Mercenaries to Conquerors: Norman Warfare in the Eleventh & Twelfth-Century Mediterranean

By Paul Brown

When a band of Norman adventurers arrived in southern Italy to fight in the Lombard insurrections against the Byzantine empire in the early 1000s, few would have predicted that within a generation these men would have seized control of Apulia, Calabria and Sicily. How did they make such extraordinary gains and then consolidate their power? Paul Brown, in this thoroughly researched and absorbing study, seeks to answer these questions and throw light onto the Norman conquests across the Mediterranean. Throughout he focuses on the military side of their progress, as they advanced from mercenaries to conquerors, then crusaders. The story of the campaigns they undertook in Italy, Sicily, the Balkans and the Near East reveals their remarkable talent for war. The dominant role played by a succession of Norman leaders is a key theme of the narrative a line of ambitious and ruthless soldiers that ran from Robert Guiscard and Bohemond to Roger II and Tancred.

• Language: English
• Publisher: Open Road Integrated Media
• Release date: Oct 14, 2016
• ISBN: 9781473880108

Paul Brown

38 years old and involved in Scouts for 20 odd years, I refuse to wear a uniform because its naff. I will include something apposite here in due course. In the meantime, thanks for taking the time to read my stuff and let me apologise about you not being able to get that time back. ;-)

Book preview

First published in Great Britain in 2016 by

Pen & Sword Military

an imprint of

Pen & Sword Books Ltd

47 Church Street

Barnsley

South Yorkshire

S70 2AS

Copyright © Paul Brown 2016

ISBN: 978 1 47382 847 6

PDF ISBN: 978 1 47388 011 5

EPUB ISBN: 978 1 47388 010 8

PRC ISBN: 978 1 47388 009 2

The right of Paul Brown to be identified as the 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.

Typeset in Minion by

Mac Style Ltd, Bridlington, East Yorkshire

Printed and bound in the UK by CPI Group (UK) Ltd,

Croydon, CRO 4YY

Pen & Sword Books Ltd incorporates the imprints of Pen & Sword Archaeology, Atlas, Aviation, Battleground, Discovery, Family History, History, Maritime, Military, Naval, Politics, Railways, Select, Transport, True Crime, and Fiction, Frontline Books, Leo Cooper, Praetorian Press, Seaforth Publishing and Wharncliffe.

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]

Contents

List of Illustrations and Maps

Illustrations

Maps

Preface

What you hold in your hands is the much-delayed result of a research project I first began in 2010. However, after delivering a paper on my initial findings at a conference held at the University of Leeds in 2011 – where I had the great pleasure of meeting formidable scholars of both southern Italian history (Graham Loud, Paul Oldfield and Alex Metcalfe) and of the Crusades (Alan Murray and Conor Kostick) – progress effectively stalled. Having been waylaid both by life and by an ongoing, laborious side project (knowledge of classical texts in the medieval period), an email from Graham Loud reignited this now dormant endeavour. He informed me that Rupert Harding from Pen & Sword Books was looking for someone to write an account of Norman military exploits in the Mediterranean. I thanked Graham for passing on the query, deciding immediately to complete what I had started a few years earlier. Although I hope he will not regret it after reading this book, Graham must be thanked not only for passing on Rupert’s query, but also for the help and encouragement he has given me to date. Lastly, the delay ended up being beneficial: various ideas have had time to coalesce and be refined, and the publication of books by Charles Stanton (2011) and Georgios Theotokis (2014) have since enhanced our understanding of a subject that had received little attention by academic historians when I first began researching it.

The end result is intended to be accessible to a broader audience than I normally write for – that is to say, hopefully the book will be of interest to scholars and students, as well as general readers with an abiding interest not only in the medieval period, but in the history of warfare and the societies that have waged it. It was certainly challenging to write with a wider audience in mind, and I am grateful to Rupert Harding for helping me to make the transition (it is hoped with some success on my part). However, as an academic historian, certain approaches are still present. Detailed notes are provided towards the end of the work, albeit with comments kept to a minimum, which include the quoted excerpts from the original languages I am conversant with: Greek, Latin, and Old French (for other languages such as Arabic, I have relied on translations in French, Italian and English). The notes have not only been retained for scholars and students who would like to consult the various primary and secondary sources on which the various arguments are based, but also to encourage other interested readers to delve further into sections that take their interest.

