Beruflich Dokumente
Kultur Dokumente
Kristian Molin
Contents
Illustrations
Acknowledgements
Introduction
vii
xi
xiii
1 Warfare
2 Military Architecture
13
38
54
64
79
7 Warfare
89
8 Military Architecture
95
108
115
124
130
III
13 Warfare
137
14 Military Architecture
145
152
168
175
VI
18 Warfare
191
19 Military Architecture
203
227
236
243
253
262
25 Residences
271
26 Prisons
277
280
283
29 Trade
290
295
31 Conclusion
299
Abbreviations
308
Notes
309
Bibliography
389
Index
4os
Illustrations
Maps
1 The eastern Mediterranean
xv
Cyprus
Cilician Armenia
139
The Aegean
195
92
Figures
1 Crac des Chevaliers (from W. Muller-Wiener, Castles of the
Crusaders, trans. J. Brownjohn (London, 1966), p. 61)
2
15
21
25
27
30
97
99
8
9
10
101
146
VIII
UNKNOWN CRUSADER
CASTLES
149
177
184
205
209
216
274
293
Plates
1 Chastel Blanc: the keep (Jonathan Phillips)
2 Antioch: the city walls and the citadel (Denys Pringle)
3
Chastel Rouge: the central keep and perimeter walls (Denys Pringle)
ILLUSTRATIONS
IX
Acknowledgements
When I originally began to work on the crusades, I was fortunate enough
to have Graham Loud as my supervisor at the University of Leeds. His help
and guidance have proved invaluable over the past eight years, and I am
very grateful to him. I would also like to express my sincere thanks to
Bernard Hamilton for the kindness he has shown me. Many other people
have, directly or indirectly, made the publication of this book possible.
From my time in Canterbury, I would like to thank Michael Simmons and
Richard Bales for all their generosity and support. I am deeply indebted to
Jonathan Phillips for his considerable help, not least his suggestion that I
submit this book to Hambledon and London, where Martin Sheppard and
Tony Morris have been very supportive and above all patient. During the
past year I have been lucky to work alongside Peter Edbury, Helen Nicholson
and Denys Pringle at Cardiff University. They have provided me with much
advice on the completion of this book. I would like to express my gratitude
to them. For his help with the computing I am, as always, indebted to Jase.
The bulk of the plans and illustrations were very generously provided by
Jonathan Phillips, Juliette Constantinou, Peter Lock and, above all, Denys
Pringle. I am also grateful to Dumbarton Oaks and Deutscher Kunstverlag
for allowing me to use illustrations from R. W. Edwards' The Fortifications
of Armenian Cilicia and W. Muller-Wiener's Castles of the Crusaders.
Finally, I owe the greatest debt to my parents, Carol and Ella. Without
their love, encouragement and patience over many years, this book would
never have been completed, and it is dedicated to all of them.
Introduction
In November 1095 Pope Urban II launched the First Crusade at Clermont
in south central France. The popularity of his message turned out to be
enormous, as tens of thousands of people joined the expedition to liberate
Jerusalem from the Muslims in order to cleanse their own souls and,
hopefully, achieve salvation after death. Having suffered enormous hardships
as it travelled from western Europe through the Balkans, across Asia Minor
and into Syria, the First Crusade finally captured Jerusalem in July 1099.
During this gruelling journey its participants passed through several different
regions whose history was to become inexorably linked with that of the
future crusader states. Stretching from the Adriatic to Constantinople and
dominating most eastern Mediterranean islands, including Cyprus, the
Byzantine empire viewed the arrival of this and future Crusades with mixed
feelings. On the one hand, the empire's inhabitants were fellow Christians
and therefore grateful for any aid they might receive against the Turkish
threat to the east. On the other hand, they were of course Orthodox Greeks
rather than Catholics, and were consequently separated from western Europe
by profound differences in religion, culture and outlook, which often caused
tension or even open warfare.
After they crossed the Bosphorus the armies of the First Crusade entered
an even more alien world as they gradually began to move into Muslim
territories. During the period covered by this book, these lands can loosely
be divided into the three key regions of Asia Minor, Syria and Egypt. Asia
Minor had originally been Byzantine, but by the 10905 most of it had been
lost for good because of a series of Turkish advances achieved in the course
of the eleventh century. During the next two hundred years the most
dominant people in the region were normally the Seljuk Turks. However,
after the mid fourteenth century the various Turkish peoples of Asia Minor
were eventually unified under the rule of the Ottomans, who subsequently
expanded their power in such a spectacular fashion that by 1520 they had
created an empire stretching from North Africa to the Balkans via Egypt,
Syria, Turkey and northern Greece. But at the time of the First Crusade
and for a long time thereafter no such unity existed. The Muslims who
inhabited the other key regions of Syria and Egypt were sporadically at war
XIV
U N K N O W N CRUSADER CASTLES
both with the Turks of Asia Minor and with each other. Unlike western
Europe, where urban growth was still extremely limited, these political
struggles were largely centred around the most important cities of the
Muslim world. Inland these included Aleppo, Damascus and Cairo, all of
which had grown extremely rich because of their trade links with Asia or
Africa. Through ports such as Alexandria, Acre and Tyre they also had
economic connections with Italy and the Byzantine empire, which ensured
that the eastern Mediterranean was far richer and far more sophisticated
than any of the north European lands inhabited by most members of the
First Crusade.
The region also differed from the west in its cultural diversity. The Muslim
peoples were not just divided geographically or politically, they varied
considerably in terms of their ethnic background and according to
whether they adhered to the Sunnite or to the Shi'ite faith. The eastern
Mediterranean also had a notable population of Jews, plus several nonCatholic Christian groups, many of whom lived under Muslim rule. These
included various Orthodox communities, as well as Maronites, Nestorians
and Syrian Jacobites. The most important such group for the purposes of
this book was the Armenians. They inhabited a large part of what is now
south-eastern Turkey and northern Syria, and in particular the region to
the north west of Antioch known as Cilicia. They had consequently lived
under Byzantine rule until the Turkish incursions of the eleventh century.
Although the Greek emperors continued to enjoy some sporadic authority
over them during the next hundred years, their distance from Constantinople ensured that by the late twelfth century the Armenians had effectively
become independent. This move toward independence was greatly aided
by the local terrain, for Cilicia itself was surrounded by mountains to the
north, east and west, and by the Mediterranean to the south. These barriers
helped to protect the Armenians both from the Greeks and from the Seljuk
Turks. Like the Byzantine Greeks, they enjoyed a love-hate relationship with
the crusading newcomers, for they were fellow Christians yet they were
non-Catholics with their own language, history and cultural identity.
This was the world which the members of the First Crusade entered in
1099. Rather than feeling overawed by the rich and complex society they
had encountered, those crusaders who chose to remain after the fall of
Jerusalem wasted no time in consolidating the military gains they had made.
Former Muslim territories in western Syria and Palestine were gradually conquered and turned into four new crusader states: the kingdom of Jerusalem,
the principality of Antioch, and the counties of Tripoli and Edessa. The
westerners who undertook these campaigns were mostly French, but others
came from Italy, Germany and Spain. Former Muslim ports such as Acre
MAP i. The eastern Mediterranean, showing approximate frontiers in the early thirteenth century
XVI
now fell under Christian control and their lucrative Asiatic trade links
came to be dominated by the Italian maritime cities of Venice, Genoa and
Pisa. In the course of the twelfth and thirteenth centuries many crusader
lands were also held by the Hospitallers, Templars and Teutonic Knights.
These were Military Orders, groups of western knights who confronted
the Muslims in battle yet followed an austere and celibate lifestyle under
the protection of the pope. Like more traditional monastic orders, they grew
wealthy on the donations of land or money made by their European patrons,
and these resources enabled them to make a substantial contribution to the
military activities of the crusader states.
After these states had been established, their twelfth-century history was
largely determined by the attitude taken towards them by the Muslim rulers
of Aleppo, Damascus and Cairo. These three key cities were frequently at
war, and at such times their rulers were prepared to negotiate peace treaties
with the Latins, or even seek Christian aid against each other. Although
its field army was relatively small, this enabled the kingdom of Jerusalem
to prosper for most of the early twelfth century. However, in the north
the rulers of Aleppo gradually proved to be more pugnacious toward
Antioch and Edessa, and eventually in 1144 the latter city was recaptured
by the Muslims. By the early 11505 the entire county of Edessa had disappeared, the failed Second Crusade (1148-49) doing nothing to prevent this.
During this period Nur al-Din, the ruler of Aleppo, also managed to extend
his rule to Damascus, threatening the security of Jerusalem itself. Worse
was to come after Nur al-Din's death in 1174, for eventually his Syrian
lands were annexed by Saladin, the ruler of Egypt. This meant that the
Christians could no longer rely on the divisions amongst the Muslims to
keep them safe, for Cairo, Damascus and Aleppo were now all controlled
by one man. Saladin's empire eventually gave him such an overwhelming
military superiority over the Latins that he was able to annihilate the
crusader field army at the battle of Hattin in July 1187. Thereafter the entire
kingdom of Jerusalem, including the holy city itself, was recaptured by the
Muslims except for the heavily fortified port of Tyre. Although the county
of Tripoli and the principality of Antioch did not suffer quite as badly,
they too lost considerable amounts of land, including the corridor of
coastal land which actually connected them, when Saladin launched another
devastating campaign against them in 1188.
Despite these disasters, the crusader states still managed to survive into
the thirteenth century. The participants of the Third Crusade (1189-92),
and in particular Richard Lionheart, managed to reconquer the coastal
regions of the kingdom of Jerusalem between Tyre and Jaffa, including the
vital port of Acre. It was also at this time that Richard conquered the island
INTRODUCTION
XVII
XVIII
of their great rivals, the Venetians. This highly destructive episode marked
the beginning of the end for Cyprus as a fully independent Prankish crusader
state, for it resulted in the Genoese occupation of Famagusta until 1464,
after which the entire island fell under Venetian control from 1489 onwards.
Finally, in 1571, Cyprus was conquered by the Ottoman Turks, thereby
destroying the last Christian domain in the eastern Mediterranean whose
Prankish settlers could trace their ancestry back to the earliest crusader states
created by the members of the First Crusade.
This book is a study of the various military and non-military functions
of fortifications occupied by Latin or Armenian Christians who settled in
the Holy Land, Cyprus, Cilician Armenia and Greece between 1187 and
c. 1380. The political and military upheavals of the period ensured that
fortifications played a prominent role in its history. Yet the study of such
structures still has a number of important gaps which this book aims to
fill. There has been a tendency in the past for scholars to concentrate their
efforts on castles in the Holy Land, and in particular famous and wellpreserved structures such as Crac des Chevaliers in Syria. Although this
imbalance is being addressed by the work of scholars such as Denys Pringle,
Ronnie Ellenblum and Adrian Boas, who in recent years have excavated,
studied or recorded very many smaller crusader fortifications and domestic
structures, their studies are still focused on the Holy Land itself, and
especially on the kingdom of Jerusalem. A primary aim of this book is to
shed more light on fortifications built or occupied by Latins and Armenians
in the more obscure areas of Cyprus, Greece and Cilician Armenia.
The second, and closely related aim is to make up for the lack of studies
which deal with crusader fortifications from the thirteenth century onwards,
even though this period, as we have seen, represented a massive expansion
of crusading endeavour as Latins settled on Cyprus and in the former
Byzantine territories around the Aegean. Within the Holy Land itself there
has often been a tendency for historians to focus on the period between
1095 and the end of the Third Crusade in 1192. This is perhaps understandable, bearing in mind that some of the most famous military encounters
of the middle ages, most notably the First Crusade and Richard Lionheart's
struggles with Saladin, took place during this period. It may also reflect the
nature of the sources, and in particular the fact that William of Tyre's
well-known history of the Holy Land, which contains so many clear and
accessible accounts of castles being built, besieged or destroyed, ended in
1184. Thus for example, R. C. Smail's famous work on crusader warfare
dealt with the period from 1097 to 1193.l In recent years scholars such as
C. Marshall, whose book on warfare between 1192 and 1291 represents a
continuation of Smail's work, have begun to address this shortfall.2 However,
INTRODUCTION
XIX
this book hopes to extend the work of Marshall and others by taking a
further look at the Holy Land in the thirteenth century, and by including
the post-1291 history of Cyprus, Greece and Armenia. Hence each chapter
begins with a brief section on warfare in these regions, to show how the
nature of local fighting influenced the appearance and functions of local
fortifications.
The third aim of this book is to provide an analytical rather than a purely
descriptive study of fortifications in the eastern Mediterranean. Once again,
this is something which is already being done in the Holy Land. Alongside
purely archaeological or architectural works such as D. Pringle's recent
survey of secular structures in the kingdom of Jerusalem,3 some scholars
have also researched the ways in which castles were actually used by the
Latins, and how they fitted into the wider military and political history of
the Latin East. The earliest and most influential book of this kind was
undoubtedly Smail's Crusading Warfare, but, as we have seen, this has now
been complemented by Marshall's important study of the period after 1192.
Other books to appear in recent years which have viewed castles in the
same light include Hugh Kennedy's Crusader Castles (Cambridge, 1994),
plus numerous publications by leading archaeologists such as Pringle's The
Red Tower (London, 1986) and Ellenblum's Prankish Rural Settlement in
the Latin Kingdom of Jerusalem (Cambridge, 1998).
Beyond the Holy Land itself, the picture is very different. Studies already
exist devoted to the visible remains of fortifications in Cyprus, Greece and
Cilician Armenia,4 but apart from giving brief historical outlines of the sites
they refer to, these works are largely devoted to listing and describing
surviving structures. What these fortifications were actually used for, rather
than what they looked like, is a topic which has barely been touched upon,
and it is one of the principal aims of this book to make up for this silence.
At times it will become apparent that this can be a difficult task, because
some areas, most notably Prankish Greece and Cilician Armenia, are lacking
both in reliable historical sources and in recent subsurface archaeological
excavation. No studies exist for these areas which can compare with the
very detailed research undertaken in the kingdom of Jerusalem, where a
wealth of information from papal letters, contemporary chroniclers or surviving charters, combined with the meticulous investigations of leading
archaeologists, has sometimes made it possible to recreate the history of
certain crusader castles virtually year by year.
Despite the limited scope for this type of research outside the Holy Land,
a combination of archaeological and historical sources can shed much new
light on the functions of local fortifications. For example, during the last
thirty years the archaeologists Hansgerd Hellenkemper and Robert Edwards
XX
U N K N O W N CRUSADER CASTLES
have produced comprehensive studies which list and describe virtually every
single surviving stronghold in Cilician Armenia. Yet because there are very
few traces of any urban fortifications left in this region, they have rejected
the notion that the Armenians constructed defences of this kind, preferring
instead to live in or near remote mountain castles. I hope to show that a
closer scrutiny of the historical evidence actually indicates that the Armenians
did dwell inside cities, whose older Byzantine ramparts they were happy to
repair or rebuild. This would suggest that Edwards and Hellenkemper have
relied too much on the archaeological evidence (or its absence) to draw
their conclusions, without studying the written sources as well.5 While there
are chapters in each of the parts of this book which deal with military
architecture, it is my intention to do more than simply describe fortifications,
and to look at a variety of sources in order to highlight the ways in which
such structures were actually used to defend or conquer new land, to suppress
hostile locals or to impose central authority over recalcitrant vassals. This
is an approach which has not been used before with regard to Greece, Cyprus
and Cilician Armenia, but in the Holy Land I hope to build on the research
already carried out in this field by Smail, Marshall and others.
An exploration of strongpoints in the eastern Mediterranean can also
give us greater insight into how local Christian settlers actually lived. In
the past there has sometimes been a tendency amongst scholars to focus
on the purely military functions of fortifications, and the ways in which
they could withstand sieges, dominate strategic valleys or form elaborate
networks of intervisible castles. Whilst these topics will be discussed, it is
important to bear in mind that many years or decades might pass before
a fortress came under siege or found itself involved in any kind of warfare.
Consequently this book is intended to redress the balance by considering
the many non-military functions of strongpoints as residences, prisons,
courthouses or centres of trade and agriculture. This topic is dealt with at
length in part five, but throughout the book it will become apparent that
it is misleading to view castles and warfare as separate from other aspects
of medieval society, such as trade and farming. In part four, for example, the
numerous rural towers which were constructed by Latin settlers in Greece
will be considered. It is tempting to see these towers as evidence of chronic
local warfare. Whilst it certainly seems to be true that such structures could
provide shelter against pirates or Turkish raiders, it should be borne in mind
that their construction would have been expensive and time-consuming,
and therefore only possible at a time when local lords were relatively rich and
their estates were untouched by external attackers. In the Holy Land this
approach has already led to many important new discoveries by Ellenblum,
who has shown that Prankish settlers built countless smaller fortified or
INTRODUCTION
XXI
semi-fortified structures which were simply intended for local farming and
defence. Strongpoints of this kind had no importance whatsoever in terms
of the large-scale warfare between Richard Lionheart and Saladin which has
become so famous in the popular view of the Crusades, yet they accounted
for the vast majority of fortifications built by Christian settlers after the
First Crusade.
The vast majority of research already carried out on crusader fortifications,
as is already clear, has concentrated on the Holy Land, and in particular
on the kingdom of Jerusalem (or Acre, as it is often known after 1187). This
book nevertheless contains a chapter on that region, partly because it is the
most famous and important area of crusading endeavour, and partly because
it gives this research as a wide a scope as possible, so that comparisons can
be made between the functions of fortifications in the Holy Land and those
situated in other parts of the eastern Mediterranean. It is hoped that this
will show how the famous castles built by the crusaders in the Holy Land
were sometimes very different from, but sometimes very similar to, the
many less well-known Strongpoints constructed in Greece, Cyprus and
Cilician Armenia.
This book deals with the period from 1187 to 1380. From what has already
been said it will be clear that much work has already been done on the
Holy Land, particularly during the twelfth century, yet after 1187 the Latins
acquired far more crusader territories around the eastern Mediterranean
than they had held before the battle of Hattin. It was also a time when
Cilician Armenia gained independence from Byzantine authority, being
transformed into another Christian state which had to face the growing
pressure applied by powerful Muslim neighbours. Many more Christian
Strongpoints were built or occupied during this period than the hundred
years immediately after the First Crusade. The year 1380 seems a natural
ending point for this book for a number of political, military and cultural
reasons: Cilician Armenia fell in 1375, bringing its history as an independent
state to a close. Cyprus was invaded by Genoa in 1374, and although the
Genoese were only able to conquer a small part of the kingdom, this period
signalled the end of the island's status as a prosperous crusader state, for
its economy was in decline and its political life gradually came to be
dominated by Italians rather than by the Prankish families who had originally settled there after 1192. In Greece, meanwhile, the tide had definitely
turned in favour of the Ottoman Turks, for by the end of the fourteenth
century they had already conquered most of the Byzantine empire and
were beginning to penetrate the remaining Latin states in south-western
Greece. Technologically speaking, 1380 also seems an appropriate date at
which to stop, because by that point gunpowder was already known in
XXII
PART ONE
Warfare
The history of the crusader states in the Holy Land between 1187 and 1291
can be divided into two roughly equal parts. First, the period from 1189 to
the early 12405 witnessed a series of efforts by the Franks to restore the
losses suffered after the battle of Hattin (1187), which had resulted in the
erosion of Christian territories around Antioch and Tripoli and the loss of
all land in the south apart from Tyre. The Third Crusade (1189-92), whose
most prominent participant was undoubtedly Richard Lionheart, achieved
the reconquest of Acre and a coastal strip as far south as Jaffa.1 Five years
later German crusaders took Beirut,2 and the Embriaco lords of Gibelet
reoccupied their old castle through diplomacy.3 In 1204 another peace treaty4
confirmed Prankish possession of Jaffa (lost temporarily in 1197) ,5 Lydda,
Ramlah and Nazareth, and at about the same time the Templars and
Hospitallers consolidated their position in the county of Tripoli by strengthening their local castles and launching punitive raids against neighbouring
Muslims.6 Further gains were made during the crusade of Frederick II
(1228-29), whose treaty with the Egyptians acknowledged Christian control
over Jerusalem, Bethlehem, Nazareth, the territory of Toron and those
parts of the lordship of Sidon previously held by the Muslims.7 Finally,
during the separate but overlapping crusades of Theobald of Champagne
(1239-40) and Richard of Cornwall (1240-41), complex negotiations with
Egypt and Damascus led to the restoration of Belvoir, Toron (Tibnin),
Tiberias, Chateauneuf (Hunin), Beaufort, Saphet and Cave de Tyron.8 The
Franks also reoccupied Ascalon, whose citadel was rebuilt, and Jerusalem,
which had been lost briefly in 1239-40.'
After the successes of the early thirteenth century, the next fifty years
represented a period of sporadic retreat. In 1244 the truce with Cairo broke
down and Jerusalem was lost to the Khwarizmians, a violent tribe of nomadic
horsemen who were allied to the Egyptians and supposedly numbered
between 12,000 and 20,000 troops.10 Shortly afterwards this alliance defeated
the Franks at the battle of La Forbie, enabling the Egyptians to retake much
of southern Palestine, including Ascalon and Tiberias in 1247.11 Further
Muslim conquests were temporarily halted by the first Crusade of Louis IX
(1248-54), who refortified several sites in the Latin East, although his
U N K N O W N CRUSADER CASTLES
WARFARE
MAP 2.
a combined field army of around 22,000 men, about half of whom came
from Egypt.17 In practice, of course, the Egyptians and the Syrians spent
more time using these troops against each other than the Franks, who
prospered 'because of the constant discord of the [Muslim] princes of the
land, which was highly favourable to the Christians'.18 Alternatively, one
Muslim power might even ally itself with the Christians against another, as
was the case at the battle of La Forbie in 1244, when the ruler of Damascus
was prepared to join forces with the kingdom of Jerusalem against the
Egyptians and Khwarizmians.19 Similar problems had occurred even during
the reign of Saladin, who was only able to keep his cumbersome and
regionalised empire together because of his own forceful personality. After
the battle of Hattin Saladin often found it extremely difficult to maintain
his field army, which was largely composed of seasonal troops who served
the sultan during the summer in exchange for land or money, but then
went home for several months as soon as winter set in. Even when they
were present, some of these forces turned out to be ill-disciplined, badly
trained and more interested in loot than complex notions of holy war.
Drawn from all over Muslim Syria, they often expressed greater loyalty
toward their local rulers, such as the Zengid rivals of Saladin in the
north, than they did toward the empire-building projects of Saladin himself.
Regional loyalties and personal aspirations also affected the various members
of Saladin's own family and entourage, who openly argued over strategies
in the middle of campaigns. Between 1187 and 1192 these serious weaknesses
help to explain why Saladin failed to capture all Christian lands in the wake
of Hattin, proved incapable of defending Acre against the armies of the
Third Crusade, and ultimately had to return large parts of the coast to
Richard Lionheart. Without a well-controlled army which stayed with him
throughout the year, Saladin could not complete the task of capturing
powerful Latin strongholds like Tyre or wiping out crusader forces in battle,
being eventually forced to reach a negotiated settlement with the leaders
of the Third Crusaded
Another reason why Saladin failed to defeat the Third Crusade was his
inability to match the overwhelming naval power of the Latins, despite
his considerable efforts at building up a sizeable Egyptian fleet.21 In the
thirteenth century there were sporadic attempts to remedy this weakness,
but Prankish possession of Cyprus and the Levantine coast made it difficult
for Egyptian war galleys to reach, let alone blockade, the sea lanes which
connected the crusader states with western Europe. A notable example of
the troubles faced by the Muslims was Baybars' disastrous naval raid on
Limassol in 1271, involving eleven to fourteen Egyptian galleys. Before they
even reached Limassol these vessels were shipwrecked off the coast because
WARFARE
U N K N O W N C R U S A D E R CASTLES
an early age and were instilled with a deeper sense of loyally toward their
own sultan than most seasonal or mercenary forces. Although Mamluks
had already been used for many years, their numbers increased significantly
in the course of the thirteenth century. It is difficult to say how many
Mamluks were accumulated by each sultan. Kalavun is reputed to have
owned between 6000 and 12,000, whilst the average number for most of
his contemporaries seems to have been between 2000 and 4ooo.27 These
troops formed the core of late thirteenth-century field armies. Although
Mamluk sultans were still reluctant to undertake sieges in midwinter, they
did make all-year-round campaigning, such as Baybars' raids on the fortress
of Margat in January 1270, much easier.28
Secondly, the Muslims became increasingly adept at using mines and
catapults to capture Christian strongpoints, including Acre itself, whose
outer walls were either sapped or bombarded to the point of collapse during
the final siege of i29i.29 Earlier, the Mamluks had captured a number of
crusader castles by digging mines beneath their walls, including the
Hospitaller fortress of Margat (1285), forced to surrender when the enormous
inner tower guarding the southern tip of the fortress was in danger of
collapsing.30 Whereas the basic art of digging siege tunnels had changed
little over the centuries, more profound developments affected catapults
during the crusader period. In the course of the twelfth century more and
more large catapults, or trebuchets, came to be fired by heavy counterweights filled with sand or stones rather than by men pulling on ropes,
and as a result stronger and more accurate weapons could be built. A recent
reconstruction of a trebuchet proved capable of hurling objects (including
a piano) weighing up to 500 kilogrammes over considerable distances.3!
Ammunition weighing almost this much was used in some of the most
decisive sieges of the thirteenth century. At Acre in 1291, for example,
catapults were used which could easily launch stones weighing nearly 200
kilogrammes.32 Ammunition of this kind was deadliest if it could be aimed
against exactly the same part of a castle's ramparts again and again, thereby
literally shaking it to destruction. In order to achieve such accuracy, the
weight of the stones fired by a trebuchet needed to be calculated in advance
so that it did not vary.33 Furthermore, if it was to be capable of throwing
such large stones over a long period without falling apart from the strain,
the trebuchet itself had to extremely tall and sturdy. Indeed, the modern
reconstruction just referred to stood to a height of sixty feet.34 Hence
trebuchets needed to be handled by skilled siege engineers, and required
large amounts of money and logistical planning for their upkeep. After
Baybars had united both Syria and Egypt under his personal rule in 1260
these resources became available, enabling him to construct large numbers
WARFARE
10
UNKNOWN CRUSADER
CASTLES
WARFARE
11
Hvres tournois per year, the total annual cost of maintaining a mere hundred
French knights whom Louis had stationed in Acre came to 10,000 Hvres
tournois. This was a cripplingly expensive bill which, so the evidence suggests,
even Louis could not always afford to pay on his own.44 Thus the cost of
equipping and transporting expeditions to the East became so great that
crusading armies shrank in size, and could only be financed by the wealthiest
European lords.45
These factors, combined with the fact that most crusaders lacked any
experience of fighting in the Latin East, weakened the effectiveness of
expeditions. In 1269, for example, an Aragonese crusade to the Holy Land
enabled the Franks to launch a raid against Muslim villages near Montfort
with an army containing 130 knights.46 This force was far too small, however,
to recapture any Christian territories lost to the Mamluks, or indeed risk
a direct confrontation with Baybars, who was said to have had such a large
field army close by that one contingent alone numbered 15,000 men. Moreover, during skirmishes just outside Acre, some Aragonese leaders of the
campaign, declaring that they had come to fight for Christ, were needlessly
killed because they simply charged into the enemy ranks and were immediately cut down by Muslim troops ravaging the area. The crusade achieved
nothing permanent, wasted Christian lives, and provoked Baybars into
carrying out a damaging counter-raid against Acre.47 Temporary crusades
could therefore create more problems than they solved, for they did not
remedy the permanent shortfall of troops, and even if they did succeed in
regaining territory this had to be defended by local Franks who subsequently
found themselves even more thinly spread out than before.
On top of this there were the usual security problems common to virtually
all parts of medieval Europe. Incidents of localised raiding, violence and
crime were common, and there was also the more remote but nevertheless
real possibility of an uprising by Muslim peasants under Latin rule. Indeed,
during the years or decades which separated full-scale invasions by Saladin,
Baybars or Kalavun, this must have been the type of fighting which affected
most Christian territories. We are therefore left with the impression of a
region where Latin settlers virtually always found themselves heavily outnumbered by external aggressors, but also had to cope with the possibility
of lawlessness or rebellion within their own lands. It is against this background that the various functions of crusader fortifications should be
discussed, for castles and urban defences were relied on by the Christians
to protect their property, maintain their hold on the land and above all
make up for their lack of soldiers in the field. Unlike European crusades,
such structures lasted for decades or even centuries, and remained firmly
under the control of local settlers. They were therefore relied upon to solve
12
U N K N O W N C R U S A D E R CASTLES
the numerous military weaknesses which have been described in this chapter,
and as such became vital to the continued existence of the entire crusading
movement in the Holy Land.
Military Architecture
During the thirteenth century the Latins occupied a vast array of fortifications in the Holy Land, most of which were either built on top of, or
incorporated into, earlier Arabic, Byzantine, Roman or even older defences.
Indeed, some strongpoints, such as the vast fifth-century ramparts at
Antioch, barely contained any Prankish building work at all, whilst fortresses
like Montfort and 'Atlit, which were built on previously unoccupied or long
abandoned sites, were extremely rare. These complexities make it preferable
to divide a few Latin strongpoints into general categories depending on
size, location and function, rather than describing all fortifications built in
the area individually.
