Military History of Late Rome 602–641

By Ilkka Syvänne

“A fascinating, informative, thoughtful and thought-provoking history. . . . comprehensively detailed and documented.” —Midwest Book Review

This book provides a fresh analysis of the Roman Empire from the reign of Phocas (602-10) until the death of Heraclius (610–41), an era of unprecedented upheavals usually considered to have resulted in the end of antiquity. The usurpations of Phocas and Heraclius led to the collapse of the Roman defenses; the Persians conquered Syria, Lebanon, Palestine and Egypt; the Avars and Slavs ravaged the Balkans; the Lombards held the upper hand in Italy; the Visigoths conquered Spain. But then came one of the most remarkable reversals of fortune in human history.

Heraclius began his holy war against the fire-worshippers. He launched a very effective counterattack against the Persians; as a result the Romans not only defeated the Persians but extended their domains at their expense, and by 633 Heraclius was already restoring the Roman fortunes in the Balkans. This volume revises many of the previously held views of how this took place and what actually happened. The Roman armies seemed invincible, and then happened yet another great reversal of fortune—the Muslims began their jihad to conquer the world.

The author explains in detail why these reversals of fortune happened; what enabled the Romans to defeat the Persians; and why the Romans lost when they fought against the Muslims and, conversely, why the Muslims were so successful. This was an era of epic campaigns and battles, all of which are explained in unprecedented detail.

Includes maps and illustrations

Praise for Military History of Late Rome 425–457

“Outstanding . . . [the series] gives us a very good picture of the long process that has come to be known as the ‘Fall of Rome.’ This is an invaluable read for anyone with an interest in Late Antiquity.” —The NYMAS Review

• Language: English
• Publisher: Open Road Integrated Media
• Release date: Dec 29, 2022
• ISBN: 9781399075688

Ilkka Syvänne

Dr. Ilkka Syvänne gained his doctorate in history in 2004 from the University of Tampere in his native Finland. Since then he has written extensively about ancient and medieval warfare and his publications include: 'The Age of Hippotoxotai, Art of War in Roman Military Revival and Disaster 491-636' (Tampere UP 2004), 'The Reign of Gallienus' (Pen & Sword, 2019), the multivolume 'Military History of Late Rome' published by Pen & Sword and the critically acclaimed Caracalla. He is the co-author with Professor Katarzyna Maksymiuk of the 'Military History of Third Century Iran' (Siedlce UP, 2018) and the 'Military History of Fifth Century Iran' (Siedlce UP, 2019). He was Vice Chairman of the Finnish Society for Byzantine Studies from 2007 until 2016. He has been an Affiliated Professor of the University of Haifa since 2016. He lives in Kangasala, Finland.

Book preview

Military History of Late Rome AD 602–641

For my wife Sini, and children Ari and Nanna for their patience.

I dedicate this volume in particular to Ari.

His enthusiasm for military history and wargaming has made his father proud.

‘Men worry less about doing an injury to one who makes himself loved than to one who makes himself feared. The bond of love is one which men, wretched creatures that they are, break when it is to their advantage to do so; but fear is strengthened by a dread of punishment which is always effective. The prince must nonetheless make himself feared in such a way that, if he is not loved, at least he escapes being hated… and the prince can always avoid hatred if he abstains from the property of his subjects and citizens and from their women… But above all a prince must abstain from the property of others; because men sooner forget the death of their father than the loss of their patrimony.’

Niccolò Machiavelli, The Prince, tr. by George Bull, 96–7

The Prophet Muhammad said:

‘War is deception.’

Umar ibn Ibrahim 4.1, 4.3 in A Muslim Manual of War, tr. by Scanlon 59, 61

The Military History of Late Rome AD 602–641

Dr. Ilkka Syvänne

First published in Great Britain in 2022 by

Pen & Sword Military

An imprint of

Pen & Sword Books Ltd

Yorkshire – Philadelphia

Copyright © Dr. Ilkka Syvänne 2022

ISBN 978 1 39907 567 1

Epub ISBN 978 1 39907 568 8

Mobi ISBN 978 1 39907 568 8

The right of Dr. Ilkka Syvänne to be identified as Author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988.

A CIP catalogue record for this book is available from the British Library.

All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopying, recording or by any information storage and retrieval system, without permission from the Publisher in writing.

Pen & Sword Books Limited incorporates the imprints of Atlas, Archaeology, Aviation, Discovery, Family History, Fiction, History, Maritime, Military, Military Classics, Politics, Select, Transport, True Crime, Air World, Frontline Publishing, Leo Cooper, Remember When, Seaforth Publishing, The Praetorian Press, Wharncliffe Local History, Wharncliffe Transport, Wharncliffe True Crime and White Owl.

For a complete list of Pen & Sword titles please contact

PEN & SWORD BOOKS LIMITED

47 Church Street, Barnsley, South Yorkshire, S70 2AS, England

E-mail: [e-mail]

Website:www.pen-and-sword.co.uk

Or

PEN AND SWORD BOOKS

1950 Lawrence Rd, Havertown, PA 19083, USA

E-mail: [e-mail]

Website:www.pen-and-sword.co.uk

Contents

Acknowledgements

List of Plates

List of Maps

Preface

Abbreviations

Chapter 1 Introduction: The Roman Empire in 602

Chapter 2 Enemies and Allies

Chapter 3 Phocas the Tyrant (602–10)

Chapter 4 Heraclius (610–41)

Appendix I: The Battle of Dhu Qar in ca. 604

Appendix II: Untangling the Information in Movses Dasxuranci/Dasxurantsi

Appendix III: The Battle of Filh/Pella According to Azdi (pp.144–68)

Notes

Bibliography

Acknowledgements

First of all I would like to thank commissioning editor Philip Sidnell for accepting the proposal for this book, and Professor Geoffrey Greatrex for recommending me. Special thanks are also due to copy editor Matt Jones, Barnaby Blacker, Tara Moran, and other staff at Pen & Sword for their stellar work and the outstanding support they have given this author. I would also like to thank many of my friends and family for their support and patience.

I owe special thanks to Perry Gray whose perceptive comments have improved the argumentation significantly. I also owe big thanks to Jyrki Halme for the re-enacting photos.

None, however, are responsible for the possible mistakes that remain. Those are the sole responsibility of the author.

