was in fashion because he was the first m a n to b e c o m e a god, for which reason he was also a figure well fitted for the imperial cult, thereby stressing another link between the imperial cult and Christianity. T h e Dioscuri, the founding heroes of R o m e , b e c a m e reinterpreted as Peter and Paul. A mythological type that gains popularity is the giant, w h o is almost entirely absent during the early a n d High Empire. Closely

77

T h e House of Aion, see W.A. Daszewski, Dionysos der Erlöser. Griechische Mythen im spätantiken Cypern (Mainz, 1985) and Bowersock, Hellenism in Late Antiquity, pp. 49-53.

related groups in sub-Hellenistic style have been found wide apart: in a suburban villa in Constantinople a n d a rural villa in central Spain. 7 8 T h e y have been carved in black marble (or rather limestone) m o u n t e d on bases of white marble, a technique which together with the style indicates an Aphrodisian workshop. In Greek thought the gigantomachia symbolizes the battle between chaos and order—civilization against barbarism. In this period when the dragon, like the hellenistic ketos, one of the symbols of the Golden Age of Augustus on the Ara Paris, b e c a m e an evil creature that could swallow J o n a h , the giants came to symbolize the dark powers threatening the civilized (Christian) world. T o stress their subdued status, giants often became represented as carriers or Atlantes. 79 In this function they continue as corbel figures in R o m a n e s q u e churches. Rural life b e c a m e essential to the future world scheme. T h e myth of O r p h e u s is revived, and both Christianity and Judaism took advantage of this figure by letting Christ a n d King David usurp the role of taming the wild animals. 80 T h e G o o d Shepherd is a key figure in early Christian art, thus the landscape becomes populated with young a n d old herdsmen, again in all media. Basically, such figures are descendants of Hellenistic genre pieces, when the sculptural interest of recording the poor, the odd and other social outcasts came into being. 81 A particular figure related to this circle stands out: the old fisherman. Like the Marsyas and the giants, this is one of the few true copies that continued to be reproduced, and reflects the kind of emotion that was m u c h favoured in the multi-ethnic and multireligious Hellenistic world, for which reason they fit well into the Late Antique World. T h e original old fisherman was certainly a famous

78

Chaisemartan and Örgen, Les documents sculptés de Silahtaga, nos. 98 (old), 99 (young)—189 (fragments); Puerta, Elvira, and Artigas, "La colecciôn de esculturas hallada en Valdetorres d e j a m a r a " , no. 1 (headless) pp. 195-6 (fragments of bases). 79 LIMC IV (Gigantes), nos. 571 and 573: chamfrons (horse armour), 574: bronze part of chariot, and 590: architectural element of the early 5th-century "Palace of the Giants" in the Athenian Agora: see Hannestad, Tradition in Ixite Antique Sculpture, pp. 126-7. 80 C.f. Sister Charles Murray, Rebirth and Afterlife. A Study of the transmutation of some pagan imagery in early Christian funerary art (Oxford, 1981), chap. II (The Christian Orpheus). King David-Orpheus is rendered in the mural paintings of the synagogue at Doura-Europos (c. 244/5) and at Gaza (dated by inscription to the year 508) in mosaics (recently restored): c.f. Minerva 3 / 4 (1992), p. 3: in general see Ancient Synagogues (supra n. 21), pp. 172-7 and pp. 129-32. 81 N. Himmelmann, Über Hirten-Genre in der antiken Kunst (Opladen, 1980), and H.P. Laubscher, Fischer und Landleute. Studien zur hellenistischen Genreplastik (Mainz, 1982).

Hellenistic genre piece copied with more or less fidelity to this same original. 82 Furthermore, a great range of related adaptations have been produced. Full-scale copies exist, as well as reduced ones a n d genuine statuettes, like that afore-mentioned from C h i r a g a n . T h e original shows an old m a n carrying his catch. H e is worn out, but still muscular, and clad only in a loin cloth that is w r a p p e d a r o u n d his waist so as not to get wet, thus exposing his genitals. A certain variety is present in the a r r a n g e m e n t of the drapery, some of the late copies loyally reflecting the concept of the exposed genitals, while others have been changed to fit into the new mental climate by hiding the male nakedness. 83 This has been done elegandy in the Chiragan piece, while others have been clad in an exomis that may sometimes be arranged rather clumsily. This goes for a headless statue, a little less than full size, in the T e r m e (fig. 9).84 T h e lower part of the garment stands out at the back to look most like a stiff underskirt— rather bizarre when seen in profile, and the lack of undercutting of the g a r m e n t u n d e r l i n e s the u n h a p p y a p p e a r a n c e of the statue. Fishermen and anglers are not an u n c o m m o n motif in art of the period, but this particular type calls for an explanation. Most of the copies actually seem to have been m a d e in Late Antiquity. H o w may one explain the interest for this particular figure? It may have been reinterpretated as Peter the F i s h e r m a n . By such a reinterpretation, the statue would become equally acceptable to pagans and Christians. Each group could understand it its own way and, not least, it was not offensive to any m e m b e r of a n o t h e r g r o u p — a perfeet piece for a religiously mixed society.

82

Basically on these types Laubscher, Fischer und Landleute. Studien zur hellenistischen Genreplastik. 83 An early example of the changing attitude is to be found in a group of three statues presumably rendering the same high-ranking man of the Gallienic period, now in the Villa Doria Pamphili in Rome. O n e statue depicts him as an ordinary togatus, the two other as a huntsman. In both these statues he is semi-naked, dressed only in a gown, but against earlier representations of this type, the drapery hides his nakedness. In one of them, a sudden wind lifts a corner of the drapery to a proper position: see H. von Heintze, "Drei spätantiken Porträtstatuen", in Antike. Plastik I (Berlin, 1962), pp. 7.32‫־‬ 84 Museo JVazionale Romane. Le sculture 1,2, ed. A. Giuliano (Rome, 1981) cat. 16. T h e size of the preserved part is 1.42 m. It was found two km. outside Rome, close to the Via Prenestina, indicating a villa context.

Conclusion T o sum up: what can we deduce from the above material? Christianity as such was hardly hostile to sculpture—neither were the Christian Emperors®—but sculpture never really became a medium of Christian art in the early period. It was left to the later medieval period to create a Christian imagery within sculpture, the seated T e r m e Christ being a u n i q u e exception. As e x p o u n d e d above, sculpture was a p a g a n m e d i u m that died out with the Ancient World. C o n t r a r y to what is stated in the handbooks, it died slowly, and to a certain degree it was able to adapt to the d e m a n d s of rising Christianity. Christians of the early period, like c o n t e m p o r a r y Jews, lived with sculpture as a part of u r b a n reality. In the cities, the various religious groups were intermingled, just as they could be buried in the same graveyards in sarcophagi p r o d u c e d by the same workshops. 8 6 T h e big silver hoards of late Antiquity also bear evidence that pagans a n d Christians could be part of the same household, 8 7 a n d legislation of the same period tries to prevent Christians from converting to Judaism: 8 8 the idea of natural born-confrontation launched by the C h u r c h Fathers (and the rabbis) is not necessarily the truth! T h e n the question: why did sculpture in the r o u n d disappear? This is a very complicated question, with no simple answer, just as complicated as that of the end of the R o m a n Empire. O n e can paraphrase G i b b o n ' s words that we should not inquire "why the R o m a n empire was destroyed, we should rather be surprised that it subsisted that long." 8 9 Society was still able to p r o d u c e marble (or other types 85 As stated by Saradi-Mendelovici, "Christian Attitudes toward Pagan Monuments in Late Antiquity", p. 49: "a systematic destruction of pagan sanctuaries was never the intention of imperial policy", with references. It is worth noticing that in pictorial representations of cities in late antique and early Byzantine times temples appear as a natural element: see J . Deckers, "Tradition und Adaption. Bemerkungen zur Darstellungen der christlichen Stadt", Mitteilungen des deutschen archäologischen Instituts, Römische Abteilungen 95 (1988), pp. 303-82. 86 See above η. 13. 87 Very significant in this respect is the Esquiline Treasure (from this hill of Rome) comprising more than sixty items of silver: the domestic goods of a noble family of the mid-fourth century, predominantly pagan, but with some Christian members, as shown by the decoration of the pieces: see K.J. Shelton, The Esquiline Treasure (London, 1981). 88 For references, see Rutgers, "Archaeological Evidence for the Interaction of Jews and Non-Jews in Late Antiquity", pp. 115-18. 89 The Decline and Fall of the Roman Empire, chap. X X X V I I I (449 582‫ ־‬A.D.): "General observations on the fall of the Roman empire in the West".

of stone) sculpture in great size, such as capitals for churches or palaces or sarcophagi, and to transport such pieces over long distances w h e n needed, but sculpture in the r o u n d was a n o t h e r matter. In fact it is a problem to transport sculpture in marble because the pieces are fragile but still heavy. A n d bronze appears to have become scarce. Matters of fashion played an important role: the late antique house with its m o r e intimate design was not fitted for grand-scale sculpture. Statuettes could be placed in the several small niches, but decoration tended to become two-dimensional, thus expressed in paintings or mosaics. A n o t h e r part of the explanation could be the changing structures of society. During the High Empire, members of the wealthy middle class were, by sheer numbers, big customers for sculpture— their grave m o n u m e n t s were those threatened by dilapidation in Late Antiquity. W h e n the middle class b e c a m e drastically reduced in size, leading to a society that basically consisted of two g r o u p s — a vast lower class, the humiliores, and a very small upper class, the honestiores— the market for sculpture must accordingly have been reduced. Wealth b e c a m e concentrated a m o n g a very small g r o u p that tended to be m o r e in favour of small-scale luxury crafts. For the most part the written sources are silent, but sometimes we get a hint—like the r e m a r k of A m m i a n u s Marcellinus about how old-fashioned R o m a n senators could be: they still wished to have h o n o r a r y statues of themselves erected in public! 90 Inscriptions a n d archaeological evidence confirm that this was the case; but time was about to run out for the municipal statue in R o m a n society. Elite display took new directions. In a way, the R o m a n s never really understood sculpture in the same way as the Greeks did. T h e original portrait of Socrates rendered him in the shape of a seated old m a n d r a p e d in a himation that barely concealed his p r o t r u d i n g belly. 9 ' T o the R o m a n s , this 90 Ammianus Marcellinus xiv, 6, 8, ed. J . C . Rolfe (Cambridge, Mass., 1935 9): see Ward-Perkins, From Classical Antiquity to the Middle Ages, pp. 2 2 - 3 and chap. 2 passim: also J. Matthews, The Roman Empire of Ammianus (London, 1989): index; "statues". 91 This is also the way he is rendered in the wall painting in a Roman house in Ephesus: V.M. Strocka, Die Wandmalerei der Hanghäuser in Ephesos, Forschungen in Ephesos V I I I / 1 - 2 (Vienna, 1977), pp. 9 3 - 6 (H 2/7), pis. 94-96: date c. 100 A.D. Aproximately a hundred years later is a painted bust, part of a series of the seven wise men, in house H 2 / 2 4 , see pp. 115-17 (pi. 265), with references to similar series including a painting in the library of a certain Rusticus from the 5th cen-

same portrait preferably b e c a m e a head that would be placed on a bust or a hcrm or in a tondo. T h e body (of a portrait) was, when rendered as an entire statue, something that should merely denote the social standing of the portrayed person. T h e head of Polycleitos' Doryphoros could be used as a hcrm, as we know from the Villa of the Papyri n e a r H e r c u l a n e u m , in the same tradition as later encountered in the above-mentioned villa of YVclshbillig. By itself, this means an abstraction of the body. Slowly the Greek-founded interest in the n a k e d b o d y d i s a p p e a r e d . T h e c o p y i n g of traditional masterpieces and the continued production of works in that tradition so well represented in, say, H a d r i a n ' s villa at Tivoli, ceased with the second century, but such pieces were still collected in Late Antiquity, as we have seen, just as sculpture in the round in the Classical tradition was produced until the sixth century. But gradually such pieces tended to become reduced in size and increasingly m o r e abstract: muscular renderings, veins, and drapery acquired the a p p e a r a n c e of patterns, the portraits b e c a m e masks (or icons). T h e rendering of the naked body could be extremely exaggerated—one sometimes gets the feeling that a sculptor wished to demonstrate that he mastered the old tradition of the m e d i u m . In other cases, the figures tend to be rendered sloppy, to give the male figure an effeminate character. 9 2 In every respect, sculpture, a m e d i u m brought to its zenith by the Greeks, was behind the times. A puzzling fact is that sculpture of size also disappears in the Sassanian Empire a r o u n d this time. O t h e r art forms took over. Christianity, however, is hardly to blame for this development. Rather, Hellenism came to an end with respect to sculpture.

tury. O n the portraits of Socrates, see G.M.A. Richter, The Portraits of the Greeks (rev. R.R.R. Smith) (Oxford, 1984), pp. 198-204: and Sokrates in der griechischen Bildniskunst. Glyptothek München, ed. I. Scheibler (Munich, 1989). 92 Mathews, The Clash of Gods, deals with the aspect of the "feminine Christ", see index p. 218. Sculpture can hardly be taken into account: the phenomenon in this medium is rather a matter of style.

LE D É V E L O P P E M E N T D E S L A N G U E S R E G I O N A L E S ET L ' I N T R O D U C T I O N D'ALPHABETS DANS DES C O M M U N A U T É S ILLETTRÉES Christian H a n n i c k

D a n s la partie orientale du bassin méditerranéen il y avait à la fin de l'Antiquité, c'est-à-dire au début du 4e siècle, quatre langues à tradition littéraire importante et à diffusion suprarégionale, le latin c o m m e langue de l'administration, le grec c o m m e langue culturelle et religieuse des populations hellénisées, le syriaque dans les groupes ethniques sémitiques et au-delà, enfin le copte en Egypte. Dans les régions limitrophes de l'empire r o m a i n et de la Perse sassanide, l'aram é e n servait aussi de langue vernaculaire entre populations de différentes ethnies ainsi q u e c o m m e langue de l'administration perse j u s q u ' e n Arménie orientale et continuait en ceci le rôle tenu à l'époque a c h é m é n i d e par " P a r a m é e n impérial". 1 Au début du 5e siècle, l'expansion du christianisme dans les régions caucasiennes conduit à la création de d e u x — o u m ê m e peut-être trois—alphabets p o u r fixer les sons de langues jusque là à tradition p u r e m e n t orale, l'arménien, le géorgien et l'albanien du Caucase. La nouvelle écriture sert alors de véhicule à la traduction de la Bible en a r m é n i e n et en géorgien; une riche littérature, d ' a b o r d ecclésiastique, dans le cas de l'arménien aussi historiographique, p r e n d son essor. Le cas de l'albanien est plus problématique et il en sera traité plus tard. C o m m e n ç o n s p a r l'arménien, où les coordonnées, m ê m e si elles d é p e n d e n t de légendes, sont plus étoffées. D'après l'historiographie a r m é n i e n n e ancienne, Koriwn, Lazar de P a r p et Moïse de C h o r è n e qui r e m o n t e n t du moins en partie au 5e siècle, la création d ' u n alphabet a r m é n i e n est l'œuvre d ' u n moine lettré, Mesrop-Mastoc', qui acheva sa tâche vers 407. Ses disciples se mirent aussitôt au

1

Cf. St. Segert, Altaramäische Grammatik mit Bibliographie, Chrestomathie und Glossar (Leipzig, 1990), p. 51. Voir en général: R. Schmitt, "Die Sprachverhältnisse in den ösdichen Provinzen des Römischen Reiches", Aufstieg und .Niedergang der römischen Welt II 2 9 / 2 (Berlin-New York, 1983), pp. 554 586!

travail et traduisirent l'Ecriture Sainte du syriaque et du grec. Son disciple Koriwn rédigea sa Vie qui d e m e u r e un m o n u m e n t historique d ' i m p o r t a n c e capitale, m ê m e si le texte qui nous est parvenu n'est pas d é m u n i d'interpolations plus tardives. 2 Le cas de l'arménien est à b e a u c o u p d ' é g a r d intéressant p o u r le t h è m e qui nous occupe et doit être analysé en tenant compte aussi bien des données linguistiques que des composantes de l'histoire ecclésiastique. Il ne fait plus a u c u n doute que les Arméniens sont un peuple indo-européen. Vers 1200 avant J . C . ils sont venus en Asie M i n e u r e avec les Phrygiens, mais leur m a r c h e vers l'est ne s'est pas accomplie d ' u n j o u r à l'autre, c o m m e le m o n t r e n t les traces archéologiques qui s'échelonnent le long de leur route. Ainsi, p e n d a n t des siècles, ils ont pu assimiler b e a u c o u p d'éléments ethniques étrangers, ce qui explique des innovations phonétiques qui séparent l'arménien classique de l'indoeuropéen et qui ont pu se produire au contact avec des populations asianiques ou balkaniques qui ne nous sont plus toutes connues, et donc non pas seulement grâce au substrat des langues caucasiennes dans l'habitat définitif des Arméniens. 3 Si l'alphabet arménien, utilisé j u s q u ' à nos jours, fut créé dans la première décennie du 5e siècle, cet événement ne correspond ni à l'introduction du christianisme ni aux débuts d ' u n e culture littéraire p a r m i la population a r m é n i e n n e . En effet, 100 ans avant cette date, entre 301 et 317, le roi d ' A r m é n i e , Tiridate III (298-330), déclare le christianisme religion officielle dans son royaume, selon la tradition ecclésiastique a r m é n i e n n e , avant l'édit de Milan de l ' e m p e r e u r Constantin. 4 D ' a u t r e part plusieurs historiens et écrivains ecclésiastiques arméniens, surtout Eznik de Kolb (première moitié du 5e siècle), font m e n t i o n de la destruction de la littérature p a ï e n n e en relation avec le Mazdéisme, les m o u v e m e n t s gnostiques et la philosophie grecque antique arménienne. 5 Ceci ne signifie pas aussitôt que ces écrits, brûlés p a r les propagateurs de la foi chrétienne en Arménie, aient été rédigés en armé2

Cf. l'étude magistrale de G. Winkler, Koriwns Biographie des Mesrop-Mastoc', ÜberSetzung und Kommentar, Orientalia christiana analecta 245 (Rome, 1994). 3 H. Vogt, Linguistique caucasienne et arménienne, Studia Caucasologica 2 (Oslo, 1988), p. 117. 4 Cf. N.G. Garsoïan, "Politique ou orthodoxie? L'Arménie au quatrième siècle", Revue des Études arméniennes 4 (1967), pp. 297-320. 5 Introduction générale à ces problèmes: Histoire des Arméniens, sous la direction de Gérard Dédéyan (Paris, 1982); R.H. Kévorkian J.P. Mahé, Arménie: 3000 ans d'histoire (Marseille, 1988); K. Beledian, Les Arméniens (Turnhout, 1994).

nien dans une écriture prémesropienne. Les trop rares m o n u m e n t s épigraphiques qui semblent attester un essai d'écriture de l'arménien différent de l'alphabet mesropien, ce q u ' o n appelle l'alphabet du syrien Daniel, dont parle Koriwn dans la biographie de M a š t 0 c ' (§ 3142‫)־‬, h ne permettent a u c u n e conclusion trop hâtive à ce sujet. 7 Par contre, l'effloraison de l'arménien classique c o m m e langue littéraire et l'apogée que cette langue atteignit à "l'âge d'or", dans la première moitié du 5e siècle, avec la traduction de la Bible et la composition de traités philosophico-théologiques c o m m e le De Deo de Eznik, H donc dans la génération des disciples de Mesrop-Mastoc', mort en 440, seraient impensables, si la langue arménienne n'avait pas été cultivée, à un niveau littéraire, ou tout proche de ce niveau, déjà avant Mesrop et le début du 5e siècle. N'oublions pas que la création de l'alphabet de M e s r o p eut lieu dans une époque de déclin politique, après la perte de l'indépendance étatique et la division du pays en 387 entre R o m e et la Perse et que la dynastie locale des Arsacides d'origine perse, réétablie sous suzeraineté romaine en 218, fut liquidée en 428, donc à l'orée de "l'âge d'or". La traduction a r m é n i e n n e de la Bible, un fleuron de la littérature en arménien classique, était déjà achevée, après une révision, vers l'année 440 à la mort de Mesrop et du Katholikos Sahak. 9 Cette traduction effectuée sur le syriaque et le grec, puis revue d'après le grec selon la tradition textuelle de la C a p p a d o c e , met en évidence les courants culturels et linguistiques d'où é m a n e l'arménien classique c o m m e langue littéraire, d ' u n e part l ' O s r h o è n e syriaque et la Mésopotamie supérieure, d'autre part la C a p p a d o c e grecque. En ce qui concerne cette dernière, on se souviendra que Saint Basile de Césarée aurait eu vers 372 des relations, p a r ailleurs historiquement assez douteuses, avec le diocèse de G r a n d e Arménie. 1 0

6

YVinklcr, Koriwns Biographie, p. 98. L'ouvrage classique sur l'histoire de l'écriture arménienne demeure: A.G. Abrahamyan, Hayocï gir ev grc'utyun (Erevan, 1978). 8 L. Mariés - Ch. Mercier, Eznik de Kolb, De Deo, Edition critique du texte arménien, traduction française, Patrologia Orientalis 2 8 / 3 4 (Paris, 1959). 9 Cf. S. Brock Ch. Hannick, "Bibelübersetzungen: Die Übersetzungen ins Armenische", Theologische Realenzyklopädie 6 (1980), pp. 200 205; ainsi que les dilîérentes contributions au volume Armenia and the Bible. Papers presented to the international symposion held at Heidelberg, July 16-19, 1990, ed. Ch. Burchard (Atlanta/Georgia, 1993). 10 Cf. N.G. Garsoïan, "Nerses le Grand, Basile de Césarée et Eustathe de Sébaste", Revue des Etudes arméniennes 17 (1983), pp. 145-169. 7

Il faut encore citer un troisième jalon dans la formation de la langue littéraire arménienne, l'apport iranien, très important autant au plan de la dérivation nominale que du lexique, un apport tellement considérable que jusqu'il y a un peu plus de cent ans, q u a n d paru l'étude capitale de Heinrich H ü b s c h m a n n " Ü b e r die Struktur des Armenischen im Kreise der indogermanischen S p r a c h e n " (1875)", on considérait l'arménien c o m m e un dialecte iranien. H ü b s c h m a n n établit que l'arménien représente une langue propre, entre l'iranien et le baltoslave. Au plan dialectal les études linguistiques r e m o n t a n t à A. Meillet, E. Bcnveniste, G. Bolognesi ont m o n t r é que les e m p r u n t s iraniens en arménien, du moins pour l'époque ancienne, ont été influencé p a r le parthe, donc viennent du d o m a i n e nord-ouest de l'iranien. C e lexique c o m p r e n d aussi n o m b r e de termes relevant de la vie religieuse comme avetaran—"évangile", sarkavag—"diacre". Ces faits posent la question d'une sensible influence du christianisme perse en Arménie, vraisemblablement à l'époque arsacide et donc avant la création de l'alphabet par Mesrop, ce qui veut dire à une époque où en Arménie la langue du culte chrétien était, dans les provinces orientales voisinés de la Perse, un dialecte iranien. 1 2 Ajoutons toute de suite q u ' à côté de ces termes religieux d'origine iranienne bon n o m b r e de concepts de la sphère religieuse en arménien sont empruntés au syriaque et au grec, ce qui m o n t r e la diversité des usages linguistiques dans le vaste territoire occupé par les Arméniens.' 3 O n notera toutefois qu'il n'y a pas de doublets dans la langue classique (sauf peut-être k'ahanay, erêc'—"prêtre", le premier e m p r u n t é au syriaque, le second, bien arménien, mais calquant au plan séman-

11

Zeitschriftßir vergleichende Sprachforschung 23 (1875), pp. 5-19, repris dans H. Hübschmann, Kldne Schriften zum Armenischen, ed. R. Schmitt. (Hildesheim-New York, 1976), pp. I 45. 12 Présentation du problème par G. Bolognesi, "La tradizione culturale armena nelle sue relazioni col mondo persiano e col mondo greco-romano", Atti del Convegno intemazionale sul tema: Ιλ Persia e il mondo Greco-Romano (Roma, 1966), pp. 569-603. Réimprimé dans G. Bolognesi, Studi glottologici,fllologici,orientali (Brescia, 1990), pp. 271-318. R. Schmitt, "Iranisches Lehrgut im Armenischen", Revue des Etudes arméniennes 17 (1983), pp. 73 112. Sur l'historié de la langue arménienne voir l'initiation de M. Nichanian, Ages et usages de la langue arménienne (Paris, 1989). 13 O n consultera toujours avec profit, malgré quelques points de détail à corriger, le dictionnaire étymologique divisé d'après les couches perse, syriaque et grecque, établi par H. Hübschmann, Armenische Grammatik I: Armenische Etymologie (Leipzig, 1897) (réimpression anastatique Hildesheim, 1972).

tique le grec πρεσβύτερος et signifiant d'abord 'l'ancien' et ainsi different de Ιερεύς), de sorte que le choix de la culture d o n n a n t e , iranien, syriaque ou grec, a du se faire avant l'éffioraison de l'arménien littéraire classique et non pas selon des critères géographiques. O n sait hélas très peu de choses sur le christianisme en langue iranienne, 1 4 sauf que b e a u c o u p de martyrs l'ont illustré. Dans cet empire pluriethnique et polyglotte il y avait certes place, à côté du siège de Séleucie-Ctésiphon de culture syriaque, pour des églises de langue iranienne dans les provinces septentrionales, de la Parthie à l'Ouest à la Sogdiane à l'Est. La création d ' u n e langue littéraire dans un milieu linguistique dans lequel une langue c o m m e le grec avait atteint un tel degré d'universalité—nous sommes au début du 5e siècle, après Libanios, à l'époque des historiens ecclésiastiques Sokrates, Sozomenos, Philostorgios—n'aurait pu se faire, si l'arménien n'avait déjà acquis au début du 5e siècle un niveau qui lui permettrait d'entrer en concurrence avec le grec et le syriaque. Pour qui Eznik a-t-il rédigé en arménien sa réfutation des hérésies d'origine aussi bien hellénique que iranienne sinon pour un public cultivé, familier à la philosophie, et qui possédait le grec? U n e réfutation n'a de sens que si les catégories logiques dont elle se sert sont au moins adéquates à celles qui sont l'objet du combat intellectuel. Il n'était pas question au début du 5e siècle en Arménie d ' a m e n e r la population à une première connaissance, d ' a b o r d rudimentaire, du christianisme. Grégoire l'Illum i n a t e u r organisa déjà vers 300 u n e structure ecclésiastique en Arménie en se servant bien sûr du grec et peut-être aussi du syriaque. Eusèbe de Césarée m e n t i o n n e dans son Histoire Ecclésiastique un évêque M c r u z a n (Μερουξάνυς) en Arménie au milieu du 3e siècle (V 1, 46), retenu dans les listes épiscopales du siège d ' A r m e n i a maior, par ailleurs fort incomplètes pour cette époque, dans les années 230-260. 1 5 A l'encontre de la Bulgarie au 9e siècle dont nous traiterons encore, il s'agit ici de populations avant tout arméniennes, car il n'y a aucune

