Famine and Pestilence in the Late Roman and Early Byzantine Empire: A Systematic Survey of Subsistence Crises and Epidemics (Birmingham Byzantine and Ottoman Studies) [1 ed.] 0754630218, 9780754630210

Citation preview

It should, after all, be common ground that in a scientific subject every positive statement is simply a hypothesis. Marc Bloch, in the Introduction to his French Rural History

Antiphon, author of tragedies, 4th century BC

BIRMINGHAM BYZANTINE AND OTTOMAN MONOGRAPHS About the series:

Birmingham Byzantine and Ottoman Monographs is devoted to the history, culture and archaeology of the Byzantine and Ottoman worlds of the East Mediterranean region from the fifth to the twentieth century. It provides a forum for the publication of research completed by scholars from the Centre for Byzantine, Ottoman and Modern Greek Studies at the University of Birmingham, and those with similar research interests.

About the book:

This book presents the first analytical account in English of the history of subsistence crises and epidemic diseases in Late Antiquity. Based on an catalogue of all such events in the East Roman/Byzantine empire between 284 and 750, it gives an authoritative analysis of the causes, effects and internal mechanisms of these crises and incorporates modern medical and physiological data on epidemics and famines. Its interest is both in the history of medicine and the history of Late Antiquity, especially its social and demographic aspects. Dr Stathakopoulos thus provides both a work of reference for information on particular events (e.g. the 6th-century Justinianic plague) and a comprehensive analysis of subsistence crises and epidemics as agents of historical causation. As such he makes an important contribution to the ongoing debate on Late Antiquity, bringing a fresh perspective to comment on the characteristic features that shaped this period and differentiate it from Antiquity and the Middle Ages.

About the author:

Dionysios Stathakopoulos read Byzantine and Medieval History at the Universities of Minister and Vienna, and has carried out research at the Centre for Byzantine, Ottoman and Modern Greek Studies at the University of Birmingham. He has taught at the Institute for Byzantine and Modern Greek Studies at the University of Vienna and at the Central European University in Budapest. He has written extensively on the Justinianic Plague and also on Hagiography and Material Culture, and is currently in Vienna doing research on medicine and society in the Palaiologan period

FAMINE AND PESTILENCE IN THE LATE ROMAN AND EARLY BYZANTINE EMPIRE A SYSTEMATIC SURVEY OF SUBSISTENCE CRISES AND EPIDEMICS

BIRMINGHAM BYZANTINE AND OTTOMAN MONOGRAPHS Volume 9 General Editors

Anthony Bryer John Haldon

Centre for Byzantine, Ottoman and Modern Greek Studies University of Birmingham

FAMINE AND PESTILENCE IN THE LATE ROMAN AND EARLY BYZANTINE EMPIRE A SYSTEMATIC SURVEY OF SUBSISTENCE CRISES AND EPIDEMICS

Dionysios Ch. Stathakopoulos

| J Routledge

Taylor & Francis Group

LONDON AND NEW YORK

First published 2004 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business The edition Copyright © 2004 by Dionysios Ch. Stathakopoulos Dionysios Ch. Stathakopoulos has asserted his moral right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this work. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data Stathakopoulos, Dionysios Ch. Famine and Pestilence in the Late Roman and Early Byzantine Empire: A Systematic Survey of Subsistence Crises and Epidemics. - (Birmingham Byzantine and Ottoman Monographs). 1. Famine - Rome - History. 2. Famine - Byzantine Empire - History. 3. Epidemics - Rome - History. 4. Epidemics - Byzantine Empire - History. 5. Rome - History - Empire, 284-476. 6. Byzantine Empire - History To 527. 7. Byzantine Empire - History - 527-1081. I. Title 949.5'013 Library of Congress Cataloging in Publication Data Stathakopoulos, Dionysios Ch., 1971Famine and Pestilence in the Late Roman and Early Byzantine Empire: A Systematic Survey of Subsistence Crises and Epidemics / Dionysios Ch. Stathakopoulos. p. cm. - (Birmingham Byzantine and Ottoman Monographs; 9). Includes bibliographical references and index. 1. Medicine, Greek and Roman. 2. Famines - History. 3. Epidemics - History. 4. Plague - History. I. Title. II. Series. R135.S794 2003 363.8'0937-dc21 2003056041 Birmingham Byzantine and Ottoman Monographs Volume 9 ISBN 13: 978-0-7546-3021-0 (hbk)

Contents List of Tables Preface and Acknowledgements List of Abbreviations Introduction: Negotiating with the Dead

vii ix xi 1

PART I: TYPOLOGY OF CRISES 1

The Late Roman and Early Byzantine Empire

17

2

A Quantitative Overview

23

3

Subsistence Crises: Causes, Location, Duration and Range Nature-induced Crises Human-induced Crises Duration, Location and Range

35 36 46 53

4

Social Response Market Activity Response of Authorities Popular Reaction

57 57 62 70

5

Epidemic Diseases Introduction Smallpox Infections of the Gastro-intestinal Tract Other Infectious Diseases Mass Poisonings

88 88 91 97 100 103

Contents

vi

6

The Justinianic Plague The Chronology of the Plague The Epidemiology of the Plague Was the Justinianic Plague a Pandemic of True Plague'? Social Response

110 113 124 144 146

7

Results Mortality Shortage of Human Resources

155 155 163

8

Conclusion: 'History that Stands Still'?

166

PART II: CATALOGUE OF EPIDEMICS AND FAMINES FROM 284 TO 750 AD 9

Catalogue

Appendices I. Measures and Currency II. Famine Prices of Grain Bibliography Index

177 387 389 395 403

List of Tables

2.1 Phenomena per century 2.2 Famines and shortages per city and century 2.3 Epidemics per city and century 2.4 Famines and shortages per city and century 2.5 Epidemics per region and century 3.1 Duration and range of subsistence crises 3.2 Location of subsistence crises 5.1 Seasonality of epidemics 6.1 Seasonality of plague

23 27 30 32 34 53 54 90 142

This page intentionally left blank

Preface and Acknowledgements Since Late Antiquity one of the central parts of the liturgy consists of supplications: God is asked to deliver humans from famine and pestilence as well as earthquakes, submersion into the sea, fire, sword, enemy incursions and civil war. Famine and pestilence, a recurring pair, must have been grave and constant companions of past peoples to have gained such a prominent position in their fears. This was the intellectual point of departure for my undertaking to study the presence and the interaction of these two phenomena over the long Late Antiquity. The present book began as a doctoral thesis in Vienna (and one of its chapters, even earlier as a Master's dissertation in Minister), but by now it is something completely new. The following people have been instrumental in this process, whose shortcomings, naturally, are entirely my own. My research has profited greatly from the constant debates with the two supervisors of my thesis at the University of Vienna, Johannes Koder and Ewald Kislinger, as its objective surpassed the limits of a catalogue of such crises and concentrated in an equal effort to analyse them using tools long established in other disciplines. Their advice and creative impulse provided me with the means to pursue this attempt and often liberated me from being stuck in unfruitful dead-end cases. Without them I may very well have never brought this task to completion. For their counsel during various stages of my work I would also like to thank Horst Aspock, the late John Laskaratos, Karl-Heinz Leven, Telemachos Lounghis, Brent D. Shaw, Robert Sallares, Argyro Tataki and loannis Telelis. The library of the Institute of Byzantine and Modern Greek Studies in Vienna with its rich collection has been the ideal workplace. A short stay in Birmingham in early 2000 marked the beginning of a return to my school-day Anglo-Saxon roots and this book is partially the product of this interaction. Leslie Brubaker, Anthony Bryer, Marianna Spanaki and especially Ruth Macrides accompanied three wonderful and productive months that saw (almost) the end of the thesis and since then

x

Preface and Acknowledgements

have been present - if nowhere else - in my thoughts and plans. I (and this book in particular) owe a great deal to John Haldon. He was enthusiastic about my research at a time when even I did not know in what direction it would take or if it was really any good altogether. His encouragement was often the drive that helped me pick myself up and go back to my desk during the endless last summer. Lester Little and Michael McCormick as well as the participants of the Justinianic Plague conference in Rome led me to fully conceptualize my own understanding of this phenomenon after reading and writing about it for more than five years. Judith Herrin saved the day on more than one important occasion. For someone writing in a language which is and is not his own there is a thin line between sentences that feel right or even ring right, but are actually grammatically (or otherwise) wrong. Caroline Robinson mastered the task of making what felt right to also read right and she did this with speed and modesty. John Smedley has been the quiet force behind this project all along; without his help and advice I would still be struggling to piece things together. My thanks are also due to Kirsten Weissenberg and everyone at Ashgate. My friends Agorita Bakali, Michael Griinbart, Lutz Humpert, Daniel Meyer, Natalie Moustakli, Stratis and Sam Papaioannou, Domna Paschalidou, Vaso Seirinidou, and Maria Stassinopoulou acted (if you would pardon the analogy) like Virgil in the Underworld, offering the right amount of inspiration, guidance, advice and support that helped me to find the way out of the Inferno of writing and rewriting, through the Purgatory of proofreading to the Paradise of completion. The pleasant task of reading the product of this struggle will be shared with a certain someone from the contrada della civetta. My studies in Vienna were made possible by the generous scholarship awarded to me by Elisabeth and Achileas Kominos. Their constant presence at my side, the frequent discussions we had and the confidence they instilled in me not only enabled me to pursue this enterprise but also brought out the best in me both personally and academically. For that they have my everlasting gratitude. Last, but most certainly not least, I would like to thank my parents who supported me in all possible ways throughout my life so far. Our bond is beyond stereotypes and for that we will continue to inspire each other to strive for excellence in all fields of life.

Abbreviations AGO AG AH AM AnBoll Annales ESC BMGS BWHO Byz BZ CCL CJ CMG cos / coss CSCO CSEL CTh DHGE DOP EHB El GCS GdU GRBS HE HLRE JOB

Acta Conciliorum Oecumenicorum Anno Graecorum, i.e. Seleucid era beginning on 1 October 312 BC Hejira, beginning on 622 Annus Mundi, beginning on 1 September 5493 Analecta Bollandiana Annales: Economies, Societes, Civilisations Byzantine and Modern Greek Studies Bulletin of the World Health Organisation Byzantion Byzantinische Zeitschrift Corpus Christianorum Series Latina Codex Justinianus Corpus Medicorum Graecorum consul / consuls Corpus Scriptorum Christianorum Orientalium Corpus Scriptorum Ecclesiasticorum Latinorum Codex Theodosianus Dictionnaire d'histoire et de geographic ecclesiastiques Dumbarton Oaks Papers The Economic History of Byzantium Enyclopedie de I 'Islam Die griechischen kirchlichen Schrifsteller der ersten Jahrhunderte Geschichte des Untergangs der antiken Welt Greek, Roman and Byzantine Studies Historia Ecclesiastica History of the Later Roman Empire Jahrbuch der Osterreichischen Byzantinistik

Xll

LRE MGHAA

MGHSS MMWR OAW, Phil-hist Kl.Sb PG PL PLRE PO PS Prokopios, BV, BP,BG RAC RE

SC ZPE

Abbreviations

Later Roman Empire Monumenta Germaniae Historica Auctores Antiquissimi Monumenta Germaniae Historica Scriptores Morbidity and Mortality Weekly Report Osterreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, Sitzungsberichte Patrologia Graeca (Migne) Patrologia Latina (Migne) Prosopography of the Later Roman Empire Patrologia Orientalis Plague Studies Bellum Vandalicum, Bellum Persicum, Bellum Gothicum Reallexikon fur Antike und Christentum Paulys Realencyclopddie der classischen Altertumswissenschaft Sources Chretiennes Zeitschrift fur Papyrologie and Epigraphik

A note on Names and Place-Names Following a trend that has been gaining acceptance in the previous years, I have decided to use the following system for rendering Greek names and place-names. Standard anglicized forms of personal names (Theodore, George, John, Thessalonica) shall be used; all other names will be transcribed as literally as possible avoiding the latinized versions established mostly in the nineteenth century. Thus Prokopios and not Procopius, Nikephoros, not Nicephorus and so on.

Introduction: Negotiating with the Dead The title above derives from Margaret Atwood's essay of the same name in which the author suggests that 'not just some, but all writing of the narrative kind, and perhaps all writing is motivated, deep down, by a fear of and a fascination with mortality.'1 What would that suggest about a book on mortality, which is motivated by this intriguing relation to mortality? In any case, the voice of the dead has always been used by those who survived a particular incident to explain, justify, condemn or move to repentance. This is an attempt to seek out this voice (especially of those who have not left us a transcription of it in their writings), listen to it and record the information it offers in a more sober and detached manner than the one adopted by the contemporaries of those survivors, avoiding at the same time the style of doom and catastrophe that is so successful among many modern reports on such phenomena of the past. Subsistence crises of varying degrees and epidemics are at the centre of this enterprise. The latter have formed a respectable topic in history since the nineteenth century.2 Tastes and trends in historical topics as a rule reflect preoccupations and anxieties of the society that produces them. Indeed the interest in epidemics of the past was initially sparked by the cholera epidemics that swept through Europe in the 1830s. This fascination has been a lasting one. The actual historical discourse does not neglect big killers of the past, above all the plague (and within the history of this disease above all the Black Death). Books such as W.H. McNeill's Plagues and Peoples (New York 1976) have drawn attention to diseases as a considerable factor in human history and indirectly helped to somewhat draw the focus of historical writing away from the exclusive retelling of 1

M. Atwood, Negotiating with the Dead: A Writer on Writing (Cambridge 2002) 156. See the excellent overview of early books on the plague by P.M. Getz, 'Black Death and the Silver Lining: Meaning, Continuity, and Revolutionary Change in Histories of Medieval Plague', Journal of the History of Biology 24 (1991) 265-89. 2