List of Plates

Roman multipurpose trooper from Thrace. (Author’s drawing)

Roman footman using darts. The Strategikon (12.2.16.39–46) instructed those of the skoutatoi (shield-bearers) footmen who had lead-pointed darts (mattiobarbuli, martiobarbuli, martzobarboula, plumbatae) to throw them at the right distance (ca. 40–60 metres) from the enemy line. The darts were usually placed inside the shield five apiece. In this example I have assumed that the footman was not among the front ranks but was one of the rear-rankers who used a javelin instead of a kontarion-spear. He has thrust the javelin into the ground so that he can grasp it immediately after he has thrown the darts, but he could equally well have held it in his left hand or placed it on the ground. He uses the above-head javelin/dart technique, which was safer to use in a phalanx than the softball/cricket throwing style. The skoutatoi equipped with the kontarion probably placed their spears on the ground. The plumbata tribolata was designed to act as a sort of caltrop so that when large numbers of them were thrown the darts formed a defensive field of caltrops. (Author’s drawing)

Avar horseman. (Author’s drawing)

A bust of Phocas. (British Museum, author’s photo)

David Plates, David confronting Goliath. (Metropolitan Museum of Art, public domain)

A coin of Phocas. (British Museum, author’s photo)

Two skoutatoi equipped lightly for difficult terrain and an infantry archer. (Author’s drawing)

Three Slavic footmen. (Author’s drawing)

A foot archer using a solenarion (arrow-guide). (Author’s drawing)

A front-rank horseman charging as instructed by the Strategikon.(Author’s drawing)

A textile roundel depicting two horsemen in combat. (Metropolitan Museum of Art, New York, 6th or 7th century. Public domain)

Two Lombard lancers. (Author’s drawing)

Emperor Heraclius at the Battle of Sarus River in 625. (Author’s drawing)

Chosroes II Parwez on horseback in the Taq-i-Bostan relief. (Author’s drawing)

Three Roman heavy infantry skoutatoi ready to fight in regular terrain. (Author’s drawing)

Visigothic soldiers. (Author’s drawing)

A medallion of Heraclius ca. 1403. (Source: Goltz, 1645)

A Roman officer in light gear. (Photo © Jyrki Halme)

Chosroes II Parwez (590-628). (Author’s painting, a combination from two different sources: the Louvre bronze bust, and Hottenroth’s reconstruction of a Sasanian/Sassanid ruler)

Chosroes II Parwez. (Author’s painting after Taq-i-Bustan relief)

A coin of Heraclius (ca. 610–13). (Classical Numismatic Group, Inc)

Chosroes II Parwez at the battle of Arzamoun/Arzamun River in 604/5. (Author’s drawing)

On the first day of the battle Chosroes II Parwez was leading his men on horseback. This is proved by the fact that a Roman soldier lassoed him. It was only the quick reaction of Chosroes’s bodyguard Mushkan that saved him from being captured. Mushkan cut the cord and freed the shahanshah. On the basis of this it is probable that the first day of the battle was a cavalry battle fought between the two cavalry forces, the Roman under the eunuch Leontius and the Persian under Chosroes. It was therefore presumably on the second day that Chosroes ordered his elephants into a fortress formation (frourion), which either means hollow square/oblong array or the deployment of the elephants in close-order phalanx. It was this elephant fortress that then crushed the opposing Roman force. I have here taken the artistic liberty of placing Chosroes on an elephant for the second day of the battle. He probably would have stayed behind to lead the men from a safe distance, to avoid being lassoed again. The elephant is the famous gigantic white elephant that Chosroes used as his personal mount. The reserves are shown behind Chosroes with their spearheads pointing skywards. I have also taken the artistic liberty of placing some of the Christian concubines from Chosroes’s harem into the howdah with him, both entertaining him and offering wine made from Shiraz grapes whenever the ruler felt thirsty. To quote the comedian Mel Brooks: ‘It is good to be the king.’ Chosroes II was a man who took full advantage of his position as a King of Kings. According to Hamza al-Isfahani (tr. p.74), Chosroes had in his palace 3,000 free women and 12,000 female slaves for music, amusement and various other kinds of service. His stable had 8,500 horses for his own personal use plus the horses for his retinue. His bodyguards consisted of 6,000 men. In addition, he had 960 elephants, 12,000 mules and 20,000 Bactrian camels. It is a wonder that he managed to achieve so much when he had 15,000 females to entertain him. The equipment of the guards is taken from a Transoxanian silver dish dated seventh-eighth century. The equipment of the elephant is a free interpretation of the evidence from the narrative sources (Sassanian elephants with armour and towers) and later twelfth-thirteenth century Persian works of art that depict shields used by the Persians to protect the vulnerable elephant ears. The reconstruction of the howdah takes artistic liberties because we do not possess period evidence for its construction.

Shahrbaraz preparing his army for combat in 622. (Author’s drawing)

Heraclius preparing his army for combat in 622. (Author’s drawing)

Heraclius spearing the unhorsed Razates at the battle of Niniveh on 12 December 627. (Author’s drawing)

Arab/Muslim soldiers. (Author’s drawing)

Two coins of Heraclius (ca. 613–16, 629–32). (Classical Numismatic Group, Inc)

Khalid ibn al-Walid, the Sword of Allah. (Author’s drawing)

List of Maps

Provinces in the Notitia Dignitatum

Arabia

Takt-e Solayman

The neighbourhood of Ctesiphon

Persian Front

Upper Mesopotamia

Persian Front Eastern Section

Roman Syria

Muslim Conquest of Syria in 634–6 (principal locations)

Roman Territory facing the Muslims in 633

Caesarea Maritima (ancient and medieval)

The Nile Delta

Environs of Alexandria in the Barrington Map

North Africa

The Sea of Marmara with its surrounding regions

Bilad al-Sham

Approaches to Constantinople

City of Constantinople

Imperial Palace of Constantinople

The Avar-Roman Wars on the Balkan Peninsula (left side)

Thessalonica

Singidunum (Belgrade)