As both a writer and an avid reader of history, I have always appreciated attention to detail, and accordingly the same approach has been taken in this book. While italicized technical terms have been reduced to a minimum, some have been retained where English equivalents do not exist or are misleading. However, since military terminology is of great importance in regard to certain interpretations discussed throughout the narrative (e.g. spear types and attendant tactics), I have chosen to discuss them at length in the Appendix. Also, while some authors choose to summarize the outcomes of campaigns and battles in a few sentences, where the sources provide the requisite detail, I have chosen to devote paragraphs or even pages to them. Hopefully this method will not be too taxing for those who appreciate a lighter narrative, but the intention from the outset has been to illuminate, where possible, the various stages of battles and sieges, in addition to the tactics and equipment used in them.

The rendering of foreign names can be a difficult task, especially when there are often two or three variants in use, especially for Arabic and Turkic names. I have generally used English equivalents where they exist (e.g. ‘Basil’, instead of ‘Vasileios’ or ‘Vasili’), but not always (‘Raoul’ instead of ‘Ralph’). With Arabic names, the accenting system used by experts such as Alex Metcalfe has generally been followed. While the classical transliteration method for Greek has been employed, the letter β is rendered as ‘v’ rather than ‘b’ (it has been pronounced by Greek-speakers as vee-ta for over 1,500 years). The other exception is the transliteration of epsilon (ε) with a rough breathing: often rendered as ‘he’, since it has been silent for over a millennium, the letter has been rendered as ‘e’ (e.g. ‘elepoleis’ [‘city-taker’], not ‘helepoleis’; ‘Ellas’ [Greece], not ‘Hellas’). Lastly, since Turkic names are rendered in a variety of ways, I decided to impose a semblance of uniformity via use of the modern Turkish alphabet. Some letters may look unfamiliar at first, but the pronunciation of them is not difficult to learn: e.g. ç = ch in chapter, ş = sh in shape; hence Selçuk is pronounced as Selchook, and Tutuş as Tootoosh.

I am greatly indebted to the prodigious linguistic ability of Lynda Garland, who kindly offered to double-check my numerous translations from the Greek and Latin sources. Naturally, any errors that remain are the result of my linguistic failings, not hers. Others who need to be named are my father and father-in-law, Ian Brown and Neville Finlay, who took photos of specific locations in order to supplement those missing from my personal collection (after spraining an ankle when last in Italy, I lost a few valuable days earmarked for inspection of fortresses and battle sites). Last, but certainly not least, my partner in life, Kristen, is to be thanked profusely for not only producing the maps, but also for putting up with a bleary-eyed obsessive towards the closing stages of the writing of this book.

Paul Brown

February 2016

Chapter 1

Background

If an ethnic group is evaluated solely by the criterion of military success, the Normans were unquestionably the most successful people in eleventhcentury Europe. While best known for the conquest of the Kingdom of England in the late 1060s, both professional and popular historians have, particularly in the last few decades, increasingly turned their attention to the lesser known, but by no means less significant, Norman conquests in Italy, Sicily and Syria. While the reasons for this heightened interest are many, some of the appeal is no doubt attributable to the ‘rags-to-riches’ or ‘against-allodds’ nature of these territorial acquisitions. The conquest of England was a ‘state-sanctioned’ operation led by the Duke of Normandy himself and his most preeminent subordinates and allies, whereas the leaders responsible for conquests in southern Italy, Sicily and Syria often came from the middling ranks of Norman society; they were variously political exiles or disenfranchised sons hailing from relatively meagre patrimonies. Moreover, while the leading families participated in the duke’s invasion in 1066 according variously to tenurial obligations and the promise of increased wealth and prestige, their counterparts in the Mediterranean, despite also being clearly driven by the desire to improve their station in life, did not do so at the request of their duke. Rather, drawn initially by the offers of military service in Italy, the ‘help’, over the course of several decades, gradually became the rulers of their former employers.