First, the Franks held a number of major castles which were constructed
on a very large scale in order to withstand substantial Muslim invasion
forces. In the north, the best preserved and most famous such fortifications
were the Hospitaller strongholds of Margat and Crac des Chevaliers. The
outer walls of Margat enclosed an extensive triangular mountain spur which
dominated the coastal route between Tripoli and the principality of Antioch.
Indeed, Margat covered such a large area that it contained a small town,
situated to the north of the inner citadel and separated from it by a rock-cut
moat. The citadel itself consisted of numerous towers and buildings ranged
around an inner courtyard, many of which date from the period after 1186,
when the Hospitallers bought Margat from its original secular owner. The
most important of these structures was the keep, a huge round tower with
walls at least 5.5 metres thick which stood at the exposed southern tip of
the site and therefore guarded the likeliest point of attack.1
The layout of Margat's defences compares very closely with that of Crac
des Chevaliers, another former baronial castle which was also situated on
a mountain spur and was acquired by the Hospitallers in ii42.2 Thereafter
the Order added several flanking towers and an enormous talus to the south
and west sides of the original enclosed courtyard, and then constructed a
whole new curtain wall around the entire site. As was the case at Margat,
the keep was placed at the most vulnerable point along the south side of the
castle, where it formed the central flanking tower of the inner bailey.3 These
major alterations appear to have been undertaken after the earthquakes of
14
U N K N O W N C R U S A D E R CASTLES
1170,1200/01 and 1202/03 but to have been completed by around 1220.4 The
earthquake of 1170 was particularly devastating, for according to one Muslim
source, it left 'several fortresses of the Franks ... such as Hisn al-Akrad
(Crac des Chevaliers), Safitha (Chastel Blanc), al-Arimah [and] Arkas ...
plunged into an ocean of ruins'.5 Bearing in mind that the Hospitallers did
not acquire Margat until 1186, this suggests that the Order carried out its
building programme in the thirty years following Saladin's invasion of Syria.
This is made all the likelier by the fact that virtually identical improvements
were made by the Hospitallers at the fortress of Silifke, an Armenian
stronghold which they did not acquire until 1210.6
Not far from Crac des Chevaliers and Margat, the Templars held the
equally substantial strongholds of Tortosa and Chastel Blanc (Safita), which
were also considerably rebuilt during the crusader period. This was done
at a much earlier stage, however, for a document dating from 1152 reveals
that the Templars had probably already acquired Chastel Blanc by this stage,
and were in the process of constructing new defences at Tortosa.7 Unlike
Crac des Chevaliers and Margat, where all new Hospitaller towers were
either round or horseshoe shaped and donjons were attached to adjacent
ramparts, Tortosa was provided with a massive rectangular keep which
stood in isolation at the north-west corner of the site. During a further
phase which appears to date from the early thirteenth century, and therefore
coincided with the Hospitallers' work at Margat and Crac des Chevaliers,
this keep was provided with two flanking corner towers placed at the water's
edge. At this point the keep was also protected from potential land attacks
by the addition of two successive ditches and curtain walls to the north,
east and south and was flanked by two corner towers situated at the water's
edge. Some idea of the sheer scale of these defences can be gained from
those sections of the inner curtain wall which still survive to their original
height of 25.5 metres.8
A similar process seems to have taken place at the hill top fortress of
Chastel Blanc, for the impressive rectangular keep which stands at the
summit was probably constructed in the middle years of the twelfth century,
or perhaps after the earthquake of 1170, whilst some at least of the outer
curtain walls and elaborate gateways defending the lower slopes were probably constructed later.9 Indeed, there may have been several phases of
improvement and reparation, for we have seen that in 1200/01 and 1202/03
more earthquakes struck the region, whilst in 1218 an Aleppine invasion
force attacking Chastel Blanc supposedly 'destroyed its towers'.10 Meanwhile,
the Templars were responsible for the construction of another rectangular
keep and related defences at the fortress of al-'Arimah, which occupied
a long ridge above the coastal plain between Tripoli and Tortosa. Little
MILITARY ARCHITECTURE
15
Crac des Chevaliers. Note the double walls and complex entrance passage. The
revetments and vaulted undercrofts of the inner bailey strengthened the castle
against siege weapons and earthquakes.
16
is known about the history of this fortress, although it had probably already
been acquired by the Templars with Chastel Blanc at some point before
1152.n This group of Templar bases, along with both Margat and Crac des
Chevaliers, survived numerous Muslim incursions in the course of the
thirteenth century before finally succumbing to the Mamluks from 1271
onwards. In that year Crac des Chevaliers and Chastel Blanc fell,12 followed
by Margat in 128513 and Tortosa in i29i.14 Al-'Arimah may have been lost
in 1271, for hi 1282 a peace treaty between the Muslims and the Templars
stated that the Order still possessed some estates nearby, but that the fortress
itself now belonged to the sultan of Egypt.15
In the kingdom of Jerusalem, there were a number of major strongholds
which resembled Tortosa, Chastel Blanc, Crac des Chevaliers and Margat.
Tortosa was similar to the Templar fortress of 'Atlit (Pilgrims' Castle), in
that both could rely on the sea to protect them from the west but needed
very extensive defences against attackers approaching from inland across
the relatively flat coastal terrain. 'Atlit, which was constructed from scratch
in 1217-18 with the aid of the Teutonic Knights and members of the Fifth
Crusade, was better off than Tortosa because it actually lay on a promontory
jutting out into the sea, but it still had vast man-made defences to protect
its landward side. These consisted of an outer ditch and rampart with three
gate towers, and a much higher inner wall whose two flanking towers were
so tall that their defenders could spot potential attackers eight miles away.
Together these towers acted as the inner donjon of the castle, in much the
same way that the inner southern towers of Crac des Chevaliers did.16 Like
Tortosa, 'Atlit also proved to be so well defended that it was never taken
by force, being simply evacuated by the Templars in August i2pi.17
The appearance of both Tortosa and 'Atlit was determined by their
proximity to the sea, but there were also fortresses in the south situated
on hill tops and which required successive defensive rings to protect them
on all sides. About half a century after they completed Chastel Blanc, the
Templars rebuilt Saphet, an important Galilean castle which they had originally acquired around 1168 and which was restored to them as a ruin in
the peace treaty of 1240.18 The new fortress built after 1240 consisted of an
outer wall, approximately 22 metres high and 825 metres long, which Tan
around a second and much higher rampart dominated by a large, probably
circular keep. The exact appearance of this keep is difficult to establish, for
it was later replaced by a similar Mamluk structure before a nineteenthcentury earthquake flattened the entire fortress. It may have been one of
the seven towers along the inner rampart mentioned in De constructione
castri Saphet, an invaluable contemporary account of the castle's construction and appearance. This document indicates that the outer wall was also
MILITARY ARCHITECTURE
flanked by seven towers, and was provided with underground tunnels which
led to several casemates guarding the castle's outer moat.19 When it was
besieged by the Muslims in 1266, this massive stronghold managed to
withstand six weeks of almost constant mining and bombardment, only
capitulating once sultan Baybars had managed to sow discord amongst its
defenders.20 Such resilience helps to explain why the Military Orders had
already begun to erect similar concentric castles in the second half of
the twelfth century, particularly at Belvoir, an almost symmetrical fortress
composed of two successive walls forming a square within a square. Belvoir
was built by the Hospitallers between 1168 and 1189 on a lofty hilltop
guarding the Jordan crossings south of Lake Tiberias. In 1189 it was lost to
Saladin. Although it was theoretically restored to the Franks in 1241, it seems
unlikely that the Hospitallers actually reoccupied it during the brief period
before the Egyptian and Khwarizmian conquests of 1244-47.^
The inland castles mentioned so far had concentric defences intended to
protect them on all sides. There were also a number of important strongholds
located in more isolated surroundings with less regular fortifications. In
18
U N K N O W N CRUSADER CASTLES
1255 Alexander IV, acting on behalf of the monks who lived there, granted
the Hospitallers Mount Tabor, a strategic hill top which lay within sight of
Belvoir and to some extent must therefore have been intended to compensate
the Order for the loss of this latter castle.22 In 1211 the Egyptian sultan
al-Adil had ringed the summit of Mount Tabor with a massive curtain wall,
1750 metres in length and flanked by ten towers.23 Seven years later, following
a failed Latin attack on the castle,24 the Muslims decided to demolish this
rampart and abandon the site rather than risk a more successful Prankish
assault in the future.25 One might therefore expect that when the Hospitallers
arrived more than three decades later they would have attempted to repair
and regarrison al-Adil's wall, but there is no archaeological evidence to
suggest that plans to build new defences were ever put into action. Instead
it seems that the Hospitallers merely occupied (and perhaps fortified) the
small monastery in the south-east corner of al-Adil's castle, and simply
relied on the height of Mount Tabor's summit and the steepness of its
slopes to keep attackers out. To some extent this may have been done
because of inadequate troops and resources, for in 1263 Baybars seems to
have captured the site with such ease that more extensive man made defences
would clearly have been desirable.26
The Muslims found it far harder to capture heavily fortified mountain
strongholds such as the castle of Beaufort, situated in the mountainous
interior south east of Sidon. Beaufort could rely on sheer cliffs hundreds
of metres high to protect it from the east, whilst steep gorges deepened by
the Franks defended it from the north and north west. Consequently, the
outcrop occupied by this castle formed an isolated stronghold which only
needed flanking towers and multiple ramparts along its more exposed
southern face. This fortress, lost to Saladin in 1190, was regained by the
Franks in 1240 and garrisoned by them until Baybars conquered it for good
in 1268. During this period the Franks strengthened the castle's defences
by building a separate citadel which was situated on the southern plateau
opposite the castle's main entrance.27
Another impressive example of a mountain castle built by the Franks was
Montfort. This fortress lay at the heart of an extensive lordship held by the
Teutonic Knights a few miles north east of Acre and acted as the Order's
headquarters until it was captured by Baybars in 1271. Montfort was built
on a precipitous spur formally acquired by the German knights in 1228 and
therefore dates entirely from the thirteenth century.28 Its defences were
arranged around a large 'D' shaped keep which stood at the eastern end
of the fortified spur and was isolated by rock-cut ditches. Access to this
keep was also blocked by a wall along the north and west slopes of the
castle, although the apparent absence of any medieval remains on the south
MILITARY ARCHITECTURE
19
side suggest that this slope was considered steep enough not to require
additional man-made defences.29 This 'semi-concentric' arrangement would
not have been feasible at the more gently-sloping sites of Saphet, Crac des
Chevaliers, Margat and Belvoir.30
To the north of Montfort, the castle of Akkar (Gibelcar), situated in the
county of Tripoli, also occupied a mountain spur which only needed extensive defences where it could be approached from the neighbouring
mountains. Its keep, a relatively simple quadrangular structure, also resembled that of Montfort in that it was placed at the neck of the ridge connecting
the castle with surrounding hills. This ridge had also been severed by a
rock-cut moat, whereas the rest of the site was so inaccessible that it was
apparently only protected by a single curtain wall following the edge of the
spur.31 Perhaps the most extreme use of nature to protect a stronghold
occurred not at larger fortresses such as Akkar or Montfort but at Cave de
Tyron, a cave fortress carved into the side of a mountain opposite the town
of Sidon. Even though it only housed a tiny Prankish garrison, it must have
been virtually impregnable, for it lay hundreds of metres above sea level
and could only be approached along a path barely one metre wide.32 Both
Akkar and Cave de Tyron had been occupied by the Franks at various times
since the twelfth century, and the former was not lost until 1271, whilst the
latter seems to have been abandoned by 1260.33
Apart from castles, the Franks also built or occupied a number of citadels
intended to defend existing urban settlements, and these could often be as
powerful as individual fortresses. Even though it has already been discussed
in the context of major Syrian castles, Tortosa was strictly speaking a citadel
rather than an independent fortress, for a much older walled town had
already existed next to it before the Templars built their new keep from 1152
onwards. By contrast, 'Atlit was built from scratch by the Templars in 1217-18,
effectively becoming a citadel once a walled settlement sprang up just beyond
its outer defences. Thus the distinction between castles and citadels is not
always such an easy or useful one to make, but if we define most citadels
as strongholds which were attached to important walled settlements already
in existence before the time of the First Crusade, it is clear that the Latins
possessed a number of such structures both inland and near the sea. Along the
coast, the citadels of Arsuf, Beirut, Jaffa and Ascalon all resembled Tortosa
in that they were placed as close to the shore as possible, in order to protect
them from the west, but needed fairly elaborate defences on their landward
side. This was clearly the case at Beirut, whose citadel had been left intact
by Saladin in 1187, was recaptured and repaired by German crusaders in
1197, and then strengthened even further by the new Ibelin lord of the city
in the early thirteenth century.34 When the German pilgrim Willbrand of
2O
MILITARY ARCHITECTURE
21
John of Ibelin in 1240, and later described as 'very strong' by Joinville,48 the
urban defences were considerably improved after the Hospitallers acquired
the site in i26i.49 This again implies that whilst Arsuf must have had walls
around it for much of the thirteenth century, these were not always in a
particularly good state, and that the Ibelins, unable to foot the maintenance
bill any longer, therefore sold the lordship to a Military Order. By this point
the military situation had deteriorated to such an extent that not even the
Hospitallers could hang on to Arsuf, which fell to Baybars in 1265 along
with Caesarea. Three years later, Jaffa was also lost for good.50
22
UNKNOWN
C R U S A D E R CASTLES
The citadels of Jaffa, Arsuf and Ascalon no longer survive, but at the
latter site a combination of historical and archaeological evidence suggests
that the stronghold constructed during the Crusades of Theobald of
Champagne and Richard of Cornwall amounted to an impressive structure
with towers, ditches and double walls built out of ashlar blocks and recycled
Roman columns. It was probably situated on a mound in the north-west
corner of the city, where it is still possible to make out two approximately
rectangular lines of defence, plus the remains of a masonry glacis.51 At
Arsuf the citadel stood on an almost identical hill to that of Ascalon, being
separated from the surrounding town by a deep moat and connected to
the port via a long staircase.52 Although it has long since vanished, the
citadel of Jaffa should be placed in the same category,53 as indeed should
that of Tiberias, a pre-iiSj Prankish settlement which, like Ascalon, only
came under effective crusader rule in the thirteenth century between the
peace treaty of 1241 and the Egyptian conquests of 1247. During that brief
spell its Prankish lord, Odo of Montbliard, rebuilt the town's citadel,
which was defended by a massive wall and moat on one side but could
rely on Lake Tiberias (rather than the Mediterranean) to shield it on the
other.5*
Whilst the coastal citadels were situated at the water's edge, others made
even greater use of the sea as a means of protection. That of Caesarea,
whose thirteenth-century history has already been touched upon, stood on
a narrow promontory whose neck was defended by powerful ramparts and
a water-filled moat.55 This castle represented a scaled down version of 'Atlit,
as did the baronial fortress of Nephin, which was situated just south of
Tripoli on a small headland defended by 'twelve good towers' and separated
from the coastal plain by two rock-hewn ditches.56 For most of the twelfth
century Nephin was held by the same baronial family as that which controlled Maraclea, another coastal settlement located between Tortosa and
Margat. This site does not seem to have been particularly well fortified, for
in 1188 it was evacuated just before the arrival of Saladin, and in 1271 Baybars
probably occupied it soon after the fall of Crac des Chevaliers.57 However,
the inhabitants may have used an isolated rock, situated about fifty metres
off the coast, as a place of refuge in times of danger. At some point after
1277 Bartholomew de Ravendel, lord of Maraclea, returned to the area and
built a powerful new tower on this rock. Although Bartholomew's tower
was demolished in 1285 as part of a peace treaty with Sultan Kalavun of
Egypt,58 its design and location can be compared with that of the sea castle
at Sidon, built by Frederick II's followers during the winter of 1227-28 on
a small island in the harbour. This structure, which was connected to the
shore via a stone and wooden bridge, was gradually altered and improved
MILITARY ARCHITECTURE
23
by its crusader occupants until it was finally evacuated shortly after the fall
of Acre in i29i.59
Another tactic adopted by the Franks was to place urban citadels not at
the strongest point of the defences but at the weakest, as they had done
with the donjons of Montfort and Margat. The fortifications of such sites
tended to be more regular in design, for they could not rely on the sea to
protect them. Hence at Gibelet, situated between Nephin and Beirut, the
citadel consisted of a rectangular enclosure with four corner towers built
around an isolated central keep. It stood at some distance from the sea, so
that it could guard the land approaches to the walled town which grew up
between it and the shore.60 This fortress, which had originally been built
at the beginning of the twelfth century, was lost to Saladin in 1188 but was
then regained through negotiations in 1197 and held by the Genoese
Embriaco family at least until 1289.6! The layout of Gibelet was also copied
at Sidon, for in 1253 Louis IX constructed an irregular enclosure defended
by several flanking towers on a small motte at the most landward extremity
of the town. Excavation of this motte indicates that Louis' land castle merely
replaced an earlier twelfth-century fortification, which must therefore have
been the town's only refuge point until the sea castle was constructed in
1227-28.62
Contemporary descriptions and plans suggest that the citadels of Acre
and Tyre were also rectangular in form, and perhaps resembled a partially
preserved urban castle built at Famagusta around 1300.63 That of Tyre was
placed 'near the Sidon gate', suggesting that it defended the northern side
of the narrow causeway which connected the city with the mainland. It
was apparently reconstructed from around 1212 onwards, for in that year
John of Brienne, king of Jerusalem, bought land inside the city so that he
could finish building a new castle.64 At Acre the citadel had originally
fulfilled a similar purpose to that of Gibelet, but rapid urban growth during
the crusader period eventually left it stranded behind newer ramparts
constructed further inland. By the late thirteenth century it functioned
more as a barracks than as a strongpoint, and appears to have played no
significant role in the defence of Acre during the final siege of i29i.65
A more crucial role was played by the citadel of Jerusalem, where the
massive Herodian structure known as the Tower of David was incorporated
into a medieval stronghold built around a central courtyard and attached
to the western wall of the city. The Tower of David, which was so strong
that one contemporary described it as 'a single stone from its base up',66
clearly formed the core of this citadel, but it seems that other elements were
added or repaired by the Latins until the late 12308. Most notably, archaeologists believe that parts of the sloped masonry revetments which form the
24
U N K N O W N C R U S A D E R CASTLES
MILITARY ARCHITECTURE
25
emperors during the fifth and sixth centuries.76 Consequently, any Prankish
building work undertaken at Antioch could only have amounted to small
repairs and improvements. It has been suggested, for example, that a gate
on the south slopes of Mount Silpius, plus a rectangular enclosure on the
south-west side of the citadel itself, were crusader additions.77 This was
in marked contrast to the much more extensive crusader fortifications
26
U N K N O W N CRUSADER CASTLES
constructed at Tripoli and Tyre, whose towers, ditches and double ramparts
must have been similar to those of Acre. Tyre was also said to have had at
least three successive curtain walls and as many as five separate gateways,
plus a vast moat occasionally rilled with sea water, protecting the neck of
land which connected it with the mainland.78
Jerusalem, on the other hand, was only ringed by a single curtain wall
which generally followed the same course as the present Ottoman defences.79
After the Muslims captured Jerusalem in 1187, they initially repaired and
maintained its defences, but these were subsequently slighted by al-Muazzam,
ruler of Damascus, in 1219, as part of the same scorched earth tactics which
led to the destruction of Mount Tabor.80 The historical and archaeological
evidence suggests that, after Frederick II regained Jerusalem in 1229, the
Christians tried to rectify this situation by repairing Sion Gate and St
Stephen's (or the Damascus) Gate, located in the south and north of the
city respectively. At St Stephen's Gate remains have been found of an 'L'
shaped entrance guarded by a fairly elaborate barbican incorporating a
church and possibly a customs house, whilst at Sion Gate it seems a similar
bent entrance originally constructed by the Ayyubids gave access to the city
through the side wall of a massive guard tower.81 Post-1229 improvements
to these structures may have been carried out by the Teutonic Knights, to
whom Frederick gave several properties in the city and relied on to garrison
his new acquisition.82 Presumably they also show that the rest of the city's
gates and walls were rebuilt, or perhaps simply repaired if al-Muazzam had
actually only caused damage in places rather than flattening the entire circuit.
On the other hand, the apparent ease with which 15,000 Muslim peasants
broke into and looted Jerusalem shortly after Frederick's departure in 1229
suggests that the walls were either inadequately restored or severely undergarrisoned (or both) after that date.83 If this was the case, then life must
have been difficult for the few Latins who returned to the city between 1229
and 1244, when Jerusalem was finally lost to the Khwarizmians.84
It has been estimated that in the twelfth century the populations of Acre,
Tyre, and Jerusalem stood at around 30,ooo.85 In the thirteenth century the
former two cities probably continued to have at least this many inhabitants
living within their walls, but the population of Jerusalem between 1229 and
1244 may have only been a fraction of the pre-nS/ total, and perhaps as
low as 5000 to 10,000 people.86 In terms of population at least, it had
therefore slipped into the second rank of walled cities alongside places like
Caesarea (perhaps 4800), Sidon (perhaps 5600) and Arsuf (perhaps 36oo).87
It is clear that several other cities of a similar size were fortified in some
way during part or all of the crusader period, including Beirut, Jaffa, Ascalon,
Gibelet, Tortosa and Haifa along the coast, plus Tiberias in the interior.
MILITARY ARCHITECTURE
27
We have already seen that this latter site was returned to the Franks in 1241
and that a new citadel was subsequently built there, but it seems unlikely
that the entire city was properly recolonised, let alone refortified, during
the six years before it was lost again to the Egyptians in 1247. A similar
situation probably prevailed at Ascalon, whose citadel was rebuilt and lost
again during the very same period. By contrast, it has already been shown
that many of the other coastal settlements just mentioned were inhabited
for much longer periods in the thirteenth century, and that sporadic attempts
were at least made to build or maintain walls around them.88 However, it
is only at Caesarea that any major archaeological remains have survived,
for here it is still possible to view the extensive new defences erected by
Louis IX between 1251 and 1252. These consisted of a sturdy rampart which
was supported by a sloping masonry talus, was flanked by sixteen rectangular
towers and was protected by a moat at least seven metres wide. From the
base of the ditch, the revetment was eight metres high and the rampart,
which has not survived, was probably a further ten metres high. The city
gates penetrating this rampart were similar to those of Jerusalem, in that
they were incorporated into the sides of towers and then turned through
ninety degrees before leading into the town itself.89
28
UNKNOWN CRUSADER
CASTLES
Hence Caesarea had fairly elaborate defences from 1252 onwards, and
considering that Louis IX also carried out repairs and improvements at
Jaffa, Sidon and Haifa during this period, they too may have had similar
fortifications.90 However, it seems unlikely that the city walls of all crusader
towns were as large or as powerful as those of Caesarea. In 1231, for example,
imperialist forces attacking the Ibelin lordship of Beirut were said to have
breached the town wall in the space of a single night, whereas the actual
citadel successfully withstood a determined siege lasting several months.91
Beirut's urban defences were probably similar in scale to those of Prankish
bourgs or castle towns.92 These were new or growing settlements which had
sprung up next to most larger fortresses, including Beaufort, Saphet, Crac
des Chevaliers, Margat and 'Atlit. These sites were normally protected by
a much simpler curtain wall and ditch than that which surrounded Caesarea.
The wall around 'Atlit, for example, had a total length of about 800 metres,
but was only defended by four small towers, only one of which could
provide flanking fire for a gateway. The other entrances to the town were
far more exposed than those at Caesarea, because they led directly through
the wall rather than being incorporated into the sides of towers. All these
structures were built shortly after the completion of the castle itself in
c. 1220.93
Along with these relatively minor urban defences, the Franks also held
countless smaller castles which were dotted across the countryside. Although
their shape and size could vary greatly according to location, many were
either fortified enclosures or individual towers. A good example of the
former type can be found at Goliath, a Hospitaller property situated about
20 kilometres north of Tripoli. Goliath formed a rectangle which measured
63 metres by 56 metres, had four small corner towers, and was also provided
with a fifth salient guarding the actual gateway.94 Castles of this kind are
sometimes referred to as castrum fortifications because of their close architectural links with Roman and Byzantine forts, although they could also be
based on Arabic farmsteads of the type which had been constructed in the
East for many centuries. However, a square or rectangular enclosure is such
an obvious way of defending a low-lying site that it could just as well have
been developed by the Franks themselves.95 The same could be said for
what one scholar has termed 'courtyard buildings':96 even smaller fortified
enclosures which were built as hospitals, monastic buildings, administrative
centres or secular farmsteads. The latter category is well represented by the
Castle of Roger the Lombard, which was situated a few miles south of
Caesarea and was probably constructed by Roger shortly after the First
Crusade. It consisted of several vaulted structures built around a central
courtyard, whose outer walls were approximately 1.6 metres thick and had
MILITARY ARCHITECTURE
29
very few openings in them, creating a fortified farm house roughly 33 metres
square.97
Most other minor strongpoints took the form of individual towers; a
design which is far likelier to have come directly from the West than have
been inspired by Roman or Arabic precedents. One such tower which has
been investigated by archaeologists in recent years is Qaqun (Caco), located
on the Sharon plain about twenty-five kilometres south east of Caesarea.
The remains of this building indicate that it measured 14.53 metres by 17.65
metres, that its walls were 2.8 metres thick, and that it had two vaulted
storeys with a crenellated terrace above. There were no openings below the
upper floor, which contained numerous arrow slits, a doorway, and traces
of clay piping feeding an underground cistern. This cistern, plus a small
perimeter wall, would have enabled the defenders to withstand a limited
siege.98 A virtually identical castle in the county of Tripoli has survived at
Chastel Rouge, whose inner tower measured 14 metres by 16 metres.99 Other
structures of this kind, such as the tower of Tukhlah in the vicinity of
Chastel Blanc, may have stood in isolation, although it seems likelier that
their surrounding perimeter wall has simply disappeared.100
This outline of crusader fortifications in the Holy Land is by no means
exhaustive, but it does enable us to identify two distinct trends in military
architecture. First, it is clear that smaller castles and urban fortifications
changed relatively little in the course of the twelfth and thirteenth centuries.
Goliath: a typical
courtyard or castrum
castle.
30
All of the minor towers and fortified enclosures already mentioned date
from before 1187, and although some, such as Goliath, were repaired or
even rebuilt after that date, it seems that no efforts were made to improve
their defences or radically alter their overall design before they were lost
to the Mamluks between the 12608 and 12905. Similarly, it has been suggested
that simpler urban defences, such as those which surrounded 'Atlit, consisted
of little more than a straightforward perimeter wall and outer ditch, and
appear to have been almost totally unaffected by advances in military
architecture. As we shall see, this confirms that these sites were never
intended to withstand major Muslim incursions involving tens of thousands
of troops.
On the other hand, it was clearly hoped that much more powerful Prankish
strongpoints, such as the city walls of Acre, the citadel of Tortosa or the
fortress of Crac des Chevaliers, would be strong enough to deal with
aggressors of this kind. As a result, such places had to be upgraded constantly
in order to deal with the growing strength of large Muslim armies, and in
particular the far more powerful trebuchets which became available to the
Ayyubids and Mamluks from the late twelfth century onwards. As we have
seen, these weapons could hurl rocks weighing hundreds of kilogrammes
with great accuracy and over relatively long distances, and so castles had
to be adapted to cope with the new threat.101
MILITARY ARCHITECTURE
31
The most obvious way to do this was to build fortifications on a vast scale
and to make exposed walls as thick as possible. Indeed, it has already been
noted that the outer walls of Tortosa and Saphet alone were more than
twenty metres high, and that the keep of Margat had walls 5.5 metres thick.