Balkans

City of Dvin

Modern Reconstructions of Ravenna

Lombard Italy

Italian places

Neighbourhood of Rome

City of Rome

Spain

Jerusalem

Ascalon

Salona

Roman administration in 602

The Military and Palatine Administration in ca. 602

Military deployment c. 602

Standard infantry formations

Other standard infantry formations

Marching formations in difficult terrain

Large cavalry army

The cavalry formations for armies of 5,000–15,000 men

The cavalry formation for armies less than 5,000/6,000 men

Slavic tactics

Nomadic battle formation

Scythian drill

Germanic cavalry charge

The two standard Persian cavalry formations

Mid-7th century Arab-Islamic army

Eastern Front 603–5

Campaigns in the northern sector in 604–7

The counter-attack of Bonosus in 609

Battle of Alexandria in about 609 and Siege of Alexandria by Persians in about 615–9

The military campaigns in 610

Siege of Caesarea (Cappadocia) in 611–2

Persian conquests in 613–22

The Avar invasion of the Duchy of Friuli in about 610

Heraclius’s first counter-offensive against the Persians in 622–3

Heraclius vs. Shahrbaraz in 623, phases 1–2

The Persarmenian theatre during the offensive 624–5

Heraclius’s surprise attack in 625

Heraclius’s offensive in 624–5

The Battles of Euchaita, Ophlimos Mons and Lycus River in about July-August 626

The Siege of Constantinople in 626

The city of Derbend, the Walls of Derbend Pass, Caspian Gates

Heraclius’s operations in Iberia and Persarmenia in 626–8

Roman campaigns against Persians in 626–8

The Battle of Niniveh on 12 December 627

The first Muslim operations

634 offensives

Movements of the Muslim armies 633 to the end of May 634

Map of Wadi Yarmuk

The first major battle of Yarmuk. The likely battle formations

The last battle of Yarmuk. The likely battle formations

Siege of Damascus

Battle of Fihl/Pella

Battle of Marj al-Rum

Battle of Emesa 637

Muslim Conquest of Egypt; the initial phase

Battle of Fihl/Pella Day 1; phases 1–4. Azdi version

Battle of Fihl/Pella Day 2; Azdi version

Preface

The intention of this final volume in the eight-book series is to present an overview of all the principal aspects of Roman military history during the years 602–641. In the first half of the sixth century Justinian I had reconquered most of the west, but the haughtiness of Justin II almost undid all of these achievements. Tiberius II and Maurice restored the situation, but the Roman Empire faced its next challenge when Phocas overthrew Maurice and he was in his turn overthrown by Heraclius. This took place against the background of a fight for survival against the reinvigorated Persian Empire. Heraclius saved the Roman Empire from collapse only to face a crushing defeat at the hands of the newly emerging superpower of Muslims. This era saw an apocalyptic outburst of fighting that resulted in the end of the classical world and the rise of the new Islamic superpower.

This text follows the same principles as previous volumes. It includes direct references to the sources only when my conclusions can be considered controversial or when necessary for some other reason. I have also not included descriptions or analyses of the sources used and their problems, because there exists expert literature devoted to this subject. When I refer to some chronicle, the exact point of reference can be found in the annalistic dating even when I do not always state it in the narrative.

When I refer to Spain I mean the whole of the Iberian Peninsula including Lusitania (modern Portugal). The aim has been to present all events in chronological order, but this approach has been sometimes abandoned for the sake of clarity so that some events in certain parts of the Empire (e.g. Italy, Spain, Balkans, North Africa, Egypt, the East) are dealt with in separate chapters. This is what Procopius and Theophylact did. As far as the language, transliteration, and titles are concerned I have usually adopted the easiest. I have used the transliterations the Romans used or those which are otherwise commonly used. I have adopted the practice of the Oxford UP and used capital letters for all offices which could be held by only one person at a time.

All illustrations, drawings, maps, diagrams etc. have been drawn and prepared by the author unless stated otherwise. I have used the Barrington Atlas as the principal source for the maps.

Abbreviations

Chapter One

Introduction: The Roman Empire in 602

¹

The general situation in 602

The inheritance of Maurice to his successor was an Empire that had defeated all its external enemies. The Persians and Avars had been crushed, and at the time of his death Maurice’s armies were poised to put an end to the Slavic incursions once and for all. Had Maurice not been so greedy and stubborn his armies would have undoubtedly followed this with further operations in Italy against the Lombards and in Spain against the Visigoths so that all of the reconquests of Justinian I would have been secured for years to come. It is even possible that the Muslim conquests would not have taken place, because the generals and soldiers of Maurice had proved themselves superior to all others. The Roman Empire was the wealthiest empire on earth with the best armed forces in existence. Furthermore, Maurice left his successor Phocas I an imperial treasury full of money. It was thanks to this that Phocas I was able to bribe everyone and raise new armies to replace those lost. However, this was not the entire picture. The Roman Empire was divided from within. Phocas was a usurper. His operatives had failed to kill the entire family of Maurice. Most importantly, Maurice’s son Theodosius had managed to flee to the court of Chosroes II Parwez who owed his position to the help that Maurice had given him. There were many officers and soldiers in the East who felt that Theodosius was the only legitimate successor of Maurice.

Roman society, administration and military in 602

At the top of Roman society was the Emperor (the Augustus). He was assisted by the imperial administrative machinery and the staff of the imperial household. In practice there also existed an unofficial hierarchy which was based on closeness to the Emperor. The most important of these were those in the Imperial Palace including for example the imperial family, eunuchs and commanders of bodyguard units (in particular the Comes Excubitorum). There still existed senates in Rome and Constantinople, but the former had lost most of its importance because it was no longer the place where the Emperor resided. Emperors could use the Senate of Constantinople when they wanted to involve members of the upper classes in their decisions, but they could also decide without consulting it. At the beginning of the seventh century, Roman imperial administration was divided into three sections: 1) Military (Foederati, Comitatenses, Limitanei); 2) Palatine; 3) Imperial and fiscal administration.