Those responsible for the majority of the conquests in the Mediterranean hailed from the Hauteville family, whose military campaigns in the Mediterranean world will be the focus of this book. On the progenitor of these remarkable mercenaries, conquerors, and statesmen, Tancred of Hauteville (c. 980–1041), unfortunately little is known. Despite the concerted efforts of the (almost certainly) Norman chronicler Geoffrey Malaterra to indicate otherwise, at the time of his death Tancred of Hauteville was, at best, a relatively insignificant subordinate of the duke of Normandy, better known to posterity as ‘William the Conqueror’ (r. 1035–87). The siring of sons was clearly of paramount importance, for a total of twelve were borne to Tancred from two wives, Murielle and Frédesende. Four sons would later become counts in Apulia, and one of them, from Tancred’s second marriage, Robert ‘the Cunning’ (Guiscard), became duke of Apulia, Calabria, and Sicily (r. 1059–85). Robert’s son Marc, better known by his nickname ‘Bohemond’, proved to be the greatest of the commanders of the First Crusade (1096–99), during which he established a formidable principality centred on Antioch in northern Syria. But Bohemond would not be the only famous grandson of Tancred of Hauteville. Roger II, the son of Guiscard’s younger brother from the same mother, Count Roger I, became the first king of Sicily, founding a political institution that survived until Giuseppe Garibaldi’s conquest of Sicily in 1860. Another Tancred, the last of those whom his biographer Raoul of Caen called the ‘Guiscardians’, and whose name evidently commemorated his great-grandfather’s, lived much of his life under the shadow of his maternal uncle Bohemond. He was, however, more successful than his uncle when it came to extending the frontiers of the principality of Antioch, doing so at the expense of the neighbouring Byzantine, Armenian, and Selçuk powers. Upon his death, Tancred, as regent of Antioch (r. 1100–03, 1104–12), was in equal measure feared and admired by both his allies and enemies.¹

Southern Italy

The first region of the Mediterranean that the Normans would make their own at the beginning of the eleventh century was a cultural, political, religious and ethnic melting-pot. The inhabitants of modern-day Puglia (Apulia), Basilicata, and Calabria were subjects of the Byzantine emperor. The solider emperor Nikephoros II Phokas (r. 963–9) reorganized the administrative structure of these regions, which had previously been divided into two ‘themes’ governed by generals invested with both civil and military authority. Apulia, formerly the theme of Longobardia, became the ‘katepanate of Italy’. A katepanate was a larger, more important administrative province of the empire, consisting of themes. The other province, the theme of Calabria, was retained, and a new one was created in Loukania (Basilicata). While now under the direct authority of the katepano stationed at Bari in Apulia/Longobardia, the other governors nevertheless exercised a degree of administrative autonomy. Campania, to the west of Apulia, was ruled by three Lombard princes based at Salerno, Capua, and Benevento. However, in the same region, the maritime cities of Amalfi, Naples, and Gaeta were governed by dukes, who despite wielding independent rule, all maintained strong economic and political links with the Byzantine empire. Lastly, although the final imperial possession of Taormina resisted capture until 902, Sicily had effectively changed from Byzantine to Muslim hands by the middle of the ninth century.²

Such was the intricate, overlapping political landscape, but even more intricate was the ethnic one. Apulia/Longobardia, aside from the Greekspeaking south, was predominantly peopled by Lombards. Originally based in Pannonia (Hungary), the Lombards established a kingdom in northern Italy in 568, but their rule was restricted to southern Italy after Charlemagne’s capture of the capital Pavia in 774. A tenth-century Salernitan chronicler referred to the ‘German tongue, long ago spoken by the Lombards’, reflecting that they had gradually integrated with the local population, who greatly outnumbered them. They were therefore essentially Italians, although many, most notably the ruling class, continued to identify strongly with the Lombard heritage even though they spoke an Italian, rather than Germanic, dialect. This continuing ethnic pride is evidenced in one of the local annals written by the Lombards of Bari, which concludes the entry with: ‘and it has been 400 years since the Lombards entered Italy’. Similarly outnumbered by Lombards in much of northern and central Apulia, the Byzantines were content to let the locals be ruled in accordance with ancestral laws – first codified by King Rothair in 643 and substantially revised by King Liudprand (r. 712–44) – under their own officials. Positions within the imperial administration were also open to Lombards, although the office of katepano was restricted to nobles from Constantinople. While some historians have tended to style Lombard leaders who rebelled against the katepano as ‘patriots’ wanting to free themselves from servitude, to use the words of Chris Wickham, they were generally loyal and the exceptions to this general rule were ‘ad hoc hostile responses to individual administrators’.³