Famine and Pestilence

2

political, diplomatic and military events. This was both necessary and fruitful, judging from the absence of any mention of the Justinianic Plague in one of the most important books on Byzantine history, G. Ostrogorsky's History of the Byzantine State (first German edn 1943, third edn 1963, English translation Oxford 1968). However, we are now dealing with the opposite extreme. Works such as D. Keys' Catastrophe (London 1999) and M. Baillie's Exodus to Arthur (London 1999) suggest that natural phenomena, such as the Dust-veil event (see No. 92 in the Catalogue), comets or the advent of the plague are the driving forces of historical causation.3 This book is an attempt to find equilibrium between these two extremes: to place famines and epidemics within the overall context of the political history of Late Antiquity and to discuss to what extent and along with which other phenomena they were responsible for changes within this period. This is the first book to examine Late Antiquity from the point of view of the subsistence crises and epidemic diseases that occurred during this time frame. It is based on an exhaustive catalogue recording all such phenomena from 284 to 750 that has been assembled especially for this study and forms its second part. It is meant to offer a complementary view of this period beyond the turmoil of political and military history, and comments on themes such as continuity and change of social and economic structures and mentalities. The phenomena recorded and analysed in this book may have taken place in a world quite different than the one we inhabit today, but they are far from being irrelevant to our everyday discourse. In some developing countries of the modern world such phenomena continue to occur and to be perceived in the same manner as thousands of years ago. The recent outbreak of malaria and cholera along with widespread occurrence of starvation during the flood-induced crisis in Mozambique in late FebruaryMarch 2000 is only one characteristic example that clearly illustrates the validity of such a suggestion.4 Even the extreme news about the occurrence

3

This deterministic neo-historicism which sets epidemics in the role traditionally occupied by kings and generals is, of course, not new. A negative peak in the production of this sort is occupied by F.F. Cartwright's Disease and History (London 1972) which is advertised with lines such as 'Malaria, smallpox, and other plagues conquered the Roman Empire. Haemophilia was at least partly responsible for the Russian Revolution' and so on. 4 Anyone familiar with the recent publication of Living with Risk: a Global Review of Disaster Reduction Initiatives (Preliminary version, July 2002, United Nations, Secretariat

Introduction

3

of cannibalism in poverty-stricken Moldova conveys images that are not essentially different than the bleak reality reflected in Late Antique famines as discussed in this book.5 The developed world may now be free from the menace of catastrophic subsistence crises, but it is still endangered by epidemic diseases migrating, intentionally or unintentionally, from less developed countries. Diseases such as diphtheria or tuberculosis have resurfaced mostly through dissemination from countries of the former Eastern bloc.6 However, the most serious threat derives from bioterrorism. Even before the dramatic events of September 11 2001 a vivid debate on the use of contagious diseases as biological weapons had evolved.7 The U.S. government conducted an exercise which revolved around the alleged release of Yersinia pestis, the causative agent of the plague, in aerosol form at the Denver Performing Arts Center in Colorado.8 As the exercise was terminated after four days, conflicting reports ranged between 950 and 2,000 deaths, about 4,000 cases of pneumonic plague, while infected patients had been reported as far as London and Tokyo. After September 11 the debate on deadly diseases as biological weapons naturally took a grim, realistic turn. The deliberate spread of anthrax failed to develop into an epidemic, but fear of plague and/or smallpox was great. In August 2002 the state of Israel conducted a large-scale smallpox inoculation campaign, a disease which had been triumphantly declared eradicated by the World Health Organisation in 1979,9 while the United States government has been for International Strategy for Disaster Reduction) will instantly recognise an important number of structural similarities between this and the present book. 5 See the report by K. Connolly, 'Cannibalism is symptom of Moldova's decline', The Guardian 5 April 2001. 6 See A. Eftstratiou, et al., 'Diphtheria', in D. Armstrong and J. Cohen, eds, Infectious Diseases, vol. II (London et al. 1999) Section 6. 34.12. 7 See T.V. Inglesby, et al., 'Anthrax as a Biological Weapon: Medical and Public Health Management', Journal of the American Medical Association 281 (1999) 1735-45; D. A. Henderson, et al., 'Smallpox as a Biological Weapon: Medical and Public Health Management', Journal of the American Medical Association 281 (1999) 2127-37; T.V. Inglesby, et al., 'Plague as a Biological Weapon: Medical and Public Health Management', Journal of the American Medical Association 283 (2000) 2281-90. 8 The exercise, termed TOPOFF, took place in May 2000. It is described in detail in T.V. Inglesby, et al., 'A Plague on Your City: Observations from TOPOFF', Biodefence Quarterly 2 (2000) 1-15. See also B. Vastag, 'Experts Urge Bioterrorism Readiness', Journal of the American Medical Association 285 (2001) 30-32, for the evaluation of this exercise and the fears expressed about the possibility of such actions taking place in reality. 9 P. Brennan, 'Israel Preparing for Worst: Begins Smallpox Vaccinations', NewsMax.com 19 August 2002. For more information on smallpox see Chapter 5.

4

Famine and Pestilence

pursuing a policy of alert regarding the infection since late 2002.10 It becomes obvious that the study of epidemics and subsistence crises of the past is firmly related to the actual anxieties and debates of the present. Epidemics and subsistence crises may in fact constitute the basis of this book. Its scope, however, is not to collect and discuss them per se, but to place them in a discernible context and to deprive them somewhat of their uniqueness as a means to analyse how their actual presence was perceived in the longue duree of Late Antiquity. Epidemics and famines have often been set in narratives as a pair since ancient times.11 Thucydides made a pun concerning the use of both words as contained in a Delphic oracle that stated that 'a Dorian war shall arrive and an epidemic'.12 As the oracle was transmitted the word (loimos) was understood as meaning pestilence, since such a disease was ravaging Athens at the time. But Thucydides assumed that if some years later a famine (limos) were to emerge in connection with a war, people would interpret the oracle as meaning that a famine would occur together with the military actions. Although we cannot be certain if both words were pronounced identically in the fifth century BC (as Thucydides' remark would suggest), in the period we are about to study they certainly were. Furthermore, the ideological context for the joint use of epidemics and famines was set in the so-called 'Synoptical Apocalypse' (Matt. 24, Luke 21), where both phenomena, together with earthquakes and warfare, constitute the signs of the imminent Judgement Day.13 As we shall see in the course of this work these two phenomena did in fact often occur side by side and were frequently interconnected; nevertheless the phonetic and most certainly the eschatological context they were embedded in favoured their joint recording. At this stage it is essential to define and discuss the objects of study around which this book evolves. The first are subsistence crises. According to their intensity and effects they may be divided into two major categories. A shortage 'is a short-term reduction in the amount of available foodstuffs, See now H. Epstein, 'Bugs without Borders', The New York Review of Books, January 16, 2003, 20-23 and the information at the Centers for Disease Control and Prevention at: http://www. bt.cdc.gov/agent/smallpox/index.asp 11 Cf. H. Hunger, Die hochsprachliche profane Literatur der Byzantiner (Byzantinisches Handbuch V/l-2 = Handbuch der Altertumswissenschaften XII/5, 1-2. Munich 1978) 1269 n. 69; Patlagean, Pauvrete, 85. 12 Thucydides, Historiae, ed C. Hude (Leipzig 1925) II 54 (177). 13 Luke 21:11; Matt. 24:7.

Introduction

5

as indicated by rising prices, popular discontent, hunger, in the worst cases bordering on starvation'. Famine 'is a critical shortage of essential foodstuffs leading through hunger to starvation and a substantially increased mortality rate in a community or region'.14 In ideal cases our sources include information that explicitly points to the gravity of a particular subsistence crisis, therefore making its distinction possible. However, there is a large number of cases where the recorded data do not suffice in order to do so. As a result the characterization can only be made using modern scholarly criteria such as those given above. In this process the margin of misinterpretation grows larger in opposite analogy to the availability of information. Ultimately the distinction between more and less serious subsistence crises is a result of the personal judgement of the reviewer. Throughout the book I have decided to characterize as famines a number of subsistence crises for which, although no mortality is recorded, the overall accompanying circumstances led me to believe it had been a serious one. The next object of study is the occurrence of infectious diseases that is the invasion and multiplication of micro-organisms in a host.15 There are three types of outbreak for such diseases. Endemics signify the constant occurrence of a disease in a particular locality or population, a disease with geographical but not chronological limitations. Epidemics indicate the massive, unexpected, chronologically and geographically limited occurrence of a particular disease. Finally, pandemics denote the massive occurrence of a disease with chronological, but no geographical limitation. The interaction of these aspects, chronological and geographical dispersion, is the decisive factor for the differentiation of these forms of occurrence. It should be noted that the three types are not equally represented in the sources. Pandemics are rare, a fact which probably reflects the actual reality. Endemic diseases were as a rule not recognized and therefore not recorded as such in our material; they are discernible only through a longterm observation of disease-related incidents in a particular region. Epidemics, on the contrary, are explosive events that attract the attention of contemporaries and as such they are recorded. Furthermore, when outbreaks of these diseases occur for the first time, as perceived within the 14

Both definitions taken from Garnsey, Famine, 6. This and the three subsequent definitions (endemics, epidemics, pandemics) are taken from H. Moegle, 'Allgemeine Epidemiologie der Zoonosen', in Bakterielle Zoonosen bei Tier und Mensch. Epidemologie, Pathologic, Klinik, Diagnostik und Bekampfung (Stuttgart 1993) 1-14. 15

6

Famine and Pestilence

mind of a particular author, we can safely expect an amount of detailed description to be given. On further occasions, as a disease grows to be more familiar and its occurrences obtain some regularity, information in the sources becomes increasingly sparse. This is certainly the case for the plague, as we shall see in the course of the book. To identify infections recorded in our sources with ones that have been described and analysed by modern epidemiology is certainly not an attempt to speculate on the presence or absence of these diseases in the period in question. Furthermore it is not intended to use modern data about these epidemics and uncritically transpose this data to the past. Caution and an ever accompanying sense of relativity are necessary.16 Nevertheless we cannot use caution and relativity ad absurdum denying any and every remark on past infections, which is seemingly what an actual trend in the history of medicine does. At first we must stress that the scope of the history of medicine and that of a historical work dealing among other things with medical issues are not identical. The aim of this book is not to establish a retrospective diagnosis of past epidemics per se, but to place them in a frame that will be discernible, and therefore suitable for interpretation today. If we refrain from issuing statements about any past medical conditions, knowing and accepting a priori that they cannot be considered anything but plausible hypotheses, we will be thrown back into a state where our work will be limited to merely critically reproducing the contents of our sources. In this way we will not offer answers or solutions to the problems connected with the information given in those sources, but will only - in the best case - be asking more questions. Yet, this has been done over a very lengthy period of time so far. I believe that the fragmentary and often audacious answers furnished by historians today are more valuable in this particular field than the extremely cautious, and thus often very limited, scope of argumentation coming from the natural sciences. This is mainly a difference in the level of discourse and it certainly cannot be easily overcome.17 If historians were made to meet the

G. Rath, 'Moderne Diagnosen historischer Seuchen', Deutsche Medizinische Wochenschrift 81 (1956) 2065-9, passim', K.-H. Leven, 'Krankheiten - historische Deutung versus retrospective Diagnose', in N. Paul, T. Schlich, eds, Medizingeschichte: Aufgaben, Probleme, Perspektiven (Frankfurt and New York 1998) 153-85, passim. It is noteworthy that this discussion is beginning to manifest itself within the history of medicine itself as scholars are asking whether history of medicine as a discipline must be expected to fulfil criteria of the conventional medical science, cf. C. Hummel, Das Kind und

Introduction

1

same criteria as doctors, then the history of disease would be silenced. By working with hypotheses, however, historians offer plausible explanations for many of the past's catastrophic epidemics by incorporating in their work the accounts of contemporary witnesses along with the remains of the material civilization of the period. Their results fill in a vast space that would otherwise be left empty, since modern medicine would often choose not to comment on it on account of the lack of data, or data as modern medicine perceives it. The chronological frame of the book has been set between Diocletian's ascent to power in 284 and 750.18 The choice of the first date as the conventional beginning of Late Antiquity is obvious and it has already proved its efficacy in a number of important works on this period such as A.H.M. Jones' The Later Roman Empire (1964), Alexander Demandt's Die Spatantike (1989) and Averil Cameron's The Later Roman Empire (1993). The final date needs some explanation as it has been selected in order to include the last, dramatic outbreak of the Justinianic Plague that ravaged the Mediterranean around 750. The emergence of Islam and the loss of vast territories for the Empire around the mid-seventh century are usually perceived as the demarcation between the Early and the Middle Byzantine period. Indeed from a political, economical and social point of view we can safely admit that they fulfil this function, as has been shown by John F. Haldon in his Byzantium in the Seventh Century. Hence it would seem preferable to limit the discussion on this period to the years around the middle of the seventh century. But our perspective is different inasmuch as it is centred on phenomena whose perception and effects remained largely unchanged from Antiquity until the advent of the technical innovations brought by the Industrial Revolution. Epidemics and famines were often quite dramatic events that had an immediate impact on the lives of the contemporary populations. Yet, their significance emerges from the fact that they occurred - in various forms and combinations - over and over again. As such, their study must be attempted over a long period of time only to realize that they remain almost immotile or immovable within this frame.19

seine Krankheiten in der griechichen Medizin (Medizingeschichte im Kontext 1. Frankfurt am Main 1999) 7. 18 For a chronological overview of the period see chapter 1. 19 Cf. F. Braudel, 'La longue duree', Annales ESC 4 (1958) 734; furthermore the discussion in E. Le Roy Ladurie, 'History that Stands Still', in The Mind and Method of the

8

Famine and Pestilence

An additional reason for the termination of the period we will study in 750 is that no information on epidemics and/or famines is included in the major Byzantine sources from 750 to the beginning of the ninth century. The texts in question are: Theophanes, the Patriarch Nikephoros, early Byzantine Short Chronicles, the Vita of St. Philaretos and the Parastaseis Syntomai Chronikai - as these are considered the most important sources for this particular period.20 The second half of the eighth century witnessed the intensification of the Iconoclast movement. We can assume that our sources, largely written by members of the opposing party to the imperial politics against icons, would have recorded any crisis that may have occurred in this period and used it as a means to illustrate the expression of divine wrath against the imperial policy that opposed icons, as they had done with the last outbreak of the Justinianic Plague. The absence of records must indicate the absence of such phenomena within this period. Late Antiquity is a very rewarding period for the study of subsistence crises and epidemics. During these five centuries the early Byzantine Empire reached its widest expansion and highest population density only to experience a sharp decline in both of them. As such it incorporates a phase of expansion followed by one of decline giving us the opportunity to examine if and how these crises influenced or were influenced by the demographic, economic and social movement of that period. The geographical frame of the book largely includes the territory under the political control of the Late Roman and Early Byzantine state: Italy, the Balkans, Asia Minor, Syria, Palestine, Egypt and North Africa. This is of course a tremendous simplification, since during the course of the period in question many of these regions were invaded, captured and subsequently governed by other political entities. Italy was controlled by the Ostrogoths from the late fifth century, and after a brief Byzantine interlude under Justinian I, was de facto lost to the Lombards save for a few strongholds of imperial power. Syria, Palestine, Egypt and North Africa were captured by the Persians in the first decades of the seventh century, regained by the Byzantines and subsequently completely lost to the Arabs from the 630s and 640s. Nevertheless all these territories have been included in my study as they constituted part of the Empire during the greater part of the period in question. Epidemics and, to a lesser extent, famines may only be Historian (Chicago 1981) 1-27, passim, and E. Carpentier, 'Autour de la peste noire: famines et epidemics dans 1'histoire du XlVe siecle', Annales ESC 17 (1962) 1062-92. 20 Cf. Karayannopulos and Weiss, 318^0.