Roman Armed Forces in 602

In 602 Roman armed forces consisted of:

1) The imperial bodyguards (excubitores commanded by the Comes Excubitorum ; scholae commanded by the Comes Scholariorum; protectores domestici commanded by the Comes Domesticorum; it is also possible that they included the so-called optimates because both were included in the Strategikon and then later in the Thema of Opsikion; the other possibility is that they were commanded by the praesental magistri ; similarly it is possible that they included personal bucellarii employed by the Emperor)

²

2) The non-praesental units posted in the capital or nearby (the Walls /teiche under a Praetor; numeri/noumera marines under praefecti ; vigiles/vigla and urbaniciani under the Praefectus Urbi)

3) The spies (kataskopoi) who served under the Magister Officiorum

4) The praesental forces (central field armies) called stratiotai (soldiers) commanded by the praesental magistri (these may have included the optimates)

5) The comitatenses (field armies) also called stratiotai commanded by the regional magistri or exarchi

6) The limitanei (frontier forces) also called stratiotai commanded usually by the duces

7) bucellarii (private retainers) – command structure depended on the employer

8) foederati (federates) and other mercenaries commanded by the Comes Foederatorum

9) The Imperial Fleet of Constantinople under a praefectus , archegos or nauarchos who was a subordinate of a praesental magister or strategos

10) The provincial fleets each under a praefectus or archegos each of whom was a subordinate of a regional magister or strategos

11) allies (symmachoi)

12) civilian paramilitary forces.

The divisions between the different categories of troops were already blurred, but as we shall see, the introduction of the themes removed the last vestiges of these. The praesental armies, comitatenses, limitanei and foederati were scattered throughout the Empire. The structure of the Armenian field army was also a special case. Most of its cavalry forces consisted of the feudal retinues of the Armenian nobles. The only real difference between the regular stratiotai and foederati/foideratoi was that the former served under regular terms of service while the latter were soldiers who served under a contract which could vary. The foederati were mercenary troops employed by the state. The bucellarii were likewise mercenaries, but their employers were wealthy private persons. It is possible that the imperial bucellarii employed by the Emperor had already been attached to state structures because the Strategikon considers them a special category of troops. A further development of this probably took place under Heraclius when he attached the bucellarii of Priscus to his forces so that it is probable that these came to form the core of the later Thema of Boukellarion (however, it is also possible that these later became the optimates of the Thema of the Opsikion if another group of imperial bucellarii existed before this). The difference between the regular foederati and symmachoi was also blurred. My own educated guess is that it concerned the terms of their contracts: the regular foederati were expected to serve where ordered while the terms of the symmachoi stipulated the place where they served. This, however, is merely my guess.

The generals

The Emperor was still the supreme commander of all armed forces and Maurice had taken personal command of the army when necessary. This, however, was rare. Emperors usually left command to their subordinates. The Emperor was at liberty to choose anyone he wanted for any position and he could also create special commands at his whim. The next in descending order below him were the praesental magistri, the regional exarchi and magistri, the honorary magistri, and below them the comites and duces in charge of frontier forces. In practice, however, the Comes Excubitorum, scribones and eunuchs of the bedchamber were often held in higher esteem than the magistri because of their closeness to the Emperor so they were often given special commands in which they outranked the others.

The mercenary federates officially had one overall commander, the Comes Foederatorum, but in practice they were detached to serve under other commanders. The Federate Arabs (considered as symmachoi) were a special case within this category of troops because they were commanded by their own tribal leaders/sheiks each of whom held the title of phylarchos.

On the basis of the Strategikon we know that this system was modified according to the needs of the moment so that the Emperor chose a single commander called strategos who was assisted by hypostrategos in command of a single theatre of war. This commander could even be given temporary imperial powers as strategos autokrator (dux imperator, general-emperor) if this was considered necessary. The subordinates of the strategos and hypostrategos were merarchai (commanding a max. of 6–7,000 men) and moirarchai (commanded a max. of 3,000 men).

The military hierarchy of the bucellarii (private retinues) varied according to their size. The greater households included the following: 1) the overall commander (efestōs tē oikia, majordomo); 2) a treasurer (optio); 3) officers (doruforoi/doryforoi); 4) privates/soldiers (hypaspistai). Some of the commanders placed these in command of regular forces and the regular officers were required to obey them. See also Military History of Late Rome Volumes VI–VII.

The Strategikon (1.3.23–4) names taxiarchoi as moirarchai (commanding a maximum of 3,000 men) of the optimates. In the previous volume we found taxiarchoi several times in important commands in the text of Theophylact, which may mean that the strategoi often appointed members of the optimates to important positions in the same manner as Belisarius had used his bucellarii. This, however, is uncertain because we do not know if Theophylact used the term in its technical sense, but it is still something that readers should keep in mind when the title appears in the text.

Units Structures and Grading

In 602 the regular Roman forces were a reflection of their past. The Roman army consisted of two basic types of units: a) units created before the third century (legions, alae, turmae, auxilia); b) units created after the third and fourth centuries (vexillationes, cunei, auxilia palatina, scholae). The units created before the third and fourth centuries (legions and auxiliaries) retained most of their old structures (infantry tribunes and centurions; cavalry tribunes, centurions and decurions). There were also temporary units of varying sizes that were detached from their mother units with temporary commanders who could be praepositi, tribunes, duces, centurions etc. These ad hoc temporary groupings had come to be known in military slang as throngs of men (Latin sing. drungus pl. drungi; in Greek sing. drouggos/droungos, pl. drouggoi/droungoi) in the third century,³ but the first attested use of the new official title drungarius (in Greek drouggarios/droungarios) for such a throng dates from the early seventh century. The droungos was both an irregular group of soldiers and an irregular unit order in combat in the Strategikon and it retained these meanings at least until the tenth century. Therefore it is possible that the title of droungarios was already in official use at the turn of the seventh century. The grading of the units created during the third and fourth centuries and later presented a simplified system which was further streamlined by the Strategikon to become the core structure of the Thematic armies. The Strategikon represented official military doctrine, so even if the units and officers retained their official titles in practice they were now grouped as in the Strategikon as follows. The list is based on the comparisons of grading by J. Haldon (1999, 109) and R.I. Frank. As can be seen, the Strategikon streamlined and rationalized even the newer simplified structures:

Conditions of service

The bulk of the regular forces (stratiotai and limitanei) consisted of volunteers. Traditional conscription was used only when there were not enough volunteers available. Soldiers could get honourable discharge (honesta missio) after 20 years of service, but they could only get the full privileges of a veteran (emerita missio) after 24 years. Soldiers were entitled to be relieved of duty (causaria missio) on the grounds of wounds, injuries or old age if they met the requirements for this. The terms of voluntary treaty soldiers, the foederati and bucellarii, varied greatly according to the contract they had signed.