Calabria, like southern Apulia, was almost entirely populated by speakers of the official Byzantine language (Greek). In the Campanian region, again the Lombards formed the dominant ethnic contingent, although Greek was still known among some of the élite in the maritime cities. Naturally, since it had been under Muslim rule for over a century by the time the Normans arrived in southern Italy, Sicily possessed large numbers of Arabic-speakers, perhaps comprising up to two-thirds of the island’s population. The non-Muslim inhabitants generally spoke Greek, and were mainly concentrated in the north-eastern area of the island, the Val Dèmone.⁴

The complex political and ethnic makeup of the south was reflected in religion too. First of all, Christians were split between those who recognised the ecclesiastical authority of the Roman patriarch, and those who instead acknowledged his Constantinopolitan counterpart. Both the inhabitants of Campania and the Lombards of Apulia followed the Latin liturgy, and those of southern Apulia, Calabria, and Sicily observed the Greek one. Prior to the rule of the Byzantine emperor Leo III (r. 717–41), the papacy possessed ecclesiastical jurisdiction in southern Italy. However, on account of the Roman Church’s opposition to Leo’s policy of iconoclasm – i.e. prohibition against the worship of icons – the incensed emperor transferred this jurisdiction to the patriarch of Constantinople. Such a move was greatly resented by the papacy, and it was in the eleventh century that the papal curia began to reassert its rights more forcefully in the region. But another important power bloc in western Europe also lay claim to southern Italy: the Germanic emperors of the Salian dynasty, who much like their Ottonian predecessors shared Charlemagne’s belief that the south belonged to the western, not eastern, emperor (as had generally been the case until the deposition of the last western emperor Romulus in 476).⁵

Determining the precise nature of military administration in the regions ruled by Lombards and Byzantines is a difficult task. The former, originally referred to collectively as an ‘army’ rather than a people, fielded a force consisting of ‘freemen’ serving under the leaders of prominent families (farae). Freemen, more frequently identified as ‘soldiers’ (arimanni), were expected not only to provide military service, but also to equip themselves. Those freemen who were wealthier than others were required by the laws of King Aistulf (r. 749–56) to serve as heavy cavalry (i.e. to equip themselves with horses, armour, shields and lances). The men of middling means (minores homines) fought as light cavalrymen (horses, shields and lances, but no armour), and the less financially fortunate as infantry (shields, bows and arrows). These troop types and their equipment were determined by the amount of land held, and while Lombard society had theoretically consisted of a single ‘class’ – there are no explicit references to an aristocracy in the various laws – there were effectively three tiers based on prestige and wealth (farae, arimanni and minores homines). As is evident in a charter of Prince Radelchis I of Benevento (r. 839–51), a century later these property-based groups had become more precisely defined as ‘nobles’, ‘the middling’ and ‘peasants’ (reiterated in a 992 charter with similar terminology). Soldiers were levied from these classes by regional officials hailing from the upper tier of society, defined in the tenth century as consisting of the nobles and their princes, who were entrusted with taxation in addition to military matters. This system was employed throughout Campania, in addition to northern and central Apulia, which, until the Byzantine reconquest in the last quarter of the ninth century, was subject to Lombard rule.⁶

The troops deployed in pitched engagements, however, seem to have been predominantly levied from urban areas, whereas the ones tasked with defence of fortified settlements (castra) – not ‘castles’ or ‘fortresses’ as the Latin term meant in the contemporary French context – were recruited from among the rural population. The mention of ‘the cavalrymen of Salerno’, led in 1052 by Prince Guaimar IV and including among them his brother-in-law Landulf, is suggestive of an élite corps drawn exclusively from the upper echelons. Substantial lay estates are known to have furnished mounted troops, as did those in the ecclesiastical sphere. As for light cavalrymen, in former times recruited from the middling ranks, little is known. However, they almost certainly continued to exist and, as will be argued in the next chapter, can be identified with those known in Apulia as conterati. Lastly, it seems evident that in the eleventh century Lombard infantry continued to consist mainly of missile troops: in addition to the traditional archers, there are various references to the use of slingers (e.g. Prince Gisulf II of Salerno deployed men armed with ‘sling and bow’ in the 1050s). Supplementing these various troop types from the ninth century onwards, when needed, were mercenaries. Arabic-speakers augmented Lombard armies during the civil war of 839–49 and, much like their eleventh-century Norman counterparts, while militarily effective they were rapacious and difficult to control.⁷