Introduction

9

discussed in a wide geographical frame, since only in this way do their movements and effects for a network of regions become evident. In order to pursue the task of analysing such phenomena the material had to be collected, as there was no similar work to draw upon. This collection is the catalogue comprising the second part of this book and the backbone of this enterprise. The analysis in Part I was only possible based on the bulk of accumulated information of the catalogue, which may also serve as an independent work of reference for anyone interested in subsistence crises and epidemics of that period. Both parts are interconnected through cross-referencing. A number of studies served as models for the catalogue. Although they dealt with certain aspects of this topic, none of them, however, was exhaustive in either its chronological or geographical perspective, compared to the scope of the present work. Such works include Ruggini's study on subsistence crises in Italy (Economia e societa nell' 'Italia Annonaria', Milan 1961), Casanova's work on famines and epidemics in the Egyptian source material ('Epidemic e fame nella documentazione greca d'Egitto', 1984) and primarily Patlagean's pioneering book on poverty in Byzantium (Pauvrete economique et pauvrete sociale a Byzance 4e-7e siecles, Paris and The Hague 1977: 73-112), in which she has devoted a chapter on famines and epidemics as causes of mortality. Kohn's article in the RAC ('Hungersnot', 1994) provided ample material on famines in Late Antiquity up until the sixth century with particular emphasis on Italy. An additional note should be made of Telelis' doctoral thesis on climatic phenomena in the Byzantine Empire from 300 to 1500 (Meiecopo^oyiKd cpaivojieva Kai icAi|ia axo Bu^dvuo, loannina 1995)21, as the author has collected data on a large number of climate-induced subsistence crises within that particular period. Supplementary material was gathered from several works referring to the Justinianic Plague.22 The groundbreaking article by Biraben and Le Goff published in 1969 represents the first - and so far the only - effort to establish a chronology of the waves of the pandemic drawing on a vast amount of source material; as such its importance is considerable. Additional texts referring to the plague were procured by Allen (The "Justinianic" Plague', 1979) and 21

In press, Athens 2003. For a detailed overview of literature on the plague see my critical study 'The Justinianic Plague revisited', BMGS 24 (2000) 256-76. Discussion in this section will be therefore limited to a brief presentation of the most important works. 22

Famine and Pestilence

10

Turner (The Politics of Despair', 1987). Due to the lack of information referring to the Byzantine realm after the mid-seventh century (see below) additional data has been supplied by Arabic sources. These texts, written mostly after the ninth century but often preserving older, non-extant material, offer us ample information on those matters. The list of plagues that befell Islam (as a rule comprising five outbreaks before the Black Death) begins at the year 6 AH (627-28 AD) and terminates with the violent pandemic wave of 744-49. The information we can draw from these sources is unique. It concerns mostly the regions that had been recently captured by the Arabs in Syria, Persia and Palestine, but we can safely assume that these outbreaks will have reached, either through warfare, or commerce, the realms of the Byzantine Empire. The Arab sources will be cited in translation; furthermore their discussion will be limited to the results of the thorough studies of M. Dols and L.I. Conrad, as the leading authorities on the epidemics in the Islamic world. After an initial corpus of crises had been collected from the above studies, data was supplemented by a direct search in the source material. Databases such as the Thesaurus Linguae Graecae and the Data Base of Byzantine History of the Centre of Byzantine Research in Athens were utilized as a point of departure, but their yield was quite limited. Finally, a selection of texts considered as valuable for including information on the subject matter was made and these were excerpted in detail.23 The distribution of texts and the amount of information provided by them within the period from 284-750 is neither continuous nor homogenous. The book begins with 284 but the catalogue's first entry dates from 304-05. The Diocletianic era is indeed quite poor as far as primary sources are concerned. The principal writers of that period (Aurelius Victor, Lactantius, Eutropios, Eusebios and the Chronicle of Jerome) include no information concerning epidemics and/or famines for the period between 284 and ca 300. There is a steady flow of works from the fourth century onwards, reaching higher density in the sixth century. However, the seventh and eighth centuries are characterized by their lack of sources. From the few contemporary authors, Theophylaktos Simokattes ends with the ascent of Phokas in 602, the Chronicon Paschale reaches 628 and after 23

These are (in chronological order): Philostorgios, Ecclesiastical History; Ps.-Joshua Stylites; Prokopios, History of the Wars', Marcellinus Comes; John Malalas; Agathias, Histories; Euagrios, Ecclesiastical History; John of Ephesos, Ecclesiastical History; Chronicon Paschale; Theophanes; Agapios of Menbidj; Michael the Syrian, Chronicle. For full references see the Bibliography.

Introduction

11

that Theophanes and Patriarch Nikephoros, both authors of the ninth century, remain the almost exclusive sources of information. There are, of course, a number of other texts, mostly of local character, that offer supplementary data (the Miracles of St. Demetrios, George Pisides, John of Nikiu, Sebeos, John of Biclaro, Jacob of Edessa et al.). However, from the point of view of this study, the information they have to offer is often very limited, since epidemics and famines were probably not among their first rank of interest.24 Apart from consulting the traditional written sources, an effort has been made to include epigraphical and papyrological material. The database Phi, which contains most edited papyri to date along with alterations taken from the ten volumes of the Berichtigungsliste der griechischen Papyrusurkunden aus Agypten (Berlin and Leipzig 1922, and Leiden 1958 -98) as well as a large number of Greek inscriptions, was used, but the yield again was quite limited. The works of Casanova on papyrology and Durliat ('La peste du VF siecle', Paris 1989) on epigraphies were of particular importance in providing additional material from these disciplines. As both epigraphical and papyrological texts are often not securely dated caution is necessary, as one of the most important premises of this work is to include securely dated incidents. In those cases where this is not provided the material has not been used. Each catalogue entry has been arranged as follows. The first line provides the basic information of the entry: Number, Date, Location, and Type of phenomenon followed immediately by a presentation of the exact passage(s) of the source(s) referring to the phenomenon. In those cases when later sources merely repeat the information provided by older ones, these will not be cited or used. For example, Prokopios, writing over a century later, adds nothing to the information given by Jerome, Philostorgios, Sozomenos and Socrates on the Sack of Rome in 410 (No. 45) and as such he is not included. Bibliographical references dealing with the phenomenon are also presented in this section. The titles included therein are not exhaustive, but were chosen based on the criteria of 24

On source-related problems see J. Haldon, 'The works of Anastasius of Sinai: A key source for the history of seventh-century east mediterranean society and belief, in A. Cameron and L.I. Conrad, eds, The Byzantine and Early Islamic Near East I: Problems in the Literary Source Material (Princeton, New Jersey 1992) 125-9. For a detailed overview of the written sources of the period see L. Brubaker and J. Haldon, Byzantium in the Iconoclast Era (ca 680 - 850): The Sources. An Annotated Survey (Birmingham Byzantine and Ottoman Monographs 7. Aldershot 2001) 163-307.

12

Famine and Pestilence

providing new or additional information on the particular subject. After that the content of the source(s) is paraphrased or cited entirely. The most important section is the one that follows, where various aspects of the data are discussed. An effort has been made to date all incidents in a thorough and clear manner and to present them in a way that will be discernible to the reader. In all possible instances adopting established dates without questioning their veracity has been avoided. Furthermore an effort has been made to present the general context of each crisis within this section. Other issues that are reflected on include the identification of stereotypical elements and the explanation of noteworthy, technical details as mentioned in the texts. Under the section 'Related with' I have noted possible relations between the crisis in question with previous or later ones. The catalogue may appear at first to be a collection of various crises that seem to be unique, to occur randomly and without any particular pattern. The data, when placed in its overall context, can withstand modern interpretation and yield results. The first part of the book is an attempt to show that the material included in the catalogue is not amorphous, but can be analysed according to morphological criteria with the scope to reveal those regularities that pervade it. 'Regularity', a term often used by historians pertaining to the Annales-school (Fr.: regularite), is to be preferred over the terms 'norm' or 'historical law'. It designates patterns that occur repeatedly over a long, basically homogeneous period, a 'longue duree\ without, however, a given, normative periodicity. Regularities become apparent from the collection and the interpretation of data; they are not rigid and clearly defined a priori, but try to describe as closely as possible various phenomena that have a long presence in a given period. The general aim of the first part is to make these regularities evident, to collect and compare them, and show their evolution within the time frame of Late Antiquity. Each catalogue entry is broken down into the smaller morphological units that compose it; similar cases are collected and types of crises as they arise from this application are then formulated. Those units may refer to the causes and results of each crisis, taking at the same time into consideration the reaction of the people that were affected by it. A significant number of sources include plenty of information on the phenomena they have recorded and as such provide us with sufficient material for discussion. An even larger number of texts, however, only mention the occurrence of a subsistence or epidemic crisis epigrammatically. The data procured by each such source is of quite limited use by itself; it is impossible to draw conclusions from such a small amount

Introduction

13

of information. If, however, discussion were to be limited to only those phenomena that are well represented in the sources, a misleading, patchy image would emerge. The material's limitations can be mastered by arranging such texts in a way that will reveal their innate similarities and by grouping and collecting these common factors together. The methodological device that underlies this enterprise is borrowed from the brilliant work by V. Propp, Morphology of the Folktale?5 Just as Propp had tabulated folktale motifs by breaking down narration motifs in abstract forms that overlap in similarity, I also developed an analogous system. I drew up several narration motifs covering the whole spectrum of crises from their causes, the social responses to them and the effects they had, and tabulated the actual recorded crises from the catalogue according to these motifs. As a result the similarities between seemingly unique and unconnected phenomena became instantly apparent. These common points of reference were then summarized and discussed, revealing the possible structures that pervade phenomena such as famines and epidemics recorded over a period of almost five centuries. In this way material of, very often, little qualifiable value could be included in the survey, broaden the spectrum of the discussion and make more abstract patterns of crises discernible. The use of this methodological device has been instrumental in the preparation of this work, although any visible traits have been edited out of its final form since they would only complicate and burden the reader. The material basis that a scholar of Ancient and/or Byzantine history has to work with is fragmentary. There are well and less well documented periods, topics that are mentioned or not by our sources, stereotypes that obscure the historicity of certain statements - in short, data that at its best enables us a more or less partial view of these past realities. Furthermore we must reflect on the fact that in most cases the information recorded in the sources is closely interwoven with the author's interpretation, often extending beyond a critical separation of the historical fact from the stereotypical, edifying fiction. As a result we cannot anticipate producing laws of universal validity from this material. We can, however, expect that certain patterns concerning the causes, effects and reaction within the

25 Second edition, revised and ed. with a Preface by L.A. Wagner (Indiana University Research Center in Anthropology, Folklore and Linguistics 10. Bloomington 1968, Reprint Austin 1998).

14

Famine and Pestilence

phenomenon, or combinations of the above that are common in many cases will emerge - in short, that a typology of crises will be revealed.

PARTI TYPOLOGY OF CRISES

Chapter 1

The Late Roman and Early Byzantine Empire Late Antiquity was a period characterized by movement. The Roman Empire was still the most powerful state encircling the Mediterranean, the mare nostrum, but after some two centuries of stability and prosperity after Augustus it was thrust into a prolonged crisis in the third century.1 Some 26 emperors, elevated to power chiefly by the army, succeeded one another in half a century characterized by anarchy, inflation and general social disarray. Diocletian (284-305) was the first ruler to undertake drastic reforms in an effort to stabilize the Empire. Conventionally the history of the Late Roman Empire begins with his reign. He divided the realm under Roman control first into two, then into four parts, to be ruled by a joint team of two augusti and two caesares. This division of the Empire into East and West was a de facto recognition of the different developments in each separate region; as time went by the division deepened and resulted in a permanent demarcation from the end of the fourth century onwards. Other significant reforms included the reorganization of the system of the provinces (they were increased in number and size), the army and the coinage. The gold coin created at the time, the solidus, was to remain the standard monetary issue throughout Late Antiquity. Regarding the specific 1 This is only a brief summary of Late Antiquity intended to provide a background to the phenomena analysed in the book. As such, discussion is limited to those aspects that are relevant to the book's objects of study. For more extensive, detailed examinations of this period see (in chronological order): J.B. Bury, A History of the Later Roman Empire from Arcadius to Irene (395 A.D. to 800 A.D.), 2 vols (London 1899, Reprint Amsterdam 1966); ibid., HLRE- P. Brown, The World of Late Antiquity; AD 150-750 (London 1971); Demandt, Spatantike', Av. Cameron, The Later Roman Empire: AD 284-430 (Fontana History of the Ancient World. London 1993); B.H. Rosenwein, A Short History of the Middle Ages (Ontario 2002).