The square layout of this tower's internal rooms meant that in some areas
this figure rose to ten metres.102 Overall strength could also be improved
by ensuring that the individual blocks of stone within walls were as large
as possible, and sometimes this could be achieved by reusing the Cyclopean
masonry of classical or pre-classical times. At 'Atlit for example, stones
quarried or recycled locally were so large that they 'could barely be pulled
in a cart by two oxen'."13 The extent to which such blocks were cut and
dressed depended on the hardness of the stone, although some masonry,
such as that used at Tortosa, may deliberately have been left with a bossed
outer face in order to strengthen it against catapults.i04
The damage caused by sappers and trebuchets could also be reduced by
bonding together stones with molten lead or iron clamps; a method which
is known to have been used at Beirut, Sidon, 'Atlit and Maraclea.105 In
addition, numerous fortifications had Roman columns incorporated into
their ramparts, including the citadels of Ascalon, Tiberias 106 and Caesarea,
where they had been 'placed horizontally in the body of the wall, in such
a way that they had nothing to fear from sapping, and could not fall, even
if they were undermined'.107 Muslim sappers were also kept out by placing
curtain walls on top of sloping revetments; a technique which has already
been referred to at Caesarea, Crac des Chevaliers and elsewhere.108 Similarly,
in 1285 the foundations of one of Margat's outer towers, which stood at
the southern tip of the castle just below the actual keep, formed such a
solid lump of masonry that even after it had been brought down by sapping
it could not be circumvented. Indeed, one Muslim source wrote that as a
result 'the night passed in great confusion, for the use of catapults was
made impossible by what had happened, and everything that could be done
with mines had been done'. The Hospitallers still surrendered shortly
afterwards, for they realised that the keep itself was now in danger of
collapsing.109 The only definite way of escaping such a fate was to build
castles on solid rock, which must have been one reason why the barren
promontory occupied by 'Atlit was so attractive to the builders of this
fortress.110
The growing power of trebuchets affected the appearance of fortifications
in other ways. In the thirteenth century round or horseshoe towers, whose
shape was not as vulnerable to artillery fire as the sharp corners of square
salients, grew in popularity. These were constructed at Margat and Crac
des Chevaliers around 1200, and possibly at Saphet in the early 124OS.111 It
32
UNKNOWN CRUSADER
CASTLES
is extremely likely that the former two castles were heavily influenced by
the work of Armenian military architects, who had already recognised the
advantages of rounded towers long ago.112 It is also interesting to note that
even when thirteenth-century Prankish castle builders did continue to use
quadrangular towers, these often took on a rectangular rather than a square
shape, so that walls exposed to artillery fire became more elongated with
greater distances between vulnerable corners. This development can be seen
very clearly in the early thirteenth-century fortifications erected by the
Templars at Tortosa and 'Atlit.113
As the range of trebuchets increased, there was also greater emphasis on
trying to prevent such weapons, or indeed other forms of attack such as
mines, ladders and crossbows, from getting close enough to a strongpoint
in the first place. This was the primary role of castle ditches. At Saphet,
for example, the ditches were more than fifteen metres deep and more
than thirteen metres wide.114 The moat at 'Atlit even seems to have had a
stone counterscarp which rose above ground level, forming a parapet where
archers could be posted to prevent attackers from even reaching its outer
lip.115 Several coastal strongpoints such as Nephin also had water-filled
ditches;116 a method which was occasionally used at inland fortresses too,
most famously along the southern front of Crac des Chevaliers.117 Furthermore, an account of the siege of Acre in 1291 indicates that one tower
along the city walls had wooden hoardings with huge spikes attached to
its base, suggesting that even if most ditches were dry, those of larger
strongpoints could still contain obstacles which were lethal to infantry and
cavalry alike.118
It is also from the siege of Acre that we hear of complex outworks, largely
constructed during the last twenty years of Prankish rule and composed of
a succession of ditches, earthworks and stone or wooden barriers.119 Such
structures tended to be most elaborate in the vicinity of gateways, which,
as we have already seen at Jerusalem and elsewhere, were often incorporated
into the side walls of towers so that they were not exposed to a direct
assault or bombardment by trebuchets.120 These entrance towers could
themselves be equipped with portcullises, arrow slits and murder holes,
making it extremely hazardous for attackers to enter castles even after their
outer doorways had been breached. In 1276, for example, Templar troops
besieging Nephin managed to reach the main gate safely, only to find
themselves trapped after their opponents inside the castle lowered the
portcullis behind them.121 Bearing in mind that Nephin was still a relatively
small castle, it is easy to understand why Mamluk armies often preferred
to keep their distance and rely on mining or bombardment against considerably larger strongholds such as Crac des Chevaliers, whose labyrinthine
MILITARY ARCHITECTURE
33
main entrance passage was so heavily defended that a direct assault against
it would have resulted in very heavy casualties.122
Strong walls and deep ditches were a passive way of dealing with besiegers,
but archers and counter-artillery enabled garrisons to strike back at their
opponents. In order to maximise the effectiveness of garrison troops who
were firing bows or crossbows, inner ramparts were built much higher than
outer ones, whilst inner and outer towers alternated so that one wall or
tower did not block the field of fire of another. Most towers and some
ramparts also had rooms or shooting galleries incorporated into them, so
that defenders were not simply restricted to firing from the tops of walls.
The majority of these, defenders were straightforward archers or crossbowmen, but some operated much larger crossbows or trebuchets intended to
destroy the trebuchets of their attackers. Another explanation for the sheer
size of some thirteenth-century Prankish towers, therefore, is that they
needed to be large enough to accommodate trebuchets on their roofs, which
acted as flat, spacious fighting platforms supported by massive vaults. Behind
these towers, inside the inner courtyards of castles, further vaults and
undercrofts were needed to provide shelter against incoming artillery, support the outer ramparts against enemy bombardment, and take the strain
of the castle's own trebuchets being fired continuously. The effect of this
type of warfare on the appearance of a stronghold is perhaps most apparent
at 'Atlit, which, as we have seen, was protected by a ditch with its own
parapet for archers, an outer wall flanked by three towers, and a higher
inner rampart whose massive double towers were probably designed to
accommodate artillery weapons. Any attackers approaching this Templar
stronghold across the open land to the east would therefore have been met
by a shower of lethal bolts, arrows or large rocks being hurled at
them.123 Indeed, in 1220 al-Muazzam, ruler of Damascus, had to call off his
attempted siege of 'Atlit because this was precisely the kind of reception
which greeted his forces.124 Nor was this the only place where the Templars
used such weapons, because they installed them at Jaffa after they acquired
this lordship in 1266 and employed 'crossbowmen with large crossbows' to
guard the outer moat at Saphet.125 It is also possible that in 1188 Saladin
was prevented from capturing the citadel of Tortosa because of the Templar
crossbowmen stationed within it, implying that this Order was particularly
skilled in the use of catapults and archers126
It is also in this context that we should view many of the improvements
carried out at major strongholds in the late twelfth or early thirteenth
centuries. At Tortosa, for example, we have seen that a straightforward mid
twelfth-century rectangular keep had two flanking towers and elaborate
outer defences added to it around 1200; a project which was clearly intended
34
to provide more flanking fire for the keep and to stop attackers armed with
trebuchets from getting too close to it.127 During their occupation of Beaufort
from 1260 onwards, the Templars were also responsible for the new citadel
constructed on the plateau opposite the castle. This was done to ensure
that attackers would not deploy siege weapons in this area, although it still
failed to prevent Baybars from doing just that when he captured Beaufort
in 1268.128 We have seen that large parts of the inner defences and the entire
outer circuit walls at Crac des Chevaliers were also completed around 1200,
so that the fortress could withstand all but the most determined sappers
and powerful trebuchets.129 Parallel developments took place in the Muslim
world, for in the early thirteenth century massive new towers which were
better able both to withstand and to accommodate trebuchets were added
to the citadel of Cairo and numerous other major strongpoints.130
Finally, the most active defence of all entailed leaving the safety of a castle
and launching minor raids, or sorties, against besieging forces in order to
throw them off guard and hopefully destroy their siege weapons. This tactic
required doorways, or posterns, to be incorporated into walls and towers
which were large enough to make surprise attacks but small enough to
defend easily. At Crac des Chevaliers, for example, there were a number of
posterns situated along the outer ramparts, plus another small gateway at
the north-west corner of the inner ward. Consequently the garrison could
conduct an active defence even after the outer walls had fallen, and could
continue the fight until virtually every building in the fortress had been
captured.131
If a siege reached this point, however, one disadvantage for the defenders
of a major strongpoint with multiple or concentric fortifications was that
outlying structures could actually provide shelter for attackers. In 1220
concerns that this would happen explain why the Templar defenders of
'Atlit, on hearing that a large enemy force was approaching, responded by
partially demolishing the outlying watch tower of Destroit before they
retreated inside their fortress.132 In 1265, when Baybars carried out an
extensive raid in the vicinity of Acre, its garrison did the same thing and
chose to destroy many neighbouring buildings rather than watch the Muslims use them in an attack on the city itself.133 In the same year something
similar actually took place at Caesarea. Once Baybars had broken through
the walls of this settlement, he used the city's cathedral as a useful vantage
point from which to shoot at Franks still holding out in the citadel.134 In
another very interesting incident relating to Acre, some sources noted that
during the final siege of 1291 the Latins preferred to leave the city gates
open. This was done so that early on in the campaign the Franks were free
to launch surprise raids against the Muslims assembling outside and to
MILITARY ARCHITECTURE
35
prevent them from getting too close to the walls.135 By the late thirteenth
century the very largest crusader castles and cities should no longer be
regarded as passive refuge points, but rather as centres of a very active and
flexible defence which could sometimes radiate for many hundreds of yards
into the surrounding countryside.
Whilst this type of warfare ultimately decided the fate of the Holy Land,
it is important to note that only a few major strongpoints such as Acre,
Tyre, Crac des Chevaliers and 'Atlit were intended to cope with it. The
massive fortifications which have been described at these sites were very
expensive to build and maintain, only constructed reluctantly in response
to the growing size of Muslim armies and the greater effectiveness of enemy
siege weapons. The vast majority of Prankish strongholds dotted across the
countryside were not intended to withstand such forces, and generally
consisted of much smaller towers or enclosures whose size and shape varied
considerably according to their location and the wealth of their owners.
Similarly, virtually all the castles built in Cyprus before 1291 or in Greece
after the Fourth Crusade were relatively small and poorly constructed,
because warfare in these regions did not involve tens of thousands of troops
or highly sophisticated trebuchets.136
A determining factor was therefore economic. The Ayyubids and even
more so the Mamluks could draw on the enormous wealth of their territories
(and especially Egypt) to raise very large invasion forces and construct
extremely powerful siege weapons. Fortunately for the Holy Land, the
spending power of the Military Orders and the great wealth of crusaders
such as Louis IX meant that it was possible for a small number of sites like
'Atlit to be constructed in response. To put the cost of building major
fortifications into perspective, it should be remembered that the Templars
supposedly spent 1,100,000 Saracen bezants rebuilding Saphet in the 12405,
and then spent a further 40,000 bezants per year simply maintaining
it.i37 This was at a time when a local knight's fee 'ranged in worth from
400 to looo Saracen bezants per annum'. In Prankish Greece the relative
poverty and humble background of most Latin settlers simply made this
kind of expenditure impossible. Indeed, this point even extended to the
quality of the stonework used inside castles, for beautifully ornate structures,
such as the gallery of the great hall at Crac des Chevaliers, which was built
by highly skilled craftsmen in an ornate mid thirteenth-century gothic style,
had no parallel in Greece apart from the royal residence of Chlemoutsi.
Most castles in Greece were built from recycled or very poorly cut stones
without any concession to luxury or decoration.139
Another reason for the impressive appearance of sites like 'Atlit was
location. The Latins' reliance on the sea as a means of communication with
36
western Europe meant that many key settlements were located in relatively
exposed parts of the coastal plain, where massive man-made defences were
needed to make up for the flat terrain. Once again, therefore, the Franks
were driven by necessity rather than by choice, but some scholars have
tended to equate size with strength, regardless of the circumstances. It has
been argued that Montfort, for example, 'was hardly as impressive' as Saphet
because it lacked the perfect concentric defences of the latter castle.140 This
sentiment has been echoed by other scholars, who point to Montfort's
relatively poor masonry as proof that it was architecturally inferior; an
argument which can be backed up by looking at contemporary records
showing that the Teutonic Knights needed considerable financial assistance
to complete their castle.141 The fact that Baybars needed two attempts to
capture it, however, but only one to take Saphet, suggests that in some
ways Monfort was actually stronger. In other words, the Teutonic Knights
did not need massive concentric defences to strengthen a site which was
already relatively easy to defend because of its mountainous location.142
Further north, it should also be noted that Baybars found it easier to capture
Crac des Chevaliers than the fairly simple but very inaccessible castle of
Akkar. After he captured Akkar in 1271, Baybars boasted of his achievements
in a mocking letter to Bohemond VI, where he described 'how we transported the mangonels there through mountains where the birds think it
too difficult to nest; how patiently we hauled them, troubled by mud and
struggling against rain'. By comparison the apparently much more sophisticated defences of Crac fell in less than a month because the garrison was
under strength, lacking in food and unable to contact a relieving force
which could stop the Muslims from undermining the walls.143
Military architecture cannot be studied in isolation, for numerous wider
political, economic, geographical and military factors influenced the appearance and relative strength of castles. But there has often been a tendency
to ignore these factors, with the result that Prankish military architecture
has been regarded as a separate science which gradually evolved as the
Latins discovered new methods of defence and rejected old ones. For
example, one scholar has recently challenged William of Tyre's claim that
the fortress of Montreal was already defended by 'walls, towers, forewalls
and a moat' in 1115, because this would suggest that the Franks were
constructing elaborate double defences early in the twelfth century, long
before excavated concentric castles such as Belvoir appeared.144 This argument relies too much on the notion that the crusaders only 'discovered'
concentric defences gradually. In reality, their experience of Roman, Byzantine and Arabic fortifications at sites like Antioch and Constantinople,
combined with the self-evident advantages of providing a fortified summit
MILITARY ARCHITECTURE
37
with more than one line of defence, suggest that they were well aware of
such concepts long before the First Crusade. Similarly, historians now agree
that the castrum, or fortified quadrangular enclosure, was not 'discovered'
in the Holy Land either, for although there were numerous Byzantine or
Arabic structures of this kind for the crusaders to copy when they got there,
they are just as likely to have been influenced by Roman fortifications in
the West or simply have developed this form of castle themselves.145 Furthermore, the trebuchet itself was not a new weapon, for this technology
had already reached the eastern Mediterranean from China during the sixth
century.146 Hence the arms race of the late twelfth and thirteenth centuries
was not caused by a rash of new scientific discoveries, but rather by an
increase in wealth of states like Egypt and France, which began to enjoy
the greater liquidity and administrative skill needed to construct stronger
castles, build more powerful trebuchets and muster larger field armies. The
Byzantine empire could only build massive fortifications such as the double
land walls of Constantinople because of its enormous wealth and sophisticated bureaucracy, indicating that it was lack of money rather than
knowledge which prevented other states from emulating such feats until
the later middle ages. The same probably applied to siege weapons, for
recently one scholar writing of the eleventh century has noted that 'the
principles of siege-engines seem to have been well-known, but actually
building them was difficult and subject to frequent failure. Improvements
in siege-techniques seem to have sprung from better general organisation
rather than advances in technology'.147
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
39
40
have referred to the outer defences of the fortress, because the architectural
evidence proves that Chastel Blanc's keep predates 1218 and therefore survived the Aleppine assault. Consequently the keep may have held out in
the same way that the citadel of Tortosa defied the Muslim invasion force
of 1188, even after Saladin had sacked the surrounding town.13 To the north,
meanwhile, the fortress of Margat proved so powerful that in 1281 the
outnumbered garrison successfully withstood a determined assault by 6000
Muslim besiegers.14 Eleven years earlier Baybars had been forced to call off
another attempt to besiege this castle because of the atrocious weather.^
Perhaps the best example of the way in which major Latin strongpoints
guaranteed the continued existence of the crusader states took place in 1244.
In that year a Khwarizmian force supposedly containing up to 20,000
horsemen swept into Palestine and captured Jerusalem, implying that the
city's walls were indeed in a poor state of repair. Prankish efforts to halt
the Khwarizmians in the field ended in disaster at the battle of La Forbie,
and the Latins were far too outnumbered to prevent them from occupying
large areas of the Holy Land. In theory, this could only have been done by
constructing a continuous barrier like Hadrian's Wall or the Great Wall of
China; but, even if this had been physically possible, tens of thousands of
men would have been needed to garrison such a structure. Consequently,
for much of 1244 'the Christians only held the fortresses'.16 Inland, these
included Saphet, Montfort, Beaufort, Crac des Chevaliers and Chastel Blanc,
whilst along the coast the Franks successfully defended all their major
strongholds except Ascalon. As a result the Khwarizmians, who were a
nomadic people and lacked their own siege equipment, were able to inflict
terrible damage on the rural economy but could not make any permanent
conquests of their own. This enabled the Franks to wait securely inside
their castles until lack of food and shelter, combined with a realisation that
further progress would be almost impossible, forced their opponents to
retreat.17
These tactics could work equally well against aggressors who were more
disciplined and more capable of undertaking a siege that the Khwarizmians.
In 1220, for example, al-Muazzam appears to have reached 'Atlit, which lay
at the very heart of the kingdom of Jerusalem, without meeting any resistance. Once he got there, al-Muazzam failed to overcome the massive
fortifications, whose 4000 Christian defenders inflicted serious casualties on
his own troops.18 Consequently the Franks had prevented al-Muazzam from
conquering areas around 'Atlit without having to guard lengthy frontiers
or defend strategic roadways. During the Third Crusade, Saladin used a
similar strategy against Richard I. By ravaging areas east of Jaffa, but at the
same time strengthening Jerusalem's defences, he made it impossible for
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
41
the crusaders to win back their former capital.19 For his part, Richard knew
that it would be suicidal to proceed into a desolate wasteland and besiege
a city without adequate food, water or shelter, and so he was obliged to
retreat. Thus it was far more important to defend individual strongholds
than the countryside which surrounded them, for outlying areas could easily
be reoccupied provided that larger castles and cities successfully withstood
a temporary invasion.20 It will be seen that in Greece, the Catalans, Franks
and Greeks all used exactly the same tactics to withstand hostile incursions.21
By maintaining just a handful of extremely powerful castles, the Franks
could also force their opponents to abandon some campaigns before they
had even begun. In 1221 a large Muslim army gathered at Horns in the
hope of launching an attack against the county of Tripoli which would
distract Christian efforts during the Fifth Crusade. However, eventually it
was decided to cancel the offensive and march south to Egypt, because the
Muslims 'reflected that the castles of the Hospitallers or the Templars could
not easily be captured in a short time'.22 Similar worries may explain why
the Mongols, having conquered all of Muslim Syria, did not invade the
kingdom of Jerusalem in 1260. It has been argued that they did not do so
because they hoped to forge an alliance with the Franks against the Egyptians,
but it is just as possible that a fear of Prankish military might caused the
Mongols to hesitate. If this is the case, then it suggests that the Mongols,
despite supposedly numbering 20,000 men, were afraid of entering a region
defended by a series of strongholds which would probably have taken them
several years to capture. To some extent, the reign of Baybars subsequently
proved them right, for although this sultan spent most of the 12605 and
early 12705 campaigning against the Franks, even he failed to capture all of
their castles and fortified cities.23
Baybars' inability to finish the Franks off is all the more remarkable when
one bears in mind that some of his largest field armies contained thousands
more troops than anything his Christian opponents could muster. The Latins
relied on their strongest fortifications to make up for their lack of troops.
Indeed, many Prankish castles were deliberately constructed in such a way
that they could be defended by relatively few men against much bigger
besieging armies. According to one chronicler, the Tower of David was so
powerful that it only required a garrison of fifteen to twenty men.24 The
anonymous author of De constructione castri Saphet considered it a great
advantage that far more troops would be needed to attack this fortress than
to defend it. This comment seems to have been borne out by subsequent
events, for Saphet reputedly had a wartime garrison of 2200, but Baybars
may well have brought more than 12,000 men with him when he captured
this fortress in 1266.25 Although exact totals are hard to calculate, it is likely
42
that other Prankish garrisons were just as outnumbered, for in 1281 the
6000 Muslim troops besieging Margat were driven off by a mere 600
Hospitaller horsemen.26 Similarly, when Baybars occupied the citadel of
Antioch in 1268, he found that 'eight thousand fighting men' had taken
shelter inside, which suggests that unless thousands of other defenders had
already been captured elsewhere in the city, then the sultan's army could
have outnumbered the entire Christian garrison.27 These figures indicate
why the Franks rarely faced their opponents in open battle and preferred
to concentrate their meagre forces inside strongholds rather than trying to
defend the countryside.
Similar problems help to explain why the Latins seem to have found it
far harder to maintain and garrison sprawling city walls than compact castles
and citadels. We have seen, for example, that the Byzantine ramparts around
Antioch were more than eighteen kilometres long and covered an entire
mountain side. In the thirteenth century, when the city's population is likely
to have been much lower than it had been in Roman and Byzantine times,
the princes of Antioch must have found it difficult to come up with enough
troops to guard every tower, gate and postern.28 Indeed, descriptions of the
fall of the city to Baybars in 1268 imply that the defenders were quite simply
overwhelmed by the sheer amount of Muslim attackers, who were so
numerous that they 'surrounded the whole city and the citadel' before
scaling the walls with relative ease.29
Similarly, it has been suggested that the walls of Jerusalem were in a
poor state of repair during the 12305 and 12405, and this appears to be
borne out by the historical evidence, which indicates that only the citadel
was kept in good condition. Thus in 1229 15,000 Muslim peasants broke
into the city with ease but were unable to take the stronghold, which
provided shelter for the beleaguered Christians until a relieving force arrived
from Acre and drove the Muslims back into the hills.3 Around this time
a Muslim source wrote that, having recently regained the city, 'the Franks
constructed at Jerusalem a citadel which incorporated amongst its towers
that of David'.31 This was confirmed ten years later by a Latin chronicler
who commented that Jerusalem 'had not been fortified strongly except the
keep ... which was called the Tower of David'.32 Indeed, in 1239 the citadel
withstood another failed assault by al-Salih of Egypt,33 before eventually
falling to Malik an-Nasir Dawud, ruler of Kerak, after a blockade lasting
more than three weeks.34 During these two sieges, the Tower of David is
reported to have been defended by very small garrisons numbering first
twenty knights and then one knight plus seventy infantry. These campaigns
also add to the impression that Jerusalem's city walls were no longer
complete, and that the Christians had concentrated all their meagre
F O R T I F I C A T I O N S AND E X T E R N A L SECURITY
43
resources on making sure that the Tower of David was capable of withstanding a siege. If this was the case, then Dawud's campaign of 1239 may
have proved decisive, for although the Christians subsequently regained
Jerusalem by treaty, the Muslims did not return the city until they had
demolished parts of the citadel and had even managed to shift some of
the gigantic Herodian stones at the base of the Tower of David. It seems
unlikely that the Franks had the time or the resources to rebuild this
structure properly before 1244, leaving them wholly at the mercy of the
dreaded Khwarizmians.35
In a sense, therefore, the Latins only lost Jerusalem once they had lost
its citadel. Likewise, the new strongholds constructed at Ascalon and Tiberias
during the early 12405 may have represented other, less successful, attempts
to defend and even recolonise settlements without having to build extensive
new city walls straight away.36 The relative ease with which compact citadels
of this kind could be defended compared with long, sprawling circuit walls
can be illustrated further by returning to the fall of Antioch in 1268. Whereas
Baybars managed to storm the city's ramparts in a mere three days, the
citadel atop Mount Silpius only surrendered when the thousands of people
who had taken shelter there found that they lacked the supplies to survive
a protracted siege.37 On the other hand, the fate of Antioch proved that
even the most impregnable citadels could not provide adequate protection
for settlements inhabited by thousands of people. The only way to do so
and to succeed in repopulating places like Tiberias was to build vast urban
fortifications. But the fact that Antioch was lost 'because there was not in
it a force sufficient for its defence' proved that the Franks were not always
capable of maintaining existing walls, let alone constructing new ones.38
This helps to explain why they never returned to Ascalon, Tiberias or
Jerusalem after the mid 12408.
By contrast, the number of people living in the three great coastal settlements of Tyre, Tripoli and Acre probably rose in the thirteenth century
and the latter city is said to have contained 40,000 inhabitants by i2pi.39
This was partly because the gradual loss of inland territories led to a
concentration of Christians along the coast, but it also reflected the great
political and economic importance of these cities. In military terms it also
meant that adequate troops and resources could be found to guard and
maintain their walls properly, and this may explain why their citadels,
whose strategic insignificance has already been referred to, are hardly ever
mentioned in the contemporary sources.40 However, it is important to
remember that even at these, the wealthiest and most densely populated of
all the crusader cities, the task of defending the walls had to be shared out.
The Hospitallers, Templars and Teutonic Knights were especially important
44
U N K N O W N CRUSADER CASTLES
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
45
aboard a small boat carrying one hundred further troops, and soon after
the Lombards raised the siege and withdrew to Tyre.48
Help could also be sent to cities or castles which were under threat, even
if they were not actually under siege. The arrival of 130 Cypriot knights at
Acre in 1265 boosted this city's defences just as Baybars was besieging
Caesarea and Arsuf a few miles down the coast.49 Many years earlier Saladin
had also been deterred from attacking Tripoli after a contingent of Sicilian
knights arrived to help the city. Indeed, both Tyre and Tripoli presented
Saladin with a dilemma, for the longer he left them in Prankish hands, the
more reinforcements they would receive from the west; yet he was reluctant
to besiege these strongpoints as long as there were other, less powerful
castles still to be taken elsewhere.50 Their naval power therefore enabled
the Latins to protect coastal fortifications which would otherwise have been
lost to the Muslims, or in the case of Beirut, a rival western faction.
Even when a city or fortress could no longer be defended against land
attackers, lives could still be saved if its garrison had some way of escaping
by sea. In 1218 a Genoese fleet sent to relieve Caesarea, which was being
besieged by a large Damascene army, could do nothing to save the city but
did at least manage to rescue its defenders and take them to Acre.51 In 1291
the garrison of Sidon was also able to get away from Mamluk troops
besieging the town by first withdrawing into the sea castle, and then sailing
to Cyprus once any further resistance became pointless.52 Although the
citadel of Tiberias was of course situated many miles from the Mediterranean, its location next to Lake Tiberias gave its occupants the same option,
for in 1187 Raymond of Tripoli, lord of the city, advised his wife and her
companions to take to their boats if Saladin, who was at that time campaigning close by, managed to break into the castle. It does not appear that
this was done, for shortly afterwards, when the Christians had been defeated
at Hattin, Saladin occupied Tiberias and granted Raymond's wife safe passage
to Tripoli.53 At other times access to the sea really did make the difference
between life and death; a point which can be illustrated by looking at the
contrasting fate of Haifa and Arsuf, both taken by Baybars in 1265. Whereas
almost all the inhabitants of Haifa managed to flee in boats just as the
Muslims broke into the town, the Hospitallers defending Arsuf were
prevented from making contact with Prankish ships trying to assist them,
and were consequently all killed or captured.54
Such incidents confirm that proximity to the sea became an important
element in the defensive strategy of all coastal cities and fortresses. One of
the principal reasons why Richard of Cornwall decided to refortify Ascalon
in the 12405 was that it could be reached by sea if it ever came under attack.
Similarly, the site of 'Atlit was regarded as an ideal place to build a castle
46
because it had 'a naturally good harbour'.55 This was not the only stronghold
designed in such a way that retreating garrisons could virtually step straight
into waiting vessels. At Sidon there was a small jetty right next to the sea
castle, which was presumably used by the city's defenders when they fled
to Cyprus in i29i.56 Further south, the citadel of Caesarea acted as the
southern breakwater of the harbour, and must have been easy for the Genoese
relieving force to reach in I2i8.57 The entrances to many larger crusader
ports were also protected by flanking towers, usually with a chain between
them which could be raised to prevent hostile ships from gaining access.
Such structures existed at Beirut, and it is interesting to speculate whether
they played any role in the Lombard siege of 1231-32.58 Similar defences also
existed at Tyre and at Acre, where the Venetians and the Genoese periodically
fought for control over the so-called Tower of the Flies guarding the harbour
entrance.59
A combination of naval strength and powerful fortifications enabled the
Latins to hold on to their coastal territories until the very end. But on land
the situation was very different, for here it was the Muslims who were at
an advantage, and sometimes outnumbered their Prankish opponents by
as much as ten to one. This discrepancy prevented the Franks from relieving
inland strongpoints in the same way as they had done at 'Atlit in 1220.
Admittedly, there were incidents of field armies rather than seaborne forces
rescuing besieged castles or cities. In 1262 Armenian-led troops prevented
Baybars from capturing Antioch, and in 1229 Prankish knights from Acre
drove Muslim attackers out of Jerusalem. However, it is significant that the
relief of Antioch involved a large Mongol contingent rather than a purely
Christian force, and that the Muslims expelled from Jerusalem were rebels
rather than proper Ayyubid troops.60 In fact there is no evidence that
Baybars ever had to abandon a siege because a purely Latin field army
turned up and forced him to retreat. The Franks quite simply lacked the
resources to mount a major expedition of this kind into the interior, where
a Christian force of two or three thousand could easily be ambushed or
even annihilated by a far larger Muslim army.
Their overwhelming superiority on land enabled the Muslims, and especially the Mamluk sultans of the late thirteenth century, to besiege Prankish
strongpoints of the interior almost at will. Although the presence of twentyone Muslim galleys at the siege of Ascalon in 1247 suggests that they did
sometimes have enough ships at their disposal to blockade coastal sites,
such incidents appear to have been rare, and Egyptian efforts to create an
effective fleet were not very successful.61 But inland there was very little to
stop the Muslims from surrounding a strongpoint, thereby preventing the
garrison from escaping in the way that the remaining occupants of Sidon
47
had done in 1291. As a result, inland sieges could end with appalling
massacres and devastating looting sprees. Such atrocities occurred at Saphet
and also at Antioch, where virtually the entire population was either killed
or enslaved, and centuries of Byzantine culture were wiped out in a matter
of hours.62 These incidents explain why Prankish control over inland regions
crumbled decades earlier, and why the vast majority of fortifications built
or repaired during the thirteenth century were located along the coast. In
1253, for example, Louis IX decided to rebuild Sidon's defences rather than
construct a new fortress in the interior, because the local barons advised
him that such a place would be too exposed to Muslim attacks without any
access to the sea.63 In 1230 Gregory IX expressed similar concerns for the
castle of Montfort, because it was located several miles inland and was
consequently proving costly to build and difficult to defend properly.64 The
fate of the Holy Land was ultimately decided by the fact that the Latins
tended to dominate the sea and the Muslims found it easier to control the
land. Under such circumstances, the absence of an adequate Christian field
army proved decisive, for even the most powerful inland strongpoints could
not be expected to hold out indefinitely without any external assistance.