While there were detachments of both provincial and élite troops known to have been stationed at Bari in the early eleventh century, their average numerical strength remains unknown. Yet since reinforcements were regularly sent to Italy in times of crisis, it follows that there were never very many. Indeed, the élite troops mentioned were more often officers than divisions or regiments, whose role was presumably to oversee defence, military administration, training and recruitment. The imperial administration in northern and central Longobardia therefore relied predominantly on Lombards – some of whom were ‘Byzantinized’ – recruited from the abovementioned classes, and hence could theoretically field an army of heavy and light cavalry in addition to infantry. The bulk of these men were recruited in accordance with their obligation to provide military service as a form of rent, a duty first referred to in a charter witnessed at Conversano in 980. The local population was also expected to provide ships and pay a naval tax, the proceeds of which presumably paid for the maintenance of the fleet and its marines. Much like the Lombards, in times of need the Byzantine administration hired mercenaries. In 1041, for example, the katepano was said to have been ordered by Emperor Michael IV to hire experienced cavalry from his ‘land’ (i.e. Longobardia/Apulia) in order to help counter the incursions of Lombardo-Norman rebels.⁸

Recruitment and attendant obligations in the theme of Calabria were similar, although the most obvious difference is that the troops, like those in southern Longobardia, were Greek-speakers rather than Lombards. The naval tax is attested in the tenth century, and that it was particularly onerous is apparent given that the citizens of Rossano revolted when it was levied in c. 965 by the governor Nikephoros Hexakionites. The number of ships maintained in Calabria seems to have been considerable, for the fleet led by the Normans in 1061 – i.e. after the region and its capital had been capitulated – was said to have outnumbered the twenty-four Muslim ships sent to oppose its crossing to Sicily. Admittedly this fleet did include ships from Apulia, but Calabria evidently contributed the lion’s share, as there were clearly enough battle-ready ships stationed at Reggio to engage a Muslim fleet in the Strait of Messina prior to the launch of the campaign. Although the last explicit mention of the building of galleys with two banks of oars (chelandia) – used for combat and the transport of cavalry – in Calabria dates to c. 965, clearly a region afflicted by seaborne raids launched from Muslim Sicily in the ninth to eleventh centuries required the maintenance of a fleet to guard against them. That chelandia continued to be produced is confirmed by the knowledge that the Byzantines procured ships from the region in order to transport troops to Sicily (1038), as did their Norman successors in 1061, who had received the surrender of the capital Reggio in the previous year. Lastly, a Calabrian ship identified specifically as a chelandion (sing.), was captured and burnt during a naval engagement in the Adriatic (c. 1064).⁹

The Calabrian theme, originally part of the theme of Sicily, was probably divided into three tactical divisions known as tourmai; at least this was the standard subdivision recommended in various military manuals. The one that anything is specifically known about is the Saline tourma (sing.) based at Oppido in the hinterland of the south-western coast (36km north-east of Reggio), and so called owing to the town’s location in the Saline Valley (the plain of Gioia Tauro). The fortified hilltop town was founded in c. 1044, and the tourma stationed there was subdivided into three vanda/droungoi of cavalry, each under the command of a count. Where the other two tourmai were located is uncertain, but presumably one was at the capital on the south coast (Reggio), and the other at Squillace on the east coast (where Reggio’s commanders took refuge when their city fell to the Normans in 1060). The size of Byzantine units was never static, but two military manuals, the first from the 960s and the second from c. 1000, state that a vandon (sing.) comprised ‘fifty men’. Hence the number of cavalry available to the Byzantine administration in Calabria was around 450 (assuming those in the other six vanda were equivalent in size to the three known to have been stationed at Oppido). Other than that, a signatory to a 1053/4 monastic charter in the same vicinity identified himself as Cyril, commander of the elsewhere unattested unit of ‘Hungarians’, whose numerical strength is therefore unknown. Any other troops – i.e. infantry and light cavalry – were presumably levied when required from the militia residing in the villages and fortified settlements. How such troops were recruited and trained is difficult to ascertain, but no doubt the method employed in Longobardia was also used in Calabria (i.e. by officers from the professional regiments). Indeed, men called Scrivones, presumably from the guards regiment associated with the regiment of ‘the Sentinels’ (Exkouvitoi), were noted by two sources to have been present at Crotone in the 1050s.¹⁰