The usual assumption is that sixth century soldiers no longer received the yearly salary called stipendium. It had been replaced by: a) rations called annona militaris (bread, meat, wine and oil); b) capitus (fodder); c) an annual stipend to buy their uniforms and equipment. The annona and capitus were usually converted into money in peacetime. Soldiers also preferred to pocket the stipend meant for equipment meaning they were sometimes poorly equipped for fighting. Maurice had set out to correct this by converting the stipend into payments in kind. He divided the salary in three parts: clothing, equipment and gold coins. Maurice had two goals: to save money and to ensure that the men were better equipped, but this only resulted in mutinies – although it is possible that the soldiers finally accepted it in 594 (see Military History of Late Rome 565–602). The soldiers also received periodic donatives in cash. The limitanei and some of the foederati supplemented their salaries with income from their farmlands or businesses, as did every other soldier who had personal or family wealth.

The logistical system was governed by the praetorian prefects. Below each praetorian prefect (hyparchos) served men with the title vicar (vicarius, eparchos) and below them governors who gathered rations and fodder from taxpayers which were then stored in storehouses. The process of delivering provisions (or cash if converted into money) to the soldiers was in the hands of the ‘regimental quartermasters’ called actuarii and their subordinates the optiones. If the Emperor dispatched an expeditionary army or assembled large numbers of soldiers to a single theatre of war, he usually appointed a special praetorian prefect to organize the supply of the army.

Supporting services

The fighting men were supported by large numbers of other professionals so that they could perform their primary function which was to fight. The Roman state provided all of them. The state owned factories producing standardized arms, equipment and clothes. The state also educated and trained specialists for every possible need, for example workmen, carpenters, bowyers, metal workers, swordsmiths, artillerymen, artisans, medics, doctors, vets, cooks, armourers, architects, engineers, siege engineers and even chaplains to take care of spiritual needs.

Combined arms combat

The military doctrine introduced by Maurice in his Strategikon was based on existing combined arms tactics but with the cavalry as the dominant arm due to the types of enemies the Romans faced. Thanks to the reforms of Maurice the general fighting quality of all forces was superb. This included officers, who had been thoroughly trained in all forms of warfare. Under Maurice, Roman armies dominated the battlefields.

Infantry tactics in the Strategikon were based on the phalanx with its variations, but it is clear that the most commonly used battle formations were lateral phalanx, hollow square/ oblong and rearward angled half-square. Cavalry tactics were based on traditional cavalry formations with reserves posted behind. Heavy infantry phalanxes consisted of rank and file formations that had 4, 8, 16 or at most 32 ranks. The light infantry was typically posted behind the heavy infantry, but it could also be placed between the files or on the flanks. Light infantry was typically used for skirmishing and pursuit of the enemy. The infantry phalanxes consisted of: a) the mere – divisions (sing. meros) consisting of at most 6,000–7,000 men (i.e. they equalled legions); b) 2,000–3,000 men moirai (sing. moira); c) and about 200–400 men tagmata (sing. tagma). Cavalry was typically posted on the flanks and/or in front. It was also expected that the commander used reserves consisting of both infantry and cavalry. Armies with less than 24,000 footmen were deployed as three phalanxes (mere), while armies that had more than 24,000 footmen were deployed in four phalanxes (mere). If there were more than 28,000 footmen, the extra men could be placed outside the cavalry wings or wherever the strategos desired.

Infantry organization was based on multiples of the tent groups that formed a file in combat. It consisted of ten men: eight soldiers, one recruit and one servant.

The standard phalangial array was: a) the lateral phalanx arrayed as a single line if baggage train followed or as a double phalanx if baggage train did not follow. The other standard phalangial formations were: a) the hollow square/oblong; b) oblique arrays to outflank; c) rearward-angled half-square (epikampios opisthia) which was usually used against cavalry; d) forward-angled half-square (epikampios emprosthia); e) wedge (embolos/ cuneus) unit formation to break through enemy array; f) hollow wedge (koilembolos) unit formation against wedge; g) convex (kyrte); h) crescent (menoeides) to outflank; i) mixed formation (cavalry placed between infantry mere) used against enemy cavalry. The Peri Stragias/Strategikes (34) includes additional variants: a) phalanxes posted in depth of one, two or several lines of phalanxes; b) multipurpose soldiers armed with both bows and melee weapons. In combat the different units were expected to be able to manoeuvre independently of each other as the situation required. For example they could wheel to present a front in another direction, to form a wedge to break through and a hollow wedge to oppose the wedge, to open the formation for enemy cavalry/elephants to pass through, or to form a double-front or double phalanx to oppose enemies approaching from behind etc.

Infantry unit orders consisted of: a) open order for marching; b) close order (shields rim-to-rim in width) and foulkon (shields rim-to-boss in width with a shield roof) for battles in open terrain; c) kneeling foulkon (shields rim-to-rim in width with spears pointing and shield roof for three to four front ranks) against cavalry; d) multipurpose spearmen and archers against cavalry (front rank shooting at the chests of horses, following at higher trajectory, until the horses came so near that they adopted the spear wall); e) drouggos/droungos – irregular array for difficult terrain and for when speed was of the essence.

Roman heavy infantry skoutatoi (shield-bearers) consisted of: a) heavy infantry skoutatoi in its standard gear of armour (mail, scale, lamellar, muscle) and helmets (various types: ridge, segmented, one piece), shields (typically round or oval 80–90 cm in width, but rectangular shields were also used) with darts, kontarion-spear, spathion-sword and sometimes also shin guards; b) multipurpose (melee weapons and bows) armoured heavy infantry possibly equipped with extra-large shields about 1.5 metres high; c) "peltastai", i.e. lightly equipped unarmoured skoutatoi wore the spathion (medium to long two-edged swords), semi-spathion/machaira (short sword), darts (plumbatae placed inside shield), shorter spear, larger-type shield, with or without helmet.

The light infantry consisted of: a) archers; b) slingers; c) javeliners. The foot archers were also instructed to use arrow-guides that had two benefits: a) the enemy could not reuse their darts; b) the arrow-guide doubled the shooting range (to 500–600 metres). The light infantry was at its best harassing the enemy, in pursuit, and in difficult terrain.

The third category of infantry consisted of dismounted cavalry, which was equipped, like all cavalry, to fight at distance with bows and at close quarters.