As the fall of Antioch plainly illustrated, the loss of a major Christian
strongpoint represented a human tragedy as well as a territorial loss, for
crusader fortifications were intended to protect people just as much as
national frontiers. Indeed, when Baybars finally obtained the surrender of
Antioch's citadel, he found that there were many 'women and children' who
had tried to seek refuge inside it alongside the actual fighting garrison.65 But
before this final catastrophe occurred, we have seen that the walls of Antioch
protected the inhabitants of the city against many previous Muslim incursions, including Saladin's invasion of 1188. Similarly, both Acre and Tyre
had such massive defences that their citizens had little to fear from the raids
which Baybars carried out in the surrounding countryside during the i26os.66
The importance of ensuring that the walls of major settlements were properly
maintained and garrisoned can be illustrated further by returning to the
Khwarizmian invasion of 1244. Whilst Acre and Tyre do not appear to have
been threatened by this event, Jerusalem was sacked by the invaders with
the result that 7000 inhabitants were supposedly massacred.67
It has already been suggested that sites like Acre, whose city walls were
able to withstand tens of thousands of Muslim troops, were few and far
between, and that most smaller towns or bourgs were only protected by a
single ditch and rampart. Consequently, the inhabitants of smaller settlements relied on their citadels more than their town walls to protect them
from major attackers. Several incidents of this kind have already been referred
to, such as the raid made by Baybars against 'Atlit in 1265 which resulted
48
in the destruction of the town but caused no physical harm to the Christians
sheltering inside the actual fortress.68 This point can be illustrated further
by taking a closer look at the history of Sidon during the thirteenth century.
Following the completion of Sidon's sea castle by members of Frederick II's
crusade in 1228, Louis IX sent a contingent of his army to build a second
fortress and new urban fortifications there in the summer of 1253. While
this work was still going on the Muslims launched a surprise raid on the
Franks, which resulted in the death of 2000 Christians, most of whom were
killed because there was not enough room for them inside the sea castle.
As a result Louis IX decided to supervise the rest of the project in person, so
that it was completed before he returned to France the following year.69
Six years after Louis IX's departure, lulian, lord of Sidon, made a rash
incursion into territories to the east of Beaufort, which were held by the
Mongols at that time. The Mongols responded by launching a devastating
counter-attack against Sidon but, thanks to Louis's new defences, Julian
was able to hold the Mongols off at the gate just long enough for the
Christian population to retreat. Meanwhile the presence of a new land castle
to complement the older sea castle ensured that this time there was enough
space for every one to find shelter. Consequently, when the Mongols finally
broke into the empty town, all they could do was to carry out widespread
looting and dismantle the walls, but they made no attempt to attack either
citadel. Louis IX's wish that the massacre of 1253 should not be repeated
had therefore been fulfilled.70
Apart from people who inhabited an adjacent town or bourg, strongpoints
also provided shelter for those who lived in the surrounding countryside.
Regardless of whether they were ultimately intending to besiege a Latin
strongpoint or merely carry out a swift raid, Muslim forces invariably caused
devastation to the Christian territories they passed through. Cattle and slaves
were seized, crops destroyed and orchards burnt, partly to feed the army
and to reward the soldiers, and partly to inflict economic damage on the
Franks. As a result, Christians living in the countryside deliberately settled
close to castles so that they could find shelter relatively quickly. The author
of De constructione castri Saphet wrote that, once this fortress had been
completed, 10,000 people living in 260 villages recolonised an area of central
Galilee which had previously been considered too dangerous to inhabit.71
During the reign of Baybars, who frequently carried out large-scale raids
in order to destroy crops needed to feed Prankish garrisons, we often hear
of hundreds or even thousands of non-combatants taking shelter inside
fortresses. When the Mamluks captured Saphet (1266), Beaufort (1268),
Akkar (1271) and Chastel Blanc (1271), they found many such people who
had arrived from an adjacent bourg or from neighbouring villages.72
49
5O
U N K N O W N CRUSADER CASTLES
either directly or via other castles.81 In the 19505 the scholars R. Fedden and
J. Thomson wrote that groups of fortifications like this 'were sited so that
intercommunication should be possible over an extraordinarily wide area',
implying that these strongpoints formed a kind of early warning system
against possible Muslim attacks from the direction of Hama and Horns.82
Yet the possibility that the Latins deliberately created such a network seems
remote bearing in mind that some of these fortifications, such as Crac des
Chevaliers, had originally been constructed before the First Crusade when
the frontiers of the county of Tripoli were still unknown.83 Consequently it
is likely that the intervisibility enjoyed by most castles in this region was a
coincidence caused by the fact that Crac des Chevaliers, Akkar, Chastel Blanc
and al-Arimah, like most medieval strongpoints, were built on hill tops.
Similar observations can be made about numerous smaller fortifications in
the county of Tripoli. In the late nineteenth and early twentieth centuries
Rey and Deschamps argued that local structures such as Tuhklah, a sturdy
square tower near Chastel Blanc, were primarily intended as lookout posts.
Yet more recently it has been pointed out that Tukhlah was basically a
residential and agricultural complex which stood in the heart of a village
and 'would have been very badly placed as a watchtower since it is overlooked
by a much higher hill about half a kilometre to the east'.84 In later sections
it will be shown that the same point applies to virtually all other fortifications
of this type in the Holy Land, Greece and Cilician Armenia.85 Hence theories
that Christian castle builders in the eastern Mediterranean deliberately created vast networks of intervisible strongpoints seem to stem from a tendency
to look at maps rather than the actual terrain, to ignore the eclectic history
of most fortifications, and to overlook the more day-to-day domestic functions of sites like Tukhlah.
Returning to the role of fortifications as a means of sheltering civilians,
during the final decades of Prankish rule the ability of sites like Beaufort
to spot Muslim attackers or send fire signals to other castles made little
difference anyway. As Latin field armies found themselves overwhelmingly
outnumbered and could not come to the rescue of inland strongholds,
sites like Beaufort became increasingly isolated and could be besieged by
the Mamluks almost at will. Once an inland castle or city was broken
into, people sheltering there were consequently doomed because they
had nowhere to run and had no Prankish relieving forces to appeal to.
Indeed, when the Muslims began to swarm into Antioch, many of its
inhabitants found themselves trapped by their own walls, because Baybars
had deliberately sealed off all the gates in order to prevent any loot from
being carried away86
By contrast, superior Latin naval power meant that civilians sheltering
F O R T I F I C A T I O N S AND EXTERNAL S E C U R I T Y
51
52
crusaders were intended for local rather than national defence, and had
little strategic value beyond their immediate surroundings. Moreover, the
regular incursions made by Nur al-Din, Saladin and others into Christian
territories showed that frontier castles stood little chance of blocking Muslim
attacks anyway.91
At first glance, it would appear that these observations applied to the
thirteenth century just as much, and perhaps even more so because of
the frequent lack of central authority within the crusader states. It is clear
that external pressure and internal squabbling sometimes made it virtually
impossible for the Franks to pursue a common military strategy toward
their Muslim enemies. The castle of Beirut, for example, largely remained
in Christian hands for as long as it did because its lords negotiated a series
of peace treaties with the Muslims, irrespective of what their Prankish
neighbours were up to.92 Similarly, in 1266 the Templars made a treaty with
Baybars whereby the latter recognised the Order's possession of Tortosa
and Chastel Blanc in exchange for half of Jabala, a coastal settlement to
the north which the Templars shared equally with the Hospitallers. This
deal was clearly made without reference to the Hospitallers, who were so
furious that their troops ended up fighting Muslim forces sent by the sultan
to occupy his new share of the town.93 This general mood of 'every man
for himself' has led scholars to argue that Smail's view of twelfth-century
fortifications still holds true for the period after 1187. It has been written
that 'few of the Latin strongpoints were of any genuine strategic value',
apart from Margat, Crac des Chevaliers and to some extent 'Atlit.94 On a
smaller scale, many Latin strongholds such as Goliath were normally too
weak to withstand large Muslim invasion forces anyway, and were primarily
intended as local centres of defence, farming and lordship. These less overtly
military functions of crusader fortifications had hardly been looked at by
historians before Smail's time.
Although we should probably still reject Key's theory of nationwide castle
networks, it would be an oversimplification to say that during the thirteenth
century the Franks were incapable of implementing any kind of overall
strategy at all, for some larger strongholds do appear to have been intended
to interact with others. This was most notable along the southern frontier
of the kingdom of Acre, where the Latins were concerned about possible
offensives by the Egyptians but also hoped to reconquer Jerusalem or make
counter-attacks against Egypt. 'Atlit, for example, was referred to as 'the
breastwork of the city of Acre' at the time of its construction.95 This implies
that the fortress had not just been built to administer local Templar properties but had deliberately been placed along the coastal route south of Acre.
Any Egyptian troops attacking the Prankish capital would have to contend
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
53
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
55
56
UNKNOWN CRUSADER
CASTLES
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
57
58
U N K N O W N C R U S A D E R CASTLES
dimension by the findings of Ronnie Ellenblum, who argues that the Muslim
population was not actually as large as historians have previously assumed.
Ellenblum gives several reasons for this. First, he argues that the massacres
committed by crusaders during or shortly after the First Crusade meant that
many Muslims had either been killed or had simply fled. Secondly, he points
out that large parts of the Holy Land were inhabited by native Christians,
and that they often actually outnumbered the remaining Muslims, who have
themselves tended to be lumped into one indiscriminate Islamic grouping
with little regard to regional, cultural or religious variations. Thirdly,
Ellenblum's conclusion that historians have underestimated the number of
Franks settled in the countryside enables him to show that the overall Latin
population of the Holy Land, whilst still smaller than the combined number
of Muslims and native Christians, cannot have constituted less than 25 per
cent of the entire population even by the standards of older scholars.29
If we combine these findings with the arguments of Mayer and others,
we are left with a picture of a Muslim population which was both smaller
and more harshly treated than scholars assumed in the nineteenth and early
twentieth centuries. This picture tends to fit the pattern of Muslim uprisings
in the crusader states from 1187 onwards, for such rebellions did break out
from time to time, implying that the peasantry felt oppressed by their Latin
overlords, but they were rare, suggesting that the notion of a very large
and threatening Muslim population has indeed been exaggerated in the
past. Indeed, the Muslims normally only felt confident enough to rebel
during periods of Latin weakness or military defeat. Thus uprisings occurred
in 1187 and 1188, following the battle of Hatting In 1266 a large Prankish
army campaigning near Tiberias lost around five hundred troops in a
Muslim ambush, subsequently being turned upon by local Muslim peasants,
who harassed the defeated Latins as they retreated back to Acre. At this
time Prankish authority over the rural population throughout Galilee was
waning as Baybars gradually advanced from the east and south.31 Even
during the first half of the thirteenth century it was clearly the case that
the more isolated a Prankish outpost became, the more likely its inhabitants
were to face the wrath of local Muslims. We have seen that in 1229 Muslims
broke into and ransacked Jerusalem, at a time when the city had only just
been regained by the Christians and was probably inhabited by very few
Latins living behind dilapidated or incomplete walls. This attack must have
been prompted by a desire for booty, but it probably also reflected a
widespread feeling that Jerusalem should not have been handed over to the
Christians by al-Kamil.32
The value of castles as a means of dealing with this type of threat is
illustrated by the outcome of the 1229 attack on Jerusalem. According to a
FORTIFICATIONS
AND I N T E R N A L SECURITY
59
western source this incident involved no less than 15,000 Muslim peasants,
but it is unlikely that they possessed the organisation or equipment needed
to besiege any well-defended strongpoints. As a result, the Prankish defenders were able to retreat into the Tower of David and wait until a relieving
force of knights from Acre turned up and drove the Muslims out of the
city.33 On a smaller scale, the sheer proliferation of minor fortifications
throughout the crusader states helps to explain why uprisings were relatively
rare, for even if the local Muslims wished to rebel, they must have understood that their chances of inflicting permanent damage or heavy loss of
life on the Franks were remote. An isolated tower might not have had much
chance against Saladin but could easily have resisted small groups of local
Muslim troublemakers. Similarly, robbers, bandits and localised raiders had
neither the will nor the means to launch attacks against Latin fortifications.
By constructing towers, fortified farmsteads or small castles the Franks could
protect themselves, their cattle and their movable property against such
people, particularly at night. For years or even decades at a time, the vast
majority of crusader strongholds were simply intended to maintain law and
order, defend proprietorial rights and safeguard economic activities in much
the same way that fortifications did throughout late medieval Europe. This
makes the constant rebuilding of sites like Goliath much more logical than
it initially appears.
Whilst fortifications could increase security in the face of rebels or
attackers, it could also diminish the amount of control which lords enjoyed
over their own Latin vassals. The fact that crusader armies tended to be
small, whilst at the same time Muslim pressure made it necessary for certain
Latin strongholds such as Crac des Chevaliers to be extremely large, could
make it difficult for a ruler to capture the major stronghold of a vassal who
chose to rebel. In the south this dilemma is illustrated by Frederick IFs
largely unsuccessful attempts to impose his authority over the kingdom of
Jerusalem in the late 12205. In 1229 Frederick marched south from Acre to
'Atlit and ordered the garrison to hand it over. When the Templars, who,
as we have seen, were opposed to Frederick's crusade because of his arguments with the papacy and status as an excommunicate, refused to open
their gates, the emperor saw that he did not have enough troops to take
'Atlit by force and was obliged to retreat. Later, Frederick also tried to
capture the Templar headquarters in Acre, but could not do so and eventually
abandoned the city altogether.34 Indeed, although it was also a useful stepping
stone in the drive to regain Jerusalem, Frederick's construction of a new
citadel at Jaffa not long afterwards may have reflected a realisation that he
was better off building his own castle rather than trying to occupy somebody
else's, if he wanted to secure a strong base in the region.35 Such tactics did
6l
the lord of Gibelet39 During the second half of the thirteenth century these
places became involved in a far more serious baronial rebellion, although
this time the lords of Nephin were allied with the counts of Tripoli against
the Embriaco rulers of Gibelet. In the course of this conflict, the Embriacos
and their Templar allies attacked Tripoli no less than three times (1258,
1276, 1282), only being kept at bay by the city's massive defences. In 1276
the Templars also made a failed assault against Nephin, and it was not until
1282 that Bohemond VII finally captured Bertrand II of Gibelet, executed
him and occupied his castle.40
At times the apparent weakness of rulers like Frederick II and Bohemond
IV also led to internecine struggles between other rival factions. The most
famous clashes of this kind probably took place in Acre, where the Military
Orders and the Italian city states had all constructed fortified compounds
or towers in order to intimidate their rivals and protect their own interests.
The Templars' headquarters, an imposing quadrilateral citadel with four
corner towers, may have been the strongest such structure, but both the
Hospitallers and the Teutonic Knights probably held similar compounds.41 The Genoese, the Pisans and the Venetians also held their own
quarters dominated by the kind of large, isolated towers still visible in some
Italian towns.42 These structures divided Acre into numerous walled enclosures or distinct quarters, and the type of fighting which sporadically took
place there can be illustrated by giving a brief description of the war of
St Sabas, fought between Venice and Genoa in the late 12505. This conflict,
initially sparked off by a property dispute, eventually led to widespread
street fighting between the Venetians, who were aided by the Pisans, and
the Genoese, whose principal allies were the Hospitallers. The two sides
bombarded each other with crossbows and catapults mounted on top of
towers and other tall buildings, and some of these weapons were so powerful
that they could hurl rocks weighing 200 kilogrammes over considerable
distances. In these circumstances, it is hardly surprising that many houses
were destroyed and countless lives lost in the space of just a few months.
It is equally understandable that the first action taken by the Venetians
after they finally emerged victorious was to raze the Genoese quarter to the
ground.43
The war of St Sabas also had wider implications which went beyond the
city of Acre itself. Most notably, the Embriaco lords of Gibelet were Genoese,
and their long struggle with the rulers of Tripoli began in 1258, when the
fighting at Acre was at its height. These clashes may therefore have
represented a wider campaign to augment Genoa's power in the east,
and even turn Tripoli into a fortified trading post controlled solely by
the Genoese. They later established such a base at Famagusta, and after the
62
U N K N O W N CRUSADER CASTLES
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
63
to get back to Europe and avoid getting bogged down in the Latin East.
Perhaps, therefore, there has been too much of a tendency in the past to
treat crusader fortifications as a distinct group, whereas in reality their role
in internal conflicts, like their role in the maintenance of law and order,
was virtually identical to that of countless other strongpoints in Italy and
many other parts of western Europe.
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
65
Beaufort and Saphet, ensured that many of the inland areas gained at
this time remained under Christian rule until the 12605. Initial efforts to
recolonise numerous urban sites lost or destroyed after the battle of Hattin
centred around the construction of new castles, especially at Caesarea,
Sidon, Ascalon, Tiberias and Jerusalem. Further north, the tower which
Bartholomew de Ravendel constructed at Maraclea after 1277 represented
another, unusually late, attempt to .re-establish a Prankish lordship previously overrun by the Muslims.4 Just along the coast, Bohemond VI also
retook Latakia around 1261 'and built a strong new tower' there, so that
the town, which had belonged to the Muslims since 1188, returned to
Christian control for the next twenty-six years.5 It is interesting to note
that out of all these gains this last example appears to have been the only
settlement restored to Christian rule after a direct confrontation with
Muslim defenders. This suggests that the Latins were normally too outnumbered to make conquests by force, yet castles still enabled them to hold
on to areas gained through diplomacy.
Although it did not involve any territorial expansion as such, Louis IX's
visit to the Holy Land between the summer of 1250 and the spring of 1254
is also instructive. According to Joinville, Louis arrived in Acre with barely
one hundred knights still available to him out of an original force of 2800
before his defeat in Egypt.6 Yet he was still able to consolidate Latin rule
in the Holy Land and help to extend the life of the crusader states by
building new fortifications or repairing older ones at Acre, Jaffa, Sidon,
Caesarea, and Haifa.7 During this building programme Louis IX's meagre
forces were not entirely safe from enemy incursions, for in 1253 a Muslim
raid against Sidon resulted in substantial Christian casualties.8 This probably
helped motivate a Latin counter-raid against the Muslim town of Banyas
soon afterwards.9 Some Christians at least survived the Muslim attack by
sheltering inside Sidon's sea castle, constructed during the winter of 1227-28
by members of Frederick II's crusade. Similarly, when Louis IX's forces
rebuilt the walls of Jaffa, Joinville noted that the army made camp on the
fields around the much older fortress already built there by Frederick; a
location which was not only convenient but was close to a source of shelter.
Both Frederick and Louis slowly but steadily consolidated Latin power
through the construction of fortifications, with the latter being aided by
the work of his predecessor, even though Louis no longer had any hope of
actually confronting the Muslims in a pitched battle.10
There are other examples dating from this period of fortifications being
built or repaired by a crusader army as it moved deeper and deeper into
enemy territory. On such occasions castles were still being erected to secure
territorial gains for the future, but in the short term they were also needed to
66
U N K N O W N CRUSADER CASTLES
protect the army itself from potential counter-attack. Richard I's campaigns
in Palestine during the Third Crusade provide us with a classic example of
such tactics. Following the recapture of Acre in July 1191, the crusaders
immediately set about refortifying its defences so that they were no longer
as exposed to Saladin's incursions as they had been when camped outside
the city.11 Later, Acre became a springboard for Richard Fs march southwards, but the further he moved away from the city, the more vulnerable
he became to enemy attacks. In a sense, self-preservation had as much to
do with Richard's hasty refbrtification of Jaffa and Ascalon as the desire
to reconquer these cities, for his troops needed regular places of shelter
where they could rest and receive supplies from the crusader fleet.12
Moving inland from Jaffa during the autumn of 1191, Richard used similar
methods to protect his army in his unsuccessful bid to recapture Jerusalem.
In October 1191 the Templars accompanying Richard quickly rebuilt Casal
des Plains, a relatively small fortification consisting of a central keep surrounded by a walled outer enclosure. This castle, which had probably
belonged to the Templars earlier in the twelfth century, had been destroyed
recently by Saladin, but its location along the road to Jerusalem made it a
useful base for Richard. As was the case at Ascalon and Jaffa, there was
a dual incentive for the reconstruction of Casal des Plains: in the short
term it protected the members of the Third Crusade, but in the long term
the Templars were clearly staking their claim to land which had been theirs
before the battle of Hattin. These hopes did not materialise, for after Richard
retreated from the area it fell under Saladin's control, and there is no
evidence that the Templars ever returned to Casal des Plains.13
Despite these setbacks, the Templars were closely involved in thirteenthcentury efforts to regain Jerusalem and the land which connected it with
the coast, an area where they had held other properties before 1187 and had
carried out their self-appointed task of protecting pilgrims.14 Once again,
it was clearly hoped that the construction of castles would enable the
Christians to bring this region back under their control and to defend it
against the Muslims even without the presence of a large Prankish field
army. When the massive Templar fortress of 'Atlit was built from 1217
onwards along the coastal route toward Jerusalem, Oliver of Paderborn
wrote that 'the primary advantage of this building is that the assembly of
Templars ... will remain in the garrison of this fort up until the restoration
of the walls of Jerusalem'.15 Toward the end of 1243, by which time Jerusalem
had been regained through peace treaties, the Templars were planning to
reconstruct Latrun (Toron des Chevaliers), another ruined fortress they had
once held along the route to the city which originally consisted of a square
keep defended by two outer ramparts. Although this would clearly have
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
67
68
U N K N O W N CRUSADER CASTLES
confused attack on Qaqun in 1271, which may have been an extensive raid
rather than a siege.23 During the German crusade of 1197-98, Acre also
became the initial base for an attack on the isolated inland stronghold of
Toron (Tibnin), but this had to be abandoned at the approach of a Muslim
relieving force.24
Further north, Prankish efforts to regain Muslim-held territories had
equally mixed results. In 1191 Bohemond III failed to reconquer Latakia and
the neighbouring port of Jabala in an offensive which was presumably
launched from Tripoli itself.25 In 1207 Raymond Roupen, an Armenian
claimant to the throne of Antioch, granted Jabala to the Hospitallers in
order to gain their support against his political opponents, and in 1210 this
was followed by Bikisrail, another former crusader castle situated to the
north east of Margat.26 However, the Hospitallers only ever managed to
recover half of Jabala, which they were forced to share with the Templars,
who had been granted the very same settlement by Raymond's rival Bohemond IV.27 Meanwhile, their strong presence in Syria meant that both
Orders were at the forefront of many campaigns against the Muslims in
the county of Tripoli, such as the failed siege of Horns which the Hospitallers
of Crac des Chevaliers undertook in 12O7.28 To the north of Antioch, the
Templars also spent much of the thirteenth century trying to reconquer
the vast estates which they had held there before 1188, and in particular the
strategic fortresses of Baghras and Darbsak. These castles guarded two of
the most important mountain passes connecting Antioch with the Cilician
plain, but the former was occupied by the Armenians between 1190 and
1216, whilst the latter had been garrisoned by Aleppine forces in the wake
of Saladin's invasion. As a result, the Templars only retained the neighbouring stronghold of Roche Roussel (Hadjar Shoglan/Chilvan Kale) during
the early years of the thirteenth century, and it was from here that they
made an unsuccessful attack on Darbsak in 1237.29
The last phase of Christian expansion in Syria occurred after the Mongol
destruction of Aleppo in 1260, which enabled Bohemond VI to commit
'many acts of aggression against the lands of Islam' so that 'he took a
number of villages in Muslim territory', as well as the town of Latakia itself.
All these conquests appear to have been made from the city of Tripoli.30
It was also at this time that Bohemond acquired Darkoush, Kafr Dubbin
and other castles situated to the east of Antioch, which had been captured
by the Muslims in 1188 but could now be reoccupied by the Franks after
the Mongol invasion.31 Unfortunately for the Templars, a similar set of
circumstances enabled the Armenians to occupy the fortress of Darbsak.
In 1268, however, all these strongholds were lost to the Mamluks along
with Antioch.32 These events again highlight the Franks' chronic lack of
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
69
7O
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
71
during the late 12705, but in 1285 Kalavun insisted on the destruction of
this stronghold as part of a new peace treaty with the Latins. It is possible
that Kalavun, who had just captured the neighbouring Hospitaller fortress
of Margat, feared that Maraclea would be used by the Franks to raid his
newly acquired lands or even attempt to recapture Margat itself.48
To the south of Maraclea, the Templars and the Hospitallers also relied
on intimidatory raids to maintain their power in the county of Tripoli and
to keep their Muslim neighbours in check. One such neighbour was the
tribe of the Assassins, schismatic Muslims who controlled the extremely
mountainous district to the east of Latakia. The rough terrain in this area
enabled the Assassins to retain their independence for much of the thirteenth
century, but their close proximity to Margat, whose garrison could easily
launch punitive raids against them, forced them to pay the Hospitallers an
annual tribute of 1200 gold pieces and 100 bushels of wheat and barley.
The threat posed by Crac des Chevaliers toward the emir of Hama and the
ruler of Bokebais (Or Abu Qubais, a Muslim castle between Hama and
the territory of the Assassins) meant that they too had to make similar
annual payments to the Order worth 4000 and 800 gold pieces respectively.49
If they did not pay there could be grave consequences, for in 1229 the
Hospitallers carried out an extensive raid around Montferrand, a castle they
themselves had hoped to control since the twelfth century, because the emir
of Hama had not paid up.50 The following year a force of 500 horsemen
and 2700 footsoldiers, composed of both Hospitallers and Templars, launched another attack toward Hama, but this time the expedition ended iii
defeat at the hands of the emir's army.51
The Hospitallers were not deterred. Angered by the emir's persistent
refusal to pay, they organised a third offensive in 1233. This expedition was
far larger than those of 1229 and 1230 and involved 100 knights, 400 mounted
sergeants and 1500 footsoldiers led by the Hospitaller grand master, as
well as twenty-five Templar knights, eighty knights from the kingdom of
Jerusalem, thirty knights led by Bohemond V's brother Henry, and 100
knights from Cyprus. This impressive force marched overnight toward
Montferrand, enabling the Franks to make a surprise attack on its castle
town the following morning. Having sacked this settlement, they continued
to ravage the surrounding countryside before returning toward the coast,
without having encountered any Muslim resistance. Subsequently the sultan
of Damascus advised the emir of Hama to pay the money he owed, and
peace was re-established with the Hospitallers.52 Indeed, the Order appears
to have collected tribute for a further thirty years, until Baybars finally
obliged it to renounce these payments as part of a peace treaty established
in 1266.53 Until that time, the Hospitallers' arrangement with Hama can
72
U N K N O W N CRUSADER CASTLES
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
73
besieged. Indeed, when Margat came under siege a mere two years later,
600 Hospitaller cavalry troops burst out and successfully drove off the 6000
startled Muslims who were attempting to capture the castle.5 In 1256 a
comparable series of events took place at Jaffa, which was initially used as
the starting point for a successful raid against inland territories south of
Ascalon. When this raid came to an end, its participants found shelter
behind the fortifications of Jaffa, which came under siege from a Muslim
counter-attack. Shortly afterwards, the blockade was brought to a swift end
when a force of 200 knights plus 300 mounted sergeants, archers and
crossbowmen, most of whom had presumably just returned from the original
raid, rushed out and defeated the Muslims.59 Similarly, during the final
siege of Acre in 1291 the Franks initially kept their city gates open, so that
they could launch surprise raids against the enemy troops arriving on the
plain outside the city.60
Clearly, it was sometimes worth risking a direct encounter with the
Muslims, provided that the Franks had some means of protecting themselves
should the battle start to turn against them. The Rule of the Templars
advised members of the Order that, if they were defeated in the field, they
had to try to reach the nearest fortress in order to keep their casualties to
a minimum.61 Likewise, the castle of Ascalon, which had been used as a
collection point for Prankish troops on the eve of the battle of La Forbie,
became a refuge site for those few Christians fortunate enough to escape
the subsequent carnage.62 Those who did not suffered the same fate as many
of the Templars involved in the failed assault on Darbsak in 1237. This
campaign turned out to be one of the worst military disasters in the Order's
history, with most of its participants being cut down by an Aleppine
counter-offensive long before they could reach the safety of Roche Roussel
(Hadjar Choglan/Chilvan Kale).63 In 1266 Hospitallers, Templars, Teutonic
Knights and other secular troops campaigning near Tiberias also found
themselves trapped in a Muslim ambush which resulted in the loss of around
500 troops; a total which subsequently appears to have increased because
the Franks were then forced to march all the way back to Acre, a journey
of approximately fifty kilometres, while being constantly harassed by local
Muslim peasants. This humiliating defeat would surely not have been so
costly, and perhaps could have been avoided altogether, if the Franks had
been able to take refuge in a nearby castle.64
The numerous raids or sorties which took place during the thirteenth
century remind us that the Franks did not simply regard their fortifications
as passive refuge points but used the garrisons of such sites to raise additional
troops for their field armies. This strategy had backfired disastrously in the
months after the battle of Hattin, when Saladin, having annihilated a Latin
74
field army containing virtually every Christian soldier who could be found
in the kingdom of Jerusalem, overran several Prankish castles which were
virtually empty. But their lack of troops, combined with the obvious practical
advantages of relying on soldiers stationed nearby to conduct local warfare,
meant that the Latins continued to rely on such forces in the thirteenth
century.65 In times of defeat this policy clearly still carried a heavy penalty.