Apulia and Calabria had developed in line with a tenth-century Italian phenomenon known to specialists as incastellamento: the fortification of either pre-existing or newly founded ‘nucleated settlements’ (castra), as Barbara Kreutz puts it. Such settlements were the opposite of the equivalent in northern France: to rephrase Chris Wickham’s words, those who worked the land resided within the fortified areas, rather than outside them. Apulia/Longobardia was much more urbanised than Calabria; accordingly many people lived in the heavily fortified towns on the east coast (e.g. Trani in the North, Bari in the centre, and Brindisi in the south), although considerable numbers resided in settlements located in the hinterlands of these urban centres. Calabria was altogether different. With Squillace and the capital Reggio being notable exceptions, coastal regions were abandoned in the tenth century in the wake of frequent raids by Arab pirates. Henceforth the majority of the population lived in inland settlements, many of which were fortified. There were, however, a few strongholds established that would equate with the contemporary French meaning of ‘castles’ or ‘fortresses’: for example, the Rocca Niceforo (Angitola) at Maierato in the modern province of Vibo Valentia, named after the abovementioned emperor Nikephoros II Phokas, and built during his reign (960s). There was nothing novel about the building of such a lofty fortress in Italy during Phokas’ reign: Cyprus was seized in 964, after which construction of the impressive fortifications in the Pentadaktylos mountains of the island’s north began. Kantara fortress (Fig. 7), which surely dates from this period, is situated 630m above sea level and, like its equivalents in Italy, was later reinforced by French-speakers. Cyprus, like southern Italy, was a frontier region vulnerable to seaborne invasions and raids; fortified lookout posts were therefore of paramount importance.¹¹

The Byzantine Empire

The second region of the Mediterranean in which the Normans would both make a name for themselves and even conquer portions of – mainly in the short term (Balkans), although three Ionian islands were permanently seized in the 1180s – was the (Eastern) Roman or Byzantine empire. If dated to the inauguration of Constantinople (Istanbul) or ‘New Rome’ by the Roman emperor Constantine in 330, the empire had been in existence for seven centuries by the time the Normans came into direct contact with it, and rather more than that if extended to the principate of the empire’s first ruler, Augustus (r. 27 BC–14 AD). From foundation until the date traditionally assigned to the collapse of the western Roman empire (476), Constantinopolitan emperors had either been the sole or eastern rulers of the Roman empire. Despite the empire’s Latin-speaking roots, by the reign of Herakleios I (r. 610–41), Greek, the lingua franca of the eastern Mediterranean since the age of Alexander the Great’s successors, was declared as the official imperial language. Accordingly, despite referring to themselves as Romans (Romaioi) – and being called that (al-Rūm) by other cultures in the Near East, Africa and Sicily – in western Europe the Byzantines were identified in Latin as ‘Greeks’. Given the empire’s multiethnic, polyglot composition, the term could not have been more inaccurate. Moreover, unbeknown to western European contemporaries, the term Ellenes (‘Greeks’) had since the fourth century transformed from an ethnonym into a pejorative term: as the Roman Pantheon had long been dominated by gods and cults of Grecian and Hellenistic origin, its adherents were called ‘pagans’ – that is, Ellenes. Accordingly, the subjects of the Eastern Roman emperors, including the former ones in Apulia, Calabria, Sicily and Syria, will be called ‘Byzantines’ or ‘Greek-speakers’ throughout this book.¹²