Men on foot used their spears in three different ways in combat: a) as spear-wall against the enemy while the rear ranks supported the attack with missiles (this was used against both infantry and cavalry); b) the first row/rank could throw the kontarion-spear or javelin and use the sword while those behind supported the front-rankers with their spears and those behind them with missiles; c) the first three ranks placed their spears on the ground after which the two front ranks aimed their arrows directly at horses, while those behind shot their arrows at a higher angle, and all then adopted the spear wall when the enemy cavalry came close.

In 602, thanks to the reforms of Maurice, Roman cavalry forces had a superb record of success against all enemies, and they expected this to persist into the future. The cavalry consisted of two classes: a) natives trained to mastery as mounted lancers, bowmen and footmen; b) recruits from the ‘barbarian lands’ who possessed basic fighting skills but were not expected to attain Roman standards until they received thorough training.

The mounted forces accompanying the Roman infantry were divided into units of mere consisting of at most 6,000–7,000 horsemen, 2,000–3,000 horsemen moirai, and about 200–400 horsemen tagmata, arithmoi and banda (sing. bandon, flag).

Fighting in difficult terrain

⁴

Fighting in wooded, rough or otherwise difficult terrain consisted of fighting in marching formation (see the diagrams). The Romans adopted this tactic in wooded areas, rough terrain, narrow passes, and against the Slavs and Antes. In such conditions the ‘heavy infantry’ skoutatoi/peltastai did not carry heavy equipment such as helmets, long spears, and mail coats; their armament consisted only of larger shields, swords and javelins. The expectation was that the army marched without wagons and did not have many horsemen. Soldiers expected to carry only the most essential supplies and equipment.

Depending on the size of the force, the lightly equipped skoutatoi were deployed in two, three or four phalanxes (mere), each of which was only two, three, or four ranks deep. These phalanxes were deployed as columns about a stone’s throw from each other. These were then wheeled into phalanxes when the army reached open terrain. Baggage trains (consisting of mules and horses) were posted behind the phalanxes and the cavalry behind the baggage. The rear was always protected by a rearguard consisting of heavy and light infantry. The formation was also covered by a protective screen of light infantry and lightly equipped skoutatoi, and/or of a small unit of cavalry, which marched about a mile ahead of the main body. The rest of the lightly-equipped infantry, possibly with small cavalry detachments, formed a screen on the flanks in irregular 4 or 5-man groups. The vast majority of the archers were kept with the marching columns because they were useful in rough, steep, narrow, and open terrain but less useful in the thickly wooded areas.

The main striking forces were the lightly-equipped soldiers deployed outside and around the phalanx structure. These four or five man throngs (droungoi, drouggoi) consisted of three or four men armed with javelins and shields and one with a bow. The throngs were deployed at a distance from which they could still hear the bugles and trumpets of the main body, and they were deployed one after the other so that they could protect each other’s back. If the leading throng encountered enemy resistance, those behind were expected to direct their attack against the enemy’s rear if possible. If there was higher ground, their main objective was always to seize it and get above the enemy. The men of these droungoi were the crème-de-la-crème of the Roman infantry.

Cavalry warfare

The regular Roman cavalry was equipped as dual-purpose troops with bows, spears and swords and they typically wore armour (muscle, mail, scale, lamellar), helmets (ridge, segmented, single piece) and shields (smaller than the infantry variants). However, there were also ethnic units like the Heruls (unarmoured) who used their native gear. The Strategikon divided the quality and amount of armour and equipment on the basis of the unit and the soldier’s place in the battle array. The personal wealth of the soldier obviously influenced this as well. The Strategikon (1.2.35–8) required that the horses of the officers (archontes = dekarchoi, pentarkhoi, tetrarchoi), chosen men and in particular those posted in the front rank were to be equipped with: 1) iron chamfron/head-piece; 2) peytrals/breastplates of iron or quilted/padded armour, or alternatively Avar-style peytrals and neck-guards/crinets.⁵ Depending on the mission the horses could be armoured or without armour, but in pitched battles the horses of the front rankers were usually armoured. The cataphracted cavalry consisted of two basic variants: a) horse with frontal armour (peytral, crinet) and head-piece (chamfron); b) horse with full armour (crinet, peytral, flancard, crupper) and head-piece (chamfron).

Cavalry organization also consisted of multiples of ten-man tent groups which fought as files in combat. The better the unit, the less deep it was; in an elite unit there could be five troopers each of whom had a squire, which means that in such unit only five troopers were posted in a file while the squires were left behind to protect the camp or to guard spare horses. At the other end of the spectrum were the regular units in which there were eight troopers and two squires. When needed the squires could also be added to the files to make the array deeper.

The Strategikon classed the cavalry formations as follows: a) large army with over 10,000–15,000 horsemen which had three divisions with outflankers and flank guards in the first line, four divisions in the second line with fill-up tagmata/banda between them, and a third line with two units on the flanks; b) medium sized force of 5,000–15,000 horsemen which had three divisions with outflankers and flank guards in the first line, two divisions with a unit between them as support line; c) small force with less than about 6,000 horsemen that had three divisions with outflankers and flank guards in the first line and one division as support line; d) if there were over ca. 50,000 horsemen, the extra men were to be posted outside the regular array either on the flanks or as a third line. If the situation allowed, the commander was instructed to post separate units of ambushers.

Cavalry unit orders were: a) open order for marching; b) droungos-irregular for difficult terrain or when speed was necessary, say in skirmishing or pursuit; c) close order used in pitched battles. The mere divisions were divided into units of koursores (runners, skirmishers) on the flanks and defensores (defenders) posted in the centre. The koursores used the irregular array in skirmishing and pursuit while the defensores protected by using close order.

Standard cavalry assault was done at a canter/trot so that the close order could be maintained, but if the koursores had skirmished before this they could be in the droungos array. However, in practice the Romans did charge at full gallop when this was considered advantageous. If the enemy fled, the pursuit was left in the hands of the koursores while the defensores followed them behind as a reserve. If any of the front division was forced to retreat, it was expected that the divisions of the second support line moved forward to protect them. Standard battle tactics were: a) If the Romans outnumbered the enemy, they outflanked the enemy on both sides; b) If the Romans had roughly equal numbers, they outflanked on the right; c) If the Romans had fewer men, the centre meros attacked before the wings. Other standard tactics were to use stratagems like hiding the second line, or placing ambushers behind a hill or in hollows or in woods. Feigned flight was often used in conjunction with these.