14 Voir l'ouvrage classique de J . Labour, Le christianisme dans l'empire perse sous la dynastie sassanide 224 632 (Paris, 1904), ainsi que M.-L. Chaumont, IM christianisalion de l'empire iranien des origines aux grandes persécutions du quatrième úècle, Corpus scriptorum christianorum orientalium 499/Subs. 80 (Leuven, 1988). 15 G. Fedalto, Hierarchia ecclesiastica orientalis I (Padoue, 1988), p. 409; voir aussi H. Acarean, Hayoc' anjnanunneri bajaran III (Beyrouth, 1972), p. 326.

trace d ' u n e dissémination de quelque importance de l'élément ethnique grec dans l'Anatolie Orientale, à l'Est du lac de V a n et j u s q u ' à l'Araxe. U n e langue littéraire présuppose le dépassement des dialectes ou tout au moins l'émergence de l'un d'eux c o m m e Koine reconnue c o m m e telle p o u r des raisons politiques ou culturelles ou économiques. Rappelons q u ' à l'époque de la création de l'alphabet arménien, au d é b u t du c i n q u i è m e siècle, l ' A r m é n i e avait p e r d u son indépendance et était divisée entre les sphères d'influence romaine et perse. La dynastie arsacide, d'origine parthe, en Arménie Orientale, les Aršakuni, 1 6 s'éteindra en 428. Les textes originaires du 5e siècle qui nous sont transmis, bien sûr dans des manuscrits plus tardifs— la tradition manuscrite a r m é n i e n n e ne remonte guère au delà du 9e siècle, à p a r t quelques f r a g m e n t s allant j u s q u ' a u 7e siècle, voire j u s q u ' a u 5e 1 7 —ne présentent pratiquement pas de traits dialectaux. Pour autant qu'on puisse en j u g e r d'après les dialectes modernes, et ceci avec toute la précaution requise, l'arménien classique se serait développé sur la base de dialectes de la région du lac de V a n , à cette époque région centrale du territoire linguistique arménien. La tradition épigraphique dont les plus anciens m o n u m e n t s remontent au 6e siècle, présente elle, c o m m e il faut s'y attendre, des particularités locales. Ces quelques remarques devraient suffire pour caractériser la situation linguistique en Arménie à l'époque de l'émergence de l'arménien c o m m e langue littéraire en relation avec la création de l'alphabet. Il ne peut donc ici être a u c u n e m e n t question de l'extension de la culture livresque dans l'Antiquité tardive à une société illettrée. T o u t autre s'avère le cas de la Géorgie, dont il sera encore question en conclusion, dans un essai comparatif et typologique. Dès la troisième guerre romaine contre Mithridate, sous Pompejus, la Géorgie occidentale, l'antique Colchide, entre dans l'orbite romaine et demeurera de culture grecobyzantine j u s q u ' a u 10e siècle, j u s q u ' à l'unification des territoires géorgiens, unification a u j o u r d ' h u i menacée, ce qui souligne l'importance et l'actualité du p r o g r a m m e de recherches sur la transformation du m o n d e romain. Vers la m ê m e époque, dans les 16

Voir la notice qui leur est consacrée dans N.G. Garsoïan, "The epic histories attributed to P'awstos Buzand (Buzandaran Patmut'iwnk')", Harvard armenian Texts and Studies 8 (Cambridge, Massachusetts, 1979), pp. 354-355. 17 Cf. le catalogue d'une exposition au musée de Bochum du 14 janvier au 17 avril 1995, Arménien — Wiederentdeckung einer alten Kulturlandschaft, p. 239.

années soixante avant notre ère, la Géorgie Orientale, Kartli, j3asse sous protectorat romain et y restera j u s q u ' a u 4e siècle de notre ère, q u a n d le pouvoir iranien prit le dessus. Au milieu du 7e siècle, les Arabes o c c u p p e n t la Géorgie Orientale et établissent l'émirat de Tbilisi. 18 La conversion de la Géorgie au christianisme au 4e siècle1'1 et la création de l'alphabet géorgien au début du 5e siècle, peut-être l'œuvre de Mcsrop-Mastoc' d'après la tradition arménienne (Koriwn § 83 2 0 ,(93‫ ־‬eurent donc lieu à l'époque où l'influence romaine, donc de culture grecque, était p r é p o n d é r a n t e en Géorgie. L'apport iranien est moindre, et secondaire, c'est-à-dire passé p a r l'Arménie. En effet, tout qui consulte le lexique géorgien des textes bibliques anciens établis p a r Joseph Molitor 2 1 constatera que n o m b r e de concepts centraux de la terminologie chrétienne en géorgien sont e m p r u n t é s à l'arménien. Mais pas tous. Par exemple le terme équivalent à " b a p t ê m e " , βάπτυσις, βάπτισμα—mkrtufiwn en arménien—est rendu en géorgien par nateli, un calque sémantique de φωτισμός. Beaucoup reste d'ailleurs à faire dans ce d o m a i n e de la langue théologique géorgienne ancienne, 2 2 pourvu que l'on se libère de préjugés nationalistes, et de telles études fructifieraient b e a u c o u p l'histoire de la théologie byzantine, car le milieu culturel grec dans lequel le christianisme géorgien s'est développé est complexe et en partie tombé dans l'oubli ou dans l'ombre au plan grec. Rappelons ici uniquement les importantes colonies géorgiennes de Jerusalem, d'Antioche et de la M o n t a g n e Noire, de l'Olympe de Bithynie. 23 Mais dans les territoires géorgiens du Caucase eux-même, la culture littéraire n'était pas absente à l'époque de l'introduction du

18

Voir en général K. Salia, Histoire de la nation géorgienne (Paris, 1983); ainsi que O . Lortkipanidse, "Wann beginnt die Geschichte der georgischen Kultur?' 1 , Georgica 16 (1993), pp. 5-15. 19 Présentation succinte du problème par Κ. Kekelidse, Die Bekehrung Georgiens zum Christentum (Leipzig, 1928). 211 Winkler, Koriwns Biographie., p. 105. 21 J . Molitor, Glossarium ibericum in quattuor evangelia et actus aposlolorum antiquioris verswnis, Corpus scriptorum christianorum orientalium 228/Subs. 20, 237/Subs. 21 (Louvain 1962). Les volumes correspondants pour le reste du Nouveau Testament sont parus dans la même série en 1964, 1965 et 1976. 22 Voir de même l'étude capitale de J . Gippert, Iranica armeno-iberica. Studien zu den iranischen Lehnwörtern im Armenischen und Georgischen, Sitzungsberichte der phil. hist. Kl. der 0/1W 606. Veröff. der Komm. f . Iranistik 26 (Vienne, 1993). 23 A côté des études spéciales comme celle de E. Metreveli sur Jérusalem ( Tbilisi, 1962) ou de W. Djobadze sur Antioche et l'Orontide (Louvain, 1978) cf. l'esquisse de L. Menabde, Ocagi drevnegruzinskoj kul'tury (Tbilisi, 1968).

christianisme. 2 4 Rappelons ici l'inscription bilingue araméenne-grecque de Άρμάντικα—Armazis-c'ihe près de M c h e t a , datant de l'an 150 de notre ère environ et découverte en 1940 p a r G.V. Cereteli. 25 D e sorte que les premières inscriptions géorgiennes, dont la plus ancienne, celle de la basilique Sion de Bolnisi, est datée de l'an 20 du souverain sassanide Peroz, donc 478, 26 n'étonnent pas dans ce contexte géographique. Soulignons d'ailleurs que l'extension géographique des inscriptions géorgiennes et a r m é n i e n n e s — c o m m e l'a m o n tré P. M u r a d y a n dans diverses études et publications qui ont suscité b e a u c o u p de remous en Géorgie 2 7 —ne correspond pas aux frontières actuelles, et ceci dans les deux directions. Q u a n t à l'albanien du Caucase, son étude acquiert de nos jours une importance toute dictée p a r les conjonctures politiques et reliées au problème de la région de K a r a b a x / A r c a x . 2 8 Mesrop serait également l'inventeur de cet alphabet (Koriwn § 112 2 9 ,(113‫ ־‬le troisième dans le m o n d e caucasien, p o u r une langue caucasienne très distincte du groupe géorgien-mingrélien. Aucun texte de cette langue n'est connu, il ne s'est conservé qu'une liste de signes dans quelques rnanuscrits arméniens. 3 0 Q u a n d des chercheurs affiliés à l'Académie des Sciences de Baku cherchent à étudier l'ethnogénèse du peuple azerbaïdjanais en r e m o n t a n t à 'l'église albanienne' du haut Moyen Age et au catholicosat albanien—d'ailleurs de langue a r m é n i e n n e — d o n t le dernier siège était à G a n d z a s a r dans le K a r a b a x septentrional et fut liquidé au milieu du siècle dernier, ils font preuve de méconnais24

Sur le paganisme géorgien et ses sources littéraires cf. M. van Esbroeck, "La religion géorgienne préchrétienne", Aufstieg und Niedergang der römischen Welt II/18.4. (Berlin-New York. 1990), pp. 2694-2725. 25 Cf. Ocerki istorii Gruzii I: Gruzija s drevnejšich vremen do IV veka nasego erì (Tbilisi, 1989), pp. 316-317. 26 Cf. Ocerki po istorii Gruzii II: Gruzija ν IV X vekach (Tbilisi, 1988), p. 426; ainsi que la transcription du texte dans L'Encyclopédie soviétique géorgienne II (Tbilisi, 1977), pp. 454-455. 2 ' P.M. Muradyan, Hayastani vrac'eren aijanagrut'yunnerè, Armeno-Georgica 1 (Erevan, 1977) (cf. sur ce livre les articles critiques de J.P. Mahé et de V. Silogava dans Bedi Kartlisa 38 (1980), pp. 295-309 et 365-395); id., Armjanskaja èpigrqfika GruZü—Kariii i Kacheti, Armeno-Georgica 4 (Erevan, 1984); id., Armjanskaja epigrafika Gruzii—-Tbilisi, Armeno-Georgica 5 (Erevan, 1988). Voir également Ζ. Aleksidse, Atenis Sionis somf}uri çarçerebi (Tbilisi, 1978). 28 Parmi l'abondante littérature consacrée à ce sujet on ne citera que le volume rédigé par P.M. Muradjan, Κ 0svešceniju problem istorii i kul'tury kavkazskoj Albanii i vostocnych provincij Armenii I (Erevan, 1991). 29 Winkler, Koriwns Biographie, p. 108 et commentaires 340-346. 30 R.H. Hewsen, "On the alphabet of the Caucasian Albanians", Revues des études arméniennes 1 (1964), pp. 427 432.

sance des faits et interprètent faussement les sources historiques et archéologiques qui soulignent la liaison ethnique et linguistique de ce siège arménien oriental avec le reste de la c o m m u n a u t é ethnique arménienne. 3 1 De sorte que p o u r notre propos, on peut laisser la question de l'albanien du Caucase de côté. O n s'étonnera de ce que les Slaves soient passés si tard de l'illittéralité à la culture livresque; si tard si l'on tient compte du niveau culturel du milieu ambiant des Slaves après leur venue dans la péninsuie balkanique dès la fin du 6e siècle. T o u t porte à croire que les Slaves, installés dans les Sklavénies j u s q u ' a u fond du Péloponnèse, sont passés assez vite au christianisme de culture orthodoxe. O n sait peu de choses des étappes de cette christianisation, sauf qu'elle se fit dans le cadre d ' u n remaniement partiel des structures de l'organisation ecclésiastique dans les territoires concernés. 3 ‫ '־‬Par contre la conversion "officielle" des Bulgares dans les années 60 du 9e siècle, puis des Slaves de la Rus' de Kiev à la fin du 10e siècle—j'évite à dessein les termes russes ou ukrainiens!—elle a fait l'objet de maintes études et j ' e n ai, moi aussi, traité récemment dans le quatrième volume de YHistoire du christianisme.'''' Plus énigmatiques, mais pour notre propos non dénués d'intérêt, s'avèrent les efforts entrepris par l'empereur Herakleios dans la première moitié du 7e siècle afin de christianiser les Slaves établis dans, ou mieux près des villes de la côte dalmate. 34 Rappelons que ces villes relevant comme tout l'Illyricum occidental de la juridiction romaine avaient outre une population de langue latine ou r o m a n e (le dalmate) de forts éléments grecs. L'un et l'autre courants ont fortement influencé l'onomastique et la toponomastique. 3 ‫ ''׳‬Ce rappel est d'autant plus utile qu'il d o n n e les dimensions linguistiques dans le cadre desquelles le problème dont nous traiterons est à situer. C'est dans ce cadre aussi qu'il faut placer

31

Echos de cette polémique dans P.M. Muradjan, Istorija-pamjat' pokolenij. Problemy istorii Nagornogo Karabacha (Erevan, 1990). 32 Cf. A.P. Vlasto, The entry of the Slavs into Christendom (Cambridge, 1970); Lj. Maksimovic, "The christianisation of the Serbs and the Croats", The legacy of saints Cyril and Methodius to Kiew and Moscow, ed. A.E. Tachiaos (Thessaloniki, 1992), pp. 167-184. 33 Ch. Hannick, "Les nouvelles chrétientés du monde byzantin: Russes, Bulgares et Serbes", dans Evêques, moines et empereurs 610 1045, Histoire du christianisme des origines à nos jours 4 (Paris, 1993), pp. 909-939. 34 Cf. Ch. Hannick, "Die byzantinischen Missionen", Die Kirche des früheren Mittelalters I, ed. K. Schäferdieck, Kirchengeschichte als Missionsgeschichte 2 (München, 1978), p. 282. 35 Cf. C. Jirecek, L'erédita di Roma nelle città délia Dalmaiia durante il Medioevo I III, Atti e memorie delta Società dalmata di storia patria 9~11 (Rome, 1984/1986).

1'énig1T1atique centre ecclésiastique de Nin, si important dans l'histoire religieuse des Slaves, non seulement des côtes dalmates mais aussi de PIllyricum occidental. 3 6 Et c'est aussi dans ce cadre ou plutôt sur cet arrière-fond qu'il faut considérer le problème philologique et linguistique des relations entre certains textes religieux cyrilliques et glagolitiques, traduits du grec ou du latin, et transmis dans des manuscrits du l i e au 13e siècle. J e fais allusion ici à la passion des q u a r a n t e martyrs de Sébaste conservée dans l'homiliaire vieux-slave Suprasliensis du 11 e siècle ainsi que dans des fragments glagolitiques d ' u n Passionale du 13e siècle et dont j'ai essayé de retracer l'histoire textuelle dans m o n étude sur M a x i m o s Holobolos et le développement de la littérature homilétique slavonne. 37 Pas plus que dans le cas des Slaves dans le Sklaveniai du Péloponnèse, où à l'époque d'Herakleios en Dalmatie, la "conversion" des Bulgares dans les années 60 du 9e siècle sous le prince Boris est reliée à un effort de se servir de la langue slave c o m m e véhicule de la p r o p a g a t i o n de la foi chrétienne. T o u t cela s'est fait sans quitter le grec ou le latin. Rappelons que la diffusion du latin dans la Basse Antiquité, et de là sans doute j u s q u ' à l'installation des Slaves dans les Balkans, n'était pas confinée aux zones côtières de l'Adriatique, c o m m e le d é m o n t r e n t les inscriptions grecques et latinés sur le territoire bulgare actuel éditées p a r Georgi Michailov. 3 8 Il faut mentionner la couche aristocratique protobulgare, donc d'originc altaïque, en Bulgarie, une couche d é m o g r a p h i q u e assez restreinte dont la langue n'est connue que par quelques dtres de fonctions civiles ou militaires et p a r des éléments onomastiques dans les célèbres inscriptions funéraires protobulgares rédigées en grec. 39 Cette langue qui ne disposait pas d'alphabet propre ne jouait aucun rôle officiel dans l'Etat bulgare, dont les débuts sont fixés d'après les premiers traités entre la Bulgarie et Byzance dans le dernier quart du 7e siècle sous le chan Asparuch et l'empereur byzantin Constantin

3(1

Vlasto, The entry, pp. 194 sq. Ch. Hannick, Maximos Holoboros in der kirchenslavuchen homiletischen Literatur, Wiener byzantinùtische Studien 14 (Vienne, 1981), p. 194. 38 Cf. un aspect de cette question dans G. Michailov, "Aspects de l'onomastique dans les inscriptions anciennes de Thessalonique", Ή Θεσσαλονίκη μεταξύ ανατολής καί δύσεως (Thessaloniki, 1982), pp. 69 84. 39 C o n s u l t e r à ce sujet l ' a m p l e bibliographie de l'article de VV. S w o b o d a , "Protobulgarzy", Slownik staroiitnoki stowiaùskich IV (Wroclaw, 1970), pp. 366-368. 37

IV. O n peut done négliger cette composante dans les considérations qui suivent. 40 Revenons à l'époque de Boris vers 860, qui est aussi l'époque de Photius et de Cyrille et M é t h o d e , les apôtres des Slaves. C o m m e je l'ai déjà souligné dans plusieurs travaux, la "conversion" de la Bulgarie sous Boris, qui est bien plus l'élévation du christianisme au rang de religion officielle de l'Etat bulgare, doit être considérée dans le cadre de l'histoire ecclésiastique de l'lllyricum, occidental et oriental. O r dans le territoire de la Bulgarie d'alors il y avait déjà n o m b r e d'évêchés relevant du patriarcat de Constantinople, aussi bien dans la partie orientale qu'occidentale du pays, après le coup de force de l'empereur Léon III en 7 3 2 / 7 3 3 qui arrache l'lllyricum occidental, et ainsi donc le vicariat de Thcssalonique, à la juridiction romaine. 4 1 Photius fait d'ailleurs lui aussi état dans sa correspondance avec les pontifes romains de cet acte de l'empereur qu'il ne peut approuver au plan canonique et qui d u r e r a de facto j u s q u ' e n 923. 42 Les évêchés sur le territoire bulgare à l'époque de Boris étaient bien sûr de langue grecque et ceci non seulement pour la population grecque ou hellénisée des côtes de la M e r Noire. Dans la Iongue épître de Phodus à Boris de l'année 865/866, 4 3 il n'est aucunement question de problèmes linguistiques, inexistants, puisque le grec était la langue ecclésiastique de la Bulgarie. O n relie volontiers la personne de Boris à l'activité des apôtres des Slaves Cyrille et M é t h o d e , en soulignant que Boris accueillit les disciples de M é t h o d e après la mort de celui-ci en 885. Soulignons d ' a b o r d q u ' u n e éventuelle rencontre entre M é t h o d e et Boris relève de la légende. Par ailleurs on s'étonne q u ' u n souverain d ' u n pays à 40

Cf. K.H. Menges, "Die Funktion der türkischen Sprache im Balkanraum und ihre mögliche Rolle bei der Formierung der heutigen slavischen Nationalsprachen", Sprachen und Nationen im Balkanraum. Die historischen Bedingungen da Entstehung der heutigen Nationalsprachen, ed. Ch. Hannick, Slavistische Forschungen 56 (Cologne, 1987), p. 28. 41 Cf. P. Schreiner, "Das Christentum in Bulgarien vor 864", Das Christentum in Bulgarien und auf der übrigen Balkanhalbinsel in der Spätantike und im frühen Mittelalter, eds. V. Gjuzelev R. Pillinger, Miscellanea bulgarica 5 (Vienne, 1987), pp. 51 61; H.G. Döpmann, "Bulgarien als Treffpunkt von östlichem und westlichem Christentum in frühbyzantinischer Zeit", Spätantike und frühbyzantinische Kultur Bulgariens zwischen Orient und Okzident, ed. R. Pillinger (Vienne, 1986), pp. 57 61. 42 Hannick, La christianisation, pp. 925 sq. 43 D. Stratoudaki-White J.F. Berrigan, The patriarch and the prince. The letter of Patriarch Photios to Khan Boris of Bulgaria (Brooklyn/Massachusetts, 1982); Hannick, IM christianisation, p. 926.

forte population slave, c o m m e la Bulgarie l'était, éleva le christianisme au rang de religion officielle justement dans les mêmes années où Constantin-Cyrille et M é t h o d e créaient un alphabet slave et diffusaient ainsi la religion chrétienne en traduisant l'évangile en vieuxslave, et ne se soucia a u c u n e m e n t de se servir de ce véhicule pour atteindre l'un de ses buts, l'émancipation de la Bulgarie du siège patriarcal de Constantinople. Q u a n t à l'accueil que Boris réserva aux disciples de M é t h o d e en 886, la Vie grecque de Clément d ' O h r i d par l'archevêque Théophylacte d ' O h r i d est ici une source de premier choix qu'il faut lire avec grande attention. 4 4 Il ne s'agit pas, c o m m e veulent le suggérer des représentations pictographiques bulgares plus patriotiques encore qu'historicisantes, d ' u n bataillon de moines animés d ' u n zèle apostolique invincible, mais de trois personnes, plutôt accablées p a r les soucis et la maladie (§ X V 43: Milev 116). L'un, Angelarij, meurt quelques jours ou semaines après son arrivée à Pliska (§ X V I 51: Milev 122). Les deux autres, Clément et N a u m , sont loin de pouvoir créer une académie ecclésiastique dans la capitale bulgare, mais sont relégués en logement dans une maison particulière, par ailleurs chez des gens probablement aristocrates de souche protobulgare c o m m e le trahit le n o m de Eschatzes = Eschac, tandis que Angelarij s'était établi chez Ceslav {ibid.). Peu après, Boris enjoignit Clément et N a u m à quitter la capitale bulgare et les envoya dans les confins occidentaux de son pays, dans la région de Kutmicevica (§ X V I I 53: Milev 124), dont le site précis reste énigmatique. 4 5 Avant de passer à l'activité linguistique de Cyrille et M é t h o d e , cherchons à préciser les données dans le cadre desquelles un alphabet slave fut créé. U n des adhérents du m o u v e m e n t méthodien en Bulgarie dans le dernier quart du 9e siècle, ou mieux au début du 10e siècle, le moine C h r a b r , c'est-à-dire la personne qui se cache sous ce pseudonyme, rappelle dans son traité sur l'alphabet slave qu'avant sa création p a r Cyrille on se servait parfois des lettres grecques complétées p a r quelques traits p o u r rendre les sons propres au slave. 46 L'auteur de la Vie de Constantin-Cyrille rapporte aussi que Cyrille était conscient du fait q u ' a v a n t lui plusieurs avaient essayé 44

A. Milev, Grückite zitija na Kliment Ochridski (Sofia, 1966), BHG 355. Cf. T. Sübev, Samostojna narodnostna cürkva υ srednovekovna Bülgarija (Sofia, 1987), p. 178 (à l'Ouest du lac d'Ochrid); A. Delikari, Der Hl Klemens und die Frage des Bistums von Velitza (Thessalonike, 1997). 4(1 K.M. Kuev, Cemozizec Chrabür (Sofia, 1967), p. 188. 45

sans succès de créer un alphabet p o u r la langue slave (VC XIV). Ces essais sont à mettre en relation avec la pré-cyrillique, une écriture mal organisée et incomplète qui n'eut guère de diffusion, dont rien n'est conservé mais qui explique le succès postérieur de la vraie cyrillique. 47 C'est m a i n t e n a n t un fait indéniable, reconnu presque par tous, que l'alphabet inventé p a r Cyrille n'est pas celui qui porte son n o m mais bien la glagolite, remplacée très tôt, dès la fin du 9e siècle, p a r la vraie cyrillique, d ' a b o r d en Bulgarie orientale, vraisemblablement sur la base de la pré-cyrillique. Ce fait d é m o n t r e p a r ailleurs l'honneur dont Boris entoura les disciples de Méthode! Laissons de côté la question épineuse du cadre géographique dans lequel Constantin-Cyrille j u s q u ' à sa mort en 869 et M é t h o d e jusqu'en 885, mis à part leurs n o m b r e u x voyages, n o t a m m e n t à R o m e , ont déployé leur activité. De nombreux arguments, malgré maintes contradictions, portent à croire que cette région, la future éparchie de M é t h o d e , avait son centre dans l'lllyricum occidental, donc au sud du D a n u b e . Qpoi qu'il en soit, cette activité n'eut pas de succès durable dans le territoire primitif où elle fut déployée. Par contre elle fut continuée en Bulgarie et, plus tard, en Russie kiévienne ainsi que dans les régions habitées p a r les Serbes et les Croates. En Bulgarie du 9e et 10e siècle, c'est-à-dire dans la région pour laquelle on dispose de plus de données historiques et linguistiques assurées, il est évident que le slave c o m m e langue ecclésiastique ne jouait q u ' u n rôle secondaire. O n a déjà noté que la conversion de Boris vers 863 se fit dans un cadre linguistique exclusivement grec, en prolongement d ' u n e tradition chrétienne multiséculaire, elle aussi grecque. Par ailleurs, si l'on retient c o m m e digne de foi la lettre de Rasdslav à l'empereur Michel III (VC X I V ) qui m a r q u e les débuts de l'activité de Cyrille et M é t h o d e et en d o n n e les fondements, il n'est pas question en " M o r a v i e " de christianisation mais d'approfondissement de la foi chrétienne. "Alors que notre peuple a rejeté le paganisme et s'attache à la foi chrétienne, nous n'avons pas de maître capable d'expliquer dans notre p r o p r e langue la vraie foi chrétienne" précise l'auteur de la Vita Constantini XIV. 4 8 Certains critiques 49