18

Famine and Pestilence

focus of this book, one of Diocletian's measures was of particular interest. To deal with inflation he issued a decree fixing the maximum prices for vital commodities and services. The result was not the expected one: goods vanished from the market and shortages ensued, turning the edict into a dead letter. Constantine I (324-37) was the son of one of the principal caesares and after a career within this system of tetrarchy he rose to the throne after defeating all his opponents. Constantine's reign represents a turning point: by moving the imperial residence and capital of the Empire from Rome to Constantinople (the official dedication took place in 330) and by showing a marked (though not exclusive) preference for the Christian religion, he laid the foundation of what has been named the Byzantine Empire. There was now a clear shift of power towards the East. Situated strategically at the intersection between Europe and Asia, both by water and by land, Constantinople was designed to become a major urban centre: lavish public buildings and churches were built while the emperor took measures to attract a growing population, such as the distribution of free bread from 332. Constantine continued Diocletian's reform work, so that today scholars have trouble in ascribing specific measures to the one or the other ruler. The Roman Empire still occupied a huge territory stretching from Britain to North Africa and from Spain to Mesopotamia. The Eastern frontier bordering the Sassanid Empire of Persia was in continuous flux; both states were in more or less a constant state of warfare punctuated by occasional periods of peace. This was to remain so until the destruction of the Persian state by the Arabs in the seventh century. Frontiers to the West and the North were persistently menaced by a number of peoples; as a rule the Eastern part of the Empire suffered less than its Western counterpart. The short interlude of the pagan Julian (361-63), the last ruler who was a blood relative of Constantine, did not stop the successful course of Christianity. The Church emerged as a growing economic and social institution in the fourth century. Theodosios I (379-95) put an end to the tolerance of religions other than the Christian faith: pagans, Jews but also dissident Christian groups were gradually marginalized in a process that culminated in the reign of Justinian I (527-65). Meanwhile developments in Eurasia seemed to follow the domino principle: under the pressure of the Huns, large numbers of Goths entered the imperial realm. The fifth century saw these Germanic peoples dominate the political scene in both parts of the Empire. In the East, Germanic

The Late Roman and Early Byzantine Empire

19

officials were integrated into the government and occupied important positions in the state machinery up until 474 when they were liquidated by the Isaurians, a mountainous people from Asia Minor, and their leader the emperor Zeno (476-91). Huns and Goths were either defeated or successfully held back from the Eastern territories and pushed towards the West. As a result, the West was overrun by these peoples and the Roman state ceased to exist. In 410, Alaric's Visigoths sacked Rome and then marched on through Gaul and conquered Spain. The Vandals crossed over to Africa, captured Carthage in 439 and established their own kingdom. In 476 the last nominal western emperor, Romulus Augustulus, was deposed by the Germanic chieftain Odoacer. Another chieftain, the Ostrogoth Thedoderic was sent against Odoacer; he managed to defeat him and conquer Italy, which he ruled, nominally, in the Emperor's name. Gaul was captured and governed by the Franks and the Burgundians. The demarcation between East and West as it had been set by Theodosios I in 395, running roughly all the way from Belgrade to Libya, was now a true frontier separating two distinct entities. The Eastern Empire was an unbroken continuation of the Roman state; Greek was the dominant language and the official dogma followed the orthodoxy of the ecumenical councils. The West was now divided into several Germanic kingdoms; Latin was the dominant language and Christianity was as a rule of the Arian faith. The Eastern provinces in Egypt, Palestine, Syria and Asia Minor went through a phase of growth and affluence. A network of important urban centres, most of them existing since Antiquity but now experiencing a period of increase, held the countryside in their firm grip. A class of great landowners accumulated wealth and power and distinguished themselves and their class through donations and crisis management. Constantinople had grown to become the largest city in the entire Mediterranean. Its ecclesiastical head, the Patriarch of the New Rome, had risen to the second position in the hierarchy of the Church just below the old Rome. Anastasios I (491-518) introduced a period of prosperity and expansion for the Eastern Empire. Through sound economic measures he managed to leave the state with an unprecedented surplus of gold coins at the time of his death. This became the basis for Justinian's I (527-65) ambitious reign. Hailed rightly as the most famous and accomplished ruler of Late Antiquity, Justinian rose to power through his uncle Justin I (518-27). In his time the Empire sought to regain the lost territories in a series of long wars. The Vandal kingdom in Africa was subdued in the 530s, but the

20

Famine and Pestilence

Ostrogoths were not as easy to deal with. The Reconquista in Italy lingered on from 535 until the final defeat and extinction of the Goths in 554. At the same time there was almost constant warfare with Persia from the time of Anastasios until the 560s, although imperial victories and territorial gains were not as decisive as in the West. It was in this world unified by sea and land communications that the plague appeared in recurrent waves from 541 to 750. At the end of Justinian's reign the elaborate new construct of an Empire possessing territory all around the Mediterranean began to collapse as a result of both demographic losses (plague, long wars) and economic hardships brought about by these two factors and the large-scale building activities of Justinian. Until the advent of Herakleios (610-41) much of the regained territory had been lost: the Lombards captured and held Italy from the late 560s, the Visigoths regained the few Byzantine holdings in Spain, while the Eastern front collapsed under renewed Persian attacks. Moreover, a new force emerged in the Balkans: the Avars and the Slavs. From the 580s onwards the Slavs began their settlement of the Balkans, gradually taking almost the entire peninsula de facto out of Byzantine control for the next two centuries. Large parts of the West now enjoyed a period of stability under the Prankish kingdoms in Gaul and Germany and the Visigothic kingdom in Spain. During the reign of Herakleios territorial movement was again considerable. After initial success against the Persians, which led to the recovery of all Byzantine realms (Egypt, Palestine, Syria), the Arab expansion began in the 630s. Until the turn of the century Egypt, Palestine, Syria and Northern Africa were irrevocably lost to the Byzantine Empire, while the Sassanid state had been crushed. The seventh century was subsequently a period of massive restructuring and reorganization as the Byzantine empire fought for its survival. The massive loss of territory especially Egypt, the 'granary' of the Empire - deprived the state of considerable human resources and commodities. From then on Byzantium concentrated on Asia Minor as an almost exclusive source for both. A large-scale reorganization of the army took place in that period, first in Asia Minor, spreading then to the entire Empire. Territory was organized into administrative and military units, the themata, in which both civil and military power was concentrated in the hands of one military commander. Soldiers were from then on recruited among the free peasant smallholders, who gradually offered their military service in exchange for land that enjoyed certain privileges.

The Late Roman and Early Byzantine Empire

21

The Arab foray seemed unstoppable. They menaced Constantinople in 678 and again in 717-18 (see No. 208), failing both times to capture the city. From 711 onwards they succeeded in taking Gibraltar, crossed over to Spain and gradually conquered it. Charles Martel managed to stop their course in the battles of Tours and Poitiers in 732. The new frontiers of the Mediterranean world were now fixed for a period that extended into the next century. The Islamic realms stretched from Spain through North Africa and Egypt to Arabia, Persia, Palestine and Syria. The Byzantine Empire held Asia Minor, (more or less directly) the Balkans and some minor parts of Italy, the rest of which was held by the Lombards. A new kingdom had emerged and was recognized by the Byzantines in 681, that of the Bulgars occupying territory south of the Danube, between the state of the Avars, the Black Sea and the Balkan mountains. The Prankish kingdom covered wide areas of modern western Europe. When the Arab danger seemed to be under control, the Byzantine empire entered a slow, gradual course of recovery. During the reign of the first emperor of the Isaurian dynasty, Leo III (717-41), and the early years of his son and heir Constantine V (741-75), the constant Arab raids in Asia Minor were successfully met and some former Byzantine territory regained. The lasting civil war between the Umayyads and the ultimately victorious Abbasids (744-49) prevented them from attacking the Empire. This unrest was exploited by Constantine who used his newly-established army corps, the tagmata, to assault Arab territory successfully and capture some strongholds. It was precisely at that time that the last wave of the plague united the Mediterranean once more: from Syria to Egypt and North Africa it crossed over into Sicily and Italy and then returned via Greece to Constantinople and ran its full course back to Syria. Both states were decimated, bringing all other activities to halt. This is the final point of the long Late Antiquity discussed in this book. The Byzantine society of the early eighth century was very different from that of one century before. With the loss of the Eastern provinces the fate of urban centres as focal points in the economy and society took the final downturn. Along with the ultimate disappearance of most cities the class that sustained them vanished as well. The senatorial, landowning elite was gradually replaced by a new military and civil service aristocracy sprung from the thematic armies and the court at Constantinople respectively. The capital itself, although holding a fraction of the population it had enjoyed under Justinian, grew in importance as the

22

Famine and Pestilence

struggling state sought to address its problems through a social and cultural introspection.

Chapter 2

A Quantitative Overview The data collected in the catalogue can be used per se to date and characterize a subsistence crisis and/or epidemic that occurred within the chronological frame of this study. The main aim of this chapter is to put together and make evident certain trends that seem to emerge from the collection of the data. The conclusions reached below are based primarily on the mere number of phenomena in a given century, city or region. This can certainly lead to oversimplified results, but since a differentiated, detailed analysis of these phenomena will follow in the next chapters, emphasis has been placed here on those general lines of development as seem to emerge by the juxtaposition of the catalogue's entries. From the 222 catalogue entries recording subsistence crises, 44 cases (respectively per century: 8, 10, 11, 5, 7) deal with the occurrence of such phenomena in close succession and as such are recorded in the same entry; these figures have been added to both respective categories. Additionally, there are five entries (Nos 27, 87, 92, 93 and 210) that do not refer to either famines or epidemics, but have been included in the catalogue because of their relation to preceding or subsequent phenomena; these have not been taken into consideration in Table 2.1. Table 2.1

Phenomena per century

Century

Famines

Epidemics

Fourth Fifth Sixth Seventh 700-50 Total

31 35 37 19 12 134

14 18 52 23 17 124

Total 45 53 89 42 29 258

24

Famine and Pestilence

The column 'Famines' includes all subsistence crises, both famines and shortages. An effort has been made to make a distinction between more and less severe subsistence crises following the criteria established by Garnsey, the most important of which is a dramatic rise in the mortality rate manifest in cases of famine and absent during shortages.1 This can be and in fact has been used in a number of well-documented cases that have been characterized accordingly as famines or shortages in the catalogue. However, in a large number of other cases this mechanism cannot be used other than in an arbitrary way: where there is very little information about a crisis the scholar can only decide according to his/her judgement as to whether the particular crisis was more, or less, severe. If then the results of these categories were to be collected and presented in a chart, this would certainly present a false picture of the events. For these reasons I have decided to put all subsistence crises into one group in this and all subsequent tables. The reader is advised to consult the catalogue for a more detailed presentation of the various crises. A further note should be made regarding the division of the various phenomena according to the centuries they occurred in. This is admittedly a purely conventional partition, as the phenomena on which this study focuses occur independently of such chronological limits. However, this same feature, the fact that these boundaries are conventional, ensures that the specimen we obtain from each century is more or less random and therefore more suitable for its use in this quantitative analysis than those samples we would have obtained by marking the periods in question according to qualitative criteria (e.g., a period beginning with the outbreak of the plague in 541). The first feature to which we should draw our attention is the unequal distribution of famines and epidemics in the various centuries. The sixth century includes more or less twice as many phenomena than any other. This is at least partly due to the outbreak of the Justinianic Plague, an epidemic that caused a sharp demographical crisis throughout the whole Mediterranean, making its occurrence worth recording in a large number of sources. The seventh and the first half of the eighth centuries show a gradual decline in the number of the recorded phenomena. This certainly has to do with the extant source material, as an example shall demonstrate. Between 628 when the major, contemporary source for the seventh century, the Chronicon Paschale, stops, and 750 only 14 phenomena (Nos 183, 184, Garnsey, Famine, X, 37; see also above, Introduction.

A Quantitative Overview

25

188, 189, 190, 192, 199, 204, 206, 208, 215, 216, 220 and 221) out of 44 refer to cities and regions within the Byzantine realm. Even the Byzantine Chronicles (Theophanes, Patriarch Nikephoros) seem to record more crises occurring in the Islamic world than in the territory of the Empire. This cannot signify that Byzantium was not visited by many epidemics and/or famines in the course of the seventh and eighth centuries, merely the fact that we lack the sources that would record them. The prevalence of references to the sixth century certainly reflects both an increase in the occurrence of the phenomena in question - epidemics above all - and a more solid state of the source material compared to previous or later times. The first case is best supported by the example of the plague: scholars have not been able to offer a plausible explanation as to why its first historical outbreak occurred in 541 and why it disappeared after 750. The visitations of the pandemic certainly raised the number of the recorded epidemics in the sixth century and at the same time also the consciousness and the interest of the contemporary writers in such matters. Another possible cause for this interest was eschatological. According to the three prevailing world eras the completion of the year 6000 from the creation of the world fell between 492 and 508. This was the year which Christians held as the advent of Judgement Day.2 According to the synoptical Apocalypse (Matt. 24, Luke 21, Mark 13), the end of days would be preceded by wars, famines, pestilences and earthquakes. Therefore it is not surprising to see that authors of the sixth century developed a particular interest in recording such natural catastrophes.3 In the second case, the matter of the density of source material, the sixth century is particularly rich in texts. An unprecedented number of historical works written by contemporaries has survived, transmitting a detailed picture of this period. Ps.-Joshua Stylites, Prokopios, John Malalas, Agathias, Euagrios, Marcellinus Comes, Menander, John of Ephesos and Theophylaktos Simokattes - to name but the most important ones - have included in their works a great deal of information on the visitation of epidemics and famines. 2

E. Daley, The Hope of the Early Church: A Handbook of Patristic Eschatology (Cambridge 1991). 3 See P. Magdalino, The History of the Future and its Uses: Prophecy, Policy and Propaganda', in R. Beaton and C. Roueche, eds, The Making of Byzantine History. Studies dedicated to Donald M. Nicol [Centre for Hellenic Studies, King's College London Publications 1. London 1993] 3—34; W. Brandes, 'Anastasios ho dikoros: Endzeiterwartung und Kaiserkritik in Byzanz um 500 n. Chr.', BZ9Q (1997) 24-63.

Famine and Pestilence

26

A more differentiated look at the recorded famines can also furnish us with valuable information. Of the 134 phenomena in the catalogue only 14 appear to have been general, interregional famines, results of larger-scale climatic anomalies that had an overwhelming effect on the food production system of the whole Mediterranean. These are Nos 4, 5, 29, 47, 64, 69, 125, 148, 167, 169, 170, 182, 187 and 222. The last three phenomena refer to famines that affected the Islamic realm and have been included in the catalogue for reference. These were indeed rare events, perhaps as rare as shortages of every kind were common.4 The general trends that seem to emerge from this table concern the relation of famines and epidemics over the course of this period. In the fourth and fifth centuries famines are significantly greater in number than epidemics. This suggests a growing population that created pressure on the available foodstuffs, causing a large number of subsistence crises. This is admittedly a Malthusian argument,5 but although a computation of the Empire's population at any time is impossible, scholarly opinion does agree on the general tendency of a rising population in the late fourth, fifth and early sixth centuries, both in the larger urban centres and the countryside.6 The advent of the plague in the 540s inflicted a sharp demographical crisis. From then on the number of famines drops continually and is finally outnumbered by the recorded epidemic cases during the first half of the 4

As Garnsey, Famine, 271, rightly concludes: 'Shortage was common, but famine rare, the outcome of abnormal conditions.' 5 Cf. for example: Th. Malthus, An Essay on the Principle of Population, sixth edition eds E.A. Wrigley and D. Souden (London 1986) 28: 'Population cannot increase without the food to support it', or 309 '[...] no more human beings can grow up than there is provision to maintain. [...] by [frequently obliging] the lower classes of people to subsist nearly on the smallest quantity of food that will support life, [this] turns even a slight deficiency from the failure of the seasons into a severe dearth, and may be fairly said, therefore, to be one of the principal causes of famine.' 6 See P. Charanis, 'Observations on the Demography of the Byzantine Empire', in Thirteenth International Congress of Byzantine Studies, Main Papers XIV (Oxford 1966) 11-12; Cameron, LRE, 114, 123, 179, 188; Treadgold, History, 139$ J. Banaji, Agrarian Change in Late Antiquity: Gold, labour and Aristocratic Dominance (Oxford 2001) 16-22; J.H.W.G. Liebeschutz, Decline and Fall of the Roman City (Oxford 2001) 30-63. On the contrary, J. Koder, 'Historical aspects of recession of cultivated land at the end of late antiquity in the east Mediterranean. Evaluation of land surfaces cleared from forests in the Mediterranean region during the time of the Roman Empire', Palaeoclimate Research 10 (1994) 159-60, writes of'a demographic and economic decline which took place in various stages starting at the end of the fifth century and taking place at the latest in the first half of the seventh century, at very different points in various regions'.