Despite losing Syria, Palestine, and Egypt to the Arabs during the seventh century, the empire remained both wealthy and powerful. At the heart of Byzantine martial effectiveness were the abovementioned administrative and military districts known as ‘themes’. Developed in response to the seventh-century Arab invasions, Byzantine provinces were equipped with semiprofessional, part-time troops – locally recruited, trained, and equipped – who were supplemented when needed by professional, permanent forces stationed at the capital: the tagmata. The latter were normally heavy cavalrymen (kataphraktoi), whereas thematic troops were generally foot soldiers and light cavalrymen, although some were kataphraktoi. Lightly armoured cavalrymen, sometimes collectively referred to as ‘cavalry themes’, formed the bulk of forces in the frontier provinces, where warfare more regularly consisted of skirmishing and ambuscades rather than pitched battles. Ascertaining the tactics and equipment of eleventh-century heavy cavalry is difficult to determine precisely, let alone summarize in brief, and, as is the case with the Normans, imposing a uniform model detracts from their evident fluidity and adaptability. The heaviest form of imperial cavalrymen since the sixth century were covered in armour (head to toe), as were their horses. Armed variously with maces, lances, swords, axes and bows (Fig. 4), they advanced at a trot, and their tactical purpose was to exploit openings created by infantry and light cavalry. John Haldon theorizes that these expensive to maintain ‘oven-bearers’ (klivanophoroi) might have been phased out during the reign of Constantine X (r. 1059–67). If so, their empty ranks were presumably filled by the other form of heavy cavalrymen who were similarly armoured (yet not as heavily), but whose mounts were unencumbered by a layer of protection. Such men seem to have been a feature of field armies since at least the second half of the tenth century, and were presumably the ones who launched an irresistible charge during a battle in 971 (the heavier panoply of the klivanophoroi and their mounts was probably too weighty for the type of charge described). Evidently John I Tzimiskes’ ‘Immortals’ surged forth at full gallop with their lances couched (locked under their right arms) in this battle, for a contemporary source related that ‘violently pricking the spurs to their horses’, they charged with their ‘lances held before them’. However, imperial cavalry adapted to whatever tactical situation faced them – and they may well have been grouped into specialist divisions – for Jonathan Shepard notes that while some contingents in the eleventh century wielded throwing spears, other battalions such as the Immortals (reformed in the 1070s) charged at full tilt with lances. Importantly, the use of the cavalry spear both as a thrusting and throwing weapon is supported by contemporary pictorial evidence – that is, the Theodore Psalter dated to 1066 (Fig. 3).¹³

The medieval incarnation of the Roman empire reached the zenith of both its military power and territorial extent during the reign of Basil II ‘the Bulgar Slayer’ (r. 976–1025). As his nickname suggests, the emperor was hewn from the rock of bellicosity. In addition to the great ability of the emperor and his generals, the military successes of Basil’s reign were in no small part due to the highly effective forces left to him by his equally warlike and expansionist predecessors: Nikephoros II Phokas (r. 963–9), and John I Tzimiskes (r. 969–76), two of the greatest military commanders of the entire medieval period. At Basil’s death, the empire spanned from southern Italy in the west to Crete in the south, to the Danube in the north and to the Euphrates in the east. However, only two militarily-minded emperors reigned briefly in the half century following Basil’s death – Isaac I Komnenos (r. 1057–9) and Romanos IV Diogenes (r. 1068–71) – resulting in an incremental decline in both the numerical strength and overall effectiveness of the imperial army. By the reign of Constantine IX Monomachos (r. 1042–55), the well-endowed treasury left by Basil II had been emptied. A fiscal crisis accordingly developed, resulting in the first-ever debasement of the gold currency established by Emperor Constantine I in 312. But this controversial precedent was not enough to rectify the empire’s financial problems and, as a result, the cost-cutting officials of the imperial bureaucracy turned to the army. The war-hardened troops from the theme of Armenia, estimated by Warren Treadgold to have formed around a fifth of the entire imperial army, were demobilized.¹⁴

Mercenaries were increasingly recruited and, with southern Italy acting as the conduit, French-speakers became prominent among them for more than a century. Substantial Norman and French recruitment had technically begun before the period of financial crisis – when around 300 of them served during the Sicilian campaign of 1038–40 – yet they were to be found increasingly among Byzantine forces from Constantine IX’s reign onwards. Included among these early mercenaries were two sons from Tancred of Hauteville’s first marriage, William and Drogo, who were later to become the first two counts of the Normans in Apulia. As will become increasingly apparent in the pertinent chapters of this book, over the next