47 Cf. P. Schreiner, "Überlegungen zur Verbreitung der griechischen Sprache und der Entstehung der griechischen Schrift im bulgarischen Raum", id., Studia byzantino-bulgaHca, Miscelanea bulgarica 2 (Vienne, 1986), pp. 1 1 7 . 1 2 8 ‫־‬ 48 Traduction de A. Vaillant, Textes vieux-slaves II. (Paris, 1968), p. 19. 49 Cf. J . Vašica, L1teránú pamâlky epochy velkomoravské 863-885 (Prague, 1966), p. 222.

ont voulu reconnaître dans les termes "dans notre langue" une interpolation tardive et veulent réduire ainsi ! , activité de Cyrille et Méthode à un niveau p u r e m e n t canonique et juridique. Q u o i qu'il en soit, ni en Moravie ni en Bulgarie, la christianisation ne s'est faite p a r le truchement de la langue slave et d ' u n alphabet slave. Quels sont d'ailleurs les textes traduits en slave et leur niveau au plan culturel!50 O n notera ici une différence fondamentale avec l'Arménie. A côté des textes bibliques, et encore, il n'est question ici que d ' u n e partie du canon des livres bibliques et non de la Bible entière c o m m e en Arménie, on a en vieux-slave entre autre un fragment d ' h o m é liaire, le Suprasliensis cyrillique 51 et le Clozianus glagolitique qui, à l'époque du vieux-slave, donc j u s q u ' à la fin du 1 le siècle, n'ont guère servi de modèle p o u r des compositions propres. Bien au contraire, les homélies de Clément d ' O h r i d ou de J e a n l'Exarque dénotent un caractère littéraire de loin inférieur aux textes byzantins et patristiques passés en slave dans le Suprasliensis. O n doit donc en tirer la conclusion q u ' e n Bulgarie du 9e et 10e siècle il y avait deux couches de population slave, l'une cultivée et apte à lire et c o m p r e n d r e le grec littéraire, sinon hellénisée ellem ê m e c o m m e c'était le cas de Boris et plus encore du T s a r Syméon, l'autre de culture plus modeste et devant avoir recours à la catéchèse en langue slave et p a r là, semble-t-il, illettrée. Q u a n t au haut clergé il reste grec en Bulgarie, tout c o m m e en Russie kiévienne du 10e-11e au 13e siècle. 52 O n pourrait imaginer que l'émergence de nouvelles langues littéraires, arménien, géorgien, slave, est reliée aux besoins d ' u n e langue ecclésiastique, c'est-à-dire liturgique. D a n s le cas de l'arménien et du géorgien, puisqu'on connaît très mal la structure ecclésiastique de ces pays avant les premiers synodes nationaux, le premier ayant eu lieu à Dvin en 551, on peut retenir cette hypothèse c o m m e digne d'observation. Dans le cas du slave, en Bulgarie du 9e siècle, j'ai déjà souligné plusiers fois, 53 et ceci contrairement à l'opinion d ' u n e 50

Cf. le tableau établi par F.V. Mares, Konstantinovo kultumi dilo po 1100 letech (Prague, 1970), p. 25. 51 Hannick, Maximos Holobolos, pp. 71 sq. 52 Cf. A.E. Tachiaos, " T h e Greek metropolitans of Kievan Rus': an evaluation of their spiritual and cultural activity", Harvard Ukrainian Studies 12/13 (1988-89), pp. 430-445. 53 Cf. Ch. Hannick, "Der Einfluß von Byzanz auf die Entwicklung der bulgarisehen Kirchenmusik", Kulturelle Traditionen in Bulgarien, eds. R. Lauer - P. Schreiner, Abh. der phil. hist. Kl. der Akad. d. Wiss. Göttingen III 177 (Göttingen, 1989), pp. 91-102.

partie de mes collègues bulgares, qu'il n'y a a u c u n e trace du slave c o m m e langue liturgique r é p o n d a n t aux différents besoins avant le début du deuxième empire bulgare, d o n c à la fin du 12e siècle. Langue liturgique veut dire bien plus que la traduction des péricopes bibliques qui occupent certes une place de choix dans l'office divin, mais une place restreinte c o m p a r é e aux nombreuses hymnes entrecoupés de versets psalmiques. J e laisse de côté les compositions h y m n o g r a p h i q u e s d'origine bulgare, attribués à Constantin de Preslav 34 et r e m o n t a n t ainsi au 10e siècle, en rappelant les n o m b r e u ses hymnes grecques du patriarche Photius qui ne trouvèrent a u c u n e place dans les recueils h y m n o g r a p h i q u e s en usage dans les églises byzantines. C e n'est q u ' à la fin du 12e siècle q u ' o n a en Bulgarie des recueils d ' h y m n e s liturgiques destinés à la pratique de l'office divin, tandis q u ' e n Russie kiévienne où la culture g r e c q u e était i n c o n n u e — s a u f dans les villes grecques de C r i m é e qui ne faisaient pas partie de la principauté de Kiev—la tradition textuelle commence en ce d o m a i n e dès la fin du l i e siècle, 55 sans doute p a r le t r u c h e m e n t de lettrés d'origine bulgare. En Géorgie, le cas est plus intéressant encore, si l'on p r e n d en considération que les usages liturgiques géorgiens les plus anciens d é p e n d e n t des coutumes hiérosolymitaines. Les quelques manuscrits h y m n o g r a p h i q u e s munis de n e u m e s musicaux, de forme et de fonction i n c o n n u e s p a r ailleurs, p r é s e n t e n t u n e caractéristique digne d'attention. La division des hymnes en kola et la place des n e u m e s principaux correspond à la tradition grecque, 5 6 de sorte qu'il est évident que le chant des hymnes en géorgien, à l'époque ancienne, attesté d'après quelques manuscrits du 10e 1 l e siècle, reproduit une tradition byzantine, celle de J é r u s a l e m , avant les réformes liturgiques et musicales venues de la capitale vers le milieu du l i e siècle. Avant de quitter la poésie liturgique byzantine, et p o u r en souligner l'importance ainsi que p o u r en m o n t r e r le niveau stylistique et

Pour la Russie kiévienne cf. V. Vodoff, "Quelques questions sur la langue liturgique à Kiev au X'' et au début du XI'' siècle", The legacy of saints Cyúl and Methodius to Kiew and Moscow, ed. A.E. Tachiaos (Thessalonike, 1992), pp. 435-448. 54 G. Popov, Triodm proizvedenija na Konstantin Preslavski, Kirilo-Metodievski studii 2 (Sofia, 1985). (cf. I. Dobrev, Palaeobulgarica 11 (1987.2), pp. 122-128). 55 C11. Hannick, "Early slavic liturgical hymns in musicological context", Ricerche Slamstiche 41 (1994), pp. 9 - 3 0 . 56 Un exemple est analysé par Ch. Hannick, "Das musikalische Leben in der Frühzeit Bulgariens aufgrund literarischer Quellen des frühslavischen Schrifttums", Byzantinoslavica 49 (1988), p. 31.

linguistique, soulignons que les textes festifs principaux, traduits aussi bien en géorgien q u ' e n slave, ont été c o m m e n t é s en grec par Ioannes Zonaras et T h e o d o r o s P r o d r o m o s au 12e siècle. Ces importants commentaires, conservés en grec dans des centaines de manuscrits—rien q u ' à Vienne j e connais une douzaine de témoins des deux recensions du c o m m e n t a i r e de Z o n a r a s (par exemple Vind. theol. gr. 143, 151, 251, 238, 310)—n'ont j a m a i s été traduits et restent encore en g r a n d e partie inédits. Et si l'on peut considérer que les Kontakia de R o m a n o s le M é l o d e pouvaient être compris à l'audition, vu que le chant en rehaussait les éléments syntaxiques, ceci n'est pas le cas des n o m b r e u x canons et stichera, de là l'importance et la nécessité parfois des c o m m e n t a i r e s ( Έ ρ μ υ ν ε ί α ι ) , surtout p o u r les quelques canons festifs rédigés en trimètres ïambiques. O u t r e la Bible et la liturgie, c'est dans le c h a m p de la patristique que l'œuvre de traduction initiée p a r la création de nouveaux alphabets a atteint son apogée. T a n d i s q u ' e n Arménie et en Géorgie, dès la fin du 5e siècle et surtout au 6e siècle—en milieux géorgiens plus tard aussi—on traduit des œuvres patristiques grecques complètes c o m m e l'Asceticon Maius de Basile de Césarée, publié r é c e m m e n t p a r G. Uluhogian (CSCO), chez les Slaves rien de semblable ne se produit avant le 14e siècle, donc environ un demi millénaire après la création de l'alphabet slave à l'époque de Cyrille et M é t h o d e , mis à p a r t quelques textes c o m m e p a r e x e m p l e u n e œ u v r e de J e a n D a m a s c è n e "Εκθεσις πίστεως dont une première traduction remonte à l'époque paléobulgare, d o n c au 1 0 e ‫ ־‬l l e siècle. 57 La question de l'origine bulgare ou kiévienne de nombreuses œuvres patristiques en traduction slave reste fort discutée. 58 Il aurait pu en être a u t r e m e n t , car les besoins étaient les mêmes. U n auteur bulgare du 14e siècle, Grigorij C a m b l a k , d o n n e d a n s l'éloge de son m a î t r e , le dernier patriarche bulgare, E u t h y m e de T ù r n o v o , la raison de ce m a n q u e culturel évident: les défauts manifestes de la langue slave ecclésiastique en comparaison avec le grec. 59 Le linguiste actuel ne p o u r r a 57 Des heiligen Johannes von Damaskus "Εκθεσις άκειβύς της ορθοδόξου πίστεως in der Übersetzung des Exarchen Johannes, ed. L. Sadnik, Monumenta linguae slavicae dialecti veteris 5, 14, 16, 17 (Wiesbaden, 1967, Freiburg, 1981-1983). 58 Cf. par exemple F. Thomson, " T h e implications of the absence of quotations of untranslated greek works in original early russian littérature, together with a critique of a distructed picture of early bulgarian culture", Slavia Gandensia 15 (1988), pp. 63-91. 59 P. Rusev - K. Gülübov - A. Davidov - G. Dancev, Pochvalno slovo za Evtimij ot Grigorij Camblak (Sofia, 1971), p. 166 (§ 29).

souscrire sans réserve à cette assertion, car le vieux-slave ecclésiastique aurait été apte à remplir cette fonction, tout c o m m e l'arménien ou le géorgien. Il m a n q u a i t plutôt les traducteurs et il m a n q u a i t surtout le public p o u r qui ces textes auraient été traduits. Au 14e siècle, en Bulgarie, apparaissent les versions slaves de YHexaémeron de Georgios Pisides, 60 également c o n n u en arménien, de l'Echelle de J e a n C l i m a q u e etc. D a n s ce dernier cas, le manuscrit slave méridional le plus ancien r e m o n t e à l'année 1364 et a été copié p a r le moine T h é o d o s e près de T ü r n o v o ; il a dû toutefois exister une traduction plus ancienne passée en Russie à la fin du 12e siècle. 61 Q p a n t aux textes de l'Antiquité classique, n o n chrétiens, les Slaves n'en ont eu a u c u n e connaissance, si ce n'est les n o m s des différents auteurs chrétiens dans les florilèges, tandis q u ' u n philosophe arménien c o m m e David l'Invincible traduisit au 6e siècle Porphyrios et d'autres traités néoplatoniciens, connus en Géorgie grâce à l'activité de Arsène d'Iqalto au 12e siècle seulement. Ces quelques remarques auront mis en exergue les conditions différentes de la réception de la culture à la fin de l'Antiquité dans trois domaines linguistiques différents, Arménie, Géorgie, ainsi q u e — a v e c un décalage évident—chez les Slaves du Sud, réception mise en rapport avec la création de nouveaux alphabets dans des milieux linguistiques plurilingues et pluriculturels.

60 Cf. N. Radosevic, Sestodnev Georgija Piside i njegov slovenski prevod (Beograd, 1979). Sur les conditions de ce travail de traduction cf. V. Gjuzelev, Ucilišla, skriptom, biblioteki, znanija ν Būlgarija XIII XIV vek (Sofia, 1975). 61 B. Christova - D. Karadzova - A. Ikonomova, Būlgarski rükopisi ot XI do XVIII vek zapazeni ν Bülgarija. Svoden katalog I, (Sofia, 1982), p. 40 (No. 66); Svodnjj katalog slavjano-russkich rukopisnych knig, chranjašcichsja ν SSSR XI XIII vv, (Moscou, 1984), p. 104 (No. 62).

C O N F L I T S ET C O M P R O M I S LANGAGIERS EN O C C I D E N T L A T I N : D E LA C R I S E C U L T U R E L L E À L ' I N V E N T I O N L I N G U I S T I Q U E ( I I I e - X e SIÈCLE) Michcl Banniard

1. 2. 3. 4. 5.

Histoire culturelle et histoire langagière. Conflits et compromis (IIIe—VIIe siècle). Crises et inventions (VIII e —X e s.). Interférences culturelles et langagières. V u e schématique des phénomènes.

1. Histoire culturelle et histoire langagière a. Les transformations du m o n d e romain en Occident ont abouti, entre autres n o m b r e u x changements qui n'excluent pas la continuité, à m é t a m o r p h o s e r une c o m m u n a u t é latinophone en une pluralité r o m a n o p h o n e . ' C e processus c o n c e r n e l ' e n s e m b l e des pays de l'Occident romain: devenus latinophones au plus tard sous le H a u t Empire, ils sont devenus r o m a n o p h o n e s au plus tard à l'époque carolingienne. Autrement dit, en moins d ' u n millénaire se sont succédées deux révolutions culturelles et langagières. A la première, engagée avec la conquête du bassin occidental de la Méditérranée, a corresp o n d u l'acquisition, d ' a b o r d c o m m e langue seconde, puis c o m m e langue maternelle, du latin, langue des vainqueurs. A la seconde, engagée avec le d é m e m b r e m e n t de l'Empire, a correspondu la disparition du latin c o m m e langue c o m m u n e et son remplacement p a r des langues régionalisées. 2 C e remplacement s'est accompli au prix d ' u n déboîtage au moins double: typologique, car les traits princip a u x qui permettent de décrire la langue latine ont été largement remplacés p a r des traits neufs, suffisamment nombreux, pour que 1

S.A. D'Arco, Protostoria dette lingue romande (Turin, 1965). Id., Bassa latinilà. Il latino trà l'età tardo-antica e l'altomedioevo con particolare riguardo all'origine delle lingue romanze (3) (Turin, 1979). 2 G. Reichenkron, Historische Latein-Altromanische Grammatik, I Teil: Das sogennante Vulgärlatein und das Wesen der Romanisierung (Wiesbaden, 1965).

l'identité de la langue ait changé à travers le temps; géographique, car, outre cette évolution c o m m u n e , il s'est produit des évolutions régionales, d'où ont émergé des entités langagières distinctes les unes des autres à travers l'espace. 3 C'est de ce second aspect, le remplacement, que j e désirerais traiter a u j o u r d ' h u i , à la lumière des travaux menés depuis une trentaine d'années. b. Le problème du passage du latin aux langues romanes relève à la fois de l'histoire au sens propre, de l'histoire culturelle et de l'histoire langagière. Il est, en effet, étroitement lié à la représentation que les chercheurs se donnent de la période considérée, mais aussi aux modèles culturels et linguistiques (implicites ou explicites) qui président à la description des faits invoqués. J e laisserai naturellem e n t de côté ici le travail de renouvellement considérable qui, en histoire pure, si j'ose dire, conduit lentement mais sûrement à la reconsidération des III C -VT siècles. L'histoire culturelle n o t a m m e n t ne s'écrit plus de la m ê m e manière depuis q u ' a émergé et a été accepté le concept d'Antiquité Tardive, qui est venu remplacer fort heureusement le concept de T e m p s de la décadence, voire d'Age barbare, et j ' e n passe . . . Du coup, le premier Moyen-Age, lui-même, bénéficie de relectures plus positives: mérovingiens et carolingiens p r e n n e n t une place sans doute plus juste dans le c h a m p historique. 4 A l'aune de ce recalage cognitif, l'histoire de la culture des siècles de transition fait l'objet d'interprétations rajeunies, souvent plus constructives que par le passé: 3 la littérature latine chrétienne avec ses grands—et ses petits—auteurs a cessé d'être jugée avec p o u r unique mesure les (supposés) canons de la perfection classique. 6 3 J . Herman, "Aspects de la différenciation territoriale du latin sous l'Empire", Bulletin de la Société de Linguistique de Paris 60 (1965), pp. 53-70. Id., "La différenciation territoriale du latin et la formation des langues romanes", Actes du XVII' congrès international de linguistique et de philologie romanes 2 (Marseille, 1982), pp. 15-62. 4 La Neustrie, Les pays au Nord de la Loire de 650 à 850, Beihefte der Francia, 2 vol., ed. H. Atsma (Sigmaringen, 1989): R. McKitterick, The Prankish Church and the Carolingian Reforms (Londres, 1977). 5 H. Atsma, "Le fonds des chartes mérovingiennes de Saint-Denis, Rapport sur une recherche en cours", Le haut Moyen Age en Ile-de-France, Mémoires de la fed. des soc. hist, et arch, de Paris et de l'Ile-de-France 32 (1981), pp. 259272‫־‬: Chartae iMtinae Antiquiores, Facsimile-edition of the Im tin Charters Prior to the Ninth Century (Corpus des actes privés originaux du HMA conservés en France), ed. H. Atsma et J . Vezin, série 13 des CIA (Zurich, 1981 sqq.): Karl der Grosse, Lebenswerk und Nachleben, t. 2, Das geistige Leben, ed. W. Braunfels (Dusseldorf, 1965): R. McKitterick, The Carolinians and the Written Word (Cambridge, 1989): M. Richter, The formation of the Medieval West. Studies in the oral culture of the barbarians (Dublin, 1994). 6 J . Fontaine, Isidore de Séville et la culture classique dans l'Espagne wisigothique (2), 3 vol.

c. D a n s le domaine p r o p r e m e n t langagier, le changement de point de vue, amorcé depuis plus d ' u n quart de siècle, s'impose avec une rapidité variable en fonction des spécialistes qui y travaillent. Les questions, diverses et difficiles, peuvent se concentrer en la formule suivante: q u a n d le latin a-t-il cessé d'être une langue vivante?' Sous cette interrogation en apparence un peu naïve se pressent des énigmes: dans quelle mesure le latin était-il devenu dès le H a u t Empire la langue spontanée de tous les locuteurs? Quel était le niveau de latinité dans la R o m a n i a du IV e siècle? C o m m e n t rendre compte du bouleversement langagier que constitue l'apparition des langues romanes? Fut-il précoce? Général? Différencié dès l'origine par régions? O n entre alors de plain pied dans le domaine a p p a r e m m e n t neutre, mais en fait très controversé, de la linguistique diachronique." J e ne souhaite pas faire l'état de ces lieux philologiques ici: soulignons seulement l'extrême contradiction des théories, qui vont d ' u n e dissolution de la latinophonic dès le IV'' siècle (au moins) à la conviction d ' u n e longue survivance j u s q u ' a u IX e siècle (au plus), sinon au-delà. d. Afin de contourner ce m u r de l'incertitude, un certain n o m b r e de chercheurs, surtout européens (mais les précurseurs furent américains; 9 il y eut un seul travail pionnier sur le vieux continent), 10 ont entrepris de mettre au point des méthodes d'enquête diachronique qui ont consisté à interroger les contemporains sur leur propre

(Paris, 1983): Id., Isidore de Seville, Traité de la nature (Bordeaux, 1960), pp. 85 139: Vita sancti Martini, éd. J . Fontaine, t. 1 et 2, Paris, 1967, t. 1 (Introduction) et t. 2 (Commentaire), pp. 359-393: J . Fontaine, "De la pluralité à l'unité dans le 'latin carolingien'", Settiniane di Studi 27 (Spolète, 1981), pp. 765-818: id., "La naissance difficile d'une latinité médiévale: mutations, étapes et pistes", Bulletin de l'Association Guillaume Budé 40 (1981), pp. 360 368. 7 F. Lot, "A quelle date a-t-on cessé de parler latin en Gaule?", Archivum Latinitatis Medii Aevi 6 (1931), pp. 97 159: H.F. Muller, "When did Latin cease to be a Spoken Language in France?", The Romanic Review 12 (1921), pp. 318-334: D. Norberg, "A quelle époque a-t-on cessé de parler latin en Gaule?", Annales ESC 21 (1966), pp. 346-356: M. Richter, "A quelle date a-t-on cessé de parler latin? A propos d'une question mal posée", Annales ESC 38 (1983), pp. 4 3 9 4 4 8 ‫ ־‬: M. Van Uytfanghe, "Les expressions du type quod uulgo uoeant dans les textes latins antérieurs au Concile de Tours et aux serments de Strasbourg: témoignages lexicologiques et linguistiques de la 'langue rustique romaine'?", Zeitschrift fur romanische Philologie. 195 (1989), pp. 28-49. 8 M. Van Uytfanghe, "Le latin des hagiographes mérovingiens et la protohistoire du français", Romanica Gandensia 16 (1976), pp. 5-89. 9 H.F. Muller, A Chronology of Vulgar ÎMtin (Halle, 1929): ibid., L'époque mérovingienne. Essai de synthèse de philologie et d'histoire (New York, 1945). 10 Ph. Wolff, Les origines linguistiques de l'Europe occidentale (2) (Toulouse, 1982).

sentiment langagier." Ces méthodes font appel aux procédures de l'histoire culturelle, religieuse et littéraire, de manière classique, mais aussi aux acquis de la dialectologie, de la dialectologie sociale, et de l'ethnographie. D e ce cocktail un peu explosif est sortie une nouvelle recette que j'appellerai la sociolinguistique rétrospective. 12 Fondée principalement sur l'étude de la communication dans les siècles de transition, elle a abouti à un certain n o m b r e de résultats que je pense opportun de proposer. Cela d o n n e r a lieu à la description de deux périodes principales, avant de tenter quelques propositions plus prop r e m e n t langagières mises au service d ' u n e théorie linguistique corrélée à ce modèle historique.

2. Conflits et compromis (IIIe-VIIe

siècles)

a. L'histoire de la latinité a reposé sur une longue suite de conflits et de compromis langagiers. C e caractère, vrai p o u r toute l'histoire de la latinité classique, s'accentue p o u r celle de la latinité tardive. J e rappellerai très brièvement ici que la littérature latine s'est bâtie sur le latin parlé classique, revu et corrigé par les créateurs littéraires: auteurs comiques, poètes lyriques, poètes épiques, historiens, orateurs . . . D e Plaute à César, l'écriture latine a impliqué un travail intense sur la langue parlée, puis écrite, grâce auquel s'est élaborée une langue littéraire protéiforme, dont la caractéristique première est

11

K. Heene, "Audire, legere, uulgo: an Attempt to define Public Use and Comprehensibility of Carolingian Hagiography", latin and the Romance Languages in the Early Middle Ages, ed. R. Wright (Londres, 1991), pp. 146-163: J . Herman, Le latin vulgaire (3) (Paris, 1975): M. Richter, "Kommunicationsprobleme im lateinisehen Mittelalter", Historische Zeitschrift 222 (1976), pp. 43-80: Id., "Towards a Methodology of Historical Sociolinguistics", Folia Linguistica Historica, t. 6 / 1 (1985), pp. 41-61: M. Van Uytfanghe, "Histoire du latin, protohistoire des langues romanés et histoire de la communication", Francia 11 (1984), pp. 579-613: Id., "L'hagiographie et son public à l'époque mérovingienne", Studia patristica 16/2 (1985), pp. 5 2 - 6 2 : Id., " T h e Consciousness of a Linguistic Dichotomy (Latin-Romance) in Carolingian Gaul: the Contradictions of the Sources and of their Interpretation", Latin and the Romance Languages in the Early Middle Ages, ed. Wright, pp. 114 129. 12 M. Banniard, Viva voce: Communication écrite et communication orale du IV' au IX' siècle en Occident Latin, Etudes Augustiniennes (Paris, 1992): W. Labov, Sociolinguistique (Paris, 1976): M. Richter, "Die Sprachenpolitik Karls des Grossen", Sprachwissenschaft 7 (1982), pp. 412-437: R. Wright, Late Latin and Early Romance in Spain and Carolingian France (Liverpool, 1982): Id., " T h e Conceptual Distinction between Latin and Romance: Invention or Evolution?", latin and the Romance Languages in the Early Middle Ages, ed. Wright, pp. 103-113.

de pardeiper à une n o r m e langagière. Cette n o r m e répond aux critères de romanitas (ou de latinilas), elle-même bâtie sur les trois exigences de Yurbanitas, de Yelegantia et de la propnetas. Elle implique de manière p e r m a n e n t e une distanciation plus ou moins considérable avec la langue spontanée et ses variations multiples, autrement dit, elle repose sur la réélaboration grammaticale et stylistique du par1er quotidien: à la fois identité et distance. La seule limite formelle qui soit imposée à la puissance créatrice des auteurs est le respect du decorum: cette adaptation au sujet, au genre, aux circonstances et au public a pour corollaire l'exigence de l'intelligibilité q u a n d il s'agit de matières relevant d ' u n e communication: histoire, philosophie, et surtout art oratoire. Naturellement cette modulation langagière ne se fonde que sur une relation sociale restreinte: elle d e m e u r e largem e n t libre des exigences d ' u n e communication large. b. Ce n'est plus le cas avec l'arrivée du christianisme: 13 il n'est plus question de limiter de manière arbitraire l'extension du message chrétien. U n double m o u v e m e n t p r o v o q u e une révolution langagière dont les effets se font nettement sentir en Occident Latin à partir du III e siècle: d ' u n e part, la langue écrite des premiers m o n u m e n t s chrétiens, fortement m a r q u é e parce que ces derniers sont le résultats de traductions, la plupart du temps du grec de la koinê, diverge considérablement de la tradition littéraire romaine; d'autre part, les l o c u t e u r s / a u d i t e u r s a u x q u e l s est destiné l ' e n s e i g n e m e n t chrétien n'appartiennent pas à l'élite n o r m é e et disciplinée des latinophones. O n connaît le long débat qui s'est engagé entre la culture et la langue romaines traditionnelles et la culture inouïe et la langue changée qu'apportait avec elle la religion nouvelle, ainsi que son résultat. La langue de l'Ecriture, porteuse d u renouvellement chrétien et de l'héritage j u d a ï q u e a fini p a r conquérir une place d ' h o n n e u r aux côtés de la langue de Virgile. 14 D ' u n autre côté, Y usus (la "pratique") 13

E. Auerbach, Litemiy language and its Public in IMe Latin Antiquity and in the Middle Ages (Londres, 1965): A. Borst, Der Turmbau von Babel, Geschichte der Meinungen über Ursprung und Vielfalt da Sprachen und Völker, t. 1 2 (Stuttgart, 1957, 1958): A. Cameron, Christianity and the Rhetoric of Empire (Berkeley/Oxford), 1991: C. Mohrmann, Etudes sur le latin des chrétiens, 4 vol. (Rome, 1965-1977): A. Olivar, IM predicaciôn cristiana antigua (Barcelone, 1991). 14 J . Murphy, Rhetoric in the Middle Ages: a History of Rhetorical Theory from saint Augustine to the Renaissance (Berkeley, 1974): M. Richter, "IMina lingua-sacra seu uulgaris?", The Bible and Medieval culture, eds. W. Lourdaux et D. Verfielst (Louvain, 1979), pp. 16-34.