The Templars, Hospitallers and Teutonic Knights, who may have contributed as many as 600 knights to the battle of La Forbie, must have found
it difficult to defend their local fortresses properly after they suffered severe
casualties at the hands of the Egyptians and Khwarizmians. This applied
most to the Teutonic Knights, who are reputed to have lost all but three
of their contingent at La Forbie, presumably putting a severe strain on their
garrisons at Montfort, Acre and elsewhere.66 Perhaps losses sustained by
the Hospitallers also contributed to the fall of Ascalon in 1247.67
Sixteen years after the battle of La Forbie, a somewhat smaller Christian
army suffered another crushing defeat in southern Galilee. This expedition,
which was led by the lord of Beirut, was partly made up of Templars sent
from Acre, 'Atlit, Saphet and Beaufort.68 As a result, all these sites may well
have been seriously undermanned after the battle, although it is extremely
difficult to estimate exactly how many troops had been drawn from each
individual fortress. However, it has been calculated that Saphet alone had a
peace time garrison of 1650 men, fifty of whom were mounted knights,6^
and if similar numbers applied to their other major castles, the Templars
could have represented a substantial percentage of the 900 knights and
additional turcopoles who are said to have taken part in the battle.
Contemporary sources suggest that the Order suffered particularly heavy
casualties and had to pay a considerable ransom for the release of their
captured commander.?0 By contrast, it was probably less risky for the Latins
to take troops from strongholds which were not particularly important
strategically. It has been noted, for example, that the citadel of Acre
contributed relatively little to the overall strength of this city, and it
could probably be left virtually empty provided that surrounding ramparts
were still being guarded. From 1254 onwards, this structure acted as the
barracks for the French regiment, a standing force which had been established by Louis IX and normally contained about 100 knights, plus additional
crossbowmen and infantry. During the 12505 and 12605 this regiment
participated in several large-scale raids, including the campaign which ended
in the successful defence of Jaffa in 1256.71
Although the French regiment clearly boosted Latin troop numbers, its
contribution was not as significant as that of the Military Orders, who
participated in virtually every Christian campaign of the thirteenth century.
8. Chastel Rouge: the central keep and perimeter walls. (Denys Pringle)
11. Qaqun: the central keep, with traces of the perimeter defences on the right.
(Denys Pringle)
14. Crac des Chevaliers: distant view showing the rich farmland of the surrounding
hills. (Jonathan Phillips)
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
75
j6
THE O F F E N S I V E F U N C T I O N S OF F O R T I F I C A T I O N S
77
rebuilding Ascalon's walls, only for them to be demolished again soon after
as part of his peace treaty with Saladin.85 Indeed, both Saladin and Baybars
systematically dismantled many strongholds to prevent their recapture, and
even as late as 1270 the latter sultan virtually flattened the last remaining
defences at Ascalon. This was presumably done to deter Louis IX or the
Lord Edward from devoting their crusades to refortifying these sites and
then attacking Jerusalem or Egypt.86 Such incidents suggest that castles were
seen as far more temporary structures than they are today, and are proof
indeed of William of Tyre's famous saying that 'a castle destroyed is a castle
half-rebuilt'.?
Finally, it is important to remember that many Prankish strongpoints
along the coast were used for naval as well as land based offensives. Attacks
of this kind were most common during the frequent clashes between the
Italian city states, which tried to dominate local trade by deploying enormous
war fleets against each other. In 1264, for example, the Venetians attacked
the pro-Genoese city of Tyre with no less than fifty galleys equipped with
special boarding towers designed to overwhelm defenders positioned along
the sea walls.88 This assault proved less successful than an earlier, more
discreet raid carried out in 1242, when Venetian galleys were allowed to slip
into the harbour after their Ibelin allies had secretly lowered the chain
across its entrance.89 Both these offensives were planned and carried out
from Acre, but in 1232 Tyre itself became the starting point for another
naval raid, this time undertaken by twenty-two imperialist galleys, which
attacked Ibelin forces camped a few miles to the south at Casal Imbert.90
The previous year Lombard troops had also arrived by sea at Beirut, enabling
them to bring with them many of the building materials needed for their
fort.91 Similar episodes took place during internecine fighting in the north,
such as the fleet of fifteen galleys sent by Bohemond VII from Tripoli
against Templar Sidon in 1278. The provocation for this assault had been
the Templars' involvement with the lord of Gibelet, who was plotting to
seize Tripoli from Bohemond.92
Numerous seaborne attacks were also organised against the Muslims
during the thirteenth century. In the build up to the Fifth Crusade, for
example, the Christian fleet gathered in or near the harbour of 'Atlit during
preparations for the forthcoming expedition to Egypt.93 Although it is not
recorded where they set out from, a number of Frankish ships launched
another damaging attack on Alexandria in 1270, and made off with two
Muslim vessels.94 These incidents, as well as references to a Templar shipyard
at Acre, make it clear that coastal fortifications provided secure bases for
the maintenance and construction of ships and galleys, essential if the Latins
were to maintain their dominance over the sea.95
78
U N K N O W N C R U S A D E R CASTLES
Virtually all aggressive military activities undertaken by the Franks involved fortifications in some way. Such structures were used to extend or
consolidate Latin territories despite the numerical inferiority of Christian
field armies. At the same time, they could provide land based or seaborne
Christian forces with supplies, shelter or additional soldiers. Although some
of the strongholds mentioned in this chapter, such as Casal des Plains,
Maraclea or the temporary Lombard fort erected opposite Beirut, were
smaller in size, most were extremely large. Fortified cities like Acre, Tyre
and Tripoli, plus major castles such as 'Atlit, Crac des Chevaliers, Margal
and Saphet, have been referred to again and again. The same massive
fortifications which played such a key defensive role in preserving the
crusader states against full-scale Muslim invasions also played a prominent
part in any Latin offensives. This is hardly surprising considering that these
sites had the most troops and resources concentrated within them, and that
some, most notably Acre, provided plenty of shelter for crusaders arriving
from Europe on the eve of a campaign. Their great size and strength meant
that such strongpoints were large and secure enough for Christian armies
to retreat inside whenever they had been defeated or were being pursued
by the Muslims. Once again these observations indicate that we need to
distinguish between the much wider military role of some of the largest
crusader fortifications, which contributed to the security of the crusader
states as a whole, and most smaller strongpoints, which were normally only
built with local defence in mind.
80
THE M I L I T A R Y O R D E R S
8l
large territorial block to the north of Antioch, which may have come into
their possession as early as the 11305.
As far as the Hospitallers were concerned, the most important fortresses
in the kingdom of Jerusalem were Arsuf, Ascalon and Mount Tabor. Arsuf
was rented out to them in 1261 by Balian of Ibelin, whilst Ascalon was
entrusted to the Order soon after its reconstruction in 1241. Mount Tabor,
on the other hand, was granted to the Hospitallers by Alexander IV, acting
on behalf of the monks who lived there. This transaction is an interesting
example of growing papal involvement in the military affairs of the Latin
East, and suggests that the church as well as the nobility had problems
defending its territories. Further north, however, the key Hospitaller strongholds of Crac des Chevaliers and Margat had both been sold to the Order
by laymen rather than the church. These purchases had occured in 1144
and 1186 respectively, so that like the Templars, the Hospitallers were already
well established in the county of Tripoli before the battle of Hattin.10
In addition to these major fortresses, the Hospitallers held countless smaller
strongholds whose size and shape varied considerably. On and around the
plain of Akkar they possessed Goliath, Chastel Rouge and numerous other
sites which were not permanently lost until 1271.11 After the fall of Chastel
Blanc, Baybars occupied 'its territory, together with the forts and towers in
the neighbourhood of Hisn al-Akrad' (Crac des Chevaliers). Once Baybars
had conquered Crac, the Hospitallers abandoned several further towers,
'burning all their property that they could not remove'.12 Most of these
structures had probably been built or acquired in the twelfth century, and
many had been sold along with Margat and Crac des Chevaliers. Indeed,
when they purchased the former stronghold, the Hospitallers received the
entire town of Banyas (Valania), which lay below the slopes of Margat.13
In the kingdom of Jerusalem, it appears that many twelfth-century
Hospitaller castles were either lost for good in the period 1187-92, or were
left in ruins even if neighbouring areas were subsequently regained by
treaty. Excavations at the castle of Belmont, for example, confirm that this
site was never restored after Saladin demolished it in 1191, and the area
was later occupied by a Muslim village. No attempt was made to rebuild
Belmont even after Frederick II negotiated the return of Jerusalem itself
in 1229.14 Furthermore, archaeological investigation suggests that the
Hospitallers never regarrisoned Belvoir or the neighbouring castle of
Forbelet either. In theory this region of Galilee came under Christian rule
again during the early 12405, and until Belvoir was actually excavated in
the 19605 some historians therefore argued that this site had been reoccupied
by the Order briefly.15 By contrast, it seems that the Hospitallers did return
to some properties regained at this time which were not so remote, such
82
as Bethgibelin, located inland from Ascalon and held for a short spell
between 1240 and 1244.16 Nearer the coast, the Hospitallers also possessed
a number of other small strongpoints which were similar in design to the
tower of Tukhlah. These included La Tor de l'Opital, on the outskirts of
Tyre, and Turns Salinarum, situated near Caesarea. Buildings of this kind
were primarily intended as local watch towers and places from which to
administer neighbouring estates.17
Bearing in mind that identifying smaller Hospitaller properties can sometimes be difficult, it is hardly surprising that trying to do the same for the
Templars, whose records no longer survive, can be virtually impossible.
Some information can be gleaned, however, from contemporary descriptions
of property disputes between the Templars and the Hospitallers. It is in
this way, for example, that we know the two Orders agreed to divide the
town of Jbala between them, after the Templars disputed the way in which
it had been granted to the Hospitallers by Raymond Roupen.18 This deal
was finalised in 1233, but in 1266 the Templars relinquished their half of
the town to Baybars in exchange for a peace treaty covering Chastel Blanc
and Tortosa. Ibn al-Furat's account of this arrangement is interesting in
that it mentions the presence of a tower at Jbala, which could either have
been similar to that erected by Bohemond VI after he captured neighbouring
Latakia around 1261, or could have been built by the Templars and the
Hospitallers themselves.19 Alternatively, it may have been Jabala's fortified
Roman theatre, which was presumably the 'strong castle' Willbrand of
Oldenburg referred to when he passed by more than half a century earlier.20
In the kingdom of Jerusalem, the fortified mill at Doc represents another
site whose Templar ownership can be confirmed by studying Hospitaller
documents. Doc was located along the same river as Recordane, which
belonged to the Hospitallers, and as a result both mills became the object
of a fierce argument between the two Orders when it was proposed to build
a new dam across this vital waterway. This dispute lasted for several years
until it was finally settled in 1235.21 Hospitaller records also suggest that the
Templars occupied the Red Tower, situated to the south east of Caesarea,
even though the abbey of St Mary of the Latins (the actual owner of the
site) had originally rented out this small agricultural centre to the Hospitallers in an agreement dating from iiSp.22 This situation had probably come
about at the time of the Third Crusade, when Templars accompanying
Richard I may have garrisoned the tower without even consulting its official
tenants. Subsequent references to this stronghold indicate that it may not
have been returned to the Hospitallers until as late as 1248. Without these
scraps of evidence we would not have known about a Templar presence
on the Sharon plain lasting more than half a century!23
THE M I L I T A R Y O R D E R S
83
Elsewhere in the south the Templars seem to have been just as badly
affected by Saladin's conquests of 1187 as the Hospitallers. In the twelfth
century they held a number of castles between Jerusalem and the coast
whose role it was to secure this region for pilgrims travelling to and from
the holy city. But we have already seen that, with the exception of Casal des
Plains, which was only reconstructed briefly during the Third Crusade itself,
this chain of strongholds was never rebuilt.24 It is hardly surprising that
other Templar properties to the east of Jerusalem which had once protected
pilgrims visiting Jericho and the River Jordan were also lost for good in
nS/.25 To the south, the twelfth-century Templar fortress of Gaza had a
similar history to that of Casal des Plains, being lost after the battle of Hattin,
restored briefly by Richard I in 1192 but demolished under the terms of the
peace treaty with Saladin later that year.26 During the early 12405, at the
same time that plans were afoot to rebuild the Templar fortress of Latrun
near Jerusalem, the Order also hoped to restore Gaza. Clearly there was a
strong desire at this time to recreate the Order's former territories in the
south and strengthen crusader rule in Palestine, but the battle of La Forbie
in 1244 and the subsequent reconquest of the area by the Egyptians quickly
put paid to the entire project27 Like the Hospitallers, the Templars seem to
have had more luck in areas which were not as exposed as Gaza or as far
inland as Jerusalem. Around 'Adit, for example, Templar lands were actually
expanding during the thirteenth century as former secular lordships such
as Caymont, a castle and village which had passed to the Order by 1262,
were gradually bought up.28 Many of these lands had to be given up to the
Mamluks in a peace treaty arranged in 1283, again reflecting the rapid decline
of 'Atlit's status during the final decade of its existence.2
Apart from holding their own castles, the three Orders also helped to
garrison the major Christian cities located along the coast. At Acre, the
Teutonic Knights were expected to maintain and garrison a section of the
ramparts near the Accursed Tower, including the gate of St Nicholas. These
defences had been granted to the Order in H93.30 A very similar arrangement
was made with the Hospitallers soon after the Third Crusade, and in 1291
all three Orders fought valiantly in defence of the city.31 Outside Acre, it
has already been noted that the Teutonic Knights helped garrison Tripoli,
whilst another section of this city's ramparts which was badly damaged
during the siege of 1289 was held by the Hospitallers.32 Elsewhere along the
coast, the Hospitallers shared the burden of defending Sidon, long before
this lordship had passed out of baronial control, and guarded another tower
at Jaffa, which was likewise held by an individual lord.33 In addition, the
Teutonic Knights held two towers flanking Caesarea's town wall, and from
1229 onwards Frederick II relied on this Order to contribute to the defence
84
UNKNOWN CRUSADER
CASTLES
of Jerusalem.34 Even though they were royal cities, all three Military Orders
also carried out repairs on the fortifications of Acre and Tyre when an
imminent Mongol attack was feared in I26o.35
Such incidents highlight the importance of the Military Orders, and the
extent to which even the most powerful secular rulers relied on their help.
By the second half of the thirteenth century they were virtually alone in
having the financial resources needed to maintain large fortifications. The
fact that they often repaired or strengthened castles shortly after they had
acquired them suggests that many secular lords could no longer prevent
their fortifications from falling into disrepair. At Beaufort, for example, the
Templars constructed a whole new citadel opposite the older castle in the
space of just eight years, and at Arsuf urban fortifications erected by the
Hospitallers soon after they took over the site were said to have enraged
Baybars.36 Even after they had completed Saphet, the Templars spent 40,000
bezants each year on the upkeep of this fortress,37 whilst the vast new
defences erected at Tortosa and Chastel Blanc from the mid twelfth century
onwards, plus those built by the Hospitallers at Crac des Chevaliers and
Margat until c. 1204, must have been just as expensive to maintain. The
money required for such projects was largely drawn from the Orders'
extensive estates in Europe and the Holy Land, and clauses in the Rule of
the Templars suggest that it was then channelled into central funds specifically set aside for castles. Many strongholds, including Saphet, were also
paid for by pilgrims and crusaders.38
Although there were greater signs of strain from the 12605 onwards, it
also seems that the Military Orders often had more garrison troops at their
disposal than the nobility, particularly in early thirteenth-century Syria. The
remains of vast cisterns, undercrofts and storerooms at sites like Crac,
combined with Willbrand of Oldenburg's claim that Margat contained
enough supplies to withstand a five-year blockade, indeed suggest that the
Orders had the resources to prepare their major Syrian castles for almost
any emergency that might arise.39 Both the Templars and the Hospitallers
also enforced strict rules regarding the use and defence of castle entrances to
prevent spies and traitors from gaining access.40 This kind of meticulous
planning and discipline, as well as the obstinate refusal of the Orders to give
up the fight even as they were being driven out of Acre, suggests that the
Hospitallers, Templars and Teutonic Knights were more skilled, disciplined
and better equipped than any other Latin troops in the Holy Land.
The forward planning and sheer commitment of the Military Orders can
be contrasted with the inefficiency and poverty of many lay rulers or lords.
In 1268, for example, troops and citizens from Antioch sheltering in the
citadel were quickly forced to surrender to Baybars because, unlike Margat,
THE M I L I T A R Y O R D E R S
85
this stronghold was badly stocked with provisions and 'had neither enough
water nor enough mills' to feed everybody.41 Eighty years earlier, Saladin
had also been able to capture the secular fortress of Saone, situated along
the land route between Latakia and Antioch, in a matter of three days, even
though Margat, Tortosa, Chastel Blanc and Crac des Chevaliers had all held
out. Saone's defences were just as impressive as (or perhaps even stronger
than) those of Margat and its other neighbours to the south, but the fact
that it was held by an individual lord seems to have contributed to the
weak and demoralised state of its garrison.42 This incident proves that
the successful defence of a fortress ultimately had more to do with the
preparedness and mental state of its garrison that it did with the scale of
its fortifications. It also shows how the Templars and the Hospitallers became
virtually the only Franks to hold any inland castles in the county of Tripoli
after the battle of Hattin. Indeed, it would not be an exaggeration to say
that from 1188 onwards this crusader state owed its very existence to these
two Orders.
In the kingdom of Jerusalem the situation did not become quite so
extreme until the mid thirteenth century, when entire lordships like Sidon
were regularly handed over to the Orders. However, it would be misleading
to assume that barons here were in a stronger position than their neighbours
further north. The lords of Jaffa, for example, were only able to hold on
to their castle by making regular appeals to the papacy for more
money.43 We have also seen that during the early 12505 Louis IX refortified
Jaffa whilst staying in the Holy Land. According to Joinville the papal legate
Odo of Chteauroux, who was helping the king with this project, spent a
staggering 30,000 livres rebuilding just one particular gateway and adjoining
curtain wall. This implies that Louis himself, who erected a further two
gates and all the other ramparts, must have spent at least three times as
much money; an amount which would have been well beyond the means
of the lord of Jaffa himself.4* Similarly, it has been shown in recent years
that, by the time Caesarea fell to the Muslims in 1265, this barony 'had
disintegrated so far that the lord's influence in many parts of his own
lordship must have been minimal'. Again, this was because many lands and
properties had been sold off to the Military Orders or other institutions in
order to raise revenues and pay off ever increasing bills.45
The pattern of baronial ownership in the east also confirms that the
further inland castles lay, the more exposed they became to Muslim incursions. As a result, they were costlier to defend and more likely to get sold
off to one of the Military Orders. Jaffa, Beirut and Caesarea were the only
major strongholds in the kingdom of Jerusalem which were still in secular
hands when they were captured from the Muslims. Likewise coastal Gibelet,
86
Nephin and Maraclea were all baronial castles throughout the thirteenth
century, whilst Bohemond VI kept Latakia for himself after he recaptured
it in 1261. By contrast, the only inland stronghold held by the counts of
Tripoli during this period was Akkar.46 It is also important to remember
that virtually all the new defences built by individual lords during the
thirteenth century were situated along the coast. These included Bohemond
VI's towers at Latakia and (possibly) Jbala, plus Bartholomew's tower at
Maraclea. Early in the thirteenth century John of Ibelin had also been
granted Beirut, and subsequently carried out extensive improvements there.
These probably concentrated on the building of a new outer wall, for
although the castle itself no doubt needed repairs, there is no evidence that
Saladin had demolished anything other than the urban defences of the
town.47 Inland the only comparable building project undertaken by a secular
lord was Odo of Montbliard's completion of a new citadel at Tiberias,
although it has already been noted that this too was a coastal fortress in
the sense that it stood on the shores of Lake Tiberias.48
To conclude this chapter, it is interesting to ask whether fortifications
belonging to the Military Orders ultimately helped or hindered the crusader
states in their struggle with the Muslims. The famous military historian
R. C. Smail wrote that central authority in the Holy Land 'was progressively
weakened ... because lands, castles, powers and rights over men had continually to be conceded to Orders, which were not wholly part of the feudal
structure'.4^ This certainly seems to have been the case in the county of
Tripoli, where we have seen that the Templars and Hospitallers built or
acquired massive fortifications which were then used to administer local
lands and wage war with the Muslims with hardly any reference to outsiders.
In short, these lordships became virtually independent enclaves. It has also
been shown that both these Orders were prominent in some of the internecine struggles which took place during the thirteenth century, such as
the Templars' defiant behaviour toward Frederick II at Acre and 'Atlit. It
could certainly be argued that episodes of this kind undermined the power
of local rulers and squandered military resources which could otherwise
have been directed against the Muslims. But we have also seen that the
Military Orders often played a vital role in the defence of the crusader
states. The strength of castles like Margat and Tortosa prevented Saladin
from wiping out the Latin East in 1188, and thereafter they ensured that
the Holy Land survived for another hundred years. Ultimately, it seems fair
to conclude that the fortifications of the Military Orders did far more to
help the Christian cause than to hinder it.
PART TWO
Warfare
Thanks to its physical isolation, the kingdom of Cyprus was by far the most
successful of all the crusader states in the eastern Mediterranean, enjoying
long periods of peace and tranquillity at a time when most Latin territories
on the mainland were gradually being reconquered by the Greeks, Muslims
or Bulgars. This notwithstanding, two distinct types of warfare threatened
to undermine Cypriot stability from time to time.
First, there was always the danger that external forces would attack the
island. Most incursions of this kind were carried out by pirates, who
appeared sporadically from the 11903 onwards, and were particularly active
during the mid fourteenth century. Other, more politically motivated raids
were also organised by the Muslims of Egypt in 1220 and 1271, by the
Genoese in the early fourteenth century, and by the Turks in the 13605.
These were all temporary attacks, however, and although they could be
damaging on a local scale, they were not as alarming as the threat of a
full-scale foreign invasion.1
It was of course just such an attack which had brought Cyprus under
Latin rule in the first place, for in the summer of 1191 King Richard I of
England landed at Limassol whilst travelling east to join the Third Crusade
besieging Acre. After he arrived, he ousted the self-declared emperor Isaac
Komnenos, who had broken away from Byzantine authority in the mid
iiSos. Richard probably only saw Cyprus as a means of financing his
campaign in the Holy Land, and once he reached Acre he sold it first to
the Templars, then to Guy of Lusignan, the former king of Jerusalem, who
received the island as compensation for losing his old title to his rival
Conrad of Montferrat. As a result, Cyprus came under the control of the
Lusignans, who also completed the task of suppressing the native population,
encouraging more Prankish settlers and setting up a Latin church.2
After these dramatic changes, the likeliest threat of a new external invasion
came from the Byzantine empire, which had already made an unsuccessful
attempt to reconquer Cyprus back from Isaac Komnenos in 1187. This
possibility was quickly eradicated by the political turmoil which afflicted
the empire both before and during the Fourth Crusade, not emerging again
as a realistic threat until the reign of the Nicaean emperor Michael VIII
90
WARFARE
91
MAP 3. Cyprus
WARFARE
93
Cyprus in the sense that they disrupted Muslim trade and may therefore
have brought a temporary halt to the mid fourteenth-century economic
decline of Famagusta, Peter paid for them by imposing crippling taxes on
his own subjects. As a result he was eventually assassinated in 1369 by a
faction of Cypriot nobles, who then placed his young son Peter II (1369-82)
on the throne. This crisis had still not been fully resolved at the time of
the Genoese invasion in 1373, when Peter's widow Eleanor of Aragon initially
allied herself with the invaders against his brothers John and James of
Lusignan, who had almost certainly been involved in Peter's assassination.13
These then were the principal clashes which threatened to undermine
Cypriot stability, and it will be shown that virtually all castles built or
occupied on the island during this period were intended to deal with at
least one of them. Some types of warfare, however, were likelier to break
out than others. The Cypriot uprisings of 1191 and 1192, for example,
indicate that the threat of internal rebellion was at its greatest during the
earliest years of Prankish rule. On the other hand, the wide chronological
distribution of clashes involving fellow westerners, which broke out in 1228,
1306 and 1369, suggests that the risk of such confrontations did not alter
dramatically one way or the other. Much the same can be said for attacks
by pirates, although raids of this kind did increase markedly during the
middle years of the fourteenth century. The most dramatic change of all
in Cypriot fortunes clearly took place after 1291, when the threat of a major
external invasion, either by the Mamluks or later by the Genoese, grew
enormously. The risk of such an attack had barely existed during the first
hundred years of Prankish rule, particularly once the Fourth Crusade had
eliminated any chance of a Byzantine counter-offensive.
After 1291, the Lusignans had to face the possibility of an attack by the
kind of vast Mamluk armies which had wiped out Latin territories on the
mainland. Indeed, it has already been noted that from the reign of Baybars
(1260-77) onwards Egyptian invasion forces in the Holy Land had rarely
numbered fewer than 12,000 men. At this time Mamluk armies also improved
dramatically in terms of quality as well as quantity, for they increasingly
relied on professional soldiers, and in particular highly skilled siege engineers.
These men had at their disposal far larger and more accurate catapults (or
trebuchets) than their twelfth-century predecessors, and by using these
powerful weapons, in combination with mines and ladders, they took much
of the credit for reducing even the greatest Prankish strongpoints on the
mainland.14 By the late fourteenth century the Genoese were scarcely less
intimidating, particularly because of their great naval strength, the only
significant weakness in the Mamluk war machine. According to one source,
the invasion fleet of 1373 was made up of 'forty-three galleys plus many
94
other vessels', and 'there were brought together in the Genoese army 14,000
troops and more ...'15
These impressive totals can be contrasted with the kind of troop numbers
available to the kings of Cyprus. Although exact figures are hard to establish,
it seems that the 200 knights and 500 infantry brought to Acre by Henry
II in 1291 were roughly equivalent to the maximum field army which could
be raised without stripping Cypriot castles of their core garrisons.16 These
forces would have been adequate for dealing with pirates who had landed
on Cyprus. They would also have sufficed against any possible uprisings by
the local population, for accounts of the Greek rebellions which took place
in 1191 and 1192 suggest that their participants were badly organised and
lacking in sophisticated weaponry such as siege engines or heavy cavalry
troops.17 It is also clear that internal clashes between rival western factions
normally involved just a few hundred troops. At the battle of Agridi (1232),
for example, the Ibelins only had 233 mounted troops. Although they were
supposedly confronted by 2000 imperialist horsemen, this latter figure may
have been an exaggerated approximation given by a source favourable to
the Ibelins in order to play up the scale of their victory.18
After 1291 the Lusignan kings suddenly found themselves having to deal
with a potential invasion by more than 10,000 Mamluk (or later Genoese)
soldiers. Inevitably, these factors had a profound effect on castles. Although
the Franks built new fortifications on Cyprus before 1291, these were not
particularly extensive, and added little to the already existing network of
Byzantine strongpoints. After the fall of Acre, however, a massive building
programme, later renewed in response to the growing Genoese threat,
resulted in the construction of many new fortifications and urban defences.
This process continued well into the sixteenth century, for from 1489
onwards Cyprus came under the control of the Venetians, who upgraded
many medieval strongpoints with new ramparts designed to withstand the
artillery of the Ottoman Turks.19 As far as Cypriot fortifications were
concerned, 1291 was actually a far more significant date than 1191.
Military Architecture
During the crusader period, the key fortifications on Cyprus were to be
found at Paphos, Kyrenia, Limassol and Famagusta, situated on the coast,
at Nicosia, the inland capital of the kingdom, and at St Hilarin, Buffavento
and Kantara, three mountain strongholds in the north. This unusually small
total, plus the fact that Cyprus often remained at peace for many decades
at a time, makes it relatively easy to trace the few changes in military
architecture which took place there between 1191 and 1374. However, thanks
to the major alterations carried out by the Venetians after they gained
control over Cyprus in 1489, it is no longer always possible to study the
actual physical remains of these developments. As a result, a combination
of both archaeological and historical sources must be used in order to assess
what the island's defences were like before the fifteenth century.