A Quantitative Overview

27

eighth century. Again, this could be interpreted in the same context: the substantial population decline must have relieved the tension on the food market resulting in a smaller number of subsistence crises. Table 2.2 Famines and shortages per city and century

City

4th c.

Const. 1 Rome 8 Antioch 4 Alex. Edessa 1 Thess. Ravenna Amida Total 14 * 700-50

5th c.

6th c.

7th c.

8th c.* Total

9 6 2 1 -

4 6 1 1 1 1 2 16

4 2 2 2 1 11

2 1 3

18

20 23 6 4 2 3 2 2 62

Tables 2.2 and 2.3 make evident that the information flow on famines and epidemics was concentrated on the two largest cities in the empire, Rome and Constantinople. The other cities have been included in the tables as figures for comparison. The number of famines and/or epidemics recorded for them throughout the period in question is far too small to enable us to draw any conclusions. Furthermore, we must again assume that this weak tradition reflects a lack of sources or a lack of interest in those cities in the extant sources rather than the actual reality as to the occurrence of these phenomena. Regarding Antioch in particular we can note that we have no information on subsistence crises in the city after the 440s (No. 55). This can be explained by a series of catastrophic phenomena that hit Antioch in the first half of the sixth century: a devastating fire in 525, a terrible earthquake in 526, followed by two less strong ones in 528 and 532, its capture and pillage by the Persians in 540 and the outbreak of the plague in 542.7 After all these visitations we can safely assume that the city was left 7

In general cf. Downey, Antioch (1961) 519-29 and Charanis, 'Observations' (as in n. 6) 6-7; on the earthquakes see Guidoboni, Catalogue, Nos 203, 206 and 209; on the Persian capture and the plague see Kislinger and Stathakopoulos, Test', 79, 93.

Famine and Pestilence

28

depopulated and destroyed, never to gain the importance it had in earlier centuries. It is clear from the presentation of the data that Rome underwent a drastic status-change in the fourth century. Eight consecutive subsistence crises in one century (seven of which are concentrated on its second half) signify severe problems with the city's provisions. With the foundation of Constantinople and the direction of the Egyptian annona towards the new capital around the 330s Rome experienced a gradual, but - taking into consideration the long tradition behind this measure - nevertheless quite sudden lack of grain.8 In the same period the population in Constantinople experienced only one case of shortage between 330-37 (No. 6): there the causes had been climatic. The absence of subsistence crises in fourthcentury Constantinople indicates that the capital's growing population was as large as or, most probably, smaller than the amount the imported grain could sustain. Until the loss of Egypt and its grain to the Persians and then to the Arabs in the seventh century Constantinople remained quite safe from the menace of famine-induced mortality: one case in the fifth century (No. 61), two in the sixth (Nos 111, the first wave of the plague, and 148), one in the seventh (No. 173, as a result of the capture of Egypt by the Persians) and one in the eighth (No. 216). This last case is the only recorded siege-induced famine that Constantinople had to endure within the chronological frame of the catalogue. In a number of cases where the capital experienced sieges there have been no recorded cases of famine.9 On the contrary, Rome suffered a number of siege-induced famines that were often coupled with the outbreak of an epidemic resulting in massive mortality: Nos 37 (fourth c.), 42, 44, 44, 72 (fifth c.), 94 and 123 (sixth c.); there are no recorded cases for the seventh and eighth centuries. It is remarkable that this famine-induced mortality in Rome is concentrated in a 150-year-period between 397-98 and the 540s. This largely coincides with the long period of unrest that was unleashed upon Italy first by the Visigoths and then by the Ostrogoths. Rome depended on the African grain shipments after the Egyptian annona was directed towards

8

The Constantinopolitan annona officially began in May 332 according to the Chronicon Paschale (531). Cf. Dagron, Naissance, 530-31; Sirks, Food, 202; Durliat, Ville, 250-52. 9 In 626 (cf. Stratos II 491-542); in 674-78 (cf. Stratos V 35-45) and in 717-18 (cf. No. 208, where a famine ravaged the Arab army that had set the city under siege but not the population of the city itself; Treadgold, History, 346-9).

A Quantitative Overview

29

Constantinople.10 From 435 onward, as the Vandals conquered North Africa, but mostly in the sixth century, the fatal combination of imperial warfare against the Vandals in Africa (533-34) and against the Goths in Italy (535-44) produced a disastrous situation for the city. The Lombard incursion that followed did not apparently cause any famine-induced mortality. This can be seen as a result of the overall population decline that was brought about by the outbreak of the plague in the 540s. This is true for Rome as well as Constantinople: after the first visitations of the plague (Nos 118 and 139 for the former and 111 and 134 for the latter), the number of famines attested in those cities declined considerably. Once again this might be partly explained by the lack of sufficient sources in this period; however, the general tendency of the figures speaks in favour of a decline in the number of subsistence crises as such. By comparing the subsistence crises that occurred in Rome and Constantinople we can obtain the following picture. Rome endured twelve famines from 284 to 750 compared to seven in Constantinople. The number of shortages in both cities is 11:12, whereas Constantinople endured the largest number. As noted earlier, Rome was also visited by more famines that resulted in mortality than Constantinople (7:5). There is a clear tendency of more famines in the old than in the new Rome. In my mind this reflects an effort of the imperial government to secure the provisions (mostly grain) for the capital as an utmost priority. The emperors lived in Constantinople in close proximity to the populace, which had to be provided for; if not food riots would occur, and these were to be avoided at any cost. There is a large recorded number of such phenomena in Rome studied by Kohns, Versorgungskrisen, passim (cf. also my own Nos 12, 31, 42, 71), but merely six in Constantinople (Nos 43, 51, 59, 132, 165, 176). One case from the catalogue (No. 169) can illustrate the absolute priority of the provision of Constantinople over any and every other city in the Empire. Between 608-11, during the reign of Phokas, a famine ravaged Thessalonica. This famine was miraculously terminated by St. Demetrios, who allegedly appeared to an official in charge of the grain shipments to Constantinople and had him direct these shipments instead to Thessalonica. A wheat shortage was already manifest in the capital, probably as a result of the uprising of the Heraclids against Phokas, during which African and Egyptian grain shipments to the capital were held back.11 Contrary to what 10 11

Sirks, Food, 146-8. J.H. W.G. Liebeschutz, Decline and Fall of the Roman City (Oxford 2001) 272-6.

Fam ine and Pestilence

30

the Miracles of St. Demetrios report, the famine in Thessalonica must have been caused by the effort of Phokas to ensure the largest possible amount of grain from the area still under his control and import it to the capital. This had as a result the occurrence of shortages and/or famines in other cities that would have otherwise obtained this grain.12 Table 2.3 City

4th c.

Const. Rome 1 Antioch 2 Alex. Edessa Thess. Ravenna Amida 1 Total 4 * 700-50

Epidemics per city and century 5th c.

6th c.

7th c.

8th c.*

Total

3 3 2 8

5 4 4 2 1 1 2 2 21

2 4 1 7

2 1 3

12 13 6 3 3 1 2 3 43

The same overall remark concerning the prevalence of the recorded famines in Rome and Constantinople over the other important cities of the Byzantine Empire may be made in regard to epidemics. Again the recorded outbreaks outside those two cities are minimal. If we concentrate on the figures for Rome and Constantinople we find the picture that emerges quite different compared to the conclusions based on Table 2.2. Constantinople was visited by 12 epidemics as opposed to 13 on record for Rome. A closer look makes clear that those 12 epidemics broke out in the city in the period between 446 and 748 which gives us an average of 25 years between the various eruptions or about four outbreaks per century. Rome's average is similar: 13 outbreaks between 397-98 and 745-46 or roughly an outbreak 12

This practise continued well into modern times; cf. Th. Malthus, An Essay on the Principle of Population, 6th edn eds E.A. Wrigley and D. Souden (London 1986) writes in 1826 (p. 185): 'When Constantinople is in want of provisions, ten provinces are perhaps famished for a supply.'

A Quantitative Overview

31

every 26.5 years. But these average numbers falsify the true picture: there are periods of concentrated epidemic activity followed by long epidemicfree stages. For example: 60 years without any outbreaks in Rome between 408 (No. 42) and 467 (No. 70), 65 between 471-72 (No. 70) and 537 (No. 94), 45 to 56 between 608-15 (No. 172) and 664-65 (No. 183) and 65 years between 680 (No. 192) and 745-46 (220). Constantinople exhibits even more extremes in that aspect: 95 years between 447 (No. 61) and 542 (No. 111) and 79 between 618-19 (No. 173) and 698-99 (No. 199). All in all, Constantinople seems to have been hit by epidemics more regularly, with a shorter periodicity than Rome. It seems impossible to come up with a model to explain the situation reflected by the sources. Jean-Noel Biraben has formulated a general rule concerning the occurrence of epidemics in the various types of human habitation centres. According to him the large cities are visited by every major epidemic in their region, the small cities escape some of them, the villages are only visited by one every two or three outbreaks, whereas the remote hamlets are but occasionally hit by an epidemic.13 This picture is corroborated by the data of the catalogue, at least in the case of Constantinople where the source material is ample enough to allow us to draw some conclusions. As the capital of the Empire, Constantinople was certainly the largest and most well-frequented city of the Eastern Mediterranean in Late Antiquity. The city imported large quantities of goods from the whole territory of the Empire and served as the destination and point of departure for embassies, both foreign and from the state itself. This movement of people and goods to and from Constantinople also brought about a passive transportation of diseases. This is implied in a case concerning the outbreak of the plague in 698 (No. 199): the dredging of the Neorion port in Constantinople is followed by the eruption of the epidemic. At the same period, but probably before the outbreak in the capital, the plague had been ravaging Syria (No. 198). Constantinople was visited seven times by the plague (Nos 111, 134, 145, 160, 173, 199, 221) as opposed to five recorded outbreaks in Rome, two of which are uncertain (Nos 118, 139[?], 151, 162[?], 220). Since Constantinople was always at the centre of historians' and chronographers' interest we may attribute this discrepancy in the figures at least partly to a gradual lack of sources and/or of interest in Rome. However, the attested epidemic waves that befell Constantinople (with the exception of a gap in lj

J.-N. Biraben, 'Rapport: La peste du VIe siecle dans 1'Empire Byzantin', in Hommes et richesses I (Paris 1989) 122.

Famine and Pestilence

32

the tradition after the first quarter of the seventh century) show the unbroken, continuous presence of the plague in the city.14 The same can be said of the other, unidentified epidemics that continued to beset Constantinople from the mid-fifth century until the end of the period in question. The previous two tables dealt with the urban manifestations of famines and epidemics as recorded in the catalogue. Their sum total of 104 phenomena represents less than half of all the entries. Therefore an effort has been made to include in the following two tables those phenomena that concern cities and regions that have not been included in Tables 2.2 and 2.3. I should note that the number of famines and epidemics per region surpasses that of the total number of phenomena as found in Table 2.1. This comes as a result of the fact that a number of phenomena occurred in more than one region at the same time and as such have been added to all respective columns. Table 2.4 Region Balkans A. Minor Syria Palestine Egypt N. Africa Italy Total * 700-50

Famines and shortages per region and century

4th c.

5th c.

6th c.

7th c.

5 4 7 3 3 8 30

10 7

8 3 5 1 3 1 16 37

5 1 5 2 3 3 19

o

1 1

2 9 33

8th c.* 2 3 3 3 1 1 13

Total 30 18 23 7 13 4 37 132

The number of phenomena per region consists of both the mention of phenomena in cities of this region (e.g. Rome for Italy) and the mention of the region itself. The regions are presented clockwise beginning from the Balkans as the area to include the capital of the Empire. The Balkans includes recorded phenomena in Constantinople, Thrace, the Illyricum, Pannonia, Macedonia, Central Greece, the Peloponnese and Crete. Asia 14

For a detailed discussion, see Chapter 6: 'The Justinianic Plague'.

A Quantitative Overview

33

Minor incorporates Lazica, Bithynia, Paphlagonia, Phrygia, Galatia, Cappadocia, Lycia and Cilicia. Syria includes the mentions of Mesopotamia and the Byzantine-Persian frontier. Finally, Italy includes all those occurrences concerning cities and regions that belong to the present state of the country. Again, the regions that include the two most important cities, the Balkans and Italy, display more or less the double number of phenomena than the other ones. As far as famines are concerned they are followed by Syria, whose place at the centre of interest of Islamic writers led to a growing body of recorded phenomena from the seventh century onward. The same, but to a lesser extent, may be said of Egypt, North Africa and Palestine: they are represented in the catalogue with a minimal number of phenomena. This is certainly due to a lack of source material. The high number of entries coming from Asia Minor is certainly connected to this region's geographical position in (more or less) close vicinity to the capital. This is particularly evident in coastal areas and the neighbouring provinces of Constantinople, Bithynia and Galatia. If we turn to Table 2.5 we find the overall picture to be quite similar, although Syria surpasses both the Balkans and Italy in recorded epidemics. It is noteworthy, moreover, to note that Italy as a region would occupy a high position in these last two tables even if the number of phenomena recorded solely for Rome was excluded. This would add up to 15 famines and 15 epidemics during the period in question. On the contrary, the elevated number of phenomena in the Balkans comes as result of the Constantinopolitan entries; if we were to subtract them the respective numbers would be nine famines and nine epidemics. The grave situation that Italy had to endure in the period in question reached its peak in the sixth century, for which 31 famines and epidemics are recorded. The number of phenomena recorded in the catalogue reflects the political and military hardships that Italy underwent in the course of the fifth, sixth and early seventh centuries.