élitaire qui bornait et modelait la langue de la communication a cédé devant les exigences neuves de l'instruction massive: le sermopiscatonus (la "langue du pêcheur") accède ainsi à la dignité au sein de la latinophonie; il devient un outil langagier accepté, a d a p t é , remodelé aux III e et IV1' siècles. C'est ainsi q u ' à tous les niveaux de la culture et de la communication, l'irruption du christianisme et l'expansion des missions chrétiennes provoquent un bouleversement profond de la latinité. 15 La communication orale contrôlée p a r les intellectuels, au lieu de se restreindre à une étroite " b a n d e passante", se déploie désormais sur tous les niveaux de la langue. 1 6 Cela signifie que s'engage un développement exponentiel des interférences entre toutes les formes de l'oralité latine. Le temps des compromis est arrivé. Naturellement, ceux-ci s'établissent p a r actions et p a r réactions: entre le latin des païens et le latin des chrétiens; entre l'oralité soutenue et la parole relâchée; entre les accents urbains et les accents ruraux; entre les messages des litterati ("ceux qui savent au m o i n s lire") et les réactions des illitterati ("ceux qui ne p e u v e n t qu'écouter"); 1 7 entre le passé s'efforçant d'être stable et le présent si mobile. C'est ainsi que naît la latinité tardive en général et le latin parlé tardif en particulier: parlé d ' u n bout à l'autre de l'Empire en Occident, il varie beaucoup, non seulement en diachronie et en synchronie (comme le dirait Jérôme), mais aussi en fonction des locuteurs, des situations d'énoncé, des interrelations sociales . . . 8 ‫ י‬Tel qu'il est, il permet une communication orale rapide et générale, tant verticale qu'horizontale, entre tous les locuteurs latinophones (et ils sont très largement la majorité). La pénétration précoce et rapide du christianisme dans les provinces, puis surtout dans les campagnes parfait sans doute, mais surtout révèle la p r o f o n d e u r et l'ampleur de la latinisation. 1 9 La Romania existe: les prédicateurs l'ont rencontrée (et confortée).

15

L. Lentner, Volkssprache und Sakralsprache. Geschichte einer Lebensfrage bis zum Ende des Konzils von Trient (Vienne, 1963): Ε. Löfstedt, Late latin (Oslo, 1959). 16 Banniard, Viva voce, chap. 2. 17 H. Grundmann, "IJtteratus-Illitteratus, Die Wandlung einer Bildungsnorm vom Altertum zum Mittelalter", Archiv fur Kulturgeschichte 40 (1958), pp. 1-65. IB B. Löfstedt, "Rückschau und Aufblick auf die vulgärlateinischen Forschung", Aufstieg und Niedergang der römischen Welt, t. II, 29, 1 (Berlin, 1982), pp. 453-479: V. Väänänen, Introduction à l'étude du latin vulgaire (2) (Paris, 1967): Id., "Le problème de la diversification du latin", Aufstieg und Niedergang der Römischen Welt, t. II, 29, 2 (Berlin, 1983), pp. 480-506. 19 M. Banniard, "Normes culturelles et réalisme langagier en Lusitanie au VI e

c. Cette structure c o m m u n i c a n t e n'est pas unanimiste. La conversion de la culture antique a u x valeurs chrétiennes, c o m m e l'acceptation progressive d'une partie de l'éducation antique par les intellectuels chrétiens ne se sont pas produites sans déchirements parfois aigus. C e n'est pas sans quelque douleur que les maîtres chrétiens de la latinité tardive ont accepté l'étirement "vertical" de leur oralité de m a n i è r e à a d a p t e r leurs compétences actives de locuteurs cultivés aux compétences passives des auditeurs: cela supposait l'élaboration de codes c o m m u n s qui r e ç u r e n t la d é n o m i n a t i o n a p p r o p r i é e et significative de sermo humilis ("style terre-à-terre"). J é r ô m e a réussi à a d m e t t r e le sermo honidus ("style hirsute") des Ecritures; Augustin a insisté sur la nécessité de privilégier l'intelligibilité sur la grammaticalité. Les centaines de sermons qui ont été conservés de lui, souvent sous la forme originale d ' u n e prise de note au vol p a r quelque tachygraphe, d o n n e n t accès à ce q u ' a été cette latinité qui, bien qu'écrite, représentait la f o r m e immédiate de la langue de c o m m u nication parlée p a r le prédicateur à l'intention de ses fidèles.2" Entre la parole savante, héritée, grammaticalisée, stylisée, restreinte, et la parole innovante, restructurée, p o l y m o r p h e , expansive, s'est instaurée une nébuleuse communicationnclle 2 ' dont l'efficacité p e r d u r e r a au delà de l'Empire. Cela signifie que les c o m p r o m i s langagiers qui ont été passés du III e au V e siècle continuent de fonctionner. 2 2 D e fait, l'émergence de nouveaux genres littéraires c o m m e les Vitae sanctorum permet de comp r e n d r e et de mesurer la réalité et l'efficacité de la c o m m u n i c a t i o n latinophone après le V e siècle. L'essor du culte des saints est associé au fleurissement des Vitae. O r , ces dernières ont une fonction communicative capitale. Modeste ou prestigieuse, la Vie assume des rôles multiples: magnifier parfois certaines des figures de l'aristocratie "converties"; lancer f r é q u e m m e n t des leçons nécessaires à la masse des baptisés; d o n n e r à cette masse l'occasion de vérifier qu'elle a

siècle: Les choix de Martin de Braga", Actes du XIV Centenarì0 del Concilio III de Toledo 589-1989 (Tolède, 1991), pp. 661 676: Id., "La rouille et la lime: Sidoine Apollinaire et la langue classique en Gaule au V e siècle", Mélanges J. Fontaine, éd. L. Holtz (Paris, 1992), t. 1, pp. 413-427. 20 Banniard, Viva voce, chap. 2: M o h r m a n n , Etudes sur le latin des chrétiens, t. 1 - 4 (Rome, 1961 1977). 21 J . Herman, "Spoken and Written Latin in the Last Centuries of the Roman Empire. A Contribution to the Linguistic History of the Western Provinces", iMtin and the Romance Languages in the Early Middle Ages, ed. Wright, pp. 2 9 - 4 3 . 22 Banniard, "Normes culturelles et réalisme langagier en Lusitanie au VI1' siècle".

désormais à sa disposition un médiateur à la fois efficace, mais aussi proche d'elle, entre terre et ciel. C'est pourquoi, la lecture à h a u t e voix de la Vie du saint local contribue puissamment à construire la connivence culturelle qui s'établit entre le modèle proposé et la foule auditrice. 2 ' Aussi les rédacteurs de Vitae font-ils des efforts considérabies p o u r écrire dans un style a d a p t é au public m a j o r i t a i r e m e n t illettré qui est appelé à écouter ces récits. C e style est n o m m é : sermo tenuis, sermo breuis, sermo rusticus sont des apellations souvent justifiées qu'il convient de traduire de m a n i è r e un peu technique p a r "langage de la c o m m u n i c a t i o n générale". 2 4 Il constitue une sous-classe du vaste ensemble qu'est alors le latin tardif, sous-classe dont les caractères (les paramètres) complexes relèvent de descriptions littéraires, historiques et linguistiques. 25 d. Dans ce deuxième temps de la latinité tardive, le centre de gravité des conflits culturels et des c o m p r o m i s langagiers se déplace lentement. La société de l'antiquité tardive change; les élites se renouvellent; leur niveau culturel se modifie, parfois s'appauvrit. 2 6 Il reste toujours aux VI1' et VII e siècles des intellectuels tentés p a r la raideur hiératique des conservateurs (Avit de Vienne, E n n o d e de Pavie, Julien de T o l è d e en sont de bons exemples); mais dans la plupart des cas, les pasteurs chrétiens, si tentés soient-ils de se replier sur les valeurs sûres de la tradition langagière impériale, font la part du feu et respectent les c o m p r o m i s nécessaires. En pleine renaissance wisigothique, le savant Isidore de Séville rappelle que l'orateur ou le rédacteur chrétien doivent observer une juste mesure dans leur travail sur la 23 F. Graus, Volk, Herrscher und Heiliger im Reich der Merowinger (Prague, 1965): M. Van Uytfanghe, Stylisation biblique et condition humaine dans l'hagiographie mérovingienne (600 750) (Bruxelles, 1987). 24 Banniard, Viva voce, ehap. 5: R. Collins, "Beobachtungen zur Form, Sprache und Publikum der Prosabiographien des Venantius Fortunatus in der Hagiographie der römischen Gallien", Zellsc^'ifl fir Kirchengeschichte 92 (1981), pp. 16 .38‫־‬ ‫ '"־‬J.N. Adams, The Text and Language of a Vulgar Latin Chronicle (Anonymus Valesianus II) (Londres, 1976): P. Flobert, Les verbes déponents latins des origines à Charlemagne (Paris, 1975): Id., "La date de YAppendix Probi", Filologia e forme letterarie, Studi ojferti a F. Delta Corte, t. 4 (Urbino, 1987), pp. 299-320: Id., "Le témoignage épigraphique des apices et des I longae sur les quantités vocaliques en latin impérial", Latin vulgaire-Latin tardif II, Actes du II' colloque international (Bologne, 1988) (Tubingen, 1990), pp. 101 110. 26 M. Heinzelmann, "Studia sanctorum. Education, Milieux d'instruction et valeurs éducatives dans l'hagiographie en Gaule jusqu'à la fin de l'époque mérovingienne", Mélanges P. Riche, Haut Moyen Age, Culture, Education et Société (Paris, 1990), pp. 105-138: P. Riché, Education et culture en Occident barbare, VI' VIII' siècle (3) (Paris, 1973): Id., Ecoles et enseignement dans le haut Moyen Age (Paris, 1979).

langue: ni céder aux facilités de l'incorrection systématique, ni succ o m b e r à la fascination du beau style gratifiant pour lui, mais inaccessible à la majorité de ses destinataires.‫ '־‬Grégoire le G r a n d parle dans ses Homélies sur l'Evangile un beau latin clair et mesuré, adapté aux fidèles de la Ville, tout en laissant glisser dans ses récits hagiographiques le phrasé du latin parlé quotidien dans l'Italie des années 600. 28 Citerai-je enfin l'exemple souvent invoqué des déclarations de Grégoire de Tours? Il privilégie lui aussi la parole moins apprêtée (loquentem rusticum) contre la parole trop recherchée (philosophantem rhetorem)?29 C h a n g é sous l'Empire du ΙΙΓ' au V'' siècle, le latin tardif se modifie de nouveau aux VI e , VII e et VIII e siècles. Dans tous les domaines où l'innovation sociale implique un renouvellement langagier, la langue écrite change peu à peu: les textes de loi sont souvent les plus fidèles témoins de la m é t a m o r p h o s e (lois franques, lois gothiques, lois lombardes . . .).3° T o u s les indices culturels et langagiers indiquent q u ' u n e évolution profonde est en cours, suivant laquelle le latin sort de son propre diasystème."

3. Crises et inventions (VIIIe-Xe

siècles)

a. C o n t r a i r e m e n t à ce qui a parfois été soutenu, les compromis passés entre les détenteurs de la communication écrite et les destinataires de la communication orale ne furent pas que purement langagiers. La culture folklorique et la culture cléricale n'occupèrent pas des lieux si distincts que cela: tout au contraire la connivence entre l ' a u t e u r / é m e t t e u r des messages et les auditeurs/destinataires a été souvent profonde et cela a été r é c e m m e n t à juste titre souligné. Les Vies lues à haute voix ont servi de terrain d'expérimentation idéal à

27

Banniard, Viva voce pp. 18 sqq. Banniard, Viva voce pp. 148 sqq. 2 '•‫ י‬H. Beumann, "Gregor von Tours und der sermo rusticus", Festschrift M. Braubach (Münster, 1964), pp. 69-98. 311 R. McKitterick, The Uses of literacy in Early Medieval Europe (Cambridge, 1990). 31 T . Janson, Mechanisms of Language Change in Latin (Stockholm, 1979): Janson et S. Kiss, Tendances évolutives de la syntaxe verbale en latin tardif (Debrecen, 1982): D. Norberg, Syntaktische Forschungen auf dem Gebiete des Spätlateins und des frühen Mittellateins (Upsal, 1943): M. Pei, The Language of the Eighth Century Texts in Northern France. A Study of the Original Documents in the Collection of Tardif and other Sources (New York, 1932). 28

l'établissement d ' u n code cette fois non plus seulement langagier, mais aussi idéologique, sinon mental, partagé p a r une majorité de fidèles. Les récits de miracles sont devenus une magnifique matrice à thèmes popularisants: protéger une récolte de blé, écarter un orage, sauver une vendange, autant d'occasions de parler sinon exactement leur langage aux illettrés, du moins de mettre en scène leur propre quotidien. Les Vies mérovingiennes regorgent d'histoires familières prises sur le vif. Voici telle sainte occupée à allumer le four à pain: le texte décrit ses gestes avant d'expliquer c o m m e n t le feu gagne accidentellement le petit bois sec; un autre j o u r la m ê m e sainte est de corvée de lessive: la narration la décrit un seau à la main, un fagot de bois sous le bras, ennuyée p a r la mauvaise disposition de la crémaillière. D a n s les deux cas, le miracle est d ' u n e simplicité tout humaine: le sang-froid de la sainte sauve le fournil dans le premier cas; un coup de vent qui souffle dans la buanderie fait jaillir un bon feu pour la lessive dans l'autre cas. Tel prestigieux ermite accomplissant à cheval une tournée en Ponthieu se trouve soumis à une rude épreuve de rodéo parce que sa m o n t u r e s'emballe au m o m e n t précis où il vient de p r e n d r e dans ses bras un b a m b i n r é c e m m e n t baptisé: le récit souligne malicieusement le désarroi du héros malgré lui qui, secoué en tous sens, ne peut que hoqueter des " S e i g n e u r . . . Seigneur". Naturellement, le b a m b i n échappe au malheureux cavalier et ne doit la vie sauve q u ' à une taupinière dont la terre molle amortit sa chute (miracle!, évidemment). U n rien d'ironie perce à l'égard du saint, car la Vie précise qu'il se déplaça par la suite, non plus c o m m e un seigneur sur un cheval, mais c o m m e le Seigneur sur un âne. Ainsi le récit de miracle introduit des situations concrètes vécues et partagées p a r tous; le prodige lui-même garde son ambiguïté (il est "naturel"); les saints d e m e u r e n t parmi les hommes: ils sont humains. b. Le langage que parlent d'ailleurs ces saints est intelligible à leurs protégés, tout c o m m e l'est celui que parlent les représentents de l'Eglise, ainsi que les moines et les marginaux de la chrétienté que sont encore les ermites. Certes, ce langage varie beaucoup: entre la déclaration intime d ' u n ermite d o n n a n t quelques avis à des égarés et l'homélie solennelle d ' u n Eloi prêchant à Noyon, 3 2 existent des registres langagiers différents. O n supposera n o t a m m e n t que l'élite 3

'·‫ י־‬M. Banniard, "Latin et communication orale en Gaule: le témoignage de la Vita Elegit", L'Europe au VII' siècle: changement et continuité, ed. J . Fontaine, J . Hillgarth (Londres, 1992), pp. 58~86.

cléricale, à l'image d'Isidore, de Julien de Tolède, ou de Grégoire le G r a n d , a su préserver quelque chose d ' u n e prononciation plus soignée, surtout en des circonstances particulières, que le c o m m u n des locuteurs. Le ton, l'accent, les choix grammaticaux, les préférences lexicales, le phrasé de l'énoncé révèlent encore de discrètes démarcations sociolinguistiques. 33 Elles portent m ê m e l'amorce des clivages f r a n c h e m e n t langagiers qui menacent la communication verticale. En outre, d ' u n e région à l'autre, la dialectalisation du latin parlé tardif s'est engagée. Prenant des couleurs de plus en plus contrastées en fonction des niveaux culturels des locuteurs, elle esquisse les contours d ' u n latin régionalisé. C e p e n d a n t , le désir d'intercommunication tant verticale qu'horizontale continue, en ces siècles de changements rapides, à surmonter les obstacles que dresse l'évolution de la langue parlée. 34 A l'entrée du VIII e siècle, la latinophonie arrive à un seuil critique de son histoire, 35 au moins en ce qui concerne les fonctions communicatives. En effet les locuteurs lettrés continuent de considérer qu'il n'y a pas de différence essentielle entre la langue écrite et la langue parlée, c'est-à-dire que la forme écrite traditionnelle, la graphie latine, d e m e u r e la forme écrite de tout énoncé oral. Inversement, tout texte, pour peu qu'il soit stylistiquement adapté, lu à haute voix, continue d'être accessible aux auditeurs illettrés. 36 C'est aussi largement le cas d ' u n e partie du texte biblique (Evangiles). 37 Mais cette p e r m a n e n c e , sans être tout-à-fait artificielle, laisse désormais place à des déchirures de plus en plus étendues. 3 8 c. Cela signifie qu'entre les producteurs d'énoncés et leurs destinataires, la communication est devenue fragile. Certes, toute compréhension entre individus de niveaux culturels et par là-même langagiers différents, rencontre des limites qui ne peuvent être surmontées que

33

M. Banniard, "Le lecteur en Espagne wisigothique d'après Isidore de Seville: de ses fonctions à l'état de la langue", Revue des Etudes Augustiniennes 21 (1975), pp. 112-144: Ib., " Vox agrestis: quelques problèmes d'élocution de Cassiodore à Alcuin", Etudes Antiques, D'Hippoaate à Alcuin, n. spécial de Trames (Limoges, 1985), pp. 195-208. 34 R. McKitterick, "Latin and Romance: an Historian's Perspective", Latin and the Romance Languages in the Early Middle Ages, ed., Wright, pp. 130 145. 35 C. Battisti, "Secoli illitterati. Appunti sulla crisi del latino prima della riforma carolingia", Studi Medievali, 3e série, 1 (1960), pp. 362 396. 36 Wright, Late Latin and Early Romance in Spain and Carolingian France. 37 P. Salmon, "Le texte biblique des lectionnaires mérovingiens", La Bibbia nelVallo medioevo, Settimane di Studi 10 (Spolète, 1963), pp. 491-519. 3B M. Banniard, "Seuils et frontières langagières dans la Francia romane du VIII1' siècle", Karl Martel in seiner £eit, Beihefte zur Francia, t. 37, eds. J . J a r n u t , U. Nonn et M. Richter (Sigmaringen, 1994), pp. 171-190.

parce que les communicants ont décidé de réussir à se c o m p r e n d r e . Mais ce caractère, secondaire et relatif en situation de monolinguisme simple, s'aggrave vivement dans une phase où la société passe d ' u n monolinguisme devenu complexe 3 9 à la diglossie. 40 J e reviendrai sur cet aspect p r o p r e m e n t linguistique, mais il m e faut indiquer tout de suite que la réception d ' u n e lecture de Vita au VHP' siècle en Gaule précarolingienne suppose une solide connivence langagière et culturelie entre celui qui délivre la lecture et ses destinataires. En effet, le code oral des illettrés ne se superpose alors que très imparfaitem e n t au code, soit oral soutenu soit écrit, des lettrés. Cette fissure langagière ne peut être surmontée que si les auditeurs illettrés acceptent de faire l'effort de mobiliser leurs compétences langagières passives p o u r combler les trous d'information que provoque, dans la trame de l'énoncé lu, la discordance entre les deux registres langagiers. O r , cet effort n'est possible q u ' a u prix d ' a c c o m m o d e m e n t s importants: le latin lu ne doit pas être trop hiératique; l'accent et la diction du lecteur ne rejetteront pas dans les limbes de l'inculture la prononciation quotidienne; 4 1 l'histoire lue doit préserver les traditions locales, en répétant des thèmes familiers et déjà connus, sinon colportés d ' u n e génération de fidèles à l'autre. 4 2 En somme, les auditeurs doivent se reconnaître un peu dans les récits et dans les commentaires qui leurs sont faits: Vie de saint, mais latin de pêcheur. Cette connivence est r o m p u e dans la deuxième moitié du VIII e siècle. L'ascension des Pippinides s'accompagne d ' u n e puissante campagne de réforme, langagière, morale et idéologique. 4 3 Voici donc que les Vies de saints et les homélies (au moins) sont peu à peu purifiées de leurs impuretés non seulement grammaticales, mais aussi idéologiques, et que les lectures à haute voix se rétractent en l'espace d ' u n e génération sur une latinité b e a u c o u p plus vieille, associée à

39

R. Wright, "Complex Monolingualism 111 Early Romance", Linguistic Perspectives on the Romance Languages, Selected papers from ... the linguistic symposium on Romance languages 21, eds. W.J. Ashby et al. (Amsterdam, 1993), pp. 378-388 (aussi dans Wright, Early Ibero-Romance (Newark, Delaware, 1995)). 40 Banniard, Viva voce, chap. 9: H. Lüdtke, "Die Entstehung der romanischer Schriftsprachen", Vox Romanica 23 (1964), pp. 3-21. 41 Wright, Late Latin and Early Romance in Spain and Carolingian France. 42 M. Banniard, Genèse culturelle de l'Europe (V'-VIIT siècle) (Paris, 1989) (éd. ailemande, Munich-Leipzig, 1993), chap. 6. 43 I,. Holtz, Donat et la tradition de l'enseignement de l'art grammatical. Etude sur l'Ars Donati et sa diffusion (IV' IX' siècle) et édition entique (Paris, 1981): V. Law, The Insular latin Grammarians (Woodbridge, 1982).

des efforts de prononciation accrus: prêtres, évêques, abbés sont invités à reconquérir une pureté langagière décrétée perdue. Les auditeurs illettrés sont donc mis en présence d'énoncés qui frappent leur oreille non seulement parce qu'ils sont inattendus (leur mémoire ne recèle plus le stock d'informations nécessaire pour surmonter l'obstacle), mais surtout parce qu'ils rejettent leur propre latin d'illettrés hors des lumières de l'Eglise. Leur évêque, sinon leur curé, ne veut plus parler un langage qui, m ê m e de loin, rappelle le leur. Repoussés langagièrement, ils le sont aussi idéologiquement. C a r les vieilles bistoires sont récrites de telle manière que les saints s'éloignent d'eux (et que la main de Dieu soit omniprésente cette fois d ' u n e manière surnaturelle). C o m m e la liturgie de la messe éloigne le vulgus du choeur, la nouvelle mise en scène des saints efface les aspects familiers des vieux récits. 44 Dans ces conditions, j e voudrais ici souligner un p h é n o m è n e qui a peut-être été négligé et qui devrait conduire à une piste de recherche fructueuse: certes les lettrés écartent d'eux les illettrés, leur latin et leurs histoires. Mais ces derniers n'ont peut-être pas été les simpies objets passifs de cette volte-face. Eux aussi avaient toutes les raisons au m o n d e de rompre les ponts langagiers avec les responsables de cette nouvelle mode. L'ancienne parabole évangélique prit tout son sens: ils n'eurent plus d'oreilles pour entendre. M ê m e si certaines parties de ces messages lancés à la m o d e nouvelle continuaient d'être partiellement intelligibles, c o m m e le laisse entendre le canon 17 édicté à T o u r s par le concile de 813, la masse des locuteurs opposa son inertie vindicative aux efforts de la prédication carolingienne. Les intellectuels, du coup, se retrouvèrent à leur tour sans langue: le temps de la traduction était venu, ce dernier terme p r e n a n t alors toute sa vigueur. 4 ‫ י‬II s'agissait, en effet, de consacrer la distance désormais établie entre la langue c o m m u n e et la langue de l'élite, puisque, de l'une à l'autre, il y avait une Iongue distance à couvrir (trans—-ferre).

44

Banniard, Genèse culturelle de l'Europe (V- VIII' siècle) chap. 6: Banniard, Viva voce, chap. 7: Wright, Late Latin and Early Romance in Spain and Carolingian France. 45 Banniard, Viva voce: Wright, ImU Latin and Early Romance in Spain and Carolingian France‫־‬. Id., "The Conceptual Distinction between Latin and Romance: Invention or Evolution?": M. Zink, IM prédication en langue romane avant 1300 (Paris, 1976).