Siege warfare

Siege tactics of the period were highly sophisticated and based on Greco-Roman traditions. With the exception of the Persians all the rest of the foes the Romans faced at this time were no match for them. They employed all sorts of stone throwers and artillery pieces (ballistae, onagri, repeating ballistae, trebuchets), mounds, tunnelling, fire-bombs, ‘hand grenades’, siege towers, battering rams, borers, drills, various types of sheds, fire hoses to spread fire, etc.

Siege warfare can be divided into defensive and offensive techniques. Standard features in offensive sieges were: a) terms of surrender offered to the enemy at first to avoid a costly siege; b) surprise attacks if possible; c) the use of a traitor; d) if the enemy refused to surrender, assault in foulkon formation with ladders possibly with some sheds; e) if the first assault failed, the building of siege engines, mounds and/or mines; f) starving the defenders into submission.

Standard defensive siege techniques were: a) the building of sophisticated fortifications in places that were easily provisioned with water and food; b) scorched earth policy if there was prior information of invasion; c) the exploitation of defensive features such as walls and towers and the use of siege engines to negate the attack; d) sending a relief army; e) diversionary invasion; f) guerrilla warfare against the besiegers; g) if all else failed, the Romans offered terms of surrender.

The Roman Navy

The Romans possessed a professional navy with professional marines. It had permanent bases all over the Empire (Constantinople, Antioch/Seleucia, Alexandria, North Africa/ Carthage, Italy, Spain, frontier fleets along the Danube). In addition they posted detachments and patrols to protect sea lanes and commerce wherever needed. Thanks to their superior ship designs and professional fleet the Romans possessed naval supremacy on the Black and Mediterranean seas. Their fleets supported their land armies when possible, e.g. by carrying troops and supplies or by clearing waters of enemy vessels. The Imperial Fleet stationed at Constantinople possessed greater numbers of larger ships (bireme and trireme dromons) than the provincial fleets, thus retaining superiority in naval battles.

The workhorses of the Roman fleets were the fast galleys called dromons. There were many variants, but most consisted of the single-banked smallest variant at this time. Standard naval combat formations resembled those on land: a) single line abreast with reserves; b) the double line/phalanx with reserves; c) convex to break through the enemy centre; d) crescent to outflank; e) circle for defence. The largest ships were always placed where the Romans intended to break through the enemy array. The Romans posted in advance of these naval phalanxes two to four ships as scouting ships that were sometimes used to break up the cohesion of the enemy array by remaining in their positions. Transport ships were usually placed behind the dromons.

During this era the Slavs and Persians challenged the Roman dominance of the seas, but to no avail. The Slavic dugouts, their small rowed ships and boats, were no match for the Roman dromons or their other ships and their professional crews. The Persians appear to have used Roman resources once they got access to Roman shipwrights and sailors in Syria and Egypt, but these were no match for the professional crews of the Imperial Fleet stationed in Constantinople.

Riverine warfare required its own instructions which are included in the Strategikon and Peri Strategikes/Strategias. This was particularly important for the warfare the Romans conducted along the Danube frontier where they had to face both Slavs and Avars. In command of the fleet was either the nauarchos (the admiral) or the strategos. Depending on the size of the fleet his divisional commanders were called either moirarchs or merarchs. Ships used in riverine combat were either medium or small. A large fleet was divided into three divisions: warships led the convoy; cargo ships were in the middle; and some warships were left as rearguards. For naval combat the warships were drawn in a single line abreast covering the entire width of the river. If there were extra ships, they were posted in the second or third line with each line separated from the next by a bowshot. River crossings were conducted as follows: a small force of infantry, both heavy and light, was shipped secretly across to protect the building of the bridge which was to be built. If the crossing was contested, the admiral used his ballistae-carrying ships to clear the opposite bank.

Chapter Two

Enemies and Allies

¹

The principal enemies of the Romans at this time were the Germanic peoples (Visigoths, Lombards, Franks), Slavic peoples (Sclavenoi), Hunnic and Turkic peoples (Sabiri Huns, Bulgars, Avars, Turks), the Persians with their allies and subjects, and the rising power of the Muslims.

The Slavs and Antae

The Antae and the Slavs formed a specific type of enemy in the Strategikon of Maurice. The Slavs lived roughly in the area between Lake Neusiedler (near Vienna) and the Novae, in other words in Moesia/Pannonia, while the Antae were located somewhere at the curve of the Black Sea between the Dniester and the Dniepr rivers. Both groups are called Slavs in the following discussion.

In 602 the situation in the Balkans was calm thanks to the efforts of Maurice, but the chaos unleashed by the usurpation of Phocas would eventually reach this area too. The Strategikon did not consider the Slavs and Antae as serious threats because they were categorized as enemies that could always be engaged in pitched battles. The Slavs were poorly organized and poorly equipped so the Romans could expect to win every time they were able to collect at least a medium sized force to oppose them. Roman fortifications, however, were not serious obstacles to the mobile and lightly equipped Slavs. They avoided the main highways to avoid both the Roman armies and their fortifications. Furthermore, they could not be held in check by rivers because they could use their monoxyles (ranging in size from canoes/boats to Viking-size long boats) to cross them. Because of this the Strategikon contained instructions on how to fight in difficult terrain. Those skills were needed against the Slavs. The principal problem for the Romans was that the Slavs could amass huge hordes of warriors. Armies of 100,000 men plus the non-combatants were not unheard of. The Romans would have needed at least a medium sized force to oppose such a horde, but this they did not possess when they faced troubles in the East.

The Slavs were not a unified group but consisted of several tribes each with their own leader. Because of this the Strategikon (11.4) recommended dividing the enemy with alliances. The Strategikon also recommended using surprise attacks, but during this era the Romans did not conduct surprise attacks against the Slavs because they were preoccupied with wars elsewhere. It was now the Slavs who were invading Roman lands to settle there, and the Romans were forced to employ other Slavic peoples (Croats and Serbs) against the Slavs who had already invaded Roman territory.