Famine and Pestilence

34

Table 2.5 Region Balkans A. Minor Syria Palestine Egypt N. Africa Italy Total * 700-50

Epidemics per region and century

4th c.

5th c.

6th c.

7th c.

3 2 5 2 1 13

3 2 3 2 4 14

9 6 12 3 3 3 15 51

2 1 5 4 3 6 21

8th c.*

3 2 9 3 1 1 19

Total

20 13 34 9 11 4 27 118

Chapter 3

Subsistence Crises: Causes, Location, Duration and Range Subsistence crises were rarely unicausal. Modern famine research perceives the causes of such crises in two broad (and conflicting) categories: due to food availability decline (FAD) or food entitlement decline (FED).1 FAD-induced crises are explained as caused by natural events or a Malthusian overpopulation putting pressure on available resources. In contrary to this, FED-induced famines are a result of political and social structures that make 'people vulnerable to the impact of a natural hazard by their place in the economic, political and social processes that affect their exchange entitlements.'2 The two categories place famine causation in the antagonism between a deterministic dependence on the natural environment and the interplay of a political society that controls both food production and its distribution. The battleground for these theories are modern subsistence crises in the developing countries, while the concepts they consider rarely include pre-industrial societies; but it is precisely such a society that is the object of the present study. Indeed, as Murton suggests 'if the analysis is stretched to include a big enough area, there is always enough food to avert a famine.'3 Given the technical limitations of transport in Late Antiquity and the high costs of interregional communications,4 we should view subsistence crises of the period as a 1

See the overview by B. Murton, 'Famine', in K.F. Kiple and K. Conee Ornelas, eds, The Cambridge History ofFoodvol II (Cambridge 2000) 1411-27, esp. \4\9ff. 2 Murton, 1420; The concept of FED was first brought up in the seminal work by A. Sen, Poverty and Famines: An Essay on Entitlement and Deprivation (Oxford 1981). * Murton, 1421. Jones, LRE II 842, based on evidence from Diocletian's price edict has calculated that it cost less 'to ship grain from one end of the Mediterranean to the other than to carry it

36

Fam ine and Pestilence

result primarily of reduced food availability rather than as events that occurred chiefly because of unequal distribution or food entitlement, although the latter certainly played an important part in the survival of specific social strata in periods of such crises. For reasons of classification I have chosen to address the following subsistence crises according to their principal cause and discuss the accompanying elements in the same section. From the 134 phenomena collected in the catalogue only two-thirds (= 87) record what the sources present as their cause. Nature-induced Crises As far as subsistence crises are concerned the nomenclature of natureinduced causes is apparent. Climatic phenomena or pests that have a negative effect on the harvest will often lead to a crisis, which may affect a smaller or a larger community on a corresponding scale. Climate Drought Drought manifests itself as a lack of (timely) precipitation over a prolonged period of time. These characteristics are reflected in the phenomenon's terminology: sometimes called drought (auchmos, siccitas) and at other times a lack of rain (anombria, abrochia). As E. Le Roy Ladurie rightly put it, in Mediterranean Europe, drought is the factor that harms par excellence grain produce: in autumn it may bring sowing to a halt, in spring destroy the harvest and even deprive peasants of seed-grain needed for the next sowing season.5 Given the fact that grain was the basic staple in the diet of the period and the area in question, it becomes evident that there was a close interconnection between drought and subsistence crises. This causality is recognized by the sources, which often present shortages and famines as having occurred after and because of droughts 75 miles'. See now also M. McCormick, Origins of the European Economy: Communications and Commerce AD 300 - 900 (Cambridge 2001) 64-115. 5 E. Le Roy Ladurie, Histoire du climat depuis d'an mil (Paris 1967) II 100. More recent studies show 'that the distribution of precipitation was the main factor in grain yield', cf. P. Garnsey, 'The Yield of the Land in Ancient Greece', in B. Wells, ed., Agriculture in Greece (Stockholm 1992) 150 n. 12, with references.

Subsistence Crises: Causes, Location, Duration and Range

37

(cf. Nos 64, 74, 77, 170, 194, 211). Garnsey, working with precipitation figures from Attica for the period 1931 to 1960, argued that the percentage probability of a failure of the wheat crop because of insufficient irrigation was about 28 per cent.6 Within the cases recorded in the catalogue, drought-induced subsistence crises form the largest group in their category: 33 in a total of 44 crises with recorded natural causes. We can therefore safely assume that they constituted an extremely common phenomenon. As such they must have often escaped the attention of our sources or would have been considered too usual and regular phenomena to be worth recording. Additionally, as droughts were events that played an important and direct role in rural communities and only an indirect one in urban environments, we can also assume that our sources, which are socially and ideologically urbanoriented, will have largely neglected such manifestations as pertaining to a world outside their scope of interest. As far as their range is concerned we cannot discern one particular pattern. There were in some cases large-scale events such as in Nos 29 (Italy, Egypt and Eastern Mediterranean), 64 (Eastern and South-Eastern Asia Minor and Palestine), 74 (Northern Africa), 77 (Africa and Gaul), 169 (the whole Empire), while in some other cases the crisis was merely local: Nos 2 (Cyprus), 18 (Palestine), 35 (Gaza), 85 (Jerusalem) or 204 (Crete). The duration of these crises is rarely attested in the sources, but as E. Patlagean has justly formulated, a famine became catastrophic if the untoward conditions prevailed over at least two consecutive annual cycles, when any available surplus stocks would have been exhausted.7 This was the case in Nos 29 and 64. Rain Precipitation was an extremely important factor in Late Antique agriculture since it constituted the main source of water supply. The previous point has shown that the lack of rain over a prolonged period of time often produced subsistence crises. The termination of such crises through miraculous rainfall also stresses the importance that precipitation had for the agricultural regime of the period (cf. below Chapter 4, Section 'Miraculous Termination'). In the Mediterranean, precipitation is concentrated in the winter months, a fact also corroborated by the source

6 7

Garnsey, Famine, 10—11. Patlagean, Pauvrete, 82; Garnsey, Famine, 54.

38

Fam ine and Pestilence

material (Nos 4, 21, 80, 92, 137).8 The onset of ample winter-rain was beneficial for the crops (cf. No. 80). Excessive or out-of-season rainfall, however, could have the opposite result: the destruction of the crops (Nos 31, 188).9 In other cases extreme precipitation could results in floods (this will be dealt with in the following section). There are no recorded crises in the catalogue caused solely by rainfall. The only possible exception could be No. 52, where the shortage is presented in the source as occurring before the excessive rainfall. It is, however, plausible that the precipitation may have caused the subsistence crisis by damaging grain stocks (as in No. 151, only that then no famine ensued). As an additional element, extreme or unseasonably heavy rainfall could aggravate a crisis that had already been set in motion (cf. No. 80) Floods With the exception of the Nile flood, whose annual occurrence was extremely important for the agricultural regime of Egypt and therefore for the grain production of the Empire (cf. Nos 29, 68, 174), floods caused primarily by excessive rainfall or the melting of snow were phenomena that had a catastrophic impact on people's lives. The Nile flood is recorded in connection with some of the catalogue's entries. It began around June 20 and reached its peak between August 25 and September 5.10 While a sufficiently high water level ensured a rich harvest, the failure to reach such a height brought about serious problems and, in the worst cases, crisis. Pliny the Elder records that a rise of about 14 cubits and more would signal a good outcome of the harvest, while a drop below 12 cubits would lead to famine.11 It is in such cases that our sources record the failure of the flood, although without mentioning its 8 Cf. Y. Karmon, Israel Eine geographische Landeskunde (Wissenschaftliche Landerkunden 22. 2nd. ed. Darmstadt 1994) 29; W.D. Hutteroth, Turkei (Wissenschaftliche Landerkunden 21. Darmstadt 1982) 108, 110-12; E. Wirth, Syrien. Eine geographische Landeskunde (Wissenschaftliche Landerkunden 4/5. Darmstadt 1971) 91; C. Lienau, Griechenland. Geographic eines Staates der europaischen Peripherie (Wissenschaftliche Landerkunden 32. Darmstadt 1989) 82-3. 9 See the observations by Gregory of Nazianzos, Inpatrem tacentem 1, PG 35, 936A. 10 D. Bonneau, Le flsc et le Nil Incidences des irregularites de la crue du Nil sur la fiscalite fonder e dans I'Egyptegrecque et romaine (Paris 1972) 9 n. 2 ; W. Scheidel, Death on the Nile: Disease and Demography of Roman Egypt (Mnemosyne, Supplement 228. Leiden - Boston - Cologne 2001) 57. 11 Pliny the Elder, Natural History, Livre F, ed. J. Desanges, vol. 2 (Paris 1980) V 58 (87); furthermore cf. N. Lewis, Life in Egypt under Roman Rule (Oxford 1983) 109-10.

Subsistence Crises: Causes, Location, Duration and Range

39

height (Nos 30, 69, 176). The dependency of Rome and afterwards Constantinople on the Egyptian grain shipments was such that a large-scale failure of the harvest necessarily moved the administration to fall back on any available grain stocks. As these were often quite limited, subsistence crises ensued. Cold Extreme cold as manifested in low temperatures, frost and extensive snowfall was certainly a rare phenomenon in the Mediterranean basin, apart from the mountain regions. This is made obvious through the way that the sources portray it as something unusual and spectacular. In 443 snow fell for six months in the vicinity of Constantinople causing massive human and animal mortality (No. 58). It is not specified whether this came as a result of the cold alone or if the snow had damaged the crops and killed the draft animals resulting in a famine. This case remains singular throughout the catalogue. In 604-05 a very cold winter in Italy was one of the reasons for the destruction of the vines and that year's crops (No. 166), resulting in a shortage. The high death toll among the Byzantines and their Lazic allies, which resulted from lack of provisions and cold is not as surprising if we consider that it took place in the Caucasus mountains, in the winter, and in a region that experiences quite low temperatures during this season (No. 128). Cold killed the already-starved rural refugees who had flooded to Edessa and who were forced to spend the night on the streets (No. 80). Particularly severe winters are sometimes recorded by Byzantine sources as extraordinary catastrophic events that brought about destruction of the harvest and fruit-bearing trees as well as the death of draft animals, resulting in famine and mass mortality.12 Winds This section bears a general and somewhat misleading title as it is meant to cover those cases where untoward climatic conditions prevented the landing of food shipments (mostly grain) in the cities of the Empire. These include winds, or their absence, storms at sea and any similar phenomena that had a negative influence on navigation. In general, in the period in question, navigation of the open sea was confined to the period from April to early October.13 Even during that 12 For the winter of 763-64 cf. I. Telelis, E. Chrysos and D. Metaxas, 4 Oi TOW pD^avTivaw Ttrrycbv yia TOY 6pi|ii3 %afid)va TOD STOIX; 763/4 ji.X.', dcoScbvrj 18 (1989) 105-26; for the winter of 926-27 cf. Kaplan, Hommes, 461-2. 13 Cf. No. 17 and McCormick, Origins, 450-68; esp. 450-52.

Famine and Pestilence

40

period, however, certain patterns of winds were necessary to ensure safe passage throughout the Mediterranean.14 The grain shipments from Alexandria to Constantinople were dependent on the south wind to cross the straits and reach the city. If the wind failed to blow, they were forced to wait for favourable conditions. There are only two such cases recorded in the catalogue ranging from the 330s to 562 (Nos 6 and 138). Justinian took measures to relieve this situation by building a large granary on Tenedos; in this way the grain ships could unload their cargo there and make further trips to Alexandria to collect grain without having to wait for the south wind.15 The third trip of the grain fleet, which usually left Alexandria before the close of August16, had not reached Constantinople in November 562 (No. 138); this indicates that the problem persisted even after the measures taken by Justinian. Rome had similar difficulties to face and was also dependent on favourable climatic conditions for the landing of the grain shipments. In two cases (Nos 12, 31), both from the second half of the fourth century, the sources record subsistence crises that were caused by the failure of the grain fleet to land. The lack of further notices is certainly due to the difficulties with the source material, rather than the disappearance of the problem itself. Concerning both Rome and Constantinople, we can note that the crises resulting from these wind conditions were not as serious as the ones that were brought about through the failure or destruction of the harvest. In most cases the modest shortages were resolved as soon as conditions became favourable.

14

W. Murray, 'Do modern winds equal ancient winds?', Mediterranean Historical Review 2 (1987) 139-67; idem, 'Ancient Sailing Winds in the Mediterranean: The Case for Cyprus', in V. Karageorghis and D. Michaelides, eds, Proceedings of the International Symposium, Cyprus and the Sea (Nicosia 1995) 33-43; 15 Prokopios, Buildings, V 1, 7-12 (IV 150-51); Muller, 'Getreide', 6. 16 See Kislinger and Stathakopoulos, 'Pest', 92 with references.

Subsistence Crises: Causes, Location, Duration and Range

41

Pests The occurrence of pests such as locusts, worms, beetles, mice, rats, birds etc. was one of the principal risk-factors in pre-pesticide agriculture.17 The means the farmers had at their disposal against these pests were few and their efficacy limited, both in terms of time and range. Such measures were the lighting of fires by the fields, the burning of locusts or making noise in order to scare them off the fields.18 The frequent recurrence of miracles where holy men drive away or destroy pests would seem to indicate that the common measures seldom brought about the necessary results that a 'supernatural' intervention could and would have.19 Although many of the pests mentioned above can have seriously negative effects on crops, vine and fruit, there are only crises related to locusts and mice or rats recorded in the catalogue. Locusts There are eleven recorded cases of locust invasion in the catalogue. Other, numerous cases where locusts are mentioned without the record of an ensuing famine or epidemic have not been included in the catalogue, but will be dealt with elsewhere.20 The geographical range of the presence of the locusts is wide, with a concentration in the Near East: they were recorded in Palestine (Nos 40, 85), in Syria (Nos 30, 80, 88, 147, 167, 205, 210, 211), and as far as Trent, in Northern Italy (No. 153).