4. Interferences culturelles et langagières a. Avec les corrections nécessaires, la description de ces événements en Gaule vaut p o u r l'Espagne mozarabe. T r o p brièvement, j e dirai ici que c'est un peu plus tard sans doute qu'eut lieu la débâcle de la communication latinophone, sans doute d ' a b o r d à C o r d o u e . Pour des raisons complexes, mais où le nationalisme langagier jouait un rôle éminent, les intellectuels mozarabes p r o m u r e n t au milieu du IX e siècle une réforme du latin n o r m a l e m e n t usuel, p o u r la rédaction et la récriture des Vitae.46 Selon toute vraisemblance, la langue du chant liturgique subit également une importante révision, toujours dans le sens d ' u n retour à la vieille latinité. Les mêmes causes produisirent probablement les mêmes effets: la c o m m u n a u t é des locuteurs se détacha plus rapidement encore de cette latinité devenue obscure et hiératique. Le martyre langagier ne tenta guère plus les C o r d o u a n s que le martyre religieux. Le prestige de la langue écrite concurrente du latin, l'arabe, provoqua m ê m e peut-être une réaction accélérée de passage à l'alphabétisation en écriture, sinon en langue arabe. La langue parlée issue du latin, la novella lingua ("la langue nouvelle"), c o m m e la n o m m e l'abbé Samson ou la uernula lingua ("la langue des domestiques"), c o m m e elle semble désignée par Alvare, se replia ainsi pour au moins un siècle sur l'oralité pure. 4 7 J e me contenterai d'indiquer ici que le latin tardif parlé en Afrique dut connaître une phase analogue: mais dans son cas, la diglossie latin p a r l é / a r a b e littéral se termina p a r l'éradication de l'héritage r o m a n , l'extinction des derniers locuteurs r o m a n o p h o n e s s'étant sans doute accomplie avant le XII e siècle. 48 Enfin, dans le cas de l'Italie, la réforme carolingienne eut peu d'influence. C e fait, allié au caractère peu évolutif de la phonétique du latin tardif parlé en Italie, et peut-être conforté p a r le sens aigu des compromis de la péninsule, 4 9 explique la longue durée d'une cohabitation protéiforme entre différents niveaux de langue, peut-être j u s q u ' a u X e siècle. Toutefois, l'apparition des premières gloses en italien au milieu de ce siècle signifie

4,1

Banniard, Viva voce, chap. 8. Banniard, Viva voce, pp. 480, 483. 48 S. Lancel, "Fin et survie de la latinité en Afrique du Nord", Revue des Etudes Latines 59 (1981), pp. 269-297. 4!l C. Violante, "Le strutture organizzative délia cura d'animé nelle campagne dell'Italia centrosettentrionale, secoli V X", Settimane di Studi 28 (Spolète, 1982), pp. 963-1158. 47

que le continuum langagier s'est alors délité et que la c o m m u n i c a tion verticale est devenue au moins aussi perturbée à cette date qu'en Gaule 150 ans et en Espagne 100 ans plus tôt. 50 b. Conclure ainsi au terme d'enquêtes sociolinguistiques à une Iongue vie du latin c o m m e langue de communication générale invite les chercheurs à des révisions parfois complexes des théories jusque là admises tant en histoire culturelle q u ' e n linguistique diachronique. D a n s le premier domaine, il paraît indispensable de mieux poser le problème des rapports entre les catégories de cultures (orale/écrite; laïque/cléricale . . .), les niveaux sociaux (puissants/pauvres; paysans/ guerriers; . . .) et les espaces m e n t a u x variés (jeunes/âgés; h o m m e s / femmes; latins/germains; éleveurs/agriculteurs): s'il paraît inapproprié de renoncer à ces disjonctions c o m m o d e s et opératoires, il faut c h e r c h e r des modèles de représentation qui ne séparent pas ces ensembles de manière étanche, ni surtout binaire, p o u r plutôt les représenter c o m m e des systèmes ouverts intéragissant les uns avec les autres. 51 Cette nécessité est éclatante dans le cas précisément de ce qu'il convient d'appeler l'interaction culturelle entre l'action de la christianisation et la réaction des christianisés. Le m ê m e m o d e de raisonnement induit un recalage méthodologique important en linguistique diachronique. Celle-ci a surmonté ses propres difficultés, non en trouvant des solutions aux problèmes, mais en les escamotant. En effet, alors que l'histoire longue de la communication d o n n e à voir à un bout de la chaîne chronologique une c o m m u n a u t é latinophone (III1' siècle) et à l'autre des sous-ensembles r o m a n o p h o n e s (VIII e siècle), la linguistique diachronique a divisé son domaine en deux champs: l'un latin, réservé aux latinistes; l'autre, r o m a n , réservé aux romanistes.'‫ '־‬C h a c u n a ainsi pu élaborer ses proprès théories sans tenir compte de la continuité h u m a i n e ni à travers l'espace en synchronie (tout le monde communiquait), ni à travers le temps en diachronie (les générations de locuteurs n'on jamais cessé de se comprendre), avec p o u r résultat une représentation dichotomique du m o n d e langagier. O n a distingué artificiellement d ' u n côté

‫נ‬

" Banniard, (Bologne, 1973) 51 YV. Labov, Mass., 1994). w R. Kontzi,

Viva voce: A. Castellani, I più antichi testi italiani. Edizione e commento pp. 543 sqq. Principles of Linguistic Change, t. 1, Internal Factors (Oxford/Cambridge Die Entstehung der romanischen Sprachen (Darmstadt, 1980).

un latin littéraire n o r m é , figé, écrit et accessible à l'historien; de l'autre un latin vulgaire désordonné, évolutif, p u r e m e n t oral et donc inaccessible à l'historien de la langue. Cette bipartition a permis toutes les hardiesses en laissant une partie des philologues romanistes libres de fabriquer p a r reconstitution arbitraire (quoique raisonnée) la langue source des langues romanes, avec des prises de position parfois extrêmes q u a n t à la chronologie des phénomènes, 5 3 malgré les choix b e a u c o u p plus prudents d'aucuns d'entre eux. 54 Elle a aussi permis de décrire le c h a n g e m e n t langagier c o m m e une sorte de catastrophe naturelle au cours de laquelle l'effondrement des élites culturelles (les temps, n'est-ce-pas, étaient "barbares") a provoqué la rétraction de la latinophonie c o m m e une peau de chagrin au profit des dialectes populaires, jusque là contenus dans une vie obscure un peu larvaire. O n ne saurait mieux qualifier ce type de linguistique diachronique que par le terme de pseudo-darwinisme, dont on devine à quel point il a induit les modèles m e n t a u x des philologues, sinon des historiens du XIX 1 ' siècle. Naturellement, ce modèle interprétatif est absolument incompatible avec la reconstitution diachronique de la communication latinophone, telle qu'elle est établie par la sociolinguistique rétrospective. Cette dernière a permis d'établir solidement que l'évolution langagière a répondu à trois critères: 1. Entre lettrés et illettrés il n'y a eu aucune barrière absolue, mais des sortes de maillages distinctifs d'épaisseur et d'étendue variables; 2. Les rapports entre la langue parlée et la langue écrite ne furent pas si relâchés que l'une et l'autre aient cessé d'appartenir au m ê m e diasystème (que j'appelle le latin parlé t a r d i f — L P T ) avant le VIII e siècle; 3. La langue écrite e l l e - m ê m e subit en fait une très profonde évolution du III e au VIII e siècle, dans une mesure variable selon les auteurs, selon les genres et selon les régions. 35 Dans ces condi-

53

G. Straka, Ixs sons el les mots (Paris, 1979). P. Pouche, Phonétique historique du français, t. 2. Les voyelles (2) (Paris, 1969); t. 3, Les consonnes (Paris, 1961): A. Martinet, Economie des changements phonétiques, Traité de phonologie diachronique (3) (Berne, 1970): T h J . Walsh, "Spelling lapses in Early Medieval Latin Documents and the Reconstruction of Primitive Romance Phonology", latin and the Romance languages in the Early Middle Ages, ed. Wright, pp. 205-218. 55 Banniard, Viva voce: Id., "Language and Communication in Carolingian Europe", The New Cambridge Medieval History, t. 2, 700 900, ed. R. McKitterick (Cambridge, 1995), pp. 695-708: Richter, The formation of the Medieval West. Studies in the oral cul54

tions, il convient de chercher des modèles diachroniques qui respectent mieux la réalité sociolinguistique, tout en rendant compte de l'apparition des premières scriptae romanes. 3 6 c. J ' i r a i à l'essentiel, tout en soulignant combien cette réorientation interprétative exige l'ouverture d'importants chantiers (mais beaucoup sont d'ores et déjà engagés). O n partira de la notion de comm u n a u t é linguistique comprise en ces termes: "Il serait faux de concevoir la c o m m u n a u t é linguistique c o m m e un ensemble de locuteurs employant les m ê m e s formes. O n la décrit mieux c o m m e étant un groupe qui partage les mêmes normes quant à la langue." 3 7 A partir de ceci, on postulera que les évolutions langagières sont le résultat du travail sur sa propre langue qu'effectue la c o m m u n a u t é intégrale des locuteurs. O n admettra n o t a m m e n t que les tournures qui, attestées en latin, a n n o n c e n t les tournures romanes, ne sont pas des fautes commises p a r une majorité inculte au détriment d ' u n e minorité cultivée, mais des variantes marquées, forgées et choisies p a r n'importe quel locuteur, des tournures traditionnelles. 58 Il y a beau temps que les grammairiens et les stylisticiens latinistes ont r e m a r q u é une étrange collusion entre certaines particularités de la langue poétique et ce qu'ils appellent les vulgarismes de la langue populaire. Ils attribuent ces p h é n o m è n e s à une influence de la langue populaire sur la langue des poètes. Mais cette assertion repose sur l'idée que la langue parlée par l'élite n'est pas susceptible, elle aussi, de variations dues au désir d'expressivité et à la pression des motivations psychologiques, qu'elles soient accidentelles, sociales ou institutionnelles: or, rien ne prouve qu'il s'agisse de concessions

lure of the barbarians: Wright, Late ÍMtin and Early Romance in Spain and Carolingian France: P. Zumthor, Langue et technique poétiques à l'époque romane (XIr XIII' siècle) (Paris, 1963). % F. Sabatini, "Dalla 'scripta latina rustica' alle 'scriptae romanze'", Studi Medieval! 3 (1968), pp. 320 358: Scrìpt0ralia, Le passage à l'écrit des langues romanes, eds. M. Selig, B. Frank, J . Hartman (Tübingen, 1993): R. Wright, "Textos astrurianos de los siglos IX y X: (-;Latin bàrbaro ο romance escrito?", Lletres asturianes 41 (1991), pp. 20 34 (traduction anglaise, "Asturian texts of the ninth and tenth centuries: Barbarous Latin or Written Romance", dans Wright, Early Ibero-Rornance). Id., " O n Editing 'Latin' texts written by romance-speakers", Linguistic Studies in Medieval Spanish, eds. R. Harris-Northall et I . D . Cravens (Madison,' 1991), pp. 191-208. 5 ' Labov, Sociolinguistique•. Labov, Principles of Linguistic Change, t. 1, Internal Factors 58 M. Banniard, "L'ablatif (instrumental) et le cas régime (indirect): sur la restructuration du latin parlé tardif au protofrançais (IIP V H P siècle)", IALLES (Paris. 1995) pp. 227 92: Id., "Oralité et formes marquées: entre expressivité et changement langagier", Lingua latina 5, J . Dangel, Cl. Moussy, ed. (Paris, 1995), pp. 69 83.

poétiques à des vulgarismes. Il est plus élégant scientifiquement et sans soute plus vrai anthropologiquement d ' a d m e t t r e que les phénomènes énonciatifs d ' e m p h a s e sont partagés p a r la c o m m u n a u t é des locuteurs. C'est ainsi que le renforcement ou le remplacement des cas p a r des tournures prépositionnelles, bien attesté chez les poètes classiques, sinon chez des prosateurs c o m m e Salluste, correspond à un effet stylistique. En stade initial, le futur m o r p h è m e r o m a n n'est souvent ni plus ni moins q u ' u n e variante emphatique, une variante marquée, non pas provoquée par une faute d'ignare, mais au contraire générée p a r le travail d ' u n créateur littéraire exprimant le génie de la langue. C e ne sera q u ' a u stade final que ce qui était au départ une variante rare est devenu une forme usuelle qui chasse peu à peu la forme ancienne de l'usage parlé: entre ces deux stades, il a pu s'écouler un demi-millénaire, parfois plus. 59 P e n d a n t celui-ci, la c o m m u n a u t é des locuteurs a accompli un labeur de renouvellement langagier qui s'est traduit selon toute vraisemblance p a r des p h é n o m è n e s de polymorphisme intense. Ce dernier point rend en particulier compte de la difficulté que rencontre l'historien de la langue à déterminer quelles formes (et quels mots) sont mortes dans la langue parlée, alorqu'elles sont encore employeés p a r écrit, et, inversement quelles formes avaient émergé dans la langue parlée sans avoir encore accès à l'écriture. De plus l'attestation écrite des nouvelles tournures ne signifie plus, dans ces conditions, que les anciennes étaient disparues depuis longtemps des mémoires. 6 0 Enfin, le basculement final d ' u n système langagier à un autre peut s'effectuer sur une période assez brève. 6 ' O n aboutit ainsi, c o m m e on le voit à une représentation positive et d y n a m i q u e du changement langagier. Pour le résumer en une formule, il n ' a pas suffi de mal parler latin pour inventer le protofrançais en Gaule (la maxime vaut p o u r chaque pays de la Romania). d. J e voudrais conclure en insistant auprès de l'ESF pour que cette histoire langagière de l'Europe débouche sur une orientation, auda-

59

C. Marchello-Nizia, L'évolution du français: ordre des mots, démonstratifs, accent tonique (Paris, 1995). 00 M. Banniard, "Latin tardif et français prélittéraire: observations de méthode et de chronologie", Bulletin de la Société de linguistique de Pans 88 (1993), pp. 139-162: Beumann, "Gregor von Tours und der sermo rusticus". 61 M. Banniard, "Géographie linguistique et linguistique diachronique", Via Domitia, Annales de l'Université de Toulouse II 24 (1980), pp. 9 43.

cieuse peut-être, mais sûrement réaliste de la genèse des langues romanes. A mes yeux, mais c'est un point de vue que partagent désormais un certain n o m b r e de linguistes diachroniciens et de sociolinguistes, le passage du latin aux langue romanes signifie certes que le latin cesse au V i l l e siècle d'être une langue vivante, mais, interprété et considéré ainsi, ce passage implique que nos langues modernés ne sont pas une forme dégradée de la vieille langue de Rome, 6 2 mais, au contraire, son accomplissement. 6 3

5. Quelques vues schématiques des phénomènes a. Fin !de la communication verticale latine (schéma chronologique) France d'oïl France d'oc Espagne mozarabe Italie du Nord et du Centre Italie du Sud Afrique

** ** ** **

**

750-800. 800-850. 850-900. 900-950. p 750-800?)

b. Abrégé Chronologique du changement linguistique en Occident latin65

0—Avant 450: le latin est la langue commune. Periode

I

1—450-650: apparition et multiplication des tournures nouvelles de substitution aux tournures classiques. 2—650: seuil critique d'équilibre entre traits latins et traits romans. (Italie 750?)

Periode

II

3—650-750: polymotphisme généralisé. (Italie 650-850?) 4—700-800: abandon de la compétence active des traits classiques.

Periode

62

III

5—800 sqq.: abandon de la compétence passive. (Italie 900?)

F. Brunot, Histoire de la langue française (2), t. 1 (Paris, 1966). M. Delbouille, "Tradition latine et naissance des littératures romanes", Grundriss der Romanischen Literaturen des Mittelalters, t. 1 (Heidelberg, 1972), pp. 3 56: Id., "La formation des langues littéraires et les premiers textes", ib., pp. 560-584 et 604-622. 64 Banniard, Viva voce, p. 492. 6 ‫ י׳‬Banniard, Viva voce, p. 534. 63

c. Remplacement des formes en diachronie66

NB: + = Forme marquée;— = Forme non marquée État

Comes ciuitatis Comes de ciuitate

+*

1

État 2

Polymorph. Polymorph.

État 3

+ ** -

*: Entrée de la forme dans la langue parlée **: Sortie de la forme de la langue parlée Etat 1: Période 1. Etat 2: Période 2. Etat 3: Période 3. d. £00m sur la communication verticale en Francia porotromane (750) 67 A. Lettrés

Comp. act. Comp. pass.

+ ++ ++ +

B. Semi-lettrés

+ + + ++

C . Illettrés

+ — + + -

Comp. act. = compétences acrives (capacité à produire un énoncé semblable à l'énoncé entendu). Comp. pass. = compétences passives (capacité à comprendre l'énoncé entendu).

66

Banniard, ration du latin mes marquées: 67 Banniard, siècle" p. 189.

"L'ablatif (instrumental) et le cas régime (indirect): sur la restructuparlé tardif au protofrançais ( I I P - V H F siècle)": Id., "Oralité et forentre expressivité et changement langagier". "Seuils et frontières langagières dans la Francia romane du VIII e

TRANSFORMATION OF THE ROMAN WORLD: REFLECTIONS ON C U R R E N T RESEARCH Paolo Delogu

W h e n the coordinators of the "Transformation of the R o m a n W o r l d " p r o g r a m m e invited me to prepare some introductory words for this plenary conference, they suggested that I should give an overall review of the work done and the results acquired by the groups involved in the p r o g r a m m e during the last five years. They also insisted that I should refrain from presenting my own views on the T r a n s f o r m a t i o n ; consequently to the general relief, w h a t you are going to listen to will not be a new theory on the subject. Nevertheless, my task soon appeared pretty difficult; although I am myself a participant in the project, I did not take part in every meeting, nor had direct experience of the work done by groups other than mine. T o gather information I was obliged to rely on my own recollections and on the newsletters that have been regularly sent to every participant in the programme. 1 Consequently the result of my review will probably be similar to what a historian of the Dark Ages would have produced: a doubtful narrative, based on tendentious oral traditions a n d scanty written records. If m a n y of you do not recognize yourselves in what follows, this very fact will contribute to the appraisal of how reliable (or unreliable) our historical sources are. In order to reduce risks, I have concentrated my observations on some very general issues that have been present for every participant throughout the entire course of the research, and that, in my opinion, are worthy of further discussion. T h e first set of observations concerns the very title of the programme: " T r a n s f o r m a t i o n of the R o m a n W o r l d " . It seems to provide the research with its historiographical ideology and its explanatory key.

1

Newsletters, edited by Ian Wood, have been regularly sent to each participant in the research programme (7 issues, from September 1993 to December 1996). Scholars will be well aware of the major texts mentioned in the article: the footnotes provide the basic references.

" T r a n s f o r m a t i o n " means that history is being analyzed as a process of continuous change. In other words the title seems to reject two traditional historical conceptions: one, the use of periodisation, in particular as concerns the distinction between Late Antiquity and the Early Middle, or Dark, Ages; two, the idea of the decline and fall of the R o m a n Empire. In m o r e general terms, the title invites us to give u p categories like decay and its opposite, progress, in order to qualify historical processes. It is easy to understand the reason for the first rejection: if transformation is a continuous process through which societies change, it is meaningless a n d even misleading to split periods inside it, owing to the fact that such an operation may suggest the existence of interruptions in the process, or the concentration of transformation only in transitional periods. F u r t h e r m o r e , implicit in the choice is a hope that better and more original evaluations of facts and processes can be achieved by having recourse to a longer time scale. O n the other hand, today's historical studies (including the research made in the T . R . W . programme) are mosdy interested in social forms; they even consider culture as a function of the social organization. Consequently, j u d g e m e n t s that seem to establish hierarchies a m o n g different socio-cultural situations, like those that derive from the polarities civilization-barbarousness, d e v e l o p m e n t - u n d e r d e v e l o p m e n t , organization-disorganization, are considered with suspicion. Even m o r e suspect are j u d g e m e n t s linking cultural situations of the past to ethical or political values of the present. In all these cases, apart from any methodological warning, there arises the censure of "political incorrectness". Each cultural situation is d e e m e d significant in itself; the transformations only reveal the way in which a given society (and societies in general) work. Consequently it is meaningless, or even incorrect, to describe a sequence of cultural situations as a process of decay or progress. History must renounce any ideological orientation. If this is the historiographical ideology of the T . R . W . research prog r a m m e , two assumptions seem to a certain extent discrepant with the premises. T h e project rejects the traditional periodisation and its basis; nonetheless it indicates a limit to the transformation of the R o m a n world and puts it in the Carolingian Age. If this determination is something m o r e than a practical device to limit somehow the extent of the research, one must assume that in the Carolingian

Age the transformation of the R o m a n World reaches a critical point following which things are so different that they cannot be considered as the simple continuation of previous processes. This implies that socio-cultural, as well as economic and political conditions, of the Carolingian Age can be considered as a system definitely different from that of the changing R o m a n World: a system which can be appropriately called as first outline of m o d e r n Europe. In fact, the T r a n s f o r m a t i o n shows two faces: if its starting point is considered, it shows the R o m a n world in motion; if its conclusion is looked at, it describes the rise of Europe. T h e two faces are clearly differentiated only at the extremities of the T r a n s f o r m a t i o n ; during the process itself, they intermingle; it is u p to the orientation of the researcher to emphasise one aspect m o r e than the other. Shall we say that in the T . R . W . p r o g r a m m e periodisation appears once again, though in original form a n d covering an unusual stretch of time? But one should also r e m e m b e r that in the Twenties and Thirties, two fathers of the history of Europe, Henri Pirenne a n d Christopher Dawson, maintained that all the processes that started in the Late Antiquity, in the field of politics, economics a n d culture, found their completion and a first new synthesis in the Carolingian Age. 2 M o r e recently Peter Brown, following lines of thought close to the T . R . W . p r o g r a m m e , similarly characterized the Carolingian Age as the time in which the social a n d cultural experiences of the preceeding centuries were normalized a n d stabilized by the political powers. 3 T h e r e is only one way to avoid the conclusion that we still have recourse to periodization so as to give order to the historical sequence; and this is to assume that from Late Antiquity to the Carolingian Age, societies and cultures of the R o m a n World were transformed in such a m a n n e r that the socio-economic structure presented a single, organic evolutionary cycle. In a m o m e n t I will examine the ways in which the groups involved in the p r o g r a m m e have worked out this suggestion. Meanwhile, in order to conclude my observations on the general lines of the prog r a m m e , I would point out what in my opinion seems a second source of ambiguity. O n e readily admits that the R o m a n World can

2 H. Pirenne, Histoire de l'Europe (Brussels, 1939): C. Dawson, The Making of Europe: An introduction to the history of European Unity (London. 1932). 3 P. Brown, The Rise of Western Christendom (Oxford, 1995).

be the object of unbiased research; a research that analyses the internal structure and the way in which the imperial system worked and changed, and that is carried out from a scientific point of view, without any ideological or emotional involvement from the student. Attitudes about the idea of Europe are probably different. Obviously, E u r o p e too can be conceived of as a system of relations a m o n g various societies a n d cultures, that are kept together not by a centralized state power, as in the case of the R o m a n World, but through shared traditions a n d by the privileged circulation of information, or by links of economic interaction, so that they develop institutions and economies that are similar and have influence on one another. But for the Europeans, Europe is inescapably a value-concept, because it corresponds to a c o m m o n tradition of civilization that has not yet completely lost its force as a means of collective identification; it is still a point of reference of self-consciousness, at least for the intellectual élites of the Continent. Historical enquiry into the origins of Europe helps to strengthen the desire Europeans have to form an integrated c o m m u n i t y , largely based on cultural bases. Like the migrating Barbarian peoples, or the monastic communities of the Middie Ages, the Europeans at the end of the second Millennium seek in their past both the definition of their c o m m o n identity and the weapons to fight against those scenarios that foresee Europe only as an integrated commercial and monetary area or, on the contrary, as a cluster of small nations in egoistic rivalry with each other. 111 this sense, research cannot be indifferent, even if it is correct; this is what I consider a potential source of ambiguity, though a seminal one, in the T . R . W . p r o g r a m m e . T h e cultural ideologies or even the national point of view of each researcher can leave their traces in the research, and influence the evaluation of historical developments, even if these are being considered as aspects of the transf o r m a t i o n of the R o m a n W o r l d , given the close interconnection existing between the two faces of this history. I will return to this point at the end of the paper. In the m e a n t i m e I will survey in brief some general concepts that have been currently employed by the working groups in order to characterize and interprété the processes they went on analyzing. I will try to show some implications that each of them presents with regard to the basic theoretical problems of the p r o g r a m m e : historical continuity and the stages of the Transformation. First of all, I want to recall the concept of ethnogenesis, that was

first formulated in the historical sciences by R e i n h a r d YVenskus4 and was further elaborated by Herwig W o l f r a m a n d his Viennese school. Probably it is the most consonant tool of interpretation for scholars w h o approach the T r a n s f o r m a t i o n as a process of continuous création, in both the social and cultural field. Ethnogenesis includes and develops the concept of acculturation; this latter considers cultural exchange between consolidated groups; ethnogenesis makes cultural change coincident with the continuous shaping and reshaping of the groups. In the T . R . W . p r o g r a m m e , the concept of ethnogenesis mainly resulted in extenuating the traditional assumption of the great cultural contrast between the R o m a n s and the Germans. Instead, the notion of mutual contamination was brought into play, and frontiers were considered as vital areas where exchange was intense a n d societies first experienced new cultural forms. Stress was also put on the compatibility of the migrating societies with those living within the R o m a n Empire; consequently it was denied that the impact of the G e r m a n s h a d m a j o r destructive effects. In my opinion the new perception of the Barbarians as a constituent of the R o m a n system, not as opponents or alternatives to it, is a very stimulating orientation of the new research, although it is also a perplexing one. T h a t the invaders did not aim at destroying R o m a n society and culture was already maintained, for instance, by Fustel de Coulanges in the nineteenth century, but the assumption that barbarian culture was fairly consistent with R o m a n culture at the time the invasions took place, suggests that we have to a b a n d o n the great romantic concept of G e r m a n i c culture being the very cause of the renewal of proto-European society, and the lasting dialectical opponent of the R o m a n - M e d i t e r r a n e a n culture. O f course, this was a strongly ideological concept, but one that had extraordinary explanatory potentialities, and one that has not yet been a b a n d o n e d . For instance, G e r m a n i c culture has been commonly referred to by the working-teams who examined the change of mentalities that resulted from signs, symbols, rituals, even though they did so in more sophisticated terms than the nineteenth century Romantics. N o w the Germanic traditions are considered as one changing c o m p o n e n t of the cultural process, neither more stable nor less vital and creative than others that were equally present.