Inevitably, this applies most to the very earliest period of Prankish rule,
for at several sites it is only by studying contemporary descriptions of Richard
I's invasion that we can ascertain what fortifications existed there in the late
twelfth century. This is certainly the case at Limassol, which was the first
Cypriot settlement occupied by Richard in May 1191. It seems that the king's
initial intention in stopping here had merely been to recover some sailors
shipwrecked nearby in a violent storm. However, Isaac Komnenos's refusal
to free these people, combined with Richard's realisation that Cyprus could
prove a useful source of wealth and supplies, quickly led to confrontation
between the two men, and on 6 May the English defeated the Greeks along
the shore and occupied their castle. These events prove that a Byzantine
stronghold existed at Limassol before 1191, although the apparent ease with
which the English took control of it suggests that it cannot have been
particularly large.1
After the fall of Limassol, Isaac Komnenos fled north and east away from
Richard's advancing troops. Along the way he sought refuge briefly at
Nicosia, suggesting that it too was defended by some sort of strongpoint;
a conclusion which can be confirmed by looking at the Cypriot rebellion
of 1192, when the Templar garrison of the city also took shelter there.2
Thereafter Isaac retreated to the castle of Famagusta, but he quickly realised
that it 'would not be safe to withstand a siege', and so he 'hid himself in
96
UNKNOWN CRUSADER
CASTLES
98
U N K N O W N CRUSADER CASTLES
St Hilarin
100
MILITARY ARCHITECTURE
IDI
Kantara
102
MILITARY ARCHITECTURE
1O3
104
Genoese admiral Peter Campofregoso consulted his men and 'gave orders
that they should raise the height of the walls wherever they were low, and
he tried to bring the sea all around the place [i.e. the citadel] so as to make
it an island'.37 In 1380, during one of the many clashes between Genoa and
Venice, wooden towers were also added to Famagusta's sea walls in order
to make them taller.38 Today, it is only possible to identify one tower along
the city's more recent Venetian ramparts which seems to date from this
period, and it may therefore represent the sole visible architectural evidence
of new Genoese defences at Famagusta during the late fourteenth century.39
While these changes were taking place at Famagusta, the Franks were
also busy constructing new urban defences at Nicosia. Some earlier fortifications of this kind may have been erected under Henry II after 1291, but
more substantial construction work began during the reign of Peter I
(1359-69).4 This had clearly not been completed properly before the accession of his son Peter II (1369-82), who found that 'the walls were very
strong; but they were low, and he sent word to the country round, and
men came together and built them up with earth and stones; and they dug
out the ditch and constructed 133 platforms to fight from in addition to the
towers'.41 These improvements were continued during the late fourteenth
century by Peter II's successor, James I (i382-98).42 The overall sequence
of events at Nicosia may have been that sporadic efforts were made to build
city walls between 1291 and 1359, but because of the city's inland location
this project was not pursued as vigorously as the more urgent strengthening
of exposed coastal sites like Kyrenia, and was not finally completed until the
period of growing Genoese aggression and Turkish raiding after the 13505.
Nicosia's citadel also underwent many radical changes during this period.
By the 13505 it is difficult to ascertain what state the original castle built
between 1192 and 1212 (and possibly extended by Henry II) was in. If this
stronghold still existed, it may have been demolished by Peter I, who was
responsible for establishing a new citadel on a hill on the outskirts of the
city. This replacement basically consisted of a 'strong and impressive' tower,
whose different floors, including a dungeon in the cellar, were connected
by ladders. It was also surrounded by a large moat, which had been excavated
by slaves who 'dug the earth all day, and carried it out on their backs'.43
The Margarita Tower, as this citadel was known, never even reached completion before Peter II ordered it to be demolished, along with several other
buildings which were in the way. A new, larger castle replaced it. This fortress
lay near the Paphos gate, in the west of the city, and must have been far
larger than its predecessor, for it contained royal apartments. It was built
by Genoese prisoners of war in the space of just ten months, using recycled
masonry from the Margarita Tower and 'walls in the town which were of
MILITARY ARCHITECTURE
105
1O6
U N K N O W N CRUSADER CASTLES
MILITARY ARCHITECTURE
lOJ
Kyrenia must also have had more in common with Syrian strongpoints
than they did with their immediate neighbours on Cyprus. Most notably,
the elaborate gateway, new towers and extensive shooting galleries which
Henry built at Kyrenia, using high quality ashlar masonry, must have turned
this fortress into a Cypriot version of 'Atlit or Crac des Chevaliers. However,
the fact that the walls of both Nicosia and Famagusta later had to be
heightened using earth, stones or even wooden stockades suggests that none
of the urban fortifications on Cyprus ever attained quite the same strength
as the vast multiple ramparts of cities like Acre and Tyre.62
Finally, and on a smaller scale, the castrum castle of Sigouri can be
compared with that of Goliath in the county of Tripoli. Goliath was probably built by the Hospitallers after they acquired the site in 1127, whereas
Sigouri was constructed by James I during the i39os.63 This confirms the
suggestion that military architecture evolved according to financial and
geographical necessity rather than scientific discovery. It has been written
that 'the castrum form of the castle was exclusive to the twelfth century'.64
But the construction of Sigouri almost two centuries later indicates that it
would be wrong to assume that this extremely simple design was immediately rejected by the Latins as soon as they started to build more complex
castles such as 'Atlit. In reality Sigouri's design was eminently suited to its
exposed surroundings, whilst its size must have made it large enough to
withstand minor Genoese incursions, yet small enough to be affordable for
a relatively poor ruler like James I. Cypriot fortifications changed according
to specific military, political or financial needs rather than any clearly
defined architectural discoveries.
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
1O9
break out. Clearly, the new Prankish towers erected at Limassol, Paphos,
Famagusta and Nicosia between 1192 and 1212 were intended to achieve this
goal. These structures would have been perfectly suited for the task, for
they were sturdy enough to withstand attacking rebels, but small enough
to be built quickly and economically, and to be garrisoned by relatively few
men. Larger strongholds would have been a waste of resources against an
enemy lacking any kind of siege weapons or organised structure. This
incidentally suggests that the Byzantine predecessors to these towers really
had been dilapidated, if they could not even be relied on to keep out crowds
of rioters. Lastly, the fact that (with the probable exception of Kyrenia) no
urban fortifications existed on Cyprus prior to 1291 also tends to confirm
that at this stage the Latins were far more concerned about internal than
external security.
Returning to the Prankish towers themselves, further evidence of their
intended function may be gleaned from their location, for with the exception
of Paphos they all appear to have been situated at the centre of their
respective settlements. Although it dates from the late sixteenth century and
may therefore be unreliable, one source even claims that at Nicosia the
Byzantine castle stood near the market place. If the Franks built their new
tower on or near such a central site it must have dominated the entire
city.6 It would have become a powerful symbol of the new regime, acting
as a constant reminder to the Greeks that resistance against their new
overlords was futile. Hence the function as well as the appearance of these
structures closely resembled that of other Prankish towers on the mainland,
where many small fortifications of this type, dotted across Syria, Palestine
and later Greece, protected Latin settlers and their agricultural interests
against local criminals or rebels.7
The subsequent history of Cyprus can be used to highlight the success
of the new towers built by Guy of Lusignan, his brother and his nephew,
for there were no further widespread rebellions on the island until the
fifteenth century. Although the risk of such incidents must have receded
considerably after the 11905, it did not disappear completely. In 1359, for
example, the papal legate Peter Thomas attempted to convert Orthodox
Greeks at Nicosia to Catholicism but the meeting ended in a riot, with
many locals shouting 'death to the legate', who was only saved by the swift
intervention of royal troops.8 At about the same time the Venetians also
warned Peter I that Greeks rebelling on Crete could try to contact their
co-religionists and supporters on Cyprus.9
As had been the case in the Holy Land following Saladin's victory at
Hattin, the native population also took advantage of Prankish military
setbacks. During the Genoese invasion of 1373, for example, the Franks were
HO
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
111
112
U N K N O W N CRUSADER CASTLES
seems, have held out almost indefinitely. If this had been possible they
could have waited until supply problems, a sense of despair or the arrival
of Richard Filangieri's reinforcements had forced their opponents to raise
the blockade.
During the first six months of 1232 these were the exact circumstances
which led to the end of Hohenstaufen power on Cyprus. Aimery Barlais
and Richard Filangieri's Lombard troops failed to capture St Hilarin and
Buffavento from the Ibelins, so that when John of Ibelin himself arrived
from the mainland with an army of supporters gathered at Acre, the
imperialists found themselves fighting on two fronts. If they had not already
committed vital troops and resources to besieging St Hilarin they would
perhaps have stood a better chance at the battle of Agridi in June. In the
wake of the battle, the imperialists had no choice but to retreat to Kyrenia,
leaving their garrison at Kantara completely cut off and unable to do
anything but surrender. Once it became apparent that the Lombards at
Kyrenia were not going to receive any more help from Frederick II, they
were forced to do the same.19
These events confirm that two key factors favoured the Ibelins during
the civil war of 1228-33. First, the imperialists relied too much on foreign
troops sent by Frederick II whereas the Ibelins raised their armies locally,
so that although they did have trouble finding enough troops, they could
at least gather reinforcements more swiftly than Frederick's baillis. This
factor proved decisive in 1230 when St Hilarin and Kantara were starved
into submission, and again between 1232 and 1233, when the Ibelins succeeded in relieving St Hilarin but the imperialists failed to assist Kyrenia
in time. Secondly, it has already been noted that the troop numbers involved
in this conflict were relatively small, so that the battle of Agridi, for example,
was only fought between a few hundred mounted troops. These figures
inevitably resulted in swift battles and lengthy sieges, for it was far easier
for a small army to defend Kyrenia and the mountain castles than it was
for one to attack them. The fact that John of Ibelin enjoyed much greater
political power and popular support on Cyprus made him the clear favourite
in a war of attrition of this kind.20
The importance of Cypriot castles to internal stability came to light once
more during Amaury of Tyre's reign as 'governor' between 1306 and 1310.
As soon as Amaury, aided by the vast majority of the nobility, had deposed
King Henry II, he immediately 'sent castellans and bailies to all the towns
and castles of the kingdom of Cyprus', to prevent them from falling into
royalist hands and to guarantee the loyalty of their garrisons.21 Thereafter
Amaury's harsh rule and shaky legitimacy caused his popularity to go into
decline, and eventually brought about his murder in 1310. It was not just
FORTIFICATIONS
AND I N T E R N A L S E C U R I T Y
113
114
U N K N O W N CRUSADER CASTLES
fellow Latins were so unusual. The only two clashes of this kind which did
occur ended in victory for the Lusignans thanks to the strength of their
castles. Thus Henry I, greatly aided by his powerful vassal John of Ibelin,
triumphed over the imperialists because they lacked the resources to hold
on to the island's largest fortresses after Frederick II's departure in 1229,
whilst Henry II ultimately defeated Amaury of Tyre's faction because it
could not hope to recapture Kyrenia and Famagusta in 1310.
Whilst all Cypriot castles clearly enhanced royal power, their role varied
somewhat depending on their size and location. The isolated mountain
fortresses of St Hilarin, Buffavento and Kantara served as final refuge
points which could be relied on to tie down large numbers of attacking
troops and resources. Although they may also have helped to suppress the
native population during the 11905, they were far too remote to intimidate
any Greek settlements directly and therefore had more to do with national
defence than the Prankish towers constructed at Limassol, Paphos, Nicosia
and Famagusta. These latter strongpoints could not hope to withstand the
kind of large catapult deployed by Anceau de Brie against Kantara in 1230,
but they could maintain law and order on a local scale and acted as highly
visible symbols of the new Prankish regime. Kyrenia, on the other hand,
enhanced both local and national defence, for it overlooked a relatively
large town but also proved too powerful even for the most determined
besiegers. Similarly, Famagusta played a relatively small role in the civil war
of 122833, but in 1310, after it had been provided with extensive new urban
fortifications, the support of its citizens for Henry II helped determine the
course of Cypriot history.
10
Il6
who were at that time campaigning in Egypt. This attack supposedly resulted
in the death, capture or dispersal of at least 13,000 crusaders anchored
there.4 More than fifty years later, in 1271, Baybars made another attempt
to raid the town so that the Franks' attention would be drawn away from
his campaign against the castle of Montfort in Galilee. This latter incursion
ended in disaster when the Muslim fleet was shipwrecked on the coast; a
vivid illustration of the incompetence of Egyptian naval forces at this time.5
Even if the attack had been successful, it is clear that both this raid and its
predecessor were isolated incidents which had far more to do with events
on the mainland than on Cyprus itself.
Incursions by pirates were equally sporadic, occurring in 1302, when
corsairs from Rhodes appeared,6 and more seriously during the reign of
King Hugh IV (1324-59), who had over a hundred prisoners hanged because
'these men were thieves, pirates, robbers, who went pillaging and murdering,
and did great damage'.7 In the early fourteenth century more deliberate
attacks were carried out by the Genoese, who raided the coast around
Paphos in 1312 and 1316 because a loan they had made to the usurper
Amaury of Tyre was not fully repaid by Henry II after he regained the
throne in 1310.8 Shortly before their conquest of Famagusta, the Genoese
also targeted both Paphos and Limassol, where 'they landed and burned
the houses ... and did much damage'.9 Some years earlier the naval
campaigns of Peter I (1359-69) against the coast of Asia Minor also led to
retaliatory raids by the Turks, who 'went and pillaged the country' in the
13605.10
During such attacks, Cypriot fortifications were expected to fulfil two
basic functions. First, strongpoints located near the coast or on high ground
could act as look-out posts. Consequently their garrisons could give early
warning both to civilians wishing to take evasive action and to naval or
other troops hoping to confront the intruder. Inevitably the three mountain castles, which enjoyed extensive views over the northern, eastern and
southern coasts of Cyprus, were most significant in this respect. Some of
these strongholds were also intervisible both with each other and with
surrounding settlements such as Kyrenia. They could therefore communicate
using fire or smoke signals. Buffavento lay at the heart of this network, and
here 'they kept a look-out every night, and as soon as they spotted ships
at sea, they would signal with fire or torches to the town of Nicosia and
the castle of Kyrenia'.11 This raises the possibility that the isolated uppermost towers already described at Buffavento, Kantara and St Hilarin were
normally used as watch towers rather than as final refuge points for the
garrisons of these castles.12
On a less spectacular scale, the harbour location of the Prankish towers
ll/
Il8
UNKNOWN CRUSADER
CASTLES
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
119
The massive new defences built at Famagusta and Kyrenia were expected to
fulfil the same defensive role as earlier Prankish strongpoints on the mainland, such as Acre, Tyre, Saphet or Tortosa. The various shooting galleries,
posterns, towers and ramparts added to Kyrenia meant that this fortress
came to resemble many of the larger Prankish castles on the mainland.21
Such sites were constructed on a vast scale so that they could resist Muslim
invasion forces numbering 10,000 or even 20,000 troops. Similarly, the large
ashlar blocks used during the rebuilding of Kyrenia were clearly chosen in
response to the advanced Mamluk artillery and sapping techniques which
the Franks had come to dread on the mainland.22
The sheer scale of the enemy which now confronted Cyprus also made
it futile to defend the island as a whole, or even prevent attackers from
landing in the first place, as was sometimes possible with pirates. The
maximum field army which could normally be raised by Cypriot kings
probably only contained around 200 knights and 500 infantry.23 Like their
predecessors on the mainland, the Lusignans therefore made no effort to
defend their borders but concentrated their outnumbered forces inside a
small number of extremely powerful strongpoints, even if this meant
abandoning control over the countryside temporarily. Famagusta and
Kyrenia were deliberately strengthened to create such strongpoints, alongside
the already impregnable mountain fortresses of Kantara, St Hilarin and
Buffavento. During the civil war of 1228-33, as already shown, these castles
could be relied on to hold out almost indefinitely provided that they were
well supplied and garrisoned, forcing an attacker to devote large numbers
of men and resources to a siege which could last for many months or even
years. The relatively small size of Cyprus, combined with the long lines of
communication between it and the mainland, would have put a severe
strain on a large Mamluk army living off the land for any length of time.
The Lusignans therefore hoped to withstand an Egyptian attack in much
the same way that the kingdom of Jerusalem had survived the Khwarizmian
invasion of 1244, when lack of equipment, supplies and patience prevented
the Khwarizmians from capturing key strongholds and eventually obliged
them to withdraw.24
During the mid fourteenth century similar thinking probably induced
Peter I to carry out considerable improvements on the urban defences of
Nicosia. The Margarita Tower, Peter's replacement for Nicosia's original
Prankish tower, was built on a strategic hill on the outskirts of the city
rather than at its centre. It therefore had much more to do with guarding
the capital against external attackers than suppressing the local Greeks.25
These defences were constructed because Peter's aggressive foreign policies
had renewed the threat of a serious Muslim attack on Cyprus. His naval
12O
U N K N O W N CRUSADER CASTLES
offensives against the coastal settlements of Asia Minor help to explain why
there were so many retaliatory Turkish raids at this time, whilst Peter's
crusade against Egypt in 1365, which resulted in the temporary occupation
and partial destruction of Alexandria, revived fear of a retaliatory Mamluk
invasion. These events probably also contributed to Peter's assassination in
1369, for the king had already spent vast amounts of money on his ambitious
naval campaigns, and now had to raise yet more taxes from his resentful
vassals in order to pay for new defences against the very enemies which he
himself had provoked.26
In the end it was not the Mamluks but the Genoese who put the Lusignans'
elaborate defence plans to the test. In 1373 they invaded Cyprus in the
hope of turning the island into an exclusive Genoese domain capable of
controlling the lucrative trade routes of the eastern Mediterranean. At first
Cypriot hopes that an invasion of this kind would be defeated in a series
of lengthy sieges failed miserably, for in October the Genoese captured
Famagusta through stealth, so that their force of 14,000 men and more than
forty galleys, far from getting bogged down, remained relatively unscathed.27
Not long afterwards a Genoese contingent also succeeded in capturing
Nicosia, where Peter I's new walls were not yet tall enough to withstand a
siege.28 Once inside Nicosia the Genoese used these defences for their own
benefit. Having realised that they lacked the resources to occupy the entire
city, they only garrisoned a stretch of the ramparts 'from the Market Gate
to the Tower of St Andrew, and they made the walls higher and held the
place in great force. And the tower which stands near the Market Gate they
filled with earth and stones and made it like a castle'.29 These fortifications
protected the Genoese against two unsuccessful counter-attacks by Prankish
forces from Kyrenia. They also enabled the invaders to intimidate the entire
city, which succumbed to a long period of looting and street fighting
involving both the Genoese and the local citizens.30
The Genoese intended to conquer the rest of the island and proceeded
to fight their way north through the Kyrenia mountains, whose passes were
guarded by Bulgarian mercenaries from St Hilarin and Kyrenia. It was at
this stage that Peter Fs widow, Eleanor of Aragon, having realised that the
Genoese were far more interested in taking control over all of Cyprus than
in helping her to avenge her husband's murder, fled from their army and
rejoined the Cypriot cause.31 Once the Genoese reached Kyrenia itself, 'the
bridges were raised and the gates nailed up', and the attackers' advance was
halted in exactly the way that had been hoped.32
Despite several weeks of intense fighting, the Genoese failed to make any
progress against such a massive obstacle, whose defenders successfully bombarded or set fire to their opponents' siege weapons in the same way that
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
121
the Lombards had done in 1232-33. At one stage many Genoese troops were
killed because they charged onto a drawbridge which had deliberately been
left open, and was counterpoised in such a way that anyone standing on
it could be tipped into the moat.33 This contraption shows just how carefully
the defences had been prepared at Kyrenia and is reminiscent of the kind
of elaborate devices used to protect Prankish castles on the mainland, such
as the labyrinthine main entrance at Crac des Chevaliers.34 At Kyrenia
fighting continued after a brief period of inconclusive talks. But still the
Genoese were repelled, despite constructing a vast wooden platform which
was lashed between two galleys and used to fire into the castle from off
shore. As a result, most of the attackers withdrew to Nicosia, and a ceasefire
was agreed which effectively ended the siege in March 1374-35
These events marked a turning point in the war, for the Genoese knew
that 'if they did not have the castle of Kyrenia, they would not be able to
hold Nicosia, or the rest of the island'.36 They could not afford to bypass
Kyrenia, which was defended by the constable, John of Lusignan, and a
large contingent of foreign mercenaries, all of whom were capable of
harassing or even severing Genoese supply lines from Famagusta and
Nicosia. Indeed, in the course of the siege Bulgarian troops hiding in the
Kyrenia Mountains did just that, capturing food and siege equipment being
transported from the capital.37
This would not, of course, have been a problem if Kyrenia had been
captured, for then the Genoese would have held another secure port on
the Cypriot coast to complement their newly acquired base at Famagusta.
Indeed, bearing in mind that Genoa was one of the greatest naval powers
in the Mediterranean, and that her initial intention in attacking Cyprus had
been to dominate seaborne trade there, it made sense to attack Kyrenia
before any of the neighbouring mountain castles. By the same token, their
failure to take Kyrenia persuaded the Genoese to relinquish Nicosia, whose
low walls made it difficult to defend, and whose inland location gave it
little strategic or economic importance in its own right. A maritime power
like Genoa was just as uninterested in Nicosia as the Italian cities had once
been in Jerusalem.
After their withdrawal from Kyrenia and Nicosia, the Genoese were limited
to the city of Famagusta and its immediate surroundings, which they held
until 144-38 But the threat of further Genoese attempts to complete the
conquest of Cyprus remained, and it is in this context that the numerous
building projects by Peter II (1369-82), James I (1385-98) and Janus (13981432) should be considered. Hence the new citadel and additional urban
defences erected at Nicosia after 1374 were clearly designed to spare the
capital from another Genoese occupation, as the first one had resulted in
122
UNKNOWN CRUSADER
CASTLES
FORTIFICATIONS
AND E X T E R N A L S E C U R I T Y
123
11
F O R T I F I C A T I O N S , N A V A L W A R F A R E AND THE C R U S A D E S
125
political as well as a geographical point of view, for Cypriot troops subsequently participated in all of them.8 It should also be noted that for
much of the thirteenth century, and particularly when Cyprus and Acre
were ruled by the same man, the island's resources could still be used to
boost the weakening defences of the Holy Land, regardless of whether a
full crusade was present. As late as 1291 Henry II took an army of 200
knights and 500 footsoldiers to Acre in a final bid to save it from al-Ashraf
Khalil's besieging army.9
During these expeditions, the coastal fortifications of Cyprus were expected to provide a modicum of security and comfort for crusaders. Both
Louis IX and Frederick II, for example, may have used Limassol's castle as
a residence or as a safe place to store funds and documents. Despite the
small size of this stronghold and the naval inferiority of the Muslims, the risk
of surprise raids made it desirable for them to remain as close as possible
to this look-out post and potential shelter. During Frederick's crusade there
must have been fresh memories of the Muslim raid on Limassol harbour
which had taken place only eight years earlier.10 Similarly, it is possible that
Baybars's failed raid on Limassol in 1271, primarily launched to distract the
Latins from his campaigns in Galilee, was also intended to disrupt the Lord
Edward's crusade to the Holy Land. Baybars's raid took place in the summer
of 1271, by which time Edward had already been in Acre for some time,
but it should be noted that his brother Edmund did not arrive in the East
until September. In the meantime King Hugh III of Cyprus (1267-84),
relying on men and supplies brought from his island, had joined Edward
in the Holy Land. Baybars may therefore have hoped to intercept Edmund
or, more plausibly, sever supply lines between Cyprus and the mainland,
at the same time punishing Hugh for his involvement in the crusade. In
these circumstances the castle of Limassol, however small, must have been
a convenient stepping-stone between Cyprus, Acre and the West, and a
potential depot for vital supplies, u
Returning to the first half of the thirteenth century, it is also possible
that Paphos, the closest Cypriot harbour to Egypt, was either involved in
or had at one stage been intended as a starting point for the Fifth Crusade
against Egypt (1217-21). It may have featured in Honorius Ill's initial plans
to use Cyprus as an assembly point, in the same way that Louis IX later
did.12 In the end, these plans were shelved. Although Limassol was subsequently used to ship some supplies to Damietta from the Holy Land, it
was Acre rather than Cyprus which eventually acted as the starting point
for the campaign. But if Paphos never played any role in the Fifth Crusade,
it is difficult to understand what function the major new castle built there
by the Franks between c. 1205 and 1222 was intended to serve.
126
This fortress was situated on a hilltop overlooking the town and stood
directly above the earlier tower probably built by Guy of Lusignan around
1192. It was defended by two successive curtain walls arranged around a
central courtyard, making it far stronger than any other Prankish fortification
built on Cyprus before 1291.13 But if simple towers were deemed sufficient
to deter pirates and suppress the native population elsewhere, why did the
Franks need to construct such a large stronghold at Paphos? It has been
suggested that it was probably built by the Hospitallers between 1200 and
1204, because of its close architectural links with Belvoir, the Order's famous
concentric castle overlooking the River Jordan. Its supposed role was to
defend Cyprus against an expected Byzantine invasion, and it therefore
complemented the Templars' new coastal base at Gastria, which lay north
of Famagusta and was completed at some point before mo.14
It seems unlikely, however, that the Templars and the Hospitallers were
building major coastal fortifications at this time, partly because Gastria
amounted to little more than a walled enclosure incapable of withstanding
a major siege, and partly because the Byzantine empire itself was in such
a state of crisis during the final years leading up to the Fourth Crusade
that it was no longer able to launch a major assault on Cyprus.15 Finally,
it is odd that Willbrand of Oldenburg, who had a strong interest in crusader
fortifications and always provided descriptions of other strongpoints he saw
in the Latin East, failed to mention a large, prominent castle when he visited
Paphos in 1212.16
This makes it possible that the castle was actually built during the Fifth
Crusade, perhaps after the Muslim raid on Limassol harbour in 1220, an
event which may have led to fears of more raids, or even a full-scale Ayyubid
counter-attack against the whole of Cyprus.17 Alternatively the castle may
have been constructed during the opening years of the Crusade between
its proclamation by Innocent III in 1215 and its actual start in 1217. This is
implied by the discovery of numerous west European coins at the site, along
with a papal seal attributable to Honorius III (1216-27), a stockpile of
arrowheads and a large store of catapult ammunition. These finds have
been used to suggest that the castle was a Hospitaller base frequented by
European pilgrims, but they could just as well indicate that Paphos acted
as a stepping-stone for troops and supplies destined for Egypt.1 Its construction should therefore perhaps be attributed to Hugh I (1205-18), or
representatives of his infant son Henry I (1218-53), rather than to the
Hospitallers, although their participation in the Egyptian campaign does
not rule out the possibility that they took part in the project.^ This would
also explain why the fortress was never rebuilt after the devastating earthquake of 1222, for by then the Fifth Crusade had come to an end. It may
F O R T I F I C A T I O N S , N A V A L W A R F A R E AND THE C R U S A D E S
12/
even account for the incomplete nature of the castle's rock-cut ditches; a
perfectly understandable omission if construction of the castle had only
begun a few years before it was destroyed.20 It is interesting to speculate
whether material taken from these ditches was subsequently used in Egypt,
for according to Jacques de Vitry the Franks imported Cypriot stone to
repair the walls of Damietta.21
After the Fifth Crusade, it is unlikely that Paphos played any part in
subsequent Christian campaigns. Indeed, after Louis DCs crusade Limassol
itself gradually came to be overshadowed by Famagusta, whose economic
importance and volume of shipping grew steadily during the late thirteenth
century.22 In 1291 it was from here rather than Limassol that Henry II set
out in his final bid to save Acre.23 His failure to do so merely encouraged
the growth of Famagusta, which emerged as the richest port in the eastern
Mediterranean still in Latin hands, and therefore gained new urban defences
in preference to Limassol.24 In addition, the fall of Acre had a profound
effect on the strategic importance of Cyprus as a whole, for it now became
the new front line in a battle zone where galleys and warships were more
important than knights and horses.
In the fourteenth century Famagusta acted as a starting point for a number
of campaigns undertaken by local naval powers, including the Hospitallers
of Rhodes and the Cypriot Franks themselves. The first major expedition
of this kind took place in 1300, when the Templars, Hospitallers and Cypriots
set off from Famagusta on an offensive aimed against Alexandria. They
sailed northwards along the Syrian coastline until they reached Maraclea,
which they sacked before heading back to Cyprus. In the same year troops
from Cyprus and the Military Orders even went ashore briefly at Tortosa,
in a failed attempt to meet up with Mongol and Armenian forces.25 Thereafter the Templars also garrisoned a small castle on the island of Rouad,
just opposite Tortosa, with a force of 120 knights and 500 archers. This
stronghold must have relied on Cyprus almost entirely for its food supply,
for shortly before it surrendered to a Mamluk besieging fleet in 1302, the
Franks tried in vain to relieve it by sending ships from Famagusta. The
Franks therefore used Cyprus as a springboard for raids against the mainland
but were prevented from making any permanent reconquests there because
of their usual lack of troops and resources.26
A similar fate befell Peter Fs renewed efforts against the Muslims in
the mid fourteenth century. Peter launched several expeditions against the
mainland. These have traditionally been linked with his pious and adventurous personality, but were probably also intended to prevent Muslim
ports from killing off Famagusta's declining international trade.27 The most
famous of these was his crusade against Alexandria in 1365, which brought
128
the Latins considerable booty but they again lacked the strength to advance
into the Egyptian interioras Undeterred, Peter continued to garrison Corycus
and Satalia (Antalya), two ports along the coast of Cilician Armenia, which
he had occupied in 1360 and 1361 respectively. In addition, he attacked
several adjoining Turkish settlements such as Anamur, forcing local rulers
to pay him tribute.29
The defences of Corycus and Satalia were also strengthened, so that at
the latter site the Cypriots 'pulled down the old lodgings, and cut down
the trees and filled up the ditches, because the Turks used to hide in them
and leap out to attack people, and did much damage. And by the captain's
orders they built up the wall of the castle and also heightened the towers'.30
Thereafter these outposts were supplied from Cyprus in the same way that
the island of Rouad had been, and in 1367 Corycus, which had been
blockaded by 'a great host of Turks', was successfully relieved by a Cypriot
fleet.31 This port subsequently held out until 1448, although Satalia returned
to Turkish rule in 1373. In occupying these sites Peter may again have been
motivated by a combination of crusading and economic interests, for they
enabled him to attack the Turks but were also useful ports in the lucrative
trade network between Cyprus, Rhodes and the mainland.32
During this period, as well as during the earlier post-npi efforts to regain
the Holy Land, Famagusta's defences were intended to provide a secure
naval base for Cypriot ships. By the mid fourteenth century the city was
protected by a harbour chain and chain tower as well as its landward urban
defences.33 These structures made Famagusta the safest anchorage in Cyprus,
where major expeditions against the mainland could be prepared. In addition, they facilitated the creation of smaller regular patrols such as that,
consisting of two Cypriot galleys, which captured two pirate vessels in i325.34
Famagusta now became virtually immune to the kind of Muslim counterraids which had devastated the Christian fleet at Limassol in 1220 and had
been attempted by Baybars in 1271. It is even possible that the first major
Christian raid on the Holy Land after the loss of Acre, carried out in 1300,
coincided with the completion of at least some of Henry IPs new defences
at Famagusta. This would have released more funds for galleys and have
given Prankish vessels a definite means of escape in case of Muslim counterattacks. If this was the case, it is reminiscent of the Hospitallers' earlier
reluctance to raid Muslim targets in Syria before their own castles of Margat
and Crac des Chevaliers had been made strong enough to withstand potential
retaliatory sieges.35 Famagusta may not, however, have been as well fortified
as the Franks would have liked at this stage, for Amaury of Tyre was still
improving the city's defences ten years later.36
Beyond Famagusta, accounts of the Genoese siege of Kyrenia in 1374
129
suggest that here too a chain existed across the harbour entrance.37 Further
afield, it is clear that the Prankish outposts of Rouad, Satalia and Corycus
were themselves heavily dependant upon fortifications to prevent their
immediate recapture by the Muslims. Just as they had done in the Holy
Land, the Franks relied on castles and city walls rather than superior troop
numbers to gain new territory and intimidate their Muslim neighbours. As
late as the 13605 the defences of Famagusta, Satalia and Corycus enabled
Peter I to dominate local sea routes and to maintain the crusading tradition
of his predecessors.