17

Cf. the now lost book of Theophrastos 'On Animals which appear all at once in large numbers' as preserved in Photios, Bibliotheque, ed. R. Henry, vol. VIII (Paris 1977) 164-66, on mice and locusts; cf. Sallares, Ecology, 426. 18 Cf. Kukules, Bios, V 273; R.F. Chapman, A Biology of Locusts (The Institute of Biology's Studies in Biology 71. London 1976) 58; H. Grassl, 'Heuschreckenplagen in der Antike', in E. Olshausen and H. Sonnabend, eds, Naturkatastrophen in der Antiken Welt (Stuttgarter {Colloquium zur historischen Geographic des Altertums 6, 1996 = Geographica Historica 10. Stuttgart 1998) 444; J.M. Beyer, 'Sie legten lediglich eine nicht unubertrachtliche Menge Eier in unser Land. Heuschreckenplagen in der Spatantike und die Chronik des Josua Stylites', Byzantina 21 (2000) 57-83. 19 As example we may take the Vita of Theodore of Sykeon wherein the holy man repeatedly destroys or repels pests: Locusts (XXXVI [32], Cla [80-81], CXVb [92]), worms (CXVb [92]), beetles (CXVIIIb [95]). 20 For a detailed study of locusts, the devastation they wreaked, the measures taken against them, as well as typological considerations as to their function when mentioned in texts cf. my forthcoming article 'The occurrence of locusts in Late Antiquity. Facts and Topoi', which is based on a much wider source base than the catalogue.

Famine and Pestilence

42

The locusts recorded in Italy should probably be identified as CalHptamus italicus, a species which is known to occasionally damage grain crops during mass outbreaks21 - although our source reports that they did not harm the crops in 591. From the accounts of its recorded areas of migration and biological characteristics, the species in question was in all other cases probably the Desert Locust, Schistocerca gregaria. Since this species is one of the most studied migrant insects up to the present we can refer to some important features about its ecology and life cycle, which can be taken to apply in general to many other locusts.22 Schistocerca gregaria occupies a large area extending from Senegal through Central Africa up to Ethiopia, South Arabia, Pakistan and India, the so-called Intertropical Convergence Zone (ITCZ). This area is used as a summer breeding ground, whereas after this, in late summer, the locusts migrate as far as the North African coast and the Red Sea region, reaching in exceptional years even the inner Turkish plateau, South Armenia, Turkmenia and Uzbekistan. There they will breed again in early spring and subsequently migrate southwards to the original summer breeding grounds - in sum covering an area of some 29 million km2. This being, in the words of the principal authority 'a mainly closed circuit of swarm migration between the summer and the spring breeding areas'.23 As with all locusts, it tends to respond to increases in population density by aggregating in swarms. These swarms may extend over one hundred km2 and their numbers may easily surpass one million locusts. An example taken from a modern locust plague may be used to indicate these trends. In September/October 1967, some five million Schistocerca gregaria arrived in the Tamesna region of Mali covering an area of 25,000 km2; after breeding their number rose to ca eighty million, sixteen times that of the immigrant parent generation.24 The migration of locusts does not occur solely as a means of procuring resources but is regarded as 'a behavioural process that has ecological 21

B. Uvarov, Grasshoppers and Locusts. A Handbook of General Acridology, 2 vols (Cambridge 1966, London 1977) II 457. 22 For the whole section cf. Uvarov II 347-51; Chapman, Biology, 34-6; R.A. Farrow, 'Flight and Migration in Acridoids', in R.F. Chapman and A. Joern, eds, Biology of Grasshoppers (New York - Brisbane - Toronto 1990) 297-9. 23 Uvarov II 351. 24 Uvarov II 503-7.

Subsistence Crises: Causes, Location, Duration and Range

43

consequences'.25 Their movement follows changes in the distribution of the seasonal rains, as breeding is only possible between 80-400 isohyets.26 Lack of precipitation during the summer will cause a mass mortality of locust eggs, while an unusually high rainfall will have the same effect.27 Once a mass migration of locusts had begun, therefore, it could only be terminated due to natural, climatic reasons. In several of the recorded cases in the catalogue our sources have included information on these outbreaks that comply with modern entomological findings. In two incidents the date of the locust plague is recorded: March 500 in Northern Syria, Mesopotamia and up to West Armenia and April 722 in Syria (Nos 80 and 211). In the first case we are fortunate in having ample information about the background of that outbreak. Ps.-Joshua Stylites has recorded that the locust eggs were laid in May 499.28 The fact that the eggs hatched in March of the following year is a sign of discontinuous development, a phenomenon recorded by modern entomology. In areas with long, dry seasons locusts will lay eggs at the beginning of those seasons and these will hatch only when the next rain season begins after several months (aestivating eggs).29 However, the information provided by Michael the Syrian concerning a locust plague in Syria, during which locust eggs caused severe damage three years after they had been deposited must be considered as fictitious.30 Hibernating or aestivating eggs will hatch several months after they have been laid, but this period of discontinued development will not surpass this amount of time.31 In one case, the locust plague of April 722, we can clearly discern an incident of the spring-zone breeding migration. Another accurate observation was made regarding the outbreak in Palestine in 517-18 (No. 85). The mention of the larvae (brouchos) along with the adult locusts indicates that breeding had recently taken place in the area, while the description of the swarms as seemingly covering the entire sky is characteristic of the swarms formed by Schistocerca gregaria. The so-called cumuliform swarms can be seen from a distance of several 25

A.G. Gatehouse, 'Migration, a behavioral process that has ecological consequences', Antenna 11 (1987) 10. 26 Uvarov II 498; Farrow, 'Flight', 297. 27 Uvarov II 470, 519. 28 Ps.-Joshua Stylites XXXIII (23 Wright / 32 Trombley and Watt). 29 Uvarov II 282. 30 Michael the Syrian II 374; cf. No. 167. 31 Uvarov I 259-64; II 276-84.

Famine and Pestilence

44

kilometres 'as a slowly moving dark mass, sometimes extending to more than a thousand metres upwards [...]'.32 As this species is credited with consuming a range of 250 to 400 species of food plants,33 it is not surprising to find records of extensive damage to various plants in our sources: grain crops (Nos 80, 147, 205, 211), vines (Nos 205, 210), herbs (No. 147), vegetables (No. 147), even fruit-bearing trees such as olive and fig trees (Nos 167, 210). Taking into consideration that locusts tend to consume food equal to their own weight per day (a female Schistocerca may eat up to 1.5 g of grass per day), the damage a high-density large swarm may cause to crops and/or trees is great.34 A swarm comprising of a density of about 30m2 would consume 45 to 225 g/m2/day; if this happened in an area of 10m2 it would sum up to a total of 0.5 to 2.25 x 10 kg (roughly 2,000 tons) of vegetation per day.35 The damage brought about by locust-plague outbreaks to pastures would be equally severe, as the food consumption of large locust swarms is calculated at being about 1000 times as much as cattle would eat.36 Jerome made an additional correct observation concerning locust migration as seen in Palestine around 406 (No. 40). He relates of how these swarms were caught between wind currents (vento surgente), which brought them down on the water, where they drowned and were washed ashore the coasts of the Dead Sea and the Mediterranean.37 The losses suffered by migrating locusts when they fly over the sea or large lakes is attested in modern research and recorded for similar areas such as the Red Sea or the Persian Gulf.38 In all but one case the crises that were caused by large-scale locust invasions were shortages or famines. The most severe incidents occurred when the locust plagues were preceded or followed by other aggravating factors - in most cases drought (cf. Nos 85, 88, 167, 211). However, the occurrence of locust swarms or bands alone could also have devastating results. The first phase of the two-year crisis in Edessa, roughly from j2

Uvarov II 200. Uvarov II 95-7. 34 Chapman, Biology, 23-4. 35 Ibid. 36 Chapman, Biology, 24. J? Jerome, Commentarius in loelem, ed. M. Adriaen (CCL 76. Turnhout 1969) II 18/20(187). 38 Uvarov II 520. 33

Subsistence Crises: Causes, Location, Duration and Range

45

March to November 500, was caused solely by locusts. The larvae and young locusts that hatched with the spring rains in March made desolate a huge area from mainland North Syria, Mesopotamia and South Armenia up to the Mediterranean coast. The destruction of the harvest was felt quite soon: in April 500 a dramatic rise in grain prices (an increase of ca 650 per cent compared to the period before the crisis) illustrates the severity of the famine. Later on climatic phenomena helped sustain and even aggravate the situation, but its onset was caused by the locust plague. Other cases are less well documented, but the essential pattern is repeated: an invasion of locusts, destruction of the crops/vines/vegetables and/or fruit-trees, famine or shortage (cf. Nos 147, 167, 210). Rats/Mice Both rats and mice are commensal rodents, often depending on human food resources for their survival. They will damage crops, vegetables or fruit in the field, but will mostly feed on stored agricultural products and cause considerable damage.39 In 384-85 mice, snakes and locusts infested Antioch's victuals during a severe famine that was already ravaging the city; the presence of the pests aggravated the food crisis, but did not provoke it (No. 30). In two additional cases, both from the seventh century, the presence of rats and mice is connected to the outbreak of subsistence crises. In 604-05 mice reportedly destroyed crops in Italy that were also affected by blight (No. 166), while in 673-74 a large number of rats is recorded as being the sole cause for a severe famine in Syria (No. 187). In pre-industrial agriculture where manure was limited and the dependency on climate considerable the failure of the crops was a common phenomenon.40 Ruschenbusch used data from Greece for the years 1921— 32, when the cultivation of cereals was still practised largely without the use of manure and chemical fertilisers had not been yet put to use, to show that 39.2 per cent of the harvest for that period would have yielded results that would fail to support the peasant population.41 These had been nature-

39

Kukules, Bios, V 274; S.A. Barnett, The Rat. A Study in Behaviour (Chicago 1964) 38-44; Sallares, Ecology, 263-6. 40 On the insufficiency of available manure cf. P. Garnsey, The Yield of the Land in Ancient Greece', in B. Wells, ed., Agriculture in Greece (Stockholm 1992) 150-51, and Sallares, Ecology, 381-5. 41 E. Ruschenbusch, 'MiBernten bei Getreide in den Jahren 1921-1938 in Griechenland als Modell fur die Antike', in E. Olshausen and H. Sonnabend, eds, Naturkatastrophen in

46

Famine and Pestilence

induced crop-failures due to blight, snow, hot winds, drought, excessive precipitation, hail and pests. By way of analogy we may suppose that similar poor results would have constituted the reality of ancient and late antique agriculture. The high number and frequency of nature-induced subsistence crises we have reviewed so far corroborates this argument. Human-induced Crises The second category of crises includes those that occur due to human actions. These are generally short-term events, often, however, of an extremely intense character and with immediate and dramatic results. The most obvious and numerous ones of this category are those that occurred as a result of military action, either as a necessary evil, tolerated as a means to achieve strategic goals, or as a conscious instrument of warfare used against military or civilian targets. Military Sieges The besieging of fortified strongholds is a military practice with a very long history. In those cases when primary intensive assaults failed in their objective those strongholds would be surrounded and besieged for an extended period. Food availability, both for those inside and outside the city walls, was an issue of tremendous importance. The knowledge of such mechanisms is attested in very early texts such as the book of Deuteronomy, which describes events of the 13th century BC, although it may have been written much later, where advice is given on how to conduct a lasting siege (20:19.58ff) and what the dramatic results of a siege would be for the besieged population (28:53.30ff). In Graeco-Roman siege-warfare the relationship between famine and warfare may be summarized in the following formula: The first thing the besieger should do, if possible, is to keep the necessities, such as food and water, from getting to the people within the walls. If the besieged possess these supplies in abundance, then it is necessary to resort to siege engines and

der Antiken Welt (Stuttgarter Kolloquium zur historischen Geographic des Altertums 6, 1996 = Geographica Historica 10. Stuttgart 1998) 78-81.

Subsistence Crises: Causes, Location, Duration and Range

47

fighting.'42 Various ancient theoretical works offered practical recipes for destroying food supplies, crops, fields and trees, how to poison water reservoirs, or how to cause a pestilence among the soldiers with the use of food, water or wine.43 The point to be made is that famine was not an accidental side-effect of warfare, but a weapon put to use in an absolutely conscious way. A large number of siege-induced crises is recorded in the catalogue ranging from the mid-fourth to the mid-eighth century (Nos 11, 42, 45, 53, 72, 81, 82, 94, 96, 97, 99, 101, 113, 120, 122, 123, 171, 189, 208, 216). The duration of these sieges varied, but in general an average of about one year was the usual practice, especially for larger, well-fortified cities, while smaller ones would capitulate within about half that time. In those cases collected in the catalogue no siege surpassed a period of 14 months (cf. Nos 53, 216). The most common feature of these crises is the occurrence of famine as a direct result of the siege, as attested in all but one case (19:20). Usually this concerns the population under siege, but there some incidents where the besiegers also faced famine (Nos 53, 94, 208). Sieges would only be effective if the blockade on the import of victuals that had been imposed on the city was total. There are cases where daring efforts were organized to smuggle food into such cities (Nos 94, 120, 123, 208, 216), which were more often than not unsuccessful. The situation within the besieged cities is often described as quite dramatic. Lack of victuals caused famines of varying degrees of severity and the confinement of space created suitable conditions for the outbreak and spread of epidemics. An equally common feature in such crises was the consumption of alternative foodstuffs. A rise in the death toll often accompanied such crises, usually due to either the outbreak of disease (Nos 11, 42, 72, 208) or as a result of starvation (Nos 45, 81, 123, 171, 189, 216). Warfare Sieges represent an important, but not the only, military action that can cause crises. Warfare that disrupts the ordinary and normal cycle of agricultural work, the crossing and encampment of troops either of the 42

Maurice, Strategikon, ed. G.T. Dennis (CFHB 17. Vienna 1981) X 1 (336); Trans. Maurice's Strategikon, trans, by G.T. Dennis (Philadelphia 1984) 106. 43 Aeneas, Poliocertique, ed. A. Dain (Paris 1967) VIII (14); Julius Africanus, Fragments des Cestes, ed. J.-R. Viellefond (Paris 1932) 56-7, 80.