4

R. Wenskus, Slammesbildung und Verfassung (Cologne, 1961).

A last observation on this subject: p h e n o m e n a that can be referred to as belonging to the field of ethnogenesis were not limited to the Migration Period or to the Age of the barbarian kingdoms on R o m a n territory. T h e y went on, without interruption, into the Carolingian Age, for instance with the development of new aristocracies, and afterwards, into the tenth century, w h e n ethnic units, defined by a territory, aimed at becoming a u t o n o m o u s political organizations. Strictly speaking, ethnogenesis as the moulding process of groups that claim collective identity and have political aspirations never ceases, as events in Italy show at this very time. It is certainly possible to distinguish phases in the ethnogenetic process; the question is whether the time-span covered by the T . R . W . research project coincides with one single, distinguished phase, or if it c o m p r e h e n d s more phases and distinct developments within the former R o m a n World. W h a t I dare say is that ethnogenetic processes taken in themselves do not exhaust the general sense of the Transformation nor fully account for its proposed chronology. O t h e r concepts that were used in the research widen the perception of the historical context a n d supply additional possibilities to describe the sense a n d rhythms of its movement. O n e of them is "disruption". In a sense, it introduces a sort of qualitative determination of the historical trend. It must be said that the concept of disruption has been used without any reference to moral or cultural values, and without imposing a reading of decay on events; rather, it describes the trend of the T r a n s f o r m a t i o n in the field of material culture a n d h u m a n settlement. T h e quality a n d quantity of the commodities available to local societies as well as the n u m b e r , dimensions a n d functions of the settlements show a steady tendency to decrease during the transformation of the R o m a n World. Some of the indicators (e.g. ceramics) disappear in time; others change their functions a n d interconnections, so that at a certain m o m e n t the cultural a n d economic system that continued to work until then is seen to have broken down in its entirety a n d in fact does not exist any more. Evaluations can differ on the various aspects of this process: debates are still going on about the real relevance of imported commodities for the material life of the communities; arguments are particularly lively about the m e a n i n g of u r b a n change: if it represents deterioration a n d a b a n d o n m e n t or was the outcome of controlled transformation carried out by societies still solid a n d well organized.

In my opinion, what confirms that the trend can be interpreted as involving a loss is the objective nature of the indicators that point to quantity, to technical abilities, a n d to standards of social intégration which become lower a n d lower. But the process of disruption in the R o m a n system of material life does not extend over the whole period chosen by the research p r o g r a m m e . It reaches its fulfilment before the Carolingian Age; after its completion, p h e n o m e n a relating to material culture and setdement, take on a different nature a n d sense. T h e concept of disruption does not apply to the entire time span of the T r a n s f o r m a t i o n p r o g r a m m e . T h e third explanatory concept used to describe the processes of transformation is "disaggregation"; this term stresses the progressive separation and the loss of links a m o n g the different parts of the R o m a n Empire as basic aspects of the T r a n s f o r m a t i o n . T h e concept is not totally new: the separation of the West from the East, the splitting of the provinces f r o m one another, a n d of the Islamic from the Christian lands, all these are f u n d a m e n t a l issues in every classic description of the transition f r o m R o m a n to medieval times. T h e present use of this explanatory key is probably more complex; disaggregation not only means that the larger political and cultural areas of the Empire lost contacts; the concept points to a process of closing u p a n d isolation concerning regions, provinces a n d sub-regional territories. Perhaps the most significant trace of disaggregation is the distribution pattern of the artisanal a n d commercial products. Local production and distribution limited to restricted areas took the place of the imperial system, which had been based on concentrated production and long range distribution. T h e fading ties a m o n g the various parts of the Empire can be perceived through other evidence too. Michael H e n d y has shown that monetary usages in the provinces changed from the fifth century onwards, as a consequence of regionalization in the tax system and in military organization; on the same basis he argued that c o m m e r c e too lost its energy. 3 Disaggregation also affected culture a n d religion: Peter Brown uses the evocative term of "microchristianities" to denote

5

M. Hendy, The economy, fiscal administration and coinage of Byzantium, Variorum Reprints (Northampton, 1989): idem, "From Antiquity to the Middle Ages: economic and monetary aspects of the transition", De la Antigüedad Tardia al Medievo, Siglos IV- VIII (Madrid, 1993), pp. 325-60.

the regional Churches, each with its distinct religious and institutional features, that developed in what had once been the united Christian Empire. 6 T h e general line of interpretation is that when the imperial web that ensured the traffic of wares and the circulation of information collapsed, societies organized themselves on restricted bases; they relied essentially upon local resources—both economic a n d cultural— drawing less and less from the external ones. Behind this concept is the perception of the region as the basic structure of E u r o p e a n social organization, one that pre-existed the R o m a n period a n d survived it. For all its relevance, however, disaggregation does not explain every process that took place in the time period covered by the T . R . W . p r o g r a m m e . For that purpose, the notion should be connected with its reverse "re-integration". Indeed, from a certain point of time one can perceive new links developing a m o n g regions and sub-regions; wares circulate anew on an interregional scale, though along different lines f r o m those supported by the R o m a n Empire. O t h e r connections were created by the developing new aristocracies, through their wars as well as their political contacts. Religion had a special importance in p r o m o t i n g new connections: pilgrimages and missions were the channels through which new links were established a m o n g regions that had previously split apart; information too spread along lines which were different f r o m those of the R o m a n times. T a k e n on their own, none of these three explanatory concepts seems fully appropriate for an exhaustive description of the transformation of the R o m a n World. O n e can perhaps get nearer to the target by integrating the three. Along this line I shall now come back to a problem I pointed out before, a n d will briefly c o m m e n t on the relation between the presumed long cycle, extending from the fourth to the ninth century, and on other terms a n d rhythms of transformation that can be identified during that time. Although the T . R . W . p r o g r a m m e invites the participants to give u p the traditional distinction between Late Antiquity a n d the Early Middle Ages, the fact remains that until recently the distinction has been held as a meaningful criterion for organising historical material. It is sufficient to r e m e m b e r the names of H e n r i M a r r o u , ' Alexander

6 7

Brown, The Rise of Western Christendom. H.R. Marrou, Décadence romaine ou antiquité tardive? III'—VI' siècle (Paris, 1977).

D e m a n d t , 8 Peter Brown 9 and Averil Cameron, 1 0 leaving aside H e n r i Pirenne who had clear-cut ideas on the subject. Scholars who used this distinction do not seem to consider it as a conventional cxpedient; on the contrary they refer to basic changes of structure as h a v i n g o c c u r r e d between Antiquity a n d the M i d d l e Ages. S u c h changes concern essential aspects of the late antique system, like the unity of the Mediterranean and the imperial control over the West; the role of the aristocratic élites and also the classical style of citylife. T h e prospect of the death of the Antique World is far from dead in current studies. T h e chronological divide has usually been placed s o m e w h e r e between the second half of the sixth a n d the first half of the seventh century; 600 A.D. or a date thereabouts has taken the place of 476 A.D. as a viable limit of Antiquity. W e can now examine whether such a consolidated historiographical idea has been definitively superseded by current research; if so, what better views and ideas have been substituted for it. As far as I know, the point was not explicitly dealt with by the working groups, nor was it elaborated in the essays published in the London Catalogue of the exhibition on the T r a n s f o r m a t i o n . " C o n sequently, in order to investigate how the problem can be tackled, I will refer to a couple of essays that were not produced within the research programme, but that were well known to the working groups. O n e is the book that R i c h a r d Hodges and David Whitehouse dedicated to Mohammed., Charlemagne and the Origins of Europe.2‫ י‬T h e authors consider the period between the fifth and the ninth centuries as being characterized by the sequence of two economic and commercial systems involving the European territories. T h e first was the R o m a n system based on the Mediterranean, from which wares spread towards the internal regions of the Continent, crossing over the limes to reach D e n m a r k a n d the Baltic. This system fell into crisis from the fifth century onwards, and actually c a m e to an end in the course of the s

A. Demandt, Der Spätanlike (Munich, 1989). P. Brown, The World of luite Antiquity (London, 1971). 10 A. C a m e r o n , The Mediterranean World in Late Antiquity. A.D. 395-600 (London, 1993): idem, 77zí Later Roman Empire (London, 1993). 11 The Transformation of the Roman World. A.D. 400-900, eds. L. Webster and M. Brown (London, 1997). 12 R. Hodges and D. Whitehouse, Mohammed, Charlemagne and the Origins of Europe (Ithaca, 1983) (French translation, with revisions, edited by C. Morrisson, Mahomet, Charlemagne et les origines de l'Europe (Paris, 1996)). 9

seventh century. During this crisis another system formed, that had its centre in the N o r t h Sea a n d connections to Frisia, D e n m a r k , Britain and Frankish Austrasia; progressively it extended towards the middie of France on one side, a n d on the other towards the Baltic and, through Russia, towards the Byzantine and Moslem lands. D u r i n g the reign of C h a r l e m a g n e this second system b e c a m e the motor of the E u r o p e a n economy a n d the source of financial supply for Carolingian politics. T h a n k s to it C h a r l e m a g n e was able to carry out the commercial a n d cultural integration of a large n u m b e r of E u r o p e a n regions into a newly unified imperial system. So the two distinct cycles that followed one a n o t h e r between the fifth a n d the ninth century join to form a larger cycle stretching from R o m a n to Carolingian integration. O n e might ask if this scheme is valid also for other aspects of the Transformation. I will therefore refer to the second essay, which is The Rise of the Western Christendom by Peter Brown. Brown outlines the social use of Christianity from the third century to the Carolingian Age a n d beyond as a continuous process, without significant aspects of break-down or discontinuity. In the East as well as in the West, changing societies experienced original forms of devotion and religious life, which flowed through the local ecclesiastical institutions, a n d found audience and support from the social a n d political powers which formed in the regions of the f o r m e r R o m a n Empire. Processes of ethnogenesis and of disaggregation are implicit in Brown's reconstruction. Special attention is paid to the experiences had by the societies that lived on the frontiers of the R o m a n and of the Christian World, because they felt the constraint of past institutional schemes less, a n d so were freer to find original forms of Christianity m o r e suited to the new social order. In the eighth a n d ninth centuries, both in the East a n d in the West, patterns of religious experience, which h a d been established during the previous centuries, continued and were further elaborated; but the general sense of the tranformation now appears to be different. Religious experiences were, so to speak, normalized a n d in this form b e c a m e essential components of societies that strove for social a n d political stability, whilst coming to be located within definite territories a n d closing their frontiers. This h a p p e n e d in the East through the labours of Iconoclasm; in the West with the making of the Carolingian C h u r c h a n d its universai expansion. T w o different trends can therefore be distinguished within the rise

of western Christendom from the fourth to the ninth century: the first was characterized by creative experiences, though incoherent and dispersed; the second by concentration and normalization. Within this last one, which stretches chronologically through the eighth and ninth centuries, the Byzantine World and Western Europe a p p e a r for the first time "with unequivocal evidence"; previously they had not yet been recognizable. This chronology of change in religious processes differs slightly from the chronology of the economic systems as outlined by Hodges and Whitehouse. T a k e n together both studies suggest that the great transformation from the R o m a n to the Carolingian World comprises two phases that have different structures and meanings; in their suecession one can see the shift from a historical situation in which the R o m a n World is still the prevailing protagonist, to one which is appropriately characterized as the Rise of Europe. T r a n s f o r m a t i o n can be conceived of as a continuous process, but it seems to pass through internal phases or cycles that account for the historiographical distinction between Late Antiquity and Early Middle Ages. T h e changing historical situation can also be perceived in other fields, as for instance in political life, where from the end of the seventh century, new institutional forms and ideological concepts appear. It is easier to describe than to explain the change of the sociocultural a n d economic structure at the turn of the seventh century. Chris W i c k h a m has proposed what is probably the most organic thesis on the subject; 13 he considers the crisis of the state as being responsible for the breakdown of the ancient social order. T h e fiscal a n d a n n o n a r i a n machinery built by the R o m a n Empire was initially kept alive by the b a r b a r i a n states, but it collapsed between the fifth a n d the seventh centuries; this m e a n t the end of a system of withdrawal, distribution and investment of wealth that was the basis of the ancient social order a n d greatly influenced Mediterranean trade. In this sense the b a r b a r i a n invasions recover their importance, not as military or political catastrophes, but because the kingdoms that were created were not able to preserve state control over a society that was already striving to get rid of its fiscal burden.

13

C.J. Wickham, "The Other Transition", Past and Present 113 (1984), pp. 3-36, now in idem, Land and Power. Studies in Italian and European Social History, 400-1200 (London, 1994): idem, "Marx, Sherlock Holmes and Late Roman Commerce", The Journal of Roman Studies 78 (1988), pp. 183 193, now in idem, land and Power.

It is more difficult to characterize the structural situation that was typical for the second cycle of the Transformation. Probably one could refer to the completion of the social system that resulted from the breakdown of the R o m a n institutions. T h e new hegemonic groups which formed in the Frankish kingdoms through the Merovingian a n d the proto-Carolingian periods are the connective tissue between the two historical cycles a n d the economic and geographic settlements that were typical for each of t h e m — p r o v i d e d that those élites are not seen simply as feudal rent-keepers, but as customers of trade, and supporters of and participants in religious and cultural experiences which they helped to spread outside the Frankish space. Another explanatory hypothesis points to a less evident transformation that could have set in motion very important economic mechanisms. T h e increase of agrarian production, above all in northern regions, may have fostered population growth a n d stimulated the rebirth of trade by making larger a m o u n t s of wealth available. Speaking about aspects and rhythms of long-lasting transformations, some consideration should also be given to the role that the T . R . W . p r o g r a m m e assigned to the Carolingian Age, as marking the arrival of a significant cycle of socio-cultural transformation. T h e question arises as to whether the Carolingian Empire really succeeded in making a unified politico-institutional body out of the E u r o p e a n territories over which it extended its rule, a n d if it really managed to create an integrated a r e a w h e r e goods a n d culture circulated lastingly. Furthermore, one can w o n d e r if the Carolingian Empire really was the first coherent form of an integrated Europe. As far as political unity is concerned, it is well known that the answer is negative. T h e Carolingian unity did not last more than fifty years a n d disintegrated quickly a n d dramatically, although it left a heritage of concepts and institutions largely c o m m o n to the countries that had been part of it. N o r was the Carolingian settlement a complete and organic form of European integration. Typically, scholars w h o studied the shaping of Europe from Late Antiquity unanimously went beyond the Carolingian Age to see the tenth or even the eleventh century as a meaningful turning point in this process. O n c e m o r e I will mention Christopher Dawson a m o n g the classics, and a m o n g m o r e recent scholars, Peter Brown a n d Klavs Randsborg. 1 4 14

Κ. Randsborg, The First Millennium A.D. in Europe and the Mediterranean. An Archaeological Essay (Cambridge, 1991).

O n l y at the end of the first Millennium—to use the term that was launched by the latter—did Anglo-Saxon and Scandinavian regions become fully integrated with those regions that had been part of the R o m a n Empire and afterwards of Carolingian Europe. O n e might add that this time-lapse fits in exactly with another big pattern of transformation in the origins of Europe, a n d one that makes reference to the commercial revolution of the economy: the pattern worked out by Pirenne and more recently by R o b e r t o Sabatino Lopez, 15 who both placed the beginning of a new m a j o r cycle of the European economy in the eleventh century. This last reference helps to answer the problem of how effective the Carolingian economic integration was. T h e Carolingian system may have promoted the growth of disposable goods a n d favoured their circulation; but doubts arise concerning the durability and solidity of the movement; one might wonder what sort of influence the political disorder that caused the disintegration of the Empire had on the production and trade of commodities. Above all, one must c o m p a r e the signs of economic reorganization that a p p e a r in the Carolingian Age with the idea, steadfastly maintained by Lopez, that throughout the Early Middle Ages the European economy was underdeveloped; notwithstanding the slow recovery that went on during the ninth and tenth centuries, a truly commercialized economy seems to flourish only from the eleventh century onwards. A few remarks about the ways in which the developing shapes of E u r o p e have been described will constitute the final set of my reflections. T h e research carried on in the T . R . W . p r o g r a m m e largely adopted the historical judgement according to which the transformation of the R o m a n World ended in a shift of the geo-historical balance a m o n g the European regions. T h e R o m a n economic and cultural system was centered on the M e d i t e r r a n e a n , from which social and cultural models spread towards the continental regions of Europe. During the transformation, the centre of the E u r o p e a n system shifted to the northern regions of the Continent, while the M e d i t e r r a n e a n countries b e c a m e a periphery or were lost to E u r o p e a n intégration, as h a p p e n e d to Spain after the Islamic conquest and to Italy, at least for the parts that escaped the Carolingian rule. This is the

15

R.S. Lopez, The Commercial Revolution of the Middle Ages (Prentice Hall, 1971).

well-known Pirenne model, that has not been superseded by recent research, which has only pushed the new centre of Europe further northwards, from Austrasia to the N o r t h Sea basin. Having been born in the southern fringe of Europe, I would like to make a case for a concept that is less influenced by the centreperiphery model, a n d which is, I hope, m o r e suited to account for the general structure of Europe, during and after the Transformation. T h e Mediterranean did not become an underdeveloped area as a consequence of the disintegration of the R o m a n World. Although a contraction of social and economic life may have taken place on its shores in connection with the collapse of the R o m a n system and with Islamic expansion, from the end of the eighth century symptoms of new political and commercial vitality appear. T h e most evident of t h e m is the resumption of Islamic expansion, now towards Italy, soon followed by new Byzantine enterprises. Both these Mediterranean powers endeavoured to impose their political and economic influence on the southern regions of Europe, particularly on the Spanish a n d the Italian peninsulas. At the same time, these two countries were reached by the Frankish expansion and were involved in the Carolingian designs of imperial organization. It is evident that both Spain and Italy were just as relevant for the Franks as they were for the M e d i t e r r a n e a n powers; their importance was due not only to their strategic position, but to ideological values and probably to commercial interests. As a consequence, the two great systems that formed during the eighth a n d the ninth centuries—the Continental a n d the M e d i t e r r a n e a n system—met and confronted each other in the two southern E u r o p e a n countries. It is worth noting that neither of these systems succeeded in extending its control over the whole of either peninsula. T h e Moslems failed to conquer Spain entirely, and as a consequence the northern regions fell u n d e r the influence of the Carolingian Empire; on the other h a n d C h a r l e m a g n e did not m a n age to conquer the whole of Italy; he renounced the regions south of R o m e where the Moslems and Byzantines extended their influence. Consequently the b o r d e r stabilized across both peninsulas. Spain and Italy should therefore be thought of as frontier areas in the very sense used by Peter Brown to characterize Ireland, Britain or Frisia in the Barbarian Age. This m e a n s that they should not be considered as peripheries which received impulses a n d information from distant centres. R a t h e r they were territories in which societies experienced original situations, due to the m a n y opportunities arising

from contacts with neighbours who lived according to different cultural and economic systems. These forms of social construction were not confined within the limits of the regions where they had their origin and development; they served as suggestions or opportunities to the larger systems with which the frontier regions were connected. In my opinion a concept like this can help to explain the particular position of Spain and Italy in the process of European integration from the eighth century onwards (once more the chronology broadly coincides with that of the cycles already discussed). T h e northern regions were the standard model for overall social and cultural developmcnts, but their élites were interested in acquiring resources and experiences from the southern regions. This is true not only as regards precious wares and commodities of M e d i t e r r a n e a n origin, but also as regards old and new traditions, literary wisdom and social eustoms. It is sufficient to r e m e m b e r the importance of the papacy the biggest Late Antique institution that survived the crisis of the R o m a n W o r l d — f o r the Carolingian C h u r c h a n d Empire. N o r were contacts sought only from the northern side, but from the southern one too, even though the traces of this are more vestigial. In short, centric one throughout could help

I would propose a polycentric model instead of a monoas being better suited to interpretate the rise of Europe the transformation of the R o m a n World: a model that us to eventually get out of Pirenne's shadow.

THE TRANSFORMATION OF THE ROMAN WORLD: R E F L E C T I O N S O N FIVE YEARS O F W O R K T h o m a s F.X. Noble

Early in the present decade conversations a m o n g a small n u m b e r of scholars from several countries resulted in a proposal to the European Science Foundation for support of a project to study " T h e Transformation of the R o m a n World". Although this general problem has long attracted the attention of professional a n d a m a t e u r scholars, not to mention poets, novelists, and film-makers, this seems a particularly good time to look again into this most fascinating, perduring, and important of historical problems. Perhaps, indeed, Rome's glacial evolution into m a n y p o s t - R o m a n worlds is the single most important process in the long development of the West (and not just of the West, to be sure). But f r a m i n g the problem this way already reveals some of what is today distinctive about scholarly approaches to the period, roughly, 400 to 900. First, R o m e changed, evolved, transformed slowly, at different m o m e n t u m s in different places, and with widely different results. Second, it follows that if R o m e did not "fall" at one time or for one reason, then there would quite naturally be m a n y p o s t - R o m a n worlds. T o say that the R o m a n Empire turned into the Middle Ages just will not do any more. This is a particularly o p p o r t u n e m o m e n t for reflecting anew on R o m e ' s transformation. O n e reason relates to the chronology of scholarly study. T h e r e is today a general awareness that the Late Antique World, say the period from 300 to about 600 or 700, must be understood on its own terms a n d according to its own internal dynamics. A corresponding awareness now impels students of the early Middle Ages to situate their own period with respect to the World of Late Antiquity. Clean-cut distinctions inside the period from the fifth century to the ninth are no longer visible, but no new models are presently available. A second reason relates in interesting ways to the changes in Europe in recent years. A division into East and West that seemed p e r m a n e n t crumbled in 1989. But to the east a n d west of the old " I r o n C u r t a i n " there have in fact emerged, or re-emerged, countless national and regional entities. I

cannot help but wonder if the old idea that one R o m a n order gave way to one other order of some kind was not dealt a death blow when the E u r o p e a n order that dominated the second half of the twentieth century yielded not one or two but countless successors. Even as N A T O expands east a n d the E u r o p e a n C o m m u n i t y gains both m o r e m e m b e r s a n d greater responsibilities, we see both new Europe-wide challenges to large scale political orders a n d incredible local richness and vitality p r o m o t e d a n d publicized by regional governing and granting agenciers seeking to assure everything from historic customs to regional economies. C a n our current perceptions of the dazzling diversity of the Late R o m a n World be rooted, at least in part, in our c o n t e m p o r a r y ways of thinking about the world in which we all live? After all, Benedetto Croce did say that all history is contemporary history. A third reason relates to the nature of scholarly communication today. Every M a y some 2,500 medievalists gather at Kalamazoo. N o w about half that n u m b e r assemble in Leeds in July. E u r o p e a n research students regularly travel to universities, libraries, archives, a n d conferences outside their own countries. National traditions are still strong, but they are breaking down. Similarly, specialists in particular disciplines are slowly but surely learning the hard lesson that they must both talk and listen to practitioners of other disciplines. Interdisciplinary work is not desirable, it is fundamentally necessary. But true interdisciplinary work must be sustained over long periods of time, and applied to well defined problems, to make real advance possible. Such advance is the slowly won gain of a dawning perception that no discipline should be regarded, condescendingly, as merely ancillary to another. T h o s e who participated in this project are more familiar than I a m of its history a n d development. Surely no one w h o works in the "transformation" period has been u n a w a r e of this project. But some of those who read these words m a y not have a very full picture of just who did what, where they did it, and u n d e r what circumstances. A brief description of the scope a n d structure of the T r a n s f o r m a t i o n of the R o m a n World project may therefore be helpful as a background to the reflections that I shall offer presently. 1 Moreover, preparing ' My remarks are based on: my participation in the Isernia meeting; the seven long, informative newsletters produced under the aegis of Ian Wood; and numerous conversations over several years with involved participants. For much information and advice, and some valuable criticisms of an earlier draft of this paper (which began life as a public lecture), I wish particularly to thank Ian Wood.

these introductory remarks helped me greatly to focus my observations. Before continuing, however, I would like to record my deep gratitude to those who m a d e possible my participation in the project's final plenary conference in Isernia (28 J u n e to 3 July, 1997). Five "Working G r o u p s " were assembled, each one taking as its task the investigation of a particular aspect of the transformation. T h e groups were: Imperium, regna et gentes; Settlement in T o w n a n d Countryside; Production and D e m a n d ; T h e Transformation of Beliefs and Cultures; Power and Society. Well over 100 participants in the project hailed from m o r e than twenty countries. T h e overall work of the project was coordinated by Javier Arce, Evengelos Chrysos, and Ian W o o d . T h r e e times dozens of participants convened in pienary conferences: at Mérida in September of 1994; at Le Bischenberg in April of 1996; a n d at Isernia in late J u n e and early July of 1997. Beyond this, however, more than thirty meetings of the individual working groups took place in over thirty different venues ranging from the north of England to Istanbul, from Copenhagen to Carthage. Every working group consisted of specialists in several different academie disciplines, of people from different lands, and of impressive n u m b e r s of w o m e n and j u n i o r scholars alongside the m o r e traditional and older male professoriate. G r o u p meetings sometimes had as few as five or six participants a n d sometimes as m a n y as eighteen or twenty. Twelve or thirteen seems to have been the average n u m ber of those present. T h e sheer magnitude of the tasks before them caused two groups to effect entirely friendly splits. In 1994 G r o u p II split so that one part could focus on rural society and one on cities. In 1995 G r o u p IV divided so that some could concentrate on material culture while others could look m o r e especially at texts. D u r i n g 1997 five m a j o r exhibitions were m o u n t e d in Stockholm, London, Leiden, Bonn, and Thessaloniki. Each of these used local m u s e u m resources to illustrate some of the most important themes embedded in the transformation project. An "Exhibitions G r o u p " held meetings of their own. A stunning volume, containing eight essays that serve to introduce to a wider public m a j o r themes of the whole project and also catalogues and pictures from the exhibitions, appeared early in 1997 u n d e r the expert editorship of Leslie Webster a n d Michelle Brown of the British Museum.‫ ־‬In time eighteen or twenty

2

The Transformation of the Roman World, A.D. 400-900 (London, 1997).

volumes will be published by Brill of Leiden u n d e r the general editorship of Ian Wood. In fact, just as the Isernia meeting was taking place, the first of these volumes appeared. ' Such tabulations make for pretty dull reading. Let me test the reader's patience for just a m o m e n t longer with a brief accounting of the m a j o r themes to which each of the working groups addressed itself. G r o u p I, u n d e r the coordination of Walter Pohl of Vienna, began by looking at the legal bases for the accommodation of the barbarians in the Late R o m a n World. After intensive explorations of that theme, the group turned to signs of ethnic identity, and then to perceptions of regna gained f r o m both written and material evidence. M o r e broadly, this group asked basic questions about rhetoric, representation, a n d discourse, as well as about processes of acculturation. Finally, G r o u p I, along with all the others at Le Bischenberg, inquired into frontiers. G r o u p II, under the direction of Miquel Barcelo of Barcelona, proposed at first to study mining, archaeometallurgy, coin production, a n d the shift to "feudal" peasant settlements. Soon the work of this group turned to an exploration of new agricultural regimes, but its reports evinced a certain tension as institutional arrangements seem to take the upper h a n d over study of field morphology, tools, tilling, and the logic of settlements. W h e n G r o u p II split, one section continued to look at the logistics of p o s t - R o m a n settlements while the other section addressed itself to mining, coin production, and the sources of bullion. Later still, the first group returned to mining and the sources and availability of bullion. T h e second section, now led by Gianpiero Brogiolo, took u p the critical issue of cities, focusing on the idea of the town, the Christianisation of u r b a n areas, a n d the transformation of u r b a n topographies. G r o u p III, with Richard Hodges at its head, started out by inquiring into the possibility of a post-Pirenne paradigm for issues relating to economic change. Intense and fruitful discussions produced a resolve to study in turn three periods: first, the sixth century, the "century Pirenne forgot"; second, the decisive years a r o u n d 700; third, the high Carolingian period. Overall this group asked penetratring questions about why certain peoples and places were more or less important in creating regional economies. 3

Kingdoms of the Empire: The Integration of Barbarians in late Antiquity, ed. YV. Pohl, The Transformation of the Roman World I (Leiden, 1997).