The typical Slavic warrior was an unarmoured lightly equipped footman who carried a large oblong shield or a small shield, knife, and two to three short javelins. They were in the habit of throwing stones and using slings. The Croats were known for their use of axes. In addition to this, the Slavs were known for their use of wooden bows and arrows smeared in poison. The Strategikon therefore instructed the soldiers to prepare themselves for combat by taking an antidote. The Slavs had their own peculiar archery draw in which the cord was drawn with the three lower fingers. The Slavs also possessed small numbers of very high quality cavalry forces, which probably consisted of the retinues of wealthier Slavs. These were probably later called Druzhinas, i.e. military retinues of the voivodes (clan chieftains) and zupans (subordinates of the voivodes). The elite cavalry forces were also better equipped and could be armoured.

The Slavic peoples specialized in the use of difficult terrain, ambushes, sudden attacks, and raids, because they were aware of their relative weakness in pitched battles against the Romans. If they could not avoid open and level ground they usually used a defensive wagon laager. When the Slavs fought pitched battles they had their own peculiar style of fighting. At first they all shouted together, possibly howling like wolves, and moved forward a short distance to see if their opponents would lose their nerve. If this happened, the Slavs attacked without remorse. Slavic leaders and kings advanced in front of their men. If the enemy did not show any signs of nervousness, the Slavs turned around and ran to the woods. This flight could be authentic, but more often than not it was a feigned flight meant to induce the enemy to follow.

Roman countermeasures consisted of: a) surprise the enemy in their homes; b) engage the Slavs in open ground; c) if the Slavs occupied a strong position, the Romans attempted to lure them out with a feigned flight; d) the Romans equipped their men lightly for combat in difficult terrain. The Slavs were also vulnerable to the combination of archery, sudden outflanking manoeuvres, and hand-to-hand fighting in open ground.

The Nomadic Foe: the Turks and Avars

The principal nomadic foe the Romans faced during 602–641 was the Avars. The Avars on their own were not a serious danger, but when they cooperated with the Slavs and Persians they were, and it was this danger that the Romans faced until 626.

Typical nomad tactics were the same as always: deceit, surprise attacks, feigned flight and cutting off supplies. The Avars could put into the field armies of 60,000 to 100,000 horsemen plus the forces provided by their allies or subjects.

Wealthier Avars were equipped with a composite bow, lance (c.3–4 metres), sword, lasso, wicker shield or small shield (pelte/parma), segmented helmet (often with a nasal which demanded a straight or intermediate archery position even on horseback), and chain mail or scale or lamellar armour; their horses could also be armoured (lamellar, scale, hides, felt). The text of the Strategikon suggests that all Avars wore the full panoply of equipment,but it is still probable that the poorer Avars wore less armour than the wealthy elite.

The nomad combat formation consisted of a single line of irregular sized units behind which were reserves. These units were further divided into smaller units of about 40–50 horsemen. The nomads always posted a separate reserve as ambushers or support troops. Spare horses were placed immediately behind the first line. The baggage train was placed two to three miles behind the front line and a separate guard was delegated for it. See the attached diagram.

The Avars, Turks and Huns deployed their units in ranks and files just as the Romans, Alans and Mongols did; the armoured troops were placed in front to protect the lightly equipped. First, they charged towards the enemy while shooting arrows, and if the enemy showed any signs of disorder the attack was immediately pressed home. If the enemy withstood this the nomads retreated in irregular wedge arrays while shooting backwards. If the enemy took the bait and pursued, the nomads attempted to ambush them or outflank them. If neither of these took place, the nomads repeated this again and again until either their arrows ran out, their horses tired, or enemy resistance collapsed. The Turks and Avars were equally effective at long range with bows and at melee with spears and swords. The nomads were also known for their use of lassoes to break up the cohesion of enemy formations. They also employed a tactical manoeuvre which the Strategikon called Scythian drill, peppering the enemy with arrows (see p.22).

Roman counter-tactics were: 1) with cavalry forces the Romans charged immediately before the nomads could spread out or manoeuvre, any pursuit being conducted carefully because the Huns often placed an ambush behind their line; 2) the recommended method was to use a combined army of cavalry and infantry in which the infantry was dominant. The successes of Maurice’s forces against the Avars are the best evidence of their effectiveness in real life combat. The Romans had other advantages over the nomads too. Excluding the Sabirs, all the nomads were inept both as besiegers and when being besieged. This was still the case even after the Roman turncoat Busas taught the nomads how to build trebuchets. See the MHLR Vol. 7 for this incident.

The Light-Haired Peoples

²

The Strategikon grouped all the Germanic peoples under the same name, the Light-Haired Peoples. They also shared certain cultural traits which made this plausible. They were bold and dauntless in battle. They considered even a short retreat a disgrace. They were particularly good at hand-to-hand combat, both on horseback and on foot. Typical Germanic cavalry armaments consisted of shields, lances, and short swords slung from their shoulders. These were the basic armaments of the Lombard (Langobard), Visigoth and Frank knights. The Germanic peoples considered it the duty of every warrior to exact vengeance if their comrades were killed. Excluding the Visigothic battle array, which was formed on the basis of territorial groupings, and the Frankish territorial levies and remnants of old Roman units, the Germanic battle line was formed according to tribe, kinship and common interest. The infantry and cavalry were arrayed in ranks and files for combat so the front of the battle line was even and dense. Their attacks were always impetuous and undisciplined, both on horseback and on foot, which is not surprising because they were also known to be disobedient towards their leaders. The Germans disliked complicated manoeuvres, and despised good order, in particular on horseback. What made the Germans particularly dangerous foes was that they were ready to dismount to fight against very superior numbers if they faced trouble in cavalry combat.

The sight of an impetuous Germanic cavalry attack could scare anyone unused to it, but the Romans’ use of orderly cavalry formations was more secure. When the cavalry attack was disorderly, individual riders could flee more easily than was the case with a close order attack conducted at a trot/canter because it forced everyone to participate. However, the Romans still expected to prevail over the Germans in cavalry battles. But because they were unlikely to flee, the Germanic cavalries had less of a disadvantage than any other enemy the Romans faced, until the appearance of the Muslims, who fought just as tenaciously as the Germans.

The Strategikon advised the strategos to avoid pitched battles with the Germans, especially in the early stages. It advised the strategos to use ambushes, sneak and surprise attacks, and stratagems, which were also typical features of nomadic warfare. It was advisable to subject them to a shortage of provisions, especially of wine, and to the discomforts of heat or cold so that they would lose their willingness to fight. This by the way did not work against the Muslims which made them more dangerous than