12
THE M I L I T A R Y O R D E R S
131
The Military Orders therefore never gained the kind of power in Cyprus
which they had once enjoyed on the mainland. This can be seen by giving
a brief summary of their Cypriot castles and estates, starting with the
Templars.5 The Templars had been the first Latins to rule Cyprus after
Richard I's invasion; but, although Guy of Lusignan was later buried in a
church established by them at Nicosia, it seems unlikely that they had had
time to construct any new fortifications before their departure in 1192.0 Six
years later, however, both this Order and the Hospitallers were asked by
Innocent III to help defend Cyprus against external attackers, and in 1210
we find the first reference to the new Templar castle at Gastria, suggesting
that the Templars were quick to re-establish themselves on the island during
the 1190S.7
The fact that Gastria's true age cannot be fixed more accurately than this
confirms that it is just as difficult to reconstruct Templar properties on
Cyprus as it is everywhere else in the Latin East. In general this stems from
the dissolution of the Templars and the transfer of their lands and castles
to the Hospitallers, but in the case of Cyprus it may have been compounded
by the events of 1279, when King Hugh III confiscated or even destroyed
the Order's belongings because of its support for Charles of Anjou, Hugh's
rival for the throne of Jerusalem.8 Thanks to these disasters, archaeological
traces of Templar fortifications on Cyprus are rare and the records of which
Order held which estates prior to 1307 have remained unclear ever since
the sixteenth century. Hence Florio Bustron's list of Templar estates gained
by the Hospitallers probably contains numerous properties which already
belonged to them before 1307, whilst in 1580 it was even claimed that Gastria
had been constructed by the Egyptian invasion force of 142.6.9
Despite these difficulties, some remains of the Templar defences at Gastria
have survived to the present day. These were constructed on a small coastal
promontory to the north of Famagusta, and consisted of a single curtain
wall and rock-cut ditch protecting the site from the interior. To the north
of this promontory, it seems that a small inlet served as a harbour for the
castle. 10 Beyond Gastria, the Templars also held minor fortifications at
Khirokitia and Yermasoia, two of their estates near Limassol. At Khirokitia,
below the fifteenth-century ruins of a Hospitaller building, the remains of
what appears to have been an older Templar tower were still visible in
Enlart's day. It may have been here that the Marshal of the Order was
held in 1307, when all the Templars on Cyprus were arrested. There are no
such remains at nearby Yermasoia, although Florio Bustron noted that the
Templar Commander was similarly incarcerated here, presumably in another
small tower or fortified building, n Descriptions of Hugh Ill's seizure of
Templar properties in 1279 also confirm that the Order had towers or houses
132
U N K N O W N C R U S A D E R CASTLES
at Limassol and Paphos, whilst in the spring of 1232 the Lombards are said
to have occupied a similar building in Nicosia.12 A document dating from
1264 records that the Order later sold this place to two secular knights for
2000 bezants blanch
The Hospitallers were granted many of these properties after 1307,14
including Gastria, Yermasoia and Khirokitia, where a fifteenth-century tower
replaced its Templar predecessor.^ The Hospitallers also had their own
properties at Limassol and Nicosia, which played a small role in the war
between the Lombards and the Ibelins. Iri 1228 Frederick II had John of
Ibelin's sons imprisoned in the Hospitallers' compound at Limassol, because
it 'was strong and nearer his ships', implying that it may have been larger
and more reliable as a prison than the royal castle itself.16 The following
year the Ibelin supporter Philip of Novara described how he managed to
avoid being captured by Frederick's five baillis and slipped away to the
hospital at Nicosia along with 150 troops, plus many women and children
of the Ibelin faction. This building cannot have been very strong, because
Philip hastily equipped it with a cistern and a wooden palisade, making
sure it was stocked with biscuits and other food suitable for a long siege.
In the end these precautions were not needed, for shortly afterwards John
of Ibelin arrived from Syria and defeated Frederick's baillis in battle. This
enabled Philip to launch a sortie against his enemies, some of whom
presumably still occupied Nicosia's royal citadel.17
The principal Hospitaller castle on Cyprus was located at Kolossi, to
the west of Limassol. This estate had been granted to the Hospitallers by
Hugh I in 1210 and subsequently became the Order's Grand Commandery,
suggesting that although the impressive tower which stands there today
was built in the mid fifteenth century, it was preceded by a similar older
structure. Florio Bustron's claim that in 1307 the Hospitallers also inherited
another tower at Kolossi from the Templars should probably be dismissed
as a further example of the confusion which surrounds the latter Order's
properties.18 On the other hand, it may be that the Hospitallers did own
a second tower at Episkopi, another of their estates near Limassol. According
to Estienne of Lusignan Episkopi 'was given by the king of Lusignan
[Hugh I] to the Knights of St John of Jerusalem, who built and fortified
the castle, before the island fell into the hands of the Turks'.19
The Teutonic Knights also held estates in the vicinity of Limassol, plus
a couple of houses at Nicosia. These had been granted to the Order either
during the reign of King Aimery (1196-1205), who received his royal crown
from the German Emperor Henry VI, or during the late 12208, when
Frederick II's influence was at its height.20 Although the Teutonic Knights
held on to these properties after the defeat of the Lombards in 1233, their
THE M I L I T A R Y O R D E R S
133
134
UNKNOWN CRUSADER
CASTLES
PART THREE
Cilician Armenia
1187-1375
13
Warfare
By the late twelfth century the Armenians living in Ciucia, which had
originally been part of the Byzantine empire, had effectively become
independent. This had come about through a combination of local geography and dramatic political change. By the iiSos military pressure applied
by the Turks had caused Greek influence in eastern Asia Minor to disappear,
and during the next two decades the Byzantine empire was racked by a
series of political crises which culminated in the capture of Constantinople
by the Fourth Crusade (1204). Meanwhile, the Armenians had nothing to
fear from their Latin neighbours in the Holy Land, who were far too
preoccupied with the struggle against Saladin to have any kind of real
influence over Cuidan affairs. In addition, Cilicia's physical isolation gave
it a certain amount of protection from the Seljuk Turks of the Anatolian
interior, as well as Saladin's Ayyubid descendants in Syria. The Cilician
interior, which was dominated by a large, fertile plain, was protected by
the Mediterranean sea to the south and by a series of rugged mountain
chains to the north, east and west. These chains could only be penetrated
by a limited number of mountain passes, such as the Amanus Gates in the
east or the famous Cilician Gates which led toward Constantinople. At
times during the thirteenth century the Armenians enjoyed a certain amount
of political control beyond these passes, particularly in the coastal regions
to the west of Silifke and in the territories around Antioch. However, the
Cilician Plain and the mountains around it always formed the core of their
lands, and it was the castles situated within this region that were of central
importance.
For much of the twelfth century politics within Ciucia itself had revolved
around the struggle between various powerful Armenian families trying to
achieve local dominance. The two key dynasties involved in these clashes
were the Hethoumids, who frequently sought aid from the Byzantine empire,
and the Roupenids, who normally favoured greater independence from the
Greeks. This goal was finally achieved in 1198 when Leon II, head of the
Roupenid dynasty, overcame Hethoumid resistance and had himself
crowned as the first king of the Armenians. Thereafter Leon's new kingdom
increasingly resembled a feudal western state which prospered through
138
international trade with Europe and Asia. Below the king, the Armenian
Church flourished at cultural centres such as Lampron1 whilst the local
nobility, whose many strongholds and estates were dotted across the
countryside, grew rich on the agricultural wealth of the Cilician Plain.2
Throughout the thirteenth and fourteenth centuries, however, external
attackers did sporadically penetrate Cilician Armenia and inflict serious
damage. From the time of Leon II's rise to power until the mid thirteenth
century, the most dangerous such opponents were the Seljuk Turks of the
interior. In 1187, for example, they made a failed attack on Sis (the Armenian
capital) which was successfully repelled near the Amamis Gates.3 About
a quarter of a century later there was further fighting to the north-west of
Marash involving Kay-Khushraw I, the Seljuk sultan of Rum (1204-11),
and in 1216 his successor Kay-Kavus I (1211-20) attempted to capture the
Armenian castle of Gabn, which was probably located somewhere in
the same region. Although Kay-Kavus was eventually obliged to abandon
this campaign, he still managed to inflict heavy losses on Armenian relieving
forces sent to help Gabn, indicating that at this time the Seljuks of Caesarea
(Kayseri) posed a serious threat to the Armenians' northern frontier.4 This
period also marked the end of Armenian influence to the west of Silifke
and the Cilician Gates, for in 1218 Leon II was forced to cede the fortresses
of Loulon and Lausada, along with many neighbouring lands, in order to
secure the release of Armenian nobles captured during the fighting for
Gabn.5 Two years earlier the Seljuks had even launched an unsuccessful
attack on Silifke itself,6 and they continued to keep up the pressure under
Kay-Qubad I, whose sultanate (1220-37) witnessed the prosperous growth
of key Turkish centres such as Antalya, Anamur and Alanya.?
In the second half of the thirteenth century the chroniclers continued to
record sporadic Seljuk raids, such as that which was halted to the north
west of the Cilician Gates in 1259.8 Turkish influence over the region declined
with the arrival of the Mongols, however, who defeated the Seljuks in 1243
and turned the sultanate of Rum into a protectorate.9 Thereafter this threat
receded until the fourteenth century, when renewed fighting on the Cilician
Plain heralded a period of Turkish resurgence which eventually culminated
in the birth of the Ottoman empire.10 Until then, however, the rise of the
Mamluks ensured that from the 12605 onwards the principal danger came
not from the north but from the south. Starting with the reign of Baybars
(1260-77), successive Mamluk sultans of Egypt gradually overran the crusader
states in the Holy Land, and during this period the same process began to
take place in Cilician Armenia. Thus in 1266 Baybars sent an army onto the
Cilician Plain via the Amanus Gates, where an Armenian force led by
Hethoum I (1226-69) tried in vain to halt it.11 Subsequently the Egyptians
WARFARE
139
launched regular attacks against the area, with particularly destructive raids
taking place in 1275,1298 and 1322.12 These were followed by even more extensive campaigns in 1337 and 1375, the latter resulting in the permanent conquest
of Cuidan Armenia and its incorporation into the Mamluk empire.13
During much of this period the Armenians could theoretically look for
support from the Muslims' Mongol opponents, with whom they had wisely
fostered friendly relations since the late 1240S.14 In 1260, for example, the
Mongol conquest of Aleppo enabled them to extend their power east of the
Amanus mountains and to occupy sites like Darbsak, a former Muslim
stronghold which the Armenians continued to garrison until i268.15 But
140
WARFARE
141
142
supposed to have been at least 8000 attackers confronting them, these men
were presumably accompanied by other soldiers who were not necessarily
on horseback. Similarly, when Oschin, lord of Corycus, defeated a Turkish
incursion in 1318-19, he may have had more troops at his disposal than the
300 cavalry said to have been under his command.35 This implies that
Armenian chroniclers, like many of their Latin counterparts, tended only
to include noblemen or troops mounted on horseback, without necessarily
giving specific figures for socially inferior foot-soldiers, archers and mercenaries. Sometimes enemy troop numbers could also be exaggerated in order
to enhance the bravery of the Armenians themselves. This was clearly the
case in 1246-47 when the chronicler Constable Sempad stated that 260,000
Turks were held at bay by the defenders of Tarsus.36
Such claims make it impossible to give an accurate figure for the total
number of troops available to Armenian rulers. During the winter of 1264-65
Hethoum I, however, responding to an appeal from the Mongols, was said
to have contributed 200 knights to the failed Mongol siege of al-Bira on
the Euphrates.37 If we place this figure alongside that of the 300 cavalry
referred to previously, and if we assume that Armenian chroniclers were
at least slightly more trustworthy when they spoke about their own knights,
we are left with the impression that there were rarely more than 500
mounted troops present on Armenian campaigns. The fact that the death
of a mere twenty-one knights during the Muslim raid of 1322 was treated
as a disaster may confirm this.38 But even reaching this rather vague estimate
is easy compared with calculating the amount of infantry soldiers who
accompanied the mounted troops. If Armenian forces were anything like
their Byzantine counterparts, the infantry component of the army was
largely made up of ill-disciplined recruits whose exact numbers were not
even known to the commanders themselves.39 Under such circumstances
it is difficult to know whether we can trust the claim that 5000 troops
were involved in an internal conflict which broke out in 1220. The fact
that this rebellion involved a substantial element of the Armenian nobility
implies that kings could normally hope to gather at least this many soldiers
together, although the total may actually have been much smaller.40 The
figure of 18,000 men (including 4000 cavalry, 1000 regular infantry, 4000
crossbowmen and numerous siege engineers) given for Hethoum II's army
in 1302 also sounds suspiciously high, particularly as it comes immediately
after a romanticised account of Armenians effortlessly killing thousands of
Muslim horsemen in battle.41
The absence of any trustworthy references to the size and composition
of the royal Armenian army means that these figures should only be regarded
as suggestions. Some additional information can be gleaned from other
WARFARE
143
144
enjoyed a reputation for their expertise in siege warfare dating back to the
twelfth century, when the Franks had employed one of their countrymen
to oversee the Latin siege works at Tyre (1124) .5 A quarter of a century
earlier another Armenian had been used by the Muslims to design an
elaborate new gateway for the citadel of Cairo, presumably because it was
felt that anyone who was good at constructing siege weapons would be just
as good at building castles.si This knowledge was put to good use in Ciucia
itself, where the Armenians held many impregnable strongholds on the
rocky outcrops dotted around the Cilician Plain.
The sheer strength of many local castles, aided by the ruggedness of the
terrain, enabled the Armenians to retain their independence until as late
as 1375, long after their field armies had become too outnumbered to put
up any effective resistance against the Mamluk empire. This proved to be
a burden as well as a blessing for many Armenian rulers, for it encouraged
some nobles to rebel, safe in the knowledge that even the king would not
be able to capture their remote mountain strongholds. Problems of this
kind reached chronic levels in the fourteenth century but were evident from
the very beginning of Leon II's reign. In many ways the military situation
in Cilician Armenia resembled that of the Latin states further south. It has
been suggested the Armenians could perhaps muster a field army of 5000
men, including a core of around 500 mounted troops. They were also
renowned for their skills in siege warfare and capable of building extremely
powerful fortifications. Even allowing for the extravagance of some contemporary chroniclers, however, they were invariably outnumbered by
their Muslim opponents, and in particular by the Mamluk armies which
eventually brought about their downfall in 1375. Like the Franks, they
needed impregnable fortresses to withstand the Muslim onslaught, even
though such structures inevitably undermined central authority. This
dilemma ensured that castles played a very prominent role in the history
of Cilician Armenia, both in terms of internal politics and external security.
14
Military Architecture
Any discussion devoted to military architecture in Cilician Armenia is
immediately hampered by two difficulties. First, the history of the kingdom
was extremely badly recorded by contemporaries, particularly during the
fourteenth century, when external invasions and internal rebellions led to
almost constant political instability. Secondly, successive Roman, Byzantine,
Arab, Crusader, Mamluk and Ottoman rulers controlled this region both
before and after the lifespan of the Armenian kingdom. As a result, local
fortifications present the historian with numerous problems of dating,
attribution and interpretation. Although ruined strongholds still occupy
countless local hilltops, many have no written history, whilst others contain
such a bewildering mixture of architectural elements that it is impossible
to establish their exact age.
Despite these difficulties archaeologists studying the area have shown that
several castles do contain substantial remains of what can be termed a
distinct Armenian style. One such stronghold is Servantikar, which was
built near the Amanus Gates on a roughly triangular plateau approximately
500 metres above sea level. The castle's close proximity to such an important
route into Cilicia ensured that it witnessed a lot of warfare, for it was here
that Egyptian armies attacked in 1266, 1275 and 1299.! On the last occasion
the Muslims even sacked Servantikar, and in 1337 it appears to have been
incorporated hito the Mamluk empire as part of a peace treaty negotiated
with King Leon V (i32o-4i).2
Today the castle lies in ruins, but a contemporary source shows that even
in its heyday it needed few man-made defences: '[Servantikar] is a strong
citadel on an outcrop in a valley. Many of its sides do not have walls, as
they are naturally defended by the cliff's edge'.3 Indeed, fortifications were
only really necessary along the gently sloping east front, which was protected
by a long, sinuous curtain wall flanked by numerous round or horseshoe
towers. Any attackers were obliged to pass below these defences and expose
themselves to lethal flanking fire in order to reach the entrance in the
north-east corner. Here they would be confronted by the outer gateway,
an 'L' shaped structure incorporating two further towers and a machicolated
ceiling. Even if they overcame this obstacle, attackers would still only have
Servantikar: a typical Armenian mountain castle. Note the horseshoe towers and
absence of a central keep.
reached the lower castle. In order to capture the upper bailey, they needed
to repeat the whole process by penetrating another curtain wall and gateway.
The inner bailey itself contained numerous residential buildings and was
dominated by a cluster of four powerful towers situated at the southern
MILITARY ARCHITECTURE
147
tip of the site. These towers were attached to adjoining curtain walls but
could also act as a kind of keep or final refuge. Any efforts to capture
Servantikar by storm would therefore have proved costly, whilst the generous
provision of water cisterns within the walls ruled out the possibility of
gaining entry through a swift blockade.
Many of the features present at Servantikar were typical of Armenian
fortifications throughout this region. First, there was the castle's location on
a rocky promontory, which made it difficult to approach and impossible to
breach using mines. The castle was so inaccessible that complete ramparts
were not always needed. Where the Armenians did build curtain walls, these
consistently followed the cliff's edge and therefore enhanced the natural
strength of the site. This layout could then be repeated, for Servantikar was
only one of many Armenian sites where two or more walled baileys created
successive lines of defence. This in turn meant that towers could be used
sparingly, for the irregular shape of the fortress created plenty of natural
angles for flanking fire. Any towers which the Armenians did add were
invariably round or horseshoe shaped, thereby making them better able to
withstand earthquakes,4 catapults or battering rams. By attaching all towers
to surrounding walls, rather than providing Servantikar with a free-standing
keep, the castle's builders also strengthened its ramparts whilst at the same
time creating even more vantage points for archers stationed on the wall
walks.
Numerous other building techniques used at Servantikar confirm that
the Armenians were highly skilled in the art of siege warfare. Not only were
those stretches of the curtain wall most likely to be hit by catapults extremely
thick, their outer faces were also covered with small, square blocks of bossed
stone which would be more resilient against bombardment. Servantikar also
only had one gateway, which was flanked by two towers and could not be
attacked head on because it was placed at a right angle to the actual rampart.
The entrance was defended from above by slot machicolation, enabling
troops in the upper chamber of the gatehouse to shoot at intruders through
holes incorporated into the vaulting. All these features indicate that Servantikar had been carefully designed to maximise the defensive potential
of the site and to make the use of siege weapons against the castle virtually
impossible.5
The general appearance of Servantikar was very similar to that of Lampron,
situated near the Cuidan Gates, and Vagha, another Armenian stronghold
located in the mountains to the north of Sis. Both these castles were built
on remote spurs hundreds of metres above sea level, and were defended
by successive baileys rising toward impregnable inner citadels. Indeed, the
upper parts of Lampron could only be reached via a circuitous ramp, which
148
UNKNOWN CRUSADER
CASTLES
had been cut out of the solid rock and was barely two metres wide. As a
result, it could not be used by more than one attacker at a time.6 The
approaches to Vagha were almost equally inaccessible, being additionally
protected by several gateways, at least one of which incorporated the same
kind of slot machicolation as that used at Servantikar.7 Many of these
defensive features can also be seen in strongholds occupying outcrops on
the Cilician Plain. The best preserved such sites are Gkvelioglou,8 Tumlu9
and Yuan, all of which lie to the south and south-west of Sis. The most
famous of these was probably Yuan (Ilan Kale/The Castle of the Snakes),
whose complex inner gateway and stout horse shoe towers represent some
of the most impressive Armenian structures of the entire region.10
Through careful study and comparison scholars have concluded that these
castles were predominantly built by the Armenians. Many of the outcrops
they occupied, however, were so ideally suited to fortification that few could
have been virgin sites before the Armenians arrived. It is interesting to
speculate, for example, whether the twenty metre wide moat dividing the
castle of Lampron from neighbouring mountains had initially been dug by
the Greeks, for it bears a close resemblance to the famous rock hewn ditch
at Saone, whose original excavation has been attributed to Byzantine rather
than Latin engineers.11 Servantikar can also be compared with Kantara,
St Hilarin and Buffavento, the three great mountain fortresses built by
the Greeks on Cyprus using a similar combination of successive curtain
walls.12 Indeed, the difficulties in trying to identify the original occupants
of some sites have led different scholars to attribute Azgit, a remote mountain castle between Servantikar and Vagha, to both the Greeks and the
Armenians.13
It should also be noted that castles like Yuan and Servantikar represented
a tiny minority, for most strongholds in Cilician Armenia were far more
complex architecturally and probably only contained a few Armenian
repairs. Between Yuan and Sis, for example, the Armenians garrisoned the
classical acropolis of Anavarza (Anazarbus), a Roman settlement which
also shows evidence of early twelfth-century Prankish occupation. 1* Other
Roman or Byzantine acropolis sites reoccupied by the Armenians included
Bodrum,15 which lay a few miles north west of Servantikar, and Sis, the
capital of the kingdom, is At other castles, such as nearby Toprak (Til
Hamdoun), the paucity of Armenian fortifications was caused by later as
well as earlier occupants. In 1212 the German pilgrim Willbrand of Oldenburg wrote that Toprak was a 'good strong fortress held by an [Armenian]
nobleman',17 yet the present defences generally appear to be a Mamluk
construction dating from the fourteenth century.18 Toprak was probably
largely rebuilt after 1337, when according to the chronicler Nerses Balientz
MILITARY ARCHITECTURE
149
Yuan
the Muslims acquired much of eastern Cilicia by treaty.19 This in turn
suggests that the site had been severely damaged during previous Mamluk
raids on the area, particularly those of 1266, 1298 and 1304.2 The original
castle seen by Willbrand of Oldenburg may, however, not necessarily have
been any more Armenian than the Mamluk one which replaced it, for the
entire site occupies a much older partially man-made tell, and had probably
been fortified in some way for centuries.
150
The eclectic nature and sheer age of many fortified sites in the region
can also make it difficult to establish whether the Armenians copied their
predecessors or invented their own architectural style, which then influenced
others. For example, while the Armenians often excavated deep ditches or
built successive curtain walls to defend weak spots, this had already been
done by Byzantine engineers both locally and elsewhere. These tactics were
indeed so obvious and so commonly used to protect castles (including Latin
strongholds such as Montfort) that it would be misleading to attribute them
to any one group or period.21 It is more useful to concentrate on specific
details such as masonry types and vaulting, an approach which immediately
reveals interesting links between Armenian fortifications and crusader castles
elsewhere. The Hospitaller strongholds of Crac des Chevaliers, Margat and
Silifke are particularly significant in this respect, for all three incorporated
the kind of complex gateways and horseshoe towers built by the Armenians
at Servantikar. Silifke, which lay on the south-western fringes of the Cilician
Plain, was granted to the Hospitallers by Leon II in 1210 and remained
under their control for the next sixteen years. During that period the Order
effectively rebuilt the entire castle; a project which coincided with major
repairs at Crac and Margat.22 This link seems to have been limited to these
three castles, for in general horseshoe towers were not at all popular with
the Franks. On the other hand, other individual features such as slot
machicolation, which 'appears to be an Armenian invention', did catch
on and were used extensively in Latin castles of the period.23 It has also
been noted that the Franks sometimes employed Armenian siege engineers,
implying that they were highly regarded as military architects beyond Ciucia
itself and that some of the features incorporated into their castles were
genuine innovations.24
In other cases the Franks rejected Armenian precedents in favour of their
own designs. It has been shown that many crusader castles were less irregular
and relied more on square or rectangular towers, incorporating several
gateways or posterns rather than a single inaccessible entrance. Several
Prankish strongholds, including Tortosa, Chastel Blanc and Montfort, were
built around a large, free-standing keep. This type of structure had been
brought to the Holy Land from Europe and was not used by the Armenians.
During the thirteenth century the Latins also used smooth as well as bossed
masonry, and often built their walls on large, sloping revetments to give
them extra strength.25 These differences between Armenian and Prankish
military architecture can partly be explained in terms of local terrain. The
need to maintain sea links with the West forced the Latins to build many
castles and city ramparts on the coastal plain, where straight walls were
more logical and man-made features such as revetments compensated for
MILITARY ARCHITECTURE
151
the absence of sheer cliffs. This does not rule out the possibility that the
Armenians occupied, repaired or even constructed their own urban fortifications. There has been a tendency to assume that the Armenians only built
remote mountain castles because of the spectacular remains at sites like
Servantikar, Yuan and Vagha, but it is their very remoteness which has
helped these strongholds to survive. It would be equally mistaken to deny
that the Franks built any urban defences simply because the massive walls
of Acre and Tyre have since disappeared. While it is true that the inaccessible
location, sinuous ramparts and elaborate gateways of Servantikar represented
a distinct style, this was not the only type of fortification constructed and
used by the Armenians.
15
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
153
154
U N K N O W N CRUSADER CASTLES
ravaging the Cilician Plain. But, even after the soldiers and castles which
defended the frontiers had been overcome, the Armenians could still use
the strongholds of the interior as refuge points. Castles built on outcrops
protruding from the plain, such as Yuan or the citadel of Sis, were so strong
and so widely spread out that they could not all be captured by attackers,
who were therefore prevented from making any permanent territorial conquests. As a result, the Armenians were able to wait inside their castles until
lack of supplies forced their opponents to retreat, at which point temporarily
lost areas could be reoccupied. When describing the Mamluk invasion of
1266, Vahram of Edessa noted that Muslim forces quickly occupied the
countryside, but 'in front of the fortresses which they attacked, they failed
miserably'.9 When the Muslims returned a few years later, Leon III found
himself so outnumbered that he did not even attempt to confront them in
the open, but even so 'those who occupied the fortified places, or had
retreated to the fortresses, escaped the carnage'.10 There can be little doubt
that the strength and proliferation of Cilician castles helped the Armenians
to retain their independence until 1375 despite being so outnumbered.
During attacks castles did not simply preserve the frontiers of the kingdom, they also sheltered hundreds of civilians living in surrounding areas.
This is confirmed by accounts of what happened when attackers actually
managed to break into Armenian fortresses. In 1298 one stronghold taken
in the vicinity of Toprak was found to contain 'a large throng, composed
of peasants, farmers' wives and children'.11 After this castle had surrendered,
al-Makrizi wrote that a further 'eleven places of the Armenians similarly
fell under the control of the victors'. Although these sites were later reoccupied with the help of the Mongols, this disaster must have inflicted heavy
suffering on the local population.12 Muslim sources also contradict Vahram
of Edessa's claim that the Mamluk raid of 1266 failed to capture any local
strongholds, for according to Ibn al-Furat the castle of Amouda, which lay
on a small hilltop between Toprak and Sis, was forced to surrender at this
time. Inside the Muslims found '2200 people, both fighting men and others,
of whom the men were killed and the captives distributed amongst the
troops'.13 Willbrand of Oldenburg's remarks that Amouda provided a place
of shelter for local people and that the nearby river yielded good supplies
of fish give further weight to the impression that the castle acted as a refuge
point for many neighbouring communities.14
Interesting comparisons can be made between these events and those
which were taking place in the crusader states during the thirteenth century.