G r o u p IV, a dyarchy under Alain Dierkens of Brussels and Patrick Périn of R o u e n , decided to study "cultural c o m m u n i c a t i o n " by looking at two big problems: the transmission of R o m a n culture and the elaboration of new cultures. Soon this group determined that iconography might prove a useful optique for asking questions about changes in beliefs a n d mentalities. It was decided that the H y p o g é e des Dunes of Poitiers might provide an excellent test-case for a variety of methods and approaches. As it b e c a m e clearer that even m o n u ments speak a language, participate in a "social g r a m m a r " , it seemed desirable to let some work on the language of iconography while others, organized under M a r c o Mostert of Utrecht, explored the more traditional world of texts a n d asked about relations between text and image as well as about the transformation of tradition and the actual transmission of both knowledge and systems of thought. Meanwhile, G r o u p IVa also decided to investigate sanctity. G r o u p V, coordinated by Frans T h e u w s of Amsterdam, started by seeking to bring historians a n d archaeologists together in discussions of social organization and burial customs. Eventually this group focused on rituals of power and on places of power. Perhaps more than any other single group the m e m b e r s of G r o u p V sought to build bridges between their own studies and those of other groups. From my remarks below it will be seen that, in my view, some of these groups pursued their programs of research more rigorously than others. G r o u p I, for example, developed an exemplary evolution of themes, each one growing logically and coherently from the one before it. G r o u p III, in deciding to tackle Pirenne, set out a chronological model that perfectly bracketed M o h a m m e d and Charlemagne. Moreover, each meeting of this group undertook both a relentless critique of the work accomplished to date as well as an ambitious program of new work. Hodges' leadership and Chris YVickham's sparkling, on-going criticisms proved invaluable in keeping the group on task. G r o u p V addressed f u n d a m e n t a l questions in original a n d effective ways but seems to have been slowed and diverted by a constant awareness that its work overlapped the labors of m a n y other groups. Group II, by contrast, seems to have had considerable difficulty defining its targets. Each of its several starting points seems admirable in itself, but there were too many false and partial starts. Perhaps, in fairness, this to-ing and fro-ing can be attributed to the sheer impossibility of addressing both town a n d countryside, to tackling farming regimes as both h u m a n communities a n d technological problems,

a n d to exploring coinage from the standpoints of economic networks and of technical matters of mining and metallurgy. It is worth observing that the work of G r o u p l i b , by selecting a restricted set of urban issues for study, arrived rather quickly at coherent results. G r o u p IV is to be praised for opening u p problems of almost terrifying complexity but, perhaps, criticized for moving ever f u r t h e r from any c o m m o n points of departure or complementary conclusions. W h e n he s u m m e d up the project last D e c e m b e r in Budapest, Ian W o o d said that the T r a n s f o r m a t i o n of the R o m a n World Project h a d "in certain ways become a reconsideration of Pirenne's thesis". 4 And one working group explicity took Pirenne's model as a basis for organizing its studies. Perhaps the reader will not object, then, if I take Pirenne's justly famous book as a framework for organizing a few thoughts on what it is the project has achieved. Pirenne's thesis is so familiar that I will not tax your patience by rehearsing it in detail. As I go along, I will take up one or another aspect of that thesis as a way to structure my reactions to the prodigious labors of this project's participants. N o scheme, I know perfectly well, is adequate to the task of explaining what h a p p e n e d between the fourth century and the ninth. But already that simple statement offers me an opening. For Pirenne provided an explanation or, since we have all become post-modernists, a master narrative. His narrative had a point and a perspective. T h e point is that "without M u h a m m a d C h a r l e m a g n e would have been unthinkable". T h a t point has not stood up to the test of time a n d virtually nothing that I have read or heard suggests that anyone wishes to insist upon it any longer. But his perspective is a n o t h e r thing entirely. At the risk of some oversimplification, Pirenne's perspective was rooted in a notion of continuity. R o m e did not fall, in fact R o m e did not even decline, in the period between about 300 and 600. R o m e changed, " t r a n s f o r m e d " you might say, but in f u n d a m e n t a l respects she was still there until the onslaught of Islam forever changed the shape of the M e d i t e r r a n e a n world. Chris Wickham has called the sixth the last R o m a n century. O f course, there had been master narratives before Pirenne's. T h e

4

Now published as "Report: T h e European Science Foundation's Programme on the Transformation of the Roman World and the Emergence of Early Medieval Europe", Early Medieval Europe 6, 2 (1997), pp. 217-27, at p. 217.

most familiar of these is the one associated with Edward G i b b o n , although, and without seeking to detract from the literary or scholarly achiemcnt of G i b b o n , the view he maintained was one that had a long history before he canonized it, so to speak. If Alcuin or Bernard of Chartres were prepared to acknowledge that their own contemporaries perched on the shoulders of giants—were, we might say, the products of transformation rather than of catastrophic change— there arose with those spectacularly self-absorbed Italians of the fourteenth century a conviction that they could look to the past "as from a fixed point in time", to borrow the apposite formulation of Erwin Panofsky. For other figures of the Renaissance, the Reformation, the Scientific Revolution, a n d the Enlightemnent, there was a comforting certainty that the present was good at least in part because it represented a recapturing of what had been best in the world before the forces of barbarism a n d superstition had inexorably snuffed out the tapers of civilization. Ironically, as Paolo Delogu pointed out elegantly in one of his talks, Pirenne himself was actually less interested in explaining the early Middle Ages than he was in getting to the urbanized world of the High Middle Ages. 5 T o be sure, some G e r m a n humanists, a few Enlightenment figures such as Rousseau, and not a few Romantics protested that the R o m a n s had become effete and m o r i b u n d and that, in reality, the vigour of the barbarians breathed new life into a decadent cultural order. But no matter which of these two perspectives prevailed, the master narrative was the same. At some point in, probably, the fifth century one civilizational order was substituted for another. Pirenne was not of course the first to argue for long-term continuity in the Mediterranean world. Alfons Dopsch preceded him by some years. But there was a crucial difference between the Austrian and the Belgian. T h e former, arguing basically on economic grounds, asserted a very long decline that ran from the early days of the high Empire until the hey-day of the Carolingian world. For the latter, it was just the opposite. T h e world was awash in gold, in prosperity, until the Mediterranean was closed by the followers of the prophet. Still, G e r m a n scholars such as Josef Vogt, French writers such as Andre Piganiol a n d Ferdinand Lot, Santo Mazzarino and his Italian

5

See now P. Delogu, "Reading Pirenne again", The Sixth Century: Production, Distribution and Demand, eds. R. Hodges and W. Bowden, The Transformation of the Roman World 3 (Leiden, 1998), pp. 15-40.

colleagues, and many English historians condnued to write as if R o m e had fallen. This is certainly how American students were taught their history until very recently. This point of view, sometimes characterized as " T h e tyranny of the classicists", reflects an interpretation of western history that is at once catastrophic a n d lachrymose. Most of us just do not think that way any more. Some of us who have laboured in this vineyard surely began to wonder, a generation ago, why it was that the world of Pierre de Labriolle, Andre Chastagnol, Henri-Irenée M a r r o u , and Peter Brown had not had a more visible impact on both popular a n d academic understandings of the "Transformation of the R o m a n World". For these scholars were the true heirs of Pirenne. In speaking as they did of "Late Antiquity",—a term that seems to have appeared first a m o n g art historians in the 1890s these scholars slowly but surely built u p a mighty edifice of interpretation that m a d e it all but impossible to accept Gibbon's narrative—however m u c h we may still enjoy reading that unsurpassed story and however stunningly m o d e r n we must acknowledge some of his specific interpretations to have been. In one respect, then Pirenne has been totally vindicated. T h e R o m a n world changed, but it was not swept away. It was transformed or, perhaps, it provided the stable framework for its own transformation. T h e period from 300 to 900 evidences both change and continuity operating in still mysterious combinations. But, then, the noun transformation d e m a n d s reflection on the verb "to transform". All of the project's participants have been thinking hard about who or what transformed w h o m or what. T o speak a bit more specifically, Pirenne obviously identified the seventh century as the decisive turning point: "the period", he said, "inaugurated by the establishment of the Barbarians within the Empire inaugurated no absolute historical innovation. . . . In short, the essential character of " R o m a n i a " remained M e d i t e r r a n e a n . " H e r e is a veritable a g e n d a of questions. W a s the seventh century decisive? W h a t does it mean for the barbarians to have been established within the Empire? And who or what were these barbarians? Did " R o m a n i a " ever have an essential character? If so, was that character Mediterranean? You will forgive me if I leave on one side Pirenne's virtually metaphysical remark about "absolute historical innovation". Before I turn to an attempt to recapitulate some of the answers I have heard to these questions, let me raise a few m o r e general concerns. Pirenne was an historian of almost unparalleled imagina-

tion. But his imagination was formed by the schools and the university, by the nineteenth-century intellectual world, of which he was a product. Questions that all of you have been asking, that all of us ask today, never occurred to him. I am not referring to, say, Pirenne's argument that the G e r m a n s did not evolve an original art because they were totally u n d e r the tutelage of R o m e , unlike the insular peopies who, he believed, were innocent of R o m a n influence. In this instance, Pirenne was simply wrong. O r , to put it more charitably, scholarship has progressed dramatically in the last sixty or seventy years. W h a t I do mean is that virtually every piece of evidence that Pirenne adduced, and not Pirenne alone I hasten to add, is today contested ground. Pirenne treated Gregory of T o u r s as if he were a vast archive. Gregory was chock-full of facts and those facts, carefully extracted from Gregory's textual swamp as today swords or coins are recovered from bogs, told a talc. T h e Gregory of T o u r s m a d e familiar to us by Ian W o o d , Martin Heinzelmann, Walter Goffart, and others is simply not a Gregory that Pirenne would have recognized. Pirenne was contemporary with some of the most celebrated Monumcntists. T h e i r understanding of Quellenkritik bears little relationship to our way of reading texts. Naturally, we still depend on the texts which the Monumentists established, and we envy their skills. But once we lay the texts on our desks we read them differendy, we think differendy about how they c a m e to be composed, and we ask new questions about their audience. How, then, do we extract evidence from such texts? D o we in fact merely track down signifiers that point to meanings and representations rather than to concrete realities? Is there a world "out there" beyond those texts to which the texts somehow give us access, or is there nothing "beyond the text"? In studying Gregory's world are we, in fact, studying Gregory himself and how he chose to represent his world? Are we disentangling t h r e a d s of discourse? If so, do those threads of discourse ever join in a coherent conversation? I do not m e a n to isolate Gregory. I put him in the dock because Pirenne did so. All the same questions apply to Cassiodorus, Jordanes, Isidore, Fredegar, Bede and Paul the Deacon. Saints' Lives are no less troublesome. Charters are by no means easy to understand. And how do we read the discourse of bracteates, high crosses, door-jambs, evangelist-portraits, apse mosiacs, field shapes and tools? Pirenne was much impressed by the persistence of Latin. Since Christine M o h r m a n n

began teaching us about all the varieties of Latin produced in this period of transformation, we have, surely, become less confident than formerly about the m e a n i n g of the language we read as a cultural artifact. Archaeologists—about w h o m m o r e a n o n — h a v e handed us m u c h new evidence. For the rest, we have the same evidence that Pirenne did. Sometimes we see it differently because of the patient acquisitions of scholarship. Sometimes we see it differently because we have patiendy and systematically deconstructed a learned tradition. Let me put this a bit differently. T h e talks and discussions I have heard, as well as the newsletters a n d the wonderful volume edited by Leslie Webster and Michelle Brown, address sources repeatedly. But always in a fragmented, isolated, particularist way. I a m struck, amidst such riches, at the absence of explicitly, even philosophically, methodological considerations. This is easy for me to say. I came in late as a critic and can leave without suffering from any devastation that I might leave behind. But all of you had an opportunity to reflect systematically on how a n d why we do what we do. You have learned much. I have learned m u c h from reading and hearing you. But you have not gathered your reflections together in a way that would make it possible for those not now privy to your conversations, and those who will come long after you, to retrace your steps, to rethink your thoughts. W h e t h e r we want to or not, we can all do what Pirenne did. W e cannot always see what you did, or why or how you did it. As examples let me take Nancy Gauthier who speaks eloquently of the ideology of cities a n d of how that ideology persisted long after the sites to which it was attached had been irreversibly transformed. How, then, do we read what the texts say about cities? W h a t are those texts really saying? Mayke de J o n g offers the evocative image of "new maps in the mind". She acknowledges frankly that both we and the objects of our inquiries are enveloped in a world of texts. And yet she would see those texts as representations m o r e than as descriptions. Walter Pohl helpfully asks us to be careful with abstractions like state, bureaucracy, C h u r c h , army, economy, culture, and law. W e know, and argue knowingly, over our own understandings of these terms, but we are less certain about the meanings they carried fifteen centuries ago, or that they bear for other citizens in today's Republic of Letters. T h e differences are important and cry out for systematic exposition—although I fully realize the irony of asking post-modern points of view to produce systematic understand-

ing. I very m u c h hope that the published volumes, either in individual essays or in thought-provoking introductions, will do some of this work of intellectual systematization. T h e r e is at the core of Pirenne's book and this project an unspoken assumption. T h e R o m a n World was, well, something. Later on it became, well, something else. For a scientist to conduct an cxperiment, he or she must control one thing and then measure one or more different things against the control. Pirenne sketched the R o m a n World and then, in a sense, turned his sketch into a moving picture and turned on the projector. In fact, he assumed a kind of normative " R o m a n i t y " which would, in his telling, persist until the film was cut. This project also neglected to set down a base line understanding. Was there such as thing as a R o m a n economy? If so, what was it? In what lands did it exist? If it had a n o r m , can standard deviations be suggested? G r o u p III did attempt to provide such a baseline and, albeit in somehat more s u m m a r y fashion, so did G r o u p l i b , as well as those w h o worked on burial practices. O t h e r teams were less resolute in this effort. W h a t was R o m a n government? H o w did it work? H o w was it meant to work? I have heard a lot about bureaucracy, about fiscal regimes, about emperors and kings, about public agents of authority. G r o u p I m a d e a few gestures in this definitional direction but I have not really heard anyone say "For the sake of a r g u m e n t let us agree to start here." O n what fundamental bases did R o m a n social relations rest? I am a little surprised to see so little attention devoted to basic questions of social existence and behaviour. M a n y sessions talked about élites as if we all knew exactly who the élites were, how they got to be élite, how they maintained or lost that status. Persons lower down the social scale enter the discussion mainly as workers on one or another kind of rural estate. T h a t is one important perspective, but it is not the only one. Did men and w o m e n experience the "transformation" differently and if so, how? Until one has, even if only as a matter of hypothesis, set down some agreed-upon starting points it is hard to follow discussions of continuity, change, or transformation. Similarly, Pirenne sketched the Carolingian World. For him that sketch was sufficiently different from his earlier sketch of the R o m a n World that he could write Q E D and stop. This project has not really articulated or attained a common stopping point in either chronological or geographical terms. In some cases, I detected an attempt to compare the late antique a n d the Carolingian worlds. In other instances,

I discern efforts to explain how it is you get from one to the other. Some discussions break off well before the ninth century while others move deeply into that period. Some discussions never depart from the M e d i t e r r a n e a n littoral while others confine themselves to the far North. Even as I admire and learn from the individual contributions, I have a hard time fitting them together into a coherent picture. But let m e return now to that agenda of questions which I elicited from Pirenne, and offer some c o m m e n t s on how I perceive you to have answered them. Even if only one group explicitly addressed Pirenne, the master's shadow is everywhere. First, then, was the seventh century critical? You have answered both yes a n d no, but mainly no. In a very specific way, those w h o have addressed burial practices, a n d perhaps those w h o have studied construction techniques, have not seen this as a key period. In the history of coinage, the sixth and eighth centuries seem m o r e important. In art history this was certainly not the decisive m o m e n t . In the history of states a n d governments, the seventh century was, again, less important than the sixth or the eighth in the West, although it was decidedly crucial in the Byzantine and Islamic East. In economic history, m a n y trajectories can be traced. Some of these reach an apex between 600 and 700 while some fall to a nadir. Which is the crucial direction? W h a t were the driving forces? A n d are generalizations about tightly-defined regions even prudent? H a d there been a bit more attention here to ecclesiastical history, to monasticism, and to missions the seventh century might loom a bit larger, but these are all histories that themselves depend on h u m a n a n d material resources, political and social power, and local possibilities. In 1958 the Spoleto conference failed to achieve any consensus on the seventh century, and forty years hence we have not done too m u c h better even though we know a great deal more. 6 Pirenne's basic chronology was, then a bit too tidy. But here already is a m a j o r achievement of the project. You have demonstrated that there were m a n y transformations—some dependent, some autono m o u s — t h a t moved at varying m o m e n t u m s . For Pirenne's elegant simplicity you have substituted a troublesome, but valid, confusion. Second, Pirenne denied the critical significance of the establishment of the barbarians, those peoples assigned so prominent a role 6 Caratteri del secolo VII in Occidente, Settimane del Centro Italiano di Studi sull'alto medioevo 5 (Spoleto, 1958).

in the old narratives. Interestingly, you retain the old significance of the barbarians but you valorise that significance in fundamentally new ways. Let me emphasize two of them. In the first place, u n d e r the b r o a d notion of "ethnogenesis", you have moved forward a discussion, already a generation old now ( R e i n h a r d YVenskus' great b o o k ' a p p e a r e d in 1961), a b o u t who, exactly, these barbarians actually were. T h e y were not coherent ethnie groups, not tribes. T h e y did not have long, coherent histories. T h e y were fascinatingly and frustratingly complex h u m a n c o m m u nities that formed, u n f o r m e d , and reformed m a n y times, in different places, u n d e r varying historical situations. T h e r e was nothing predictable in the processes that, taken together, can be described by ethnogenesis. Not all Goths were Goths, as Walter Pohl says. Peter H e a t h e r reminds us that it is hard to figure out what "Gothicness" actually was—then or now. Never again will we have maps with those neat lines of invasion that Walter Goffart has been analyzing, a n d ridiculing, for some years. N o r will we have waves, floods, currents, and plagues of barbarians. N o w we have new communities that were m a d e in the age of transformation in settings that must in every case be explored and explained. T h i s work is productively underway, and Walter Pohl's group has contributed powerfully to our fresh understandings. G r o u p I has investigated who these peopies were, how they thought about themselves, and how they represented themselves. 8 These are all extremely difficult questions. It is no criticism to say that m o r e work needs to be done. In the second place, you have in devastating ways rethought the idea of "establishment". O n e set of critical insights focuses on mechanisms. As Pat Geary has remainded us all, the G e r m a n i c world was perhaps R o m e ' s last and greatest creative act. T h a t is, G e r m a n i c establishment took place, often, on R o m a n terms. Walter Goffart's thesis about the barbarians a n d the R o m a n fiscal system remains controversial, as does the thesis of J e a n Durliat. In Walter Pohl's first volume Wolf Liebeschütz provides an immensely helpful resume c u m critique of Goffart a n d Durliat. 9 Pohl himself points to the R o m a n army as an agent of assimilation. I think that ecclesiastical ' Stämmesbildung und Verfassung (Cologne, 1961). See now Strategies of Distinction: the Integration of Barbarians in Late Antiquity, ed. W. Pohl, The Transformation of the Roman World 2 (Leiden, 1998). 9 "Cities, taxes and the accommodation of the barbarians: the theories of Durliat and Goffart", Kingdoms of the Empire, pp. 135 151. 8

institutions can be drawn more fully into this discussion. Surely, the Latin language was a powerful teacher. Did places generally or buildings in particular elicit or inculcate certain types of behaviour? Groups I, IIb, IVa, a n d V m a y well provide, in their final publications, impressive answers to this question. Another set of insights points to localities. Archaeology alone can tell us, finally, where the G e r m a n s and Slavs settled. And mentioning Slavs is a reminder that your transformation does not stop on the west side of the Rhine and the south side of the D a n u b e . In this vein, however, I might have expected m o r e attention to the Balkans and to Scandinavia, and at least some to the Celtic world. T o the barbarians has been attributed, all too often, the ruralization, or to look at the matter the other way around, the de-urbanization, of the transformation period. In fact, as Nancy Gauthier, not to m e n tion particular contribuitions by other teams, or Javier Arce's excellent chapter in the Webster and Brown book, show very well, both u r b a n and rural change were long-term, complex processes in which the barbarians were not the only participants. After all, the R o m a n World was overwhelmingly rural so perhaps the key questions pertaining to the transformation period are geo-political, or functional ones. At your second plenary conference there was m u c h talk of frontiers. This perspective is fruitful—and quite natural to an American long since taught by Frederick Jackson T u r n e r that a frontier is a dynamic region and not a line on a m a p . Old narratives privileged states and assumed that states were stable entities with, at any one time, fixed boundaries. You have shown, on the contrary, that communities of people existed in remarkably fluid configurations. A n d cultures too. Artistic motifs, commercial networks, coin types, and scripts—to mention only a few examples—were no respecters of any kind of political boundaries. I thrill to think of the stunning, multilayered maps you could draw of the age of "establishment". Another issue concerns the b r o a d notion that the barbarians were rural folk. Fair enough. But you have been exploring the rural regime. Some of you have been studying the local, regional, and international networks within which the rural regimes operated. Some of you have focused on rural regimes as examples of particular kinds of political, personal, a n d legal relationships. Moreover, G r o u p II initially p l a n n e d — a most welcome decision—to ask about agriculture as such: soils, climates, plants, tools, etc. W h e n you have published

your volumes, Pirenue's way of thinking about "establishment" will have vanished. A few questions remain on the margin. I have already mentioned broad areas left out of account. I have detected little attention to historical d e m o g r a p h y — a notoriously elusive topic, to be sure. I wonder about daily life and the extent to which this has been explored as a way to view change and continuity in the transformation period. Second, we can ask about Pirenne's " R o m a n i a " . Just what was this? For Chris W i c k h a m it was an economy driven by the fiscal mechanisms of the R o m a n state. T h i s largely vanished although Jean Durliat detects more continuity than others are willing to concede. For Nancy Gauthier, it was an urban civilization and, in her finely nuanced account, an u r b a n ideology. She presents us with a neat paradox: the ancient city was transformed out of all recognition but its ideology persisted. Does Romania score or lose a point in this match? M a r c o Mostert draws a parallel that is in equal measures amusing and illuminating between zoological gardens and ways of organizing knowledge. H e says, uncontroversially, that Romania meant a certain way of thinking about, of organizing, of presenting information. Romania,s way of systematizing its intellectual heritage changed dramatically. So far, however, G r o u p I V b has not given us a m a p of the zoo. Perhaps Romania meant Empire. If so, then it is obvious that the Western Empire departed the stage between 476 and 800, but the show went on in the Byzantine east and perhaps in the caliphate too. But the ghosts of Empire were everywhere. Look at the bracteates in Leslie Webster's catalogue. T h i n k of the imperial posturing in the various G e r m a n i c kingdoms. Romania was a stern teacher. She m a d e people think about Empire even when none was immediately available. O f course, Romania did not simply vanish without a trace after a thousand years. Here your project has done some of its most interesting work. You have asked " H o w did Romania persist?" T h a t is, what were the media a n d mechanisms that communicated Romania's message to her heirs? Buildings did this—as the archaeologists are showing us every d a y ‫ ״‬and so did art works. So too, law books, wills, and charters. So did personal a d o r n m e n t s such as military, ecclesiastical, and civil clothing. Ecclesiastical ritual—whether or not it was done "right" is something that G r o u p V has been at pains to understand -was a powerful broadcasting network. T h e interpretation and

assimilation of two mighty learned traditions, the sacred and the profane, kept Romania both alive a n d vital. Romania could not have vanished but what exactly does it m e a n to say that it survived? T h i r d , was Romania Mediterranean? In one sense, the answer to this question is a simple a n d obvious "yes". But on closer inspection, things grow more complicated. T h e R o m a n Empire stretched from the Scottish frontier to Mesopotamia. A Briton living in the shadow of H a d r i a n ' s Wall had little in c o m m o n with an A r a b in the Syrian outback. T h e climate of, say, Belgic Gaul differed, and generated a different lifestyle, from the climate of R o m a n Egypt. Late Antiquity produced "Chrétientés" in profusion—Armenian, Georgian, Syriac, Coptic, O r t h o d o x , L a t i n — t h a t energetically disputed with each other. In short, the unimaginably vast and complex R o m a n Empire was neither more nor less M e d i t e r r a n e a n in any essential way than the n u m e r o u s worlds to which it bequeathed its countless legacies. In saying this, I a m merely reiterating your continual a n d necessary emphasis on diversity. If we wanted to spend a lot of time adding u p the issues for and against Pirenne, I suspect that he would not come off too badly on the whole a n d given when he wrote. But what I really hope I have r e m i n d e d you of is this: you have asked your questions in very different ways, and you have gotten wholly new answers to the old questions. At the most basic level, you have said that when Pirenne focused on continuity, he got it about right, whereas when he explained particular p h e n o m e n a he often got it wrong. So you have demolished the old narratives. H a v e you put anything in their place? No, certainly not, if that would m e a n a new grand narrative, or even a new synthesis. In a way, the picture on the front of Richard Hodges's new book on San Vincenzo 1 0 can stand as a m e t a p h o r for the point at which you have arrived. It shows the head of a prophet from a fresco—here evidencing continuity in an old, rich religious tradition. O n c e it was beautiful, this picture. T h e n , amidst violence a n d neglect, it fell a n d broke into littie pieces. Arduous labors have put a few of the pieces back into place. O n c e it a d o r n e d the a n t e r o o m to the refectory of a great monastery, but now it is displayed on a wall in a museum in Venafro. W e now know, at least, what it is supposed to be. W e know where it once was and we know something of what surrounded it. W e think 10