Famine and Pestilence

48

state or belonging to enemy groups that consciously damages crops and deprives country areas and cities of their provisions - these are all actions that have triggered off various types of crises as brought together in the catalogue. The presence of the state troops, usually connected with imminent warfare in a particular city or region, put a heavy strain on the local population.44 The order to provide wheat for 'soldier's bread' that was imposed on the population of Edessa may serve as an example. From 503 to 505 they presented a total of 2,110,000 modii of wheat, while they were forced to bake some of this bread at their own cost (No. 81). No famine is recorded, but the grain price during this period is extremely high, characteristic of times of famine (cf. Appendix II: 'Famine Prices of Grain'). The recorded crises that seemed to have been caused by such a presence are few (Nos 5, 9, 16), but we can safely assume that the phenomenon must have been more common than that. Shortages occurred as a result, although relieved by the grant of imperial help. Most cases in this category are connected with warfare, either in the form of an enemy incursion or of hostilities between the Imperial army and enemy forces. Famine, as attested in 11 of the 14 cases, was the usual result (Nos 17, 25, 26, 46, 65, 95, 98, 119, 141, 194, 208). Finally, only one case illustrates what must have been a common situation: forced to flee because of the Persian incursion in Palestine and Egypt, a large number of refugees found shelter in Alexandria, provoking a sudden rise in the city's population and subsequently famine, possibly even the outbreak of epidemics (Nos 174,175). Non-military This section concerns crises that are human-induced but not connected with military actions. They consist of political activities taken ad hoc in order to settle economic problems, and which more often than not had the opposite results. Price regulation The first category in this section deals with measures taken mostly by the Imperial government in an effort to regulate problematic market situations by establishing specific price frames for 44

Cf. Garnsey, Famine, 253.

Subsistence Crises: Causes, Location, Duration and Range

49

goods and services. The means used for this purpose were edicts of maximal prices. Diocletian was the first to issue such an edict with power throughout the whole Empire in 301.45 It will suffice to note that it did not achieve its desired goal and its practice was abandoned shortly after, not before, however, it had provoked a general scarcity of goods.46 Apparently this measure's failure did not prevent its further application, though only on a local scale. There are three such examples of such price regulations, all from the fourth century, that were imposed on the market in Antioch (Nos9, 16,30). In the first case a price reduction was introduced by Caesar Gallos in Antioch in 354 as a means to confront a shortage (No. 19). It was thought that by making goods less expensive one would increase their availability for the economically-weaker urban groups. The city council opposed this action, being not in their interest as the class who produced most of the goods that were available in the market. We are not informed as to whether prices really were reduced, but a shortage occurred - in all probability the landowners withheld their goods from the market.47 We are better informed regarding a similar crisis that occurred in the same city some years later, in 362-63 (No. 16). The situation differed in so far as the emperor himself, Julian, was present in the city along with a large army. Additionally, adverse climatic phenomena had taken their toll on food production so that when the imperial delegation arrived the city was on the verge of shortage and unrest. Julian took a number of steps to relieve the crisis in concordance with the city council, but when all else failed he issued an edict of maximum prices. The result was again the opposite of that desired: prices rose to the allowed maximum, while the shortage persisted. The emperor had large quantities of grain imported, but the landowners clearly acted antagonistically. They bought the grain they needed at the low price in the city market and sold their own grain at a much higher price to the peasants. As a result the shortage and public

45

For the most complete edition of the text so far see C. Roueche, Aphrodisias in Late Antiquity (Journal of Roman Studies Monographs 5. London 1989) No. 231, 265-318. 46 Lactantius, De mortibus persecutorum, eds. S. Brandt and G. Laubmann (CSEL 27,2. Vienna - Prague - Leipzig 1897) VII 7 (180-81); cf. W. Ensslin, 'Valerius (Diocletianus)', RE VII A, 2469-72; Stein, Geschichte,ll3', Jones, LRE, II 308$ Th. Pekary, Die Wirtschaft der griechisch-romischen Antike (Wissenschaftliche Paperbacks 9. Wiesbaden 1979) 126. 47 Durliat, Ville, 360-65.

Fam ine and Pestilence

50

discontent escalated. The edict remained in use as long as the emperor was present in the city and was probably abandoned shortly afterwards.48 The third case is similar to the previous ones (No. 31). It occurred again in Antioch in 384-85 as the comes Orientis, Ikarios, fixed the bread price at a moderate figure probably as a measure to increase its availability on the market. The bakers, however, found themselves in the impossible position of being obliged to produce bread at a low price, whereas the necessary grain was available at a much higher one. As a result they fled the city and bread disappeared from the market altogether. The decree was cancelled and through Libanios' intervention the bakers returned to the city, thus terminating the crisis.49 In all three cases we can observe a number of common elements. The effort to regulate the market by force met the dire resistance of those connected to its mechanisms: those that supplied it with the produce from their great estates and also, in the last case, the bakers who acted as intermediary between the grain producers and the consumers. Shortages of varying degrees ensued, connected with popular unrest (Nos 9, 16). We do not know how the first two crises were terminated, but to judge from the third one it seems as if the market of Antioch tended to rapidly fall back to its usual status, once the regulatory interventions had been called off. We cannot be certain whether this reflects the actual occurrences, but there are no further mentions of such edicts of maximum prices after the fourth century. In further instances the imperial legislatory body intervened in the course of the market only in or rather after extreme periods of crises to correct instances of overwhelming abuse. Such actions could be helping patrons claim back slaves that had been sustained by others in the midst of famine (No. 47), fixing the prices for individuals to buy back relatives that had been sold into slavery during a severe famine (No. 64), or settling a token with which peasants could buy back land they had sold during a period of shortage (No. 91).

48

G. Downey, The Economic Crisis at Antioch under Julian the Apostate', in P.R. Coleman-Norton, ed., Studies in Roman Economic and Social History in Honour of Allan Chester Johnson (New York 1951, Reprint 1969) 312-21; H.P. Kohns, 'Die tatsachliche Dauer des Maximaltarifs f!ir Antiochia v. J. 362', Rheinisches Museum 114 (1971), 78-83, passim; Durliat, Ville, 365-75. 49 Petit, Libanius, 118-21; Downey, Antioch (1961), 419-23; Durliat, Ville, 375.

Subsistence Crises: Causes, Location, Duration and Range

51

Grain transportation Since grain was the staple food for the largest part of the population throughout the period in question, its availability was of extreme importance. This is reflected in the constant attempts of the imperial legislation to ensure a steady import of this foodstuff for Rome and Constantinople.50 The import of grain is one of the most important recurrent themes in the writings of Symmachus, who was city prefect of Rome and therefore responsible for its provision with grain from July 384 to January 385 (cf. Nos 27, 29, 31, 33). The Gothic rulers of Italy naturally followed this policy.51 In 526 Theoderic organized the import of Spanish grain for Rome, as the usual supply from North Africa was in jeopardy (No. 89). As Rome was no longer the capital of the Empire, imperial grain shipments were rare; as such they are recorded as something exceptional as in 575-79 when Justin II sent Egyptian grain to the city when it was on the verge of famine (No. 146). Other authorities, such as the Papacy, also operated in the same direction. Pope Gregory the Great ordered fifty pounds gold's worth of grain bought and transported to Rome in 591 as the harvest had been poor and famine was feared imminent (No. 153). In this section I will refer to a number of crises that were provoked when grain failed to reach a given locality. It is important to stress that in the incidents reported here grain production had not been affected by climatic or other phenomena; shortages were due solely to the failure of transport, which occurred for a variety of reasons. The result was in all cases the occurrence of famines or shortages. During the rule of Maxentius (306-12), a famine hit Rome, possibly because the grain shipments from North Africa were withheld due to the insurrection of the vicarius Africae L. Domitius Alexander (No. 3). In 362 Rome was ravaged by famine and Julian had grain bought, collected and shipped to the city (No. 15). However, his adversary in the civil war that was ensuing in this period, Constantios, intercepted the shipments and directed them to Constantinople. The outcome of this crisis was probably positive, as Julian promised to send further shipments to the starving city; any remaining difficulties will have vanished since he ascended the throne shortly after this incident. A similar crisis broke out in 397-98 as the comes Africae, Gildo, stopped the grain shipments to Rome and a famine broke out in the city (No. 37). In 452 Marcian was informed of plans to 50

Cf. Sirks, Food, passim; Durliat, Ville, 31-279. D. Claude, 'Studien zu Handel und Wirtschaft Wirtsi im italischen Ostgotenreich', Miinstersche Beitrage zur antiken Handelsgeschichte 15 (1996) 42-75. 51

52

Famine and Pestilence

hold back in Alexandria the grain shipments destined for Constantinople. He had the grain delivered to Pelusium instead and the population in Alexandria suffered from famine (No. 67). Constantinople's dependence on Egyptian grain was considerable. The loss of Egypt to the Persians in 618-19 was instantly felt in Constantinople: a severe famine occurred that was soon coupled with an outbreak of the plague (No. 173). In saint's lives we can observe aspects of this phenomenon from somewhat different perspectives. In the Vita of Nicholas of Sion we are informed that when the peasants who usually furnished Myra with foodstuffs heard that the plague had broken out in the city, they refused to enter it and soon a severe shortage occurred as all means of subsistence were lacking (No. 110). In the Miracles of St. Demetrios there are two almost identical narratives dealing with grain transportation (Nos 157, 169). In both cases the city of Thessalonica suffered from famine when its patron, Demetrios, intervened in order to rescue his people. In both cases he appeared to men connected with the grain transportation, in the first case a captain, in the second an official under the command of the comes of Abydos (in fact one of the most important office holders within the system of grain transportation), both times in Chios, and urged them to direct shipments to Thessalonica. In the narration this was fulfilled, but we have reason to doubt whether grain reserved for Constantinople could in fact be directed to any other destination. In any case the fact that these grain shipments to the famished city could only be accomplished with the saintly aid of Demetrios clearly illustrates the importance that the import of grain had for the contemporary populations. In order to do justice to the complexity of past subsistence crises it is important to adopt multilateral causation models. It is essential to distance ourselves from pure determinism, ascribing crises merely to natural causes and at the same time not revert to the opposite model according to which these events occur solely as a result of political and economic structures. It was not distribution or availability of food that created food crises, but rather a combination of both.

Subsistence Crises: Causes, Location, Duration and Range

53

Duration, Location and Range Only a limited number of entries (97) furnishes us with enough information as to be able to address issues of duration, range and location of the recorded crises. In order to comment on the trends that may appear form the distribution of to these three factors we have established categories according to which they will be organised. As far as their duration is concerned we have divided the entries into three categories: short, for crises that lasted up to three months; mid, for those with a duration of between three and nine months, and long for those that reach or surpass one year. As far as their range is concerned we have established three categories: local, for crises that involved one particular city, town or village; regional, for those that were diffused in the territory of one particular province, and general for those involving more than one region. The results can be seen in the following table Table 3.1

Duration and range of subsistence crises

Duration Short

Local Regional General N*:6, H: 14 N: 3, H: 3 T:24 T:8 N:4, H: 15 N: 4, H: 2 Mid N: 10, H: 8 T:20 T:20 T:6 Long N:3, H:4 N: 5, H: 1 N: 5, H:T:7 T:6 T:6 Total 51 34 12 *N = Nature-induced crises, H = Human-induced crises, T = Total52 A number of trends become apparent from the juxtaposition of the data. At first we may observe that local-range crises are of overwhelmingly short- or mid-term duration (24:20:7); furthermore that these types of 52

There is in some cases a discrepancy between the numbers attributed to either nature- or human induced crises and the total sum of phenomena. This comes as a result of a number of incidents for which we do not have enough information as to place them in either category.

Famine and Pestilence

54

crises were mostly human-induced (29:10). Crises of a regional range were mostly mid-term events (8:20:6). In this type, however, nature-induced crises are clearly predominant (18:12). Finally crises with a general range are very few compared to both other types with a relation of 51:34:12 more than four times fewer than local ones. The absence of short-term general-range crises is obvious, as the diffusion of a crisis throughout a vast geographical frame required sufficient time. Again there is a clear majority of nature-induced crises over human-induced ones (9:2). Indeed only such phenomena as climatic anomalies or pests could have a negative effect on the lives of contemporary populations over vast geographical horizons, as large-scale interregional warfare was seemingly beyond the capacities of these past societies. As far as the different types of location that these crises occurred in, I opted for a model of three categories according to the type of location they occurred in: urban, rural and mixed. The category 'mixed' refers to crises that affected both urban and rural communities. One must note that this division is sometimes arbitrary as the sources do not furnish us with enough information as to be certain which of types of human habitation were struck by these subsistence crises. In the following table subsistence crises have been divided according to their causes, taking into consideration only those for which there is enough evidence as to suggest possible trends. Table 3.2 Cause

Location of subsistence crises Urban

Rural

Drought F:4, S: 6 F: 2, S: 2 Locusts F: - S: 1 Siege F : 1 7 , S:Warfare F:-, S: 2 F: 6, S: Grain F: 8, S: 5 Total F:29, S: 14 F: 8, S: 2 * F = Famines, S = Shortages

Mixed

Total

F: 13, S: 4 F: 4, S: 1 F: 8, S: 1 F: 25, S: 6

31 6 1 7 17 13 84

Subsistence Crises: Causes, Location, Duration and Range

55

The emerging trends show a clear predominance of drought-induced subsistence crises in rural and mixed environments and at the same time an equally evident preponderance of siege-induced famines in urban centres. Famines caused by warfare have a strong presence both in rural and mixed settings, whereas in urban situations famines and shortages tend to develop when the transportation of grain is disrupted. In 450 years surveyed in the catalogue I have recorded 134 subsistence crises, which would sum up at an incident every 3.3 years on average (cf. Chapter 2). We must agree with Garnsey on the fact that 'a succession of bad or mediocre harvests had a more devastating effect than a single crop failure.'53 The probability of two successive crop failures is more than three times less likely to occur than that of a single one.54 Nevertheless, small-scale farmers who usually produced merely enough food for their sustenance would be affected even by a slight deviation from the norm (cf. Nos 4, 21, 22, 64, 91, 98, 194, 207). In one case (No. 80) our source records the farmers' efforts to overcome the destruction of the wheat harvest by locusts by sowing millet, a less demanding grain, as late in the year as June and July. When this crop failed as well the crisis escalated.55 As far as the prospect of crop failure is concerned we may note an unbroken continuity since Classical Antiquity, as no technical amelioration had been achieved and dependence on exogenous factors was still considerable.56 A variety of factors could bring about the destruction of the harvest and the ensuing subsistence crises: climatic phenomena, the occurrence of a plague of locusts, or warfare. According to our sources drought was the main cause of the failure of the harvest as expressly recorded in 14 out of 29 cases (Nos 2, 21, 27, 29, 55, 74, 125, 126, 153, 167, 170, 192, 194, 219). When other factors were also present the crises were certain to become severe, as in cases where drought was accompanied by the appearance of locusts (Nos 85, 88, 211). The migration of locusts per se could likewise have a devastating effect on the harvest, but this was not as

53

Garnsey, Famine, 23. Garnsey, Famine, 17, Table 3. 55 Sallares,£co/ogy, 363,492-3. 56 T. Lounghis, EmffKOTirjarj fiv£avTivrj