Light in the Dark Ages: the Rise and Fall of San Vincenzo al Voltumo (London, 1997).

we can interpret its style in the context of other often no less fragmentary examples. It is not as beautiful as it once was. W e are not sure how to look at it. W e cannot be sure how they looked at it. A n d bits are missing, probably never to be recovered. From so extensive a series of investigations, have I detected anything missing? I have already mentioned several things: theoretical reflections on method; base-line starting points; w o m e n and gender; daily life; demography; certain broad geographical areas. I shall elaborate no more on these." Let me mention -tarnen breve—three other topics. T h e r e has been no systematic reflection on law and order. T o d a y no one would think any longer of the transformation period as dark, chaotic, anarchic, lawless. But no one has explicitly addressed the m a n i a in the transformation period for rules and order. I am thinking of the R o m a n law books and their vast dissemination and influence, the hundreds of pages of conciliar records, the dozens of monastic rules, the countless liturgical books, and the m a n y treatises on correct ruling and living‫ ־‬not to mention books on g r a m m a r , rule books if ever there were any. This period was astonishingly interested in doing things right. T h e R o m a n s had their sense of ordo. T h e Carolingians had their norma rectitudinis. Your period links them. Also, there has been little discussion of economic ideas, as distinct from economic practices. T o be sure, there has been talk of the applicability of various m o d e r n theoretical p e r s p e c t i v e s ‫ ״‬markets, surplus extraction, gift exchange. This is fine, and useful. But what of the people of the transformation period? W h a t did they think about the material realities of life? And did their ideas change over time, or in certain places, or a m o n g specific social groups? Finally, I would m e n t i o n language. H o w was it t r a n s f o r m e d ? W h a t role did it play in the transformation itself? W h a t periodization should be assigned to language? W h a t accounts for regional diversity or how can language be made to yield perspectives on regionalism? W h a t vocabularies were fashioned or refashioned? W h e n new words were invented, to what concrete realities and mental constructs did those words point? All of this is a crucial complement to textual studies, but to much more besides. My reading of more than 100 pages of newsletters tells me that the topic of language was raised m a n y

" I should add, however, that discussions in Isernia led to the formulation of a publication plan for two volumes dedicated to issues of gender.

times in different contexts but it never c a m e up in the final plenary conference. I see no evidence that it was ever addressed as a subject in its own right, 1 ‫ ־‬as distinct from its being used to shed light on other problems. Perhaps the publication plans of G r o u p IV can be adjusted to take u p linguistic issues. In concluding, I want to make two points. T o make the first I shall return to Paolo Dclogu's eloquent lecture, delivered on a fine Sunday m o r n i n g in the beautiful setting of San Vincenzo Nuovo—itself a splendidly transformed place. Professor Delogu offered some alternatives to traditional conceptual perspectives or, in simpler terms, some new scholarly vocabulary. H e invited you, and I concur, to reflect on "disruption" and "disaggregation". Significantly, he used neither of these words in a negative, pejorative way. H e finds them helpful as ways of entering discussions of what h a p p e n e d between 400 a n d 900. If carefully applied, I find Delogu's words more useful than change a n d continuity. N a n c y G a u t h i e r pointed out that for the ancient town to die, there had to be a conjunction of forces. Chris Wickham set down five or six factors which had to operate in some interrelationships in order for economic change to occur. Mayke de J o n g paraphrased Robert Markus to say that power b e c a m e "localized, circumscribed, and controllable". I myself might have used the image of "the m a x i m u m organization of the small space". But look where you will, the results of your project tend to be local and plural. I have already c o m m e n t e d on your recognition of the danger of privileging models based on the presumably superior legitimacy of the centralized state. Viewed from this angle, the transformation period would always look like a fragmenting intrusion between two periods—the High R o m a n Imperial a n d the High Carolingian—of "natural" centralization. T o view history this way is to adopt and to make universally valid the perspectives of nineteenth-century intellectuals. T h e y were participants in and commentators upon the ereation of m o d e r n — m e a n i n g here recent—states a n d it is not at all surprising that they should have sought in the remote past the necessary antecedents of their own creations. But to study history in their way is, finally, to study them. W e must make a firm decision whether we shall study nineteenth-century intellectual history or the history of the transformation period taken on its own terms. If we

12 See, however, the papers of Cameron, de Jong, Hannick and Banniard in this volume.

choose the latter, and I think you have definitively done this, then we must willingly acknowledge that life was—is?—always lived, for most people, in the small space. Large entities, political, social, or economic ones, can always be assembled like a mosaic f r o m the tesserae of small spaces. But the tesserae alone are tangibly real and the large picture is an illusion, a suggestion of what reality might be like. My final point concerns memory. I took away wonderful m e m o ries of the dancing children of Isernia, of San Vincenzo—of the site itself and of the almost paternal pride of R i c h a r d Hodges, J o h n Mitchell a n d Federico Marazzi, and of M o n t e Cassino. W h o can forget Walter Pohl's exegesis of Procopius' pigs? T h a t story will wind up in all our lectures. M e m o r y will not soon let go of M a r c o Mostert's "deconstruction" of the three walls in our lecture hall a n d his metaphorical connection of those walls with the learned traditions of the West. But I really mean the memories you are making. Leslie Webster said she was part of the " a n o n y m o u s " portion of the project. T h e n she said that 30,000 people saw the London exhibition. 30,000 peopie! W h o a m o n g us does not crave that kind of anonymity? Your publications will make memory. Your lecture halls and seminar rooms will make m e m o r y . Y o u r five ambitious exhibitions will long be r e m e m b e r e d . Leslie Webster's book, born of one of those exhibitions, immediately takes its place as the best introduction to the transformation period. T h e result will be that, at the end of the second millennium, you will change forever how we think about life in " T h e First Millennium"—the title of a book by one of your collaborators. Let us hope too that universities, learned societies, a n d national funding agencies will r e m e m b e r what can be accomplished by international teams of scholars equipped with the time and resources to undertake a p r o g r a m of research unprecedented in both its international, interdisciplinary scope and its original achievements. Knowledge is always fragmentary and tentative, at least in this world, as St. Paul reminds us. But for your courage and industry in helping the future to understand the past, in helping us to understand a world remote from o u r own a n d yet strangely like it, I applaud you, and I thank you most warmly for allowing me some small share in your work.

M O D E S O F C O M M U N I C A T I O N : AN A F T E R W O R D Ian W o o d

Re-reading the papers delivered at M é r i d a brings h o m e the importanee of the topic of " M o d e s of C o m m u n i c a t i o n " , and reinforces the accuracy of Averil C a m e r o n ' s final c o m m e n t : " T h e theme of communication, it seems to me, comes near to including the whole of what this project has been about. Above all, it implies movement, and therefore it accommodates change.'" C o m m u n i c a t i o n is, indeed, at the heart of the question of the transformation of the R o m a n World. Yet it was only at M é r i d a that the E u r o p e a n Science Foundation project on the " T r a n s f o r m a t i o n " dealt head on with the issue, a n d as T o m Noble remarked in his plenary address at Isernia, one of its chief aspects, language, "never c a m e up in the final plenary conference". 2 It is, therefore, useful to take stock, however briefly, a n d draw together some of the themes that arose out of the Mérida conference. First there is language itself. T h e period of the transformation of the R o m a n World saw m a n y important changes in this field, some of them part of a steady process of development, particularly in Latin itself, and some of them new. Within the Latin language there is the metamorphosis discussed by Michel Banniard, from "une c o m m u nauté latinophone en une pluralité r o m a n o p h o n e " . 3 T h e growing acceptance of the sermo humilis, which accompanied the christianisation of the R o m a n provinces, marked a period of linguistic compromise, in which the accessibility of language, even to the illiterate, was a matter of concern. But this in its turn led to a régionalisation of the language. A new set of fissures within the Latin language was i n t r o d u c e d with the search for linguistic purity p u r s u e d by the Carolingians and their Christian advisers, which resulted in the newly cleansed Latin of the élite being separated off from the proto-Romance of the majority of the population.

1 2 3

See above, p. 125. See above, pp. 275.6‫־‬ See above, p. 223.

As for that newly cleansed language, it b e c a m e , in Mayke de J o n g ' s definition, a M a n d a r i n : a n d as such it had two significant roles: within government, a n d as the religious language of the Latin West, m u c h as Arabic was to become the language of government and religion in the Islamic world of the southern Mediterranean. In both religion a n d government, language a n d its mastery was tied up with the organisation and control of knowledge, an issue raised by Averil C a m e r o n . 4 Christianity had a m a j o r part to play both in the initial acceptance of the sermo humilis a n d in the split between the classicising Latin of the Carolingian reformers a n d the p r o t o - R o m a n c e of the majority. Christianity also had its role in another set of linguistic developments charted by Christian Hannick, for it was the spread of Christianity, a n d m o r e particularly the translation of the Bible, which seems to have prompted the creation of Armenian and Georgian alphabets, and thus to have transformed those languages from being purely oral to being written as well. At the other end of the period of transformation, the role of Christianity in the development of a Slavonic alphabet, or rather more than one alphabet, is better known than the developments in the Caucasus, even if the history of Glagolithic a n d Cyrillic is more complex than was once thought. O t h e r languages would also have merited consideration along similar lines to those set out by Christian Hannick: there are the various G e r m a n i c languages in the West, where once again, Christianisation h a d m u c h to do with the development of written f o r m s — a n d , in the case of Old English at least, the written language came very quickly to have a legal function, thus echoing—albeit at a distance—the bipartite roles, in religion and government, of Mayke de J o n g ' s M a n darin, although here the spoken language was never confined to the élite. In the East one might add the histories of Coptic and Syriac, which were also vehicles for a flourishing Christian literature. T h e e m e r g e n c e into the limelight of languages which h a d long been eclipsed by Greek a n d Latin, and the establishment of alphabets for languages which had hitherto not been set down in writing, is certainly one notable aspect of the transformation of the R o m a n World. T h e r e is more to the issue of language, however, than its transcription and development. At the heart of Michel Banniard's model is the matter of linguistic register. T h e question of changing linguistic 4

See above, pp. 116-8, 119-20.

registers was also a matter confronted by Lennart Rydén in his consideration of the development of Greek hagiography, a n d the same question surfaced in Javier Arce's exploration of the Vitas Patrum Emeútensium, and in Ian W o o d ' s discussion of Latin saint's Lives in the early medieval West: it also lies behind Peter Brown's comments on the decline of a rhetorical culture and the rise of one dominated by the Bible and its codices. 3 T h e Bible and Christian language is central to m u c h of what is covered by these lectures. T h e complexities of language use a n d language register in a bilingual or multilingual world were what most concerned Walter Pohl— and multilingualism must have been normal in m a n y of the armies of the transformation period. Mastery of languages was thus as vital for access to and dissemination of military knowledge as it was for the control of religion and the administration. M o r e immediately it was a necessity for survival, as the stratèges Gilarkios discovered to his cost. 6 In a bilingual or multilingual society the use of language is not limited to the straightforward conveyance of information, or to the conveyance of information in the appropriate register. In a bilingual society, or even in a society that had once been bilingual, the choice of language could have a further significance. This is a point raised by consideration of a m o n u m e n t like the Ruthwell Cross, with its two languages and two alphabets, set in a landscape where the natives would have spoken a third language. 7 Professor Oikonomedes, in examining the seals of the Byzantine world, sees Latin as sometimes being employed simply because it was regarded as a noble language, even where Greek was the language of everyday and of the administration. 8 Language thus takes us beyond any straightfoiward definition of verbal communication to wider questions of discourse and representation—and thus to a whole range of issues noted by both Averil C a m e r o n and Walter Pohl. And this in turn opens up the question of social language. Here, alongside the use of words is the use of objects, as for instance belts, which, as Walter Pohl remarks, could be employed in a variety of constructions, including that of status

5

See See ' See 8 See 6

above, above, above, above,

pp. 27-8. pp. 128-30. p. 41. p. 55.

a n d perhaps of ethnicity 9 —though ethnicity, its m e a n i n g a n d construction, raises a host of other complications. In the world of the successor states of western Europe and even of the newly visible cultures of the eastern M e d i t e r r a n e a n , representation was a significant issue, and one that went alongside n a m i n g a n d categorisation. In considering the issue of modes of communication it was inevitable, therefore, that the papers given at Mérida should not be confined to the issue of language. M e a n s of communication, other than those of the spoken or written language, were from the start given a high profile. T h e question of artistic vocabulary is at the heart of Niels Hannestad's work, for a consideration of the survival of sculptural traditions is also a consideration of the development of a language of Christian iconography. Exacdy what was being conveyed by some of the recurrent elements of early Christian iconography is what concerned Beat Brenk. T h e a p p a r e n t anonymity of m u c h representation in Late Antiquity was carefully decoded, to show how it was used to present auctontas of various types to the onlooker. Anonymity does not at all necessitate uniformity. T h e mental and emotional way in which the onlooker read, or ought to have read, imagery was what primarily concerned Peter Brown, who insisted that historians should pay attention to the reader as well as the image being read. H e charts a seismic shift in the patterns of response to images in the course of the period of the transformation of the R o m a n World. T h e exact chronology of that shift varied from region to region, but that the period saw a massive set of changes in modes of communication, in both visual a n d verbal terms is not in doubt. Whatever the lacunae in this collection of papers—in terms of languages a n d other vehicles of representation not explored, and in terms of other areas of communication, such as trade (which appears only in Javier Arce's piece, 10 but which, as Evangelos Chrysos remarked, was integral to the theme of the conference)—the lectures delivered at M é r i d a sketch the main areas of a subject which is, as Averil C a m e r o n insisted, central to the whole question of the transformation from Antiquity to the Middle Ages. These lectures deserve to be taken together and treated as explorations of a single major theme.

9

See above, pp. 135-6. It is, of course, a central issue for Paolo Delogu, and an issue of significance for Thomas Noble, in their Isernia lectures. 10

INDEX

Agathias, 75, 76, 81 Albanian language, 205, 212, 213 Alcuin, 93, 96, '107, 108 Alexander the Great, 176, 177, 178, 179, 195 Alexander Romance, 195 Amman, 192, 198 Ammianus Marcellinus, 113, 202 Ambrose, 160 ff. Anglo-Saxons, 255 Antioch, 211 Aphrodisias, 184, 186, 190, 192, 195, 196, 199 Arabic language, 63, 65 Architecture Roman, 8 Visigothic, 10 Arianism and anti-Arianism, 5, 7, 9, I 1. 14 Aristotle, 195 Armenia, 209, 210 ff, 220, 221 Armenian language, 205, 206, 208, 209, 212 Armenians, 206 Asclepius, 178, 190, 191, 197 Attaleiates, Michael, 78 Augustine of Hippo. 229 Augustus, 188, 195 Baltic Sea, 252 Basil of Caesarea, 207 Basilicas, 8, 9, 179 Bilingualism, 47 ff, 61, 234 Book of Durrow, 42, 43 Boris, Khan of Bulgaria, 214, 215,

Constantine IV Coptic language, 205 Cordoba, 236 Cvril ('Apostle of the Slavs'), 215, 2 1 6 ff. Cyrillic alphabet, 217 Dionysus, 191, 197, 198 Dura Europos, 21 Epistola de litteris colendis, 65 Ethnogenesis, 114, 115, 139, 212, 246, 247 ff, 252, 271 S. Eulalia, I ff Eznik of Kolb, 206, 209 Feudalism, 262 Fidelis (bishop of Mérida), 5, 10,

11 ff. Florilegia, 117, 118, 221 Genesios, 77, 78, 80, 81 George the Monk, 76, 80, 81, 82 Georgia, 210. 219, 220, 221 Georgian language, 205, 211, 212 Gerasa - basilica, 179 Germanic peoples, 247, 272, 273 s. Gervasius, 161 ff, 166 Gospels, 72, 233 G Ū ths, 271 Gregory the Great, 18, 19, 20, 22, 30, 33, 44, 96, 168, 169, 231, 233 Gregory of Tours, 33, 94, 104, 119, 136, 231, 267

216, 218

Bulgaria, 209, 213 ff. Caracalla, 188 Carolingians, 63 ff, 106, 244, 245, 248, 252, 253, 254, 255, 256, 269, 275, 276 Cassiodorus, 117, 267 Charlemagne, 252 Chiragan,^ 187 ff, 193, 195, 200 Choniates, Niketas, 80, 81, 82 Constantine I, 206

Hadrian, 188, 203 Heraldeios, 213, 214 Herms, 193, 195 Hilarius (bishop of Aries), 158 Hrabanus Maurus, 67, 68 Hypatius (bishop of Ephesus), 17 Iconography, 28, 29, 36, 37, 40, 41, 43, 178, 180 ff. Icons, 32, 34, 44, 153 ff, 169, 175 Iranian language, 208, 209

Isidore of Seville, 117, 136, 230, 267 Islam, 174

Paul (bishop of Mérida), 3, 5, 10,

11 ff. Paulinus of Nola, 155 ff, 160, 169,

Jerome, 228, 229 Jerusalem, 211 - Holy Sepulchre, 39 J o h n of Damascus, 117 J o h n of Ephesus, 133 Jonas of Bobbio, 93, 96, 99, 100, 101

Judaism, 120, 174, 199, 201 Julian the Apostate, 181, 194 Julian of Toledo, 230, 233 Justin, 197 Justinian, 49, 84, 119, 120 Kievan Rus', 213, 217, 218, 219 Kinnamos, J o h n , 79, 81 Komnena, Anna, 79, 80, 81, 85 Koran, 63, 65 Koriwn, 205, 206, 207 s. Lawrence, 36 Leo the Deacon, 77, 85, 90 Leodgar of Autun, 97, 98, 105 Leontios of Neapolis, 87, 88, 90 Liberius (pope), 158, 159, 160 Libraries, 121 Life of Symeon the Fool, 88, 89, 90 Liudger, 108 Marcus Aurelius, 188 s. Martin of Tours, 25, 32, 157, 169 Mediterranean Sea, 251 Menorah, 177, 179 Merovingians, 97, 99, 100, 102, 103, 121, 122, 137, 140, 232, 254 Mesrop-Mastoc', 205, 207, 212 Miracles, 123, 232 Miracles of St. Demetrius, 50 Monasticism, 66, 69, 95, 97, 246 Mosaics, 38, 147, 160, 161, 162, 167, 168, 169, 179, 180, 184 Muses, 196 Naples - baptistry, 147, 148 North Sea, 252 Occitan language, 241 Odysseus, 182 Paganism, 24, 119, 120, 173 ff. Paul (apostle), 156, 158, 198

182 Peter (apostle), 198, 200 Persia, 207 Photius, 72, 73, 215, 219 Pilgrimages, 250 Plotinus,' 150, 151, 196 Poitiers - hypogeum, 39, 263 Porphyry,' 150, 151 Praeiectus of Clermont, 97, 98, 105 Priscus, 133 Procopius, 75, 76, 77, 80, 81, 85, 128 ff, 138, 140 s. Protasius, 161 ff. Prudentius, 26, 182 Psellos, Michael, 73, 74 Ravenna, 35 ff, 149 ff, 162 ff. Rhetoric, 134 Rimbert, 96, 108 Ripon, 40 Rome, 181, 182, 207 Esquiline Hill, 176 - s. Clemente, 177 Ruthwell Cross, 40 ff, 45 Sarcophagi, 174 , 177, 179, 202 Sasanian empire, 20, 21, 203 Septimius Severus, 188 Serenus (bishop of Marseilles), 18, 19, 22, 23 Severan dynasty, 188 Severus (patriarch of Antioch), 22 Severus Alexander, I 75 Siege of Mérida (712), 2 Sigillography, 48, 51 ff Skylitzes, J o h n , 77, 81 Slavs, 213, 214, 215, 272 Socrates, 195, 196, 197, 202 Sol - cult, 175, 176, 178, 179 Sulpicius Severus, 155, 157, 160, 169 Symeon Metaphrastes, 73, 74, 82 Synagogues, 179 Syriac language, 205, 206 Theodoret of Cyrrhus, 82, 182, 197 Theophanes the Confessor, 76, 77,

80, 81

Theophylact Simocatta, 76, 80, 82, 132, 134, 138 Thessaloniki Hagios Demetrios, 167, 168 Tiridates III, 206 Toledo, 4 Tours, Council of (813), 235 Trade, 253 Trajan, 188

Vatican City - s. Peter's, 179 Venice - s. Mark's, 146, 166, 169 Visigoths, 3, 4, 95, 97 Willibald, 106, 107 Zonaras, John, 78, 81, 220

LIST O F ILLUSTRATIONS

Abbildungen 1-27: 1. Venedig, San M a r c o , Porphyrtetrarchen 2. Porphyrporträt des Galerius aus G a m z i g r a d - R o m u l i a n a 3. Saloniki, Oktogon im Palast, Ehrenbogen mit Clipeus des Kaisers Galerius 4. R o m , Museo Nazionale R o m a n o , Kopf einer Togastatue, u m 400 5. Neapel, Baptisterium, Lebewesen des Menschen, u m 400 6. R o m , Museo Nazionale R o m a n o , sog. polychrome Fragmente (Christus der Bergpredigt) 7. R o m , Museo Nazionale R o m a n o , sog. polychrome Fragmente (Heilungsszene) 8. R a v e n n a , S. Vitale, Apsis, Christus 9. R a v e n n a , S. Apollinare in Classe, Apsis, Christus 10. Bonn, F.J. Dölger Institut, Bildnis eines Philosophen 1 1. Sinai, Katharinenkloster, Christusikone 12. Rom, Praetextat-Katakombe, Arkosol der Celerina, Bischof Liberius 13. R o m , Praetextat-Katakombe, Arkosol der Celerina 14. Mailand, S. Vittore in ciel d'oro, Hl. Ambrosius 15. Mailand, S. Vittore in ciel d'oro, Hl. Gervasius 16. Mailand, S. Vittore in ciel d'oro, Hl. Protasius 17. R a v e n n a , S. Vitale, Bischof Maximian 18. R a v e n n a , S. Vitale, Hl. Vitalis 19. R a v e n n a , S. Apollinare Nuovo, Märtyrinnen 20. R a v e n n a , S. Apollinare Nuovo, M ä r t y r e r 21. Thessaloniki, Hagios Demetrios, Hl. Demetrios 22. Thessaloniki, Hagios Demetrios, Hl. Demetrios mit Bischof 23. Kiev, M u s e u m , Ikone mit den heiligen Sergios und Bacchos 24. Welschbillig, Villa rustica, Vespasian 25. Welschbillig, Villa rustica, Zeus 26. Welschbillig, Villa rustica, Barbar 27. Welschbillig, Villa rustica, Bildnis eines a n o n y m e n Jünglings

Figures 1-9: 1. E n t h r o n e d statue of y o u n g Christ. R o m e , M u s e o delle T e r m e (h.: 0.72m.) 2. Young Christ, detail of 1 3. Sol, from Silahtaraga. Istanbul, Archeological M u s e u m (h.: 0.90m.) 4. D i o n y s o s , f r o m C h i r a g a n . T o u l o u s e , M u s é e S a i n t - R a y m o n d (h.: c.0.65m.) 5. T h r e e Graces. Antalya, Archaeological M u s e u m (h.: 1.79m.) 6. Togatus. Aphrodisias, depot (h.: 2.15m.) 7. Togatus, detail of 6 8. Sage. Istanbul, Archaeological M u s e u m (h.: 0.60m.) 9. Old Fisherman. R o m e , M u s e o delle T e r m e (h.: 1.42m.)

ABBILDUNGEN 1-27 FIGURES 1 - 9

Abbildungen 1-27 belong to Mit was für Mitteln kann einem physisch Anonymen Auctoritas Verliehen werden ? by Beat Brenk, pp. 143-172. Figures 1-9 belong to How did rising Christianity cope with Pagan Sculpture? by Niels Hannestad, pp. 173-203.

Abb. 2. Porphyrporträt des Galerius aus Gamzigrad-Romuliana.

Abb. 5. Neapel, Baptisterium, Lebewesen des Menschen, um 400.

Abb. 8. Ravenna, S. Vitale, Apsis, Christus.

Abb. 9. Ravenna, S. ApoUinare in Classe, Apsis, Christus.

Abb.10. Bonn, F.J. Dölgei‫ ״‬Institut, Bildnis eines Philosophen.

Abb. 11. Sinai, Katharinenkloster, Christusikone.

Abb. 12. Rom, Praetextat-Katakombe, Arkosol der Celerina. Bischof Liberius.

Abb. 13. Rom, Praetextat-Katakombe, Arkosol der Celerina.

Abb. 14. Mailand, S. Vmore in ciel d'oro, Hl. Ambrosius.

Abb. 15. Mailand, S. Vittore in ciel d'oro, Hl. Gervasius.

Abb. 16. Mailand, S. Vittore in ciel d'oro, Hl. Protasius.

Abb. 19. Ravenna, S. Apollinare Nuovo, Märtyrerinnen.

Abb. 20. Ravenna, S. Apollinare Nuovo, Märtyrer

Abb. 21. Thessaloniki, Hagios Demetrios, Hl. Demetrios.

Abb. 22. Thessaloniki, Hagios Demetrios, Hl. Demetrios mit Bischof.

FIGURES 1-9

Fig. 1. Enthroned statue of young Christ. Rome, Museo dellc Terme (h.: 0.72m.).

Fig. 2. Young Christ, detail of 1.

Fig. 4. Dionysos, from Chiragan. Toulouse, Musée Saint-Raymond (h.: c.0.65m.).

Fig. 5. Three Graces. Antalya, Archaeological Museum (h.: 1.79m.).

Fig. 6. Togatus. Aphrodisias, depot (h.: 2.15m.).

Fig. 7. Togiitus, detail of 6.

Fig. 8. Sage. Istanbul, Archaeological Museum (h.: 0.60m.).

Fig. 9. Old fisherman. Rome, Museo delle Terme (h.: 1.42m.).