Like the Armenians, the Franks relied on their strongest castles to protect
themselves and to prevent permanent territorial losses, even if their Muslim
opponents sometimes occupied the land. In 1244, for example, we have
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
155
seen that the Latins survived the Khwarizmian invasion of the Holy Land
because massive fortresses like Saphet and Pilgrims' Castle held out. More
poorly defended sites, however, were overrun, most notably Jerusalem,
whose undergarrisoned and dilapidated walls could not prevent the massacre
of thousands of Christians inside the city.15 Jerusalem shared the same fate
as Amouda because neither site proved capable of withstanding major
external invasion forces. Their tragic history can be contrasted with that of
the remote and virtually impregnable mountain strongholds situated in less
accessible locations. Vagha, for example, continued to act as the residence
of the Catholicos, or head of the Armenian Church, well into the fifteenth
century.16 The lords of Gabn, whose exact location somewhere to the north
of Sis has never been established, seem to have retained a certain amount
of independence until the same period.17 Indeed, Sis itself was not permanently conquered by the Mamluks until 1375, when the citadel was captured
and the surrounding city destroyed.18 Consequently, whilst all Armenian
castles contributed to the overall security of the region, a distinction needs
to be made between the more exposed sites such as Amouda and strongholds
situated along the mountainous northern fringes of the Cilician Plain. These
latter castles proved better able to withstand the Muslims, in the same way
that formidable crusader strongpoints like Pilgrims' Castle held out long
after more vulnerable sites like Jerusalem had been lost.
In their long struggle to survive the Muslim onslaught, the Armenians
were also aided by wider political factors which had nothing to do with the
strength of their mountain fortresses. Many of the thirteenth-century attacks
launched by the Seljuks or by the Mamluks were probably not in fact
intended to conquer the Cilician Plain. The relatively rapid destruction of
Leon II's power to the west of Silifke between 1200 and 1220 suggests that
the Seljuks certainly had the military capacity to capture Armenian fortresses
given the right circumstances. However, later setbacks, such as the arrival
of the Mongols, usually ensured that subsequent Turkish attacks amounted
to little more than temporary raids, and they did not pose a serious threat
again until the fourteenth century. The Mamluk campaigns of 1266, 1275
and 1298 were all primarily designed to secure loot rather than new lands.19
By contrast, Baybars captured several Armenian-held castles to the east of
the Amanus mountains and incorporated them into the Mamluk empire
as early as 1268, shortly after he had taken the city of Antioch from the
Franks. Many of these castles, abandoned by their former Muslim owners
during the devastating Mongol invasion of 1260, had then passed under the
control of Bohemond VI of Antioch or the Armenians.20 Perhaps, therefore,
Baybars was keen to regain these sites in order to make up for the humiliation
of 1260 and to ensure that the formerly Christian principality of Antioch
156
U N K N O W N CRUSADER CASTLES
could now be turned into a strong buffer zone against any future Mongol
aggression. All these events prove that, when it suited him, Baybars was
perfectly capable of making permanent territorial gains at the expense of
the Armenians, but instead he chose to treat the Cuidan Plain itself as a
source of tribute and booty which could be exploited from time to time.
The desire for quick booty at the expense of few casualties helps to explain
a number of late thirteenth-century incidents relating to the Cuidan Plain.
In 1265 Baybars may have called off his expedition to the area because
Hethoum I had discovered the plan and had mustered a field army in
response, thereby removing all hope of a profitable surprise raid without
any casualties. The following year Mamluk troops returned to the region,
but the fact that they only attacked relatively easy targets such as Amouda
and appear to have ignored tougher castles in the mountains suggests that
they had come to plunder rather than to conquer. In 1298 an attempt to
capture Sis was abandoned in favour of a much easier looting spree across
the Cilician countryside.21 A peace treaty between Leon III and Kalavun
dating from 1285 also indicates that for much of this period the sultans of
Egypt relied on the mere threat of punitive raids to extract large amounts
of tribute.22 Non-payment by the Armenians resulted in further Mamluk
incursions in the opening years of the fourteenth century. According to
al-Makrizi, the Muslims pillaged the slopes below the citadel of Sis in 1302
and sacked Toprak in i304.23 Two years later 2000 soldiers were sent from
Aleppo because of further Armenian refusals to pay up, but these troops
were defeated with the aid of the Mongols, who had themselves come to
Sis to claim tribute as the price of their friendship. The Egyptian authorities
reacted angrily, dispatching a further 4000 men who were finally able to
collect the outstanding payment, presumably because the Mongols had now
departed.24 These raids reflect a continuing interest in temporary plunder
rather than permanent conquest, for they involved relatively few troops and
continued to avoid difficult targets, such as the actual citadel of Sis, in
favour of somewhat easier prey like Toprak.
Any Armenian castles captured by the Muslims in this period were then
either abandoned deliberately or later given up in the face of Mongol
pressure. Despite being overrun in 1298, Toprak had been regarrisoned by
Armenian troops by 1304, for the Mamluks had not actually held on to the
castle after they had looted it. Strongholds do not seem to have been
captured for the sake of territorial gain rather than booty until later in the
fourteenth century. It is clear that 1337 was a turning point in this respect,
for in that year 'the troops of the sultan of Egypt ... entered Cilicia with
60,000 cavalry troops and besieged Ayas ... they would not leave until
the town had been delivered to them, along with all the land between the
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
157
Ceyhan river and the territory of the Arabs, land where there lay forty
castles and fortresses, each with its own lord. These were abandoned to the
Arabs voluntarily and by treaty'.25 The Ceyhan river ran through the heart
of Cilicia, from the mountainous interior around Marash to the Mediterranean coast near Ayas. Consequently, if this treaty was carried out to the
letter, the Armenians lost many of their most important castles, including
Servantikar, Yuan and Toprak. It has already been noted that the current
remains of the last fortress are predominantly Mamluk, suggesting that after
1337 Muslim policy changed from sacking Toprak to repairing it and placing
it under permanent Egyptian rule.26
Once this shift in Mamluk thinking had taken place, the Armenians were
doomed, for they could no longer pursue a policy of waiting inside their
mountain castles and then reoccupying the land once temporary Muslim
raiding expeditions had withdrawn. When the Mamluks made their final
assault on Sis in 1375, they were no longer content to sack the lower city
but besieged the citadel itself, refusing to give up until it had been captured
and the last Armenian king had been led away in captivity.27 Returning to
the treaty of 1337, it is clear that during these final decades of Armenian
independence Mamluk incursions were beginning to inflict such widespread
damage on the economy that the Muslims could demand the surrender of
castles which they had not even attacked directly. By this point one chronicler
noted that Muslim tactics had 'made a desert of the land of the Armenians',28
implying that many local garrisons no longer had the will or the resources
to resist even if they did occupy virtually impregnable mountain strongholds. Once again, the end of Cilician Armenia as an independent kingdom
can be compared with the demise of the crusader states before 1291.
These had survived temporary looting sprees such as the Khwarizmian
occupation of Palestine but could not prevent Baybars and his successors
from picking off even the strongest crusader strongholds through a
combination of devastating raids and systematic sieges.29
The ability of the Armenians to hold out for almost ninety years after
the fall of Acre should still be regarded as a great achievement. It can be
attributed to the ruggedness of the terrain and the sheer strength of castles
like Sis or Lampron. Indeed, the impressive remains of such fortresses have
led some scholars to argue that there was another factor which contributed
to the remarkable resilience of the Armenians: visual communication. The
archaeologist R. W. Edwards, for example, has written that 'most of the forts
have intervisibility which allows for rapid communication and the efficient
mustering of troops'.30 Edwards adds that 'every garrison fort in Cilicia
Pedias has intervisibility with at least two other forts in the plain and
most can communicate directly with the highland valleys'.31 This theory has
158
U N K N O W N CRUSADER CASTLES
been taken even further by the scholar Hansgerd Hellenkemper, who has
concluded that in the early thirteenth century Leon II deliberately constructed a large number of Cilician strongpoints to fit into a vast network
that eventually included Ayas, Misis, Yuan, Gkvelioglu, Tumlu, Anavarza,
Toprak, Amouda, Harunia, Ak Kale and Bodrum. These, it is argued by
Edwards and Hellenkemper, could all communicate with each other using
fire or smoke signals, thus giving early warning of an imminent Muslim
invasion.32
This conclusion can be backed up by a number of written and archaeological sources. The deliberate creation of an intervisible network required
the presence of a strong central ruler and Leon II seems the most likely
candidate. Not only did his reign (1198-1219) mark the zenith of Armenian
power in many ways,33 some chroniclers also made direct reference to his
interest in castles. Michael the Syrian wrote that 'the valorous Leon extended
his domination over seventy-two fortresses',34 whilst Vahram of Edessa
claimed that 'he constructed a number of castles and strongholds, with
which he surrounded Cilicia'.35 Even more significantly, the twelfth-century
keep at Anavarza actually has a surviving inscription which commemorates
repair work carried out there by Leon II.36 A similar feature dating from
1206 has been discovered on the walls of the sea castle at Corycus,37 whilst
some have argued that a carved relief above the inner gateway at Yuan
represents an enthroned Leon after his coronation in 1198.38 The close
architectural links between this latter fortress and several other mountain
castles, all of which have the familiar Armenian combination of horseshoe
towers and sinuous curtain walls, could also suggest that they were erected
by the same man.
Edwards himself, however, has sounded a note of caution about dating
so many strongholds to such a brief period of time. He has pointed out
that the inscription at Anavarza dates from 1188, ten years before Leon II
became king, and that both this inscription and that at Corycus only
commemorate repair work rather than new foundations. The idea that the
relief at Yilan depicts Leon II is also more of a tradition than a fact, whilst
the building style used in this and many other castles is common to all
Armenian military architecture between the tenth and fourteenth centuries.
Therefore Edwards concludes that it is an oversimplification to attribute
so many Cilician fortresses to one king, but he still argues in favour of
a deliberate intervisible network of castles created by the Armenians over a
longer period of time.39
Even if such a network had been built up over the entire two-hundred-year
period between the growth of Armenian independence in the late twelfth
century and the fall of the kingdom in 1375, numerous difficulties remain.
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
159
l60
U N K N O W N CRUSADER CASTLES
of 1198, for several major sites which are missing clearly belonged to the
royal domain, including the capital Sis, the port of Ayas and the cities of
Tarsus, Misis and Adana. Other strongpoints omitted from the list, such
as Tumlu, Yuan and Vagha, presumably also belonged to the king, but the
possibility that these were actually baronial castles which had not yet been
constructed in 1198 cannot be ruled out completely. These difficulties make
it virtually impossible to prove whether or not royal castles formed their
own intervisible network.46
It is also unclear whether such a network would have been particularly
practical anyway. The purpose of a nationwide network would have been
to provide an early warning system, so that the garrison of a fortress like
Toprak, which was situated relatively close to the Beln Pass and the
Amanus Gates, could send fire or smoke signals to warn the king and his
barons of an imminent Muslim attack. For this system to have worked,
part of each castle garrison would have been required to keep a constant
look out and be prepared to light some form of large beacon at short
notice. Even in ideal weather conditions, this would not always have been
easy. At Anavarza, for example, one scholar has noted that it is possible
to see both Yuan and Tumlu from the parapet of the keep, but the roof
of this structure can only be reached via three cramped and partially
blocked storeys, making it unsuitable for building bonfires.47 The keep was
constructed by the Franks early in the twelfth century, possibly not long
after the First Crusade, indicating that it was not intended to fit into any
subsequent Armenian defence schemes.48 The usefulness of warning systems
which rely on simple beacons is also limited because it is difficult to describe
the nature of an invading army or the direction in which it is travelling.
It is equally problematic for anyone who receives a fire signal to send
anything more than a prearranged reply, whilst the cost of maintaining
the whole network throughout the year may have been prohibitive. All
these problems appear to have persuaded the Byzantine authorities to
disband a similar system many centuries earlier. Their chain of beacons
had stretched all the way across Asia Minor to Constantinople and had as
its starting point the fortress of Loulon, which was situated near the Cilician
Gates and could therefore be used to observe Arab raids in the region. Its
abandonment long before the Greeks actually lost control over this area
suggests that it proved unworkable.49
A variety of practical and political difficulties therefore make it unlikely
that the Armenian kings were able to create a nationwide network of
intervisible strongholds. This need not rule out the possibility that individual
castles could sometimes light beacons to warn local people working in the
countryside, or that fire signals were used between a very small number of
FORTIFICATIONS
AND E X T E R N A L S E C U R I T Y
101
castles. We have already seen that a limited network of this kind proved
perfectly satisfactory on Cyprus, where the defenders of Buffavento, an
isolated mountain fortress which enjoyed extensive views over the island's
coastal areas, could apparently warn other, less elevated castles of an imminent pirate attack. In this case it is worth remembering that all the strongholds
concerned were held by the king, which presumably made it easier to
supervise the system.50 Perhaps individual Armenian castles were occasionally constructed with intervisibility in mind, especially the new strongholds
said to have been built by Leon II and Hethoum I for the defence of the
mountain passes.51
Recent research carried out on the Aegean island of Euboea indicates
that intervisibility between two or more strongpoints need not necessarily
have been deliberate. It has been shown that the numerous Latin towers
dotted across the Euboean countryside were not intended for fire signals,
as was once thought, but were used as fortified farmsteads, trading posts
and status symbols. Their location was not determined by strategic thinking
but by more mundane needs such as water supply and proximity to good
agricultural land. Any intervisibility between towers was purely coincidental,
and perhaps inevitable when one considers that people tend to build fortifications on hill tops.52 Similarly, the many small strongpoints built by the
Franks in areas such as the plain of Akkar were simply fortified agricultural
centres rather than clusters of look-out posts which had deliberately been
placed so that they could communicate with each other.53
Returning to Cilician Armenia, this point is epitomised by the fortress
of Amouda. Amouda sheltered thousands of civilians during the Mamluk
raid of 1266, whilst its location next to the Ceyhan river provided it with
plenty of fish and probably made it a useful place to collect tolls from
people travelling between Sis and the Amanus Gates. All these factors suggest
that the outcrop occupied by the castle was chosen because it lay at the
heart of a well-populated area good for fishing and farming. The keep at
Amouda was also clearly built by its thirteenth-century owners, the Teutonic
Knights, with little regard to local building techniques and on a site which
had already been fortified in some way for many centuries. In other words,
the fact that other castles can be seen from the ramparts of Amouda is
probably a coincidence, having nothing to do with the primary functions
of the stronghold as a centre for local trade, farming and defence. As was
the case in the Holy Land, most castles therefore had no wider strategic
importance beyond their immediate surroundings54
Another explanation sometimes given for the ability of the Armenians to
withstand the Muslims for so long was their dislike of urban life. This theory
has probably again been influenced by the survival of so many Armenian
102
F O R T I F I C A T I O N S AND E X T E R N A L S E C U R I T Y
163
had already been attached to the fortress. This settlement must have been
very similar to the bourgs or castle towns in the Holy Land, such as that
which stood next to the Templar fortress of 'Atlit.64 It is clear that communities of this kind existed elsewhere in Cilicia, and that even remote
mountain castles like Vagha, far from existing in complete isolation, usually
had some form of settlement located on the slopes below them.65
Despite the evidence for ruined walls at places like Tarsus, the Armenians
may not have been hostile to the idea of living in larger cities. The impression
given by Willbrand of Oldenburg that Tarsus, Misis and Adana were past
their prime may not have been caused by deliberate royal policy but could
simply have reflected the general decline in urban numbers which had taken
place since classical times. By comparison, the Greek and Armenian population of Antioch had shrunk considerably since the fifth century, even
though the Franks who ruled the city during the crusader period were
supposedly far keener on the idea of urban living. We have seen that the
vast ramparts built by Theodosius and Justinian around Antioch were a
mixed blessing to the Franks: they were extremely powerful but must have
been difficult and expensive to maintain for a city whose population may
have dropped by as much as two thirds.66 Perhaps the Roman-Byzantine
walls of cities on the Cilician Plain had been allowed to decay because it
was a waste of effort to maintain them, even though many Greeks and
Armenians still lived there.
It is also possible that the ruins seen by Willbrand of Oldenburg only
represented a temporary state of affairs, for other sources indicate that
Misis, Tarsus and Adana were all properly fortified both before and after
Wulbrand's visit of 1212. In 1151, for example, the Greek general Andronikos
Komnenos, sent into the region to try to re-establish waning Byzantine
control there, found his progress halted by the defiant Armenian citizens
of Misis, whose leader Thoros eventually defeated the Greeks outside the
city. An Armenian account of these events leave us in no doubt that the
walls of Misis were intact at this point.67 Similar evidence has survived from
1221, when Leon IFs death two years earlier led to a violent succession
dispute between his great nephew Raymond Roupen and his daughter
Isabelle, who was aided in the struggle by the powerful Hethoumid baron
Constantine.68 According to Constable Sempad this conflict came to an end
when Constantine trapped Raymond Roupen's supporters inside Tarsus.
Constable Sempad's description of this incident is worth quoting in full:
So the lords of Tarsus, having returned, re-entered their city and then, having
shut the gates, mounted the ramparts in order to fight off their pursuers. But a
man in the city whose name was Vasil made contact with [Constantine], who
104
U N K N O W N CRUSADER CASTLES
promised him whatever he wanted. The man opened the gates fully during the
night; [Constantine] and his troops penetrated the city where they pillaged
the properties of the Greeks. But the rebellious lords fled the city and climbed
up to the citadel which was strong and impregnable thanks to its defences;
eventually the skilled ... Constantine overcame them without bloodshed by
means of negotiation .. ,69
FORTIFICATIONS
AND E X T E R N A L S E C U R I T Y
165
citadel was clearly well prepared for a long siege even as late as 1375, and
had enabled the city's inhabitants to emerge unscathed from numerous
earlier hostile incursions. The present remains of Sis indicate that the slopes
below the citadel were also protected by outer walls and defences, many of
which must presumably have been built after Willbrand of Oldenburg's
visit.78
The tenacity with which the Armenians rebuilt their cities after Muslim
attacks is also evident at the ports of Ayas (Lajazzo) and Corycus. The
limited number of natural harbours along the Cilician coast gave these
settlements great economic importance as the key trade links between the
West, Lesser Armenia and the Mongol territories to the east of Ciucia,
particularly after the fall of Acre in ligi.79 These factors also made them
attractive to hostile raiders, so the Armenians were careful to maintain the
Roman, Greek and Arab defences which protected them. At both sites these
defences consisted of a land castle overlooking the town and a sea castle
built on an island in the harbour. The eclectic history of many of these
structures can best be illustrated by describing the land castle of Corycus,
which consisted of two square enclosures built in concentric fashion and
flanked by a variety of square or polygonal towers. These fortifications had
been constructed using anything that came to hand, including a Roman
triumphal arch incorporated into the inner ramparti It is clear that both
this fortress and its neighbour in the harbour were garrisoned by the
Armenians, for a plaque has survived on the latter stronghold which commemorates repairs carried out there by Leon II in 1206.81 Indeed, at Ayas
it is possible that the entire sea castle was rebuilt by the Armenians, for
the chronicler Bar Hebraeus wrote that when the Muslims attacked the city
in 1282 'they did not find in it one of its inhabitants, for they had all fled
to sea, and had gone into a new fortress which they had built out in the
sea'.82
This incident shows that the citizens of Ayas and Corycus were able to
survive enemy attacks by retreating into powerful citadels, just like the
inhabitants of Sis did. Only seven years before the raid described by Bar
Hebraeus 8000 Mamluk troops had entered Ayas 'and they killed those
whom they found inside' but were unable to reach those inhabitants who
had fled to the sea castle. These people were therefore able to escape with
their lives, although they were subsequently turned upon by Prankish pirates
who 'even carried off their cloaks' in a tragic example of Christian disunity.83
Nevertheless, Ayas as a whole recovered from both this incident and the
1282 raid; and between 1321 and 1322 the city even survived consecutive
attacks by the Mongols and the Mamluks. The latter offensive resulted in
the capture of the sea castle 'by various means and bombardment with
l66
U N K N O W N CRUSADER CASTLES
catapult',84 suggesting that the Muslims had come better prepared and were
now thinking in terms of more permanent territorial expansion. However,
according to a peace treaty with the Egyptians negotiated in 1323, the sultan
agreed to make good the damage he had caused both at Ayas and elsewhere.85
A document dating from June 1328 also records the arrangements made by
Pope John XXII to contribute 30,000 florins 'toward the rebuilding and
construction and fortification of the sea and land castles of Ayas and other
places in the kingdom of Armenia'.86 Whether or not these combined efforts
led to the total restoration of all the affected strongholds cannot be ascertained, although Ayas did not finally come under Egyptian rule until the
Mamluk campaign of 1337.87
Throughout this period the Armenians did not simply try to keep Ayas
alive by repopulating its houses and rebuilding its defences. Successive kings
also granted privileges to western merchants, who were allowed to trade
there in return for paying tolls to royal officials. Such privileges were granted
to the Genoese in 1201, 1215, 1216, 1288 and 1289, to the Venetians in 1201,
1245,1271, 1307,1321 and 1333, and to the Pisans in m6.88 Further privileges
were granted to merchants from Sicily (1331), Montpelier (1314, 1321) and
Catalonia (1293), as well as to the Florentine banking company, the Bardi
(i335)-89 Bearing in mind that these documents continued to be issued both
before and after enemy incursions, and that the last one dates from 1335,
a mere two years before the final Mamluk conquest of Ayas, it is clear that
the Armenians did their best to maintain and expand the economic life of
the city until the very end. That this policy was highly successful can be
seen from Marco Polo's description of Ayas in 1295 'as the market for all
the riches of the East. All the spicery, and the cloths of silk and gold, and
other valuable products that come from the interior are brought to that
city ... Whoever would travel to the interior takes his way by this city of
Layas'.90
This does not fit in with the idea that the Armenians ignored the cities
on the plain, preferring to live only in the mountains and the countryside.
It has been argued throughout this chapter that they actually built or
garrisoned a wide range of fortifications, including city defences, and were
prepared to live in both rural and urban settlements. The purpose of all
their major strongpoints, regardless of whether they were walled towns,
citadels or mountain castles, was to provide shelter for local troops and
civilians whilst at the same time preventing external attackers from making
permanent territorial conquests. It is true, however, that mountain castles
like Vagha and Servantikar proved most successful in achieving this, and
were therefore largely responsible for the ability of Armenian kings to retain
their independence until as late as 1375. During the first half of the thirteenth
l6j
century Leon II and Hethoum I may well have constructed new fortresses
which were specifically intended to watch and defend the key mountain
passes. The impressive remains of some of these strongholds should not,
however, lead us to believe that the Armenians only ever built mountain
castles, and that these were all specifically designed to fit into an intervisible
network. Instead Cilician Armenia seems to have had more in common
with the Holy Land, where a variety of urban defences and individual castles
continued to protect Prankish lands until Muslim armies finally overcame
them through sheer weight of numbers.
FORTIFICATIONS
AND I N T E R N A L SECURITY
169
alliance which made them the arch-enemies of Leon It's own anti-Greek
Roupenid family to the east and south.4 During the twelfth century Roupenid
efforts to overcome the Hethoumids were repeatedly thwarted by the sheer
strength of Lampron, which was unsuccessfully besieged in the early 11705
and again about ten years later.5 It was even claimed that 'all the lords of
this castle rebelled because they knew it was impregnable'.6 Bearing in mind
that Leon II's coronation list mentions another forty-four local lords, all
of whom held at least one fortress,7 it is easy to see why Leon II and his
successors feared the possibility of baronial rebellion, even if all castles were
not as strong as Lampron or their owners as fiercely independent as the
Hethoumids.
Leon II pursued a number of policies intended to strengthen his own
authority over the nobility. The royal domain itself, built up through conquest by Leon's twelfth-century Roupenid predecessors, represented a good
starting point, for sites like Sis and Ayas gave the king far greater economic
(and therefore military) power than any other individual lord. However,
the history of Lampron showed Leon that it was not enough to try to
intimidate difficult barons; their castles had to be confiscated. In 1201 he
came up with a plan to bring down the Hethoumids which relied on subtlety
rather than brute force:
With this aim in mind, he sent to Hethoum, son of Oschin, a message to trick
him: 'I want to establish a bond of friendship with you', he told him, 'and give
Phillipa, daughter of my brother Roupen, in marriage to your oldest son Oschin',
Hethoum accepted this suggestion. The celebrations for the marriage were to
take place at Tarsus. When the Hethoumids had arrived in this city with all their
relatives and children, King Leon seized them and occupied Lampron without
any bloodshed. After having imprisoned Hethoum for a while, he released him,
gave him a number of villages, and treated him with good will from then on. As
for Hethoum, he showed himself to be a loyal vassal.8
In some ways it is surprising that Hethoum fell for this trick, for as recently
as 1185 the Hethoumids, aided by their ally Bohemond III, had themselves
ensnared Leon's predecessor, Roupen III (1175-87), by inviting him to 'orgies
with women of ill repute' in Antioch.9 The ruse clearly worked, however,
and while the Hethoumids later regained Lampron, Leon's acquisition of
this one castle considerably increased his own power.
Leon's treatment of Hethoum after Lampron had been taken is also
significant, for it shows that the king was trying to score a symbolic as well
as a strategic victory. This was more than a straightforward dispute between
two rival Armenian clans, as might have been the case a century earlier; it
was an opportunity for Leon to emphasise the superiority which his new
royal status had given him. Hethoum was made to understand that he could
I/O
U N K N O W N CRUSADER CASTLES
no longer act independently, but only received his lands in return for service
and loyalty toward the king. Similarly, in 1198 the forty-five barons present
at Leon II's coronation were encouraged to attend because the new king
'attracted them with his promises, and made them his men with his grants'.10
Clearly Leon wanted these lords to understand that they did not just hold
their castles and estates by conquest or birthright but because he allowed
them to do so. When Vasil of Vaner, one of the barons who had witnessed
Leon's coronation, died without an heir in 1214, the king quickly placed his
lands under royal control so that Leon rather than anyone else could decide
the fate of these important estates.11 Such actions indicate that Armenian
rulers tried to prevent baronial uprisings by confiscating or distributing
castles and fiefs in the manner of west European kings.
The 'westernising' policies pursued by Leon II extended to other areas
of government. It was Leon, for example, who granted many privileges to
Italian merchants, thereby attracting more trade to Ayas, increasing royal
income and improving links between Cuidan Armenian and Europe.12
More significantly in terms of castles, Leon also transferred numerous
estates and strongholds to the Hospitallers and the Teutonic Knights.13
These arrangements were primarily intended to bring military assistance
against the Seljuk Turks, but when it came to the western frontier castle
of Silifke they may also have provided Leon with a means of removing an
important stronghold from another independently-minded baronial family.
Silifke was granted to the Hospitallers in 1210 along with considerable
surrounding estates which included the strongholds of Goumardias
(Camardias) and Norpert (Castellum Novum).14 Three years earlier, these
properties had belonged to a local nobleman called Henry and his three
sons Constantine, Joscelin and Baldwin. They had conspired against Leon
who had arrested them and confiscated their lands. The exact cause of this
rift is not known, although the fact that Henry was married to the sister
of John, the Armenian Catholicos, must have been a contributing factor
as Leon and John were in the midst of a violent personal feud at this
time.15 By handing Silifke over to the Hospitallers Leon was placing a key
frontier castle in the care of more trustworthy occupants who were likelier
to remain loyal to their patron.
Two important considerations indicate that this argument should be
used with caution. First, there is no evidence to suggest that other lands
and castles transferred to the Military Orders had originally belonged to
rebellious nobles. The estates of Vasil of Vaner, for example, were eventually
sold to the Hospitallers, but these properties were acquired by Leon II
because Vasil had died without an heir rather than because he had
misbehaved in some way.16 Similarly, the castles of Amouda and Harunia,
i8. Margat: the east double walls of the inner castle, with round keep at the far end.
(Jonathan Phillips)
20. St Hilarin: the Byzantine defences of the lower bailey. (Denys Pringl)
22. Kantara: view of the castle from the south east. (C. Enlart)
28. Karytaina: view of the castle from the east. (Courtaula Institute)
29. Arcadia from the south. The rounded corner tower to the right is thought to
be a Prankish addition to an otherwise Byzantine fortress. (Courtaula Institute}
30. Androusa: the keep. Note the poor quality of the masonry compared to
fortifications in the Holy Land such as Tortosa. ( Courtaula Institute)
31. Thessaloniki: the city walls. Similar Byzantine defences were occupied and
repaired by the Latins at Constantinople and elsewhere. (Courtaul Institute]
33. Corycus: view of the sea castle from the land castle. (R. W. Edwards)
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
1/1
1/2
U N K N O W N CRUSADER CASTLES
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
1/3
by the Mongols, alongside his nephew and nominated successor King Leon
IV, in 1307.27
In the fourteenth century conflicts of this kind continued to break out
with depressing regularity. In 1329 Oschin, lord of Corycus and father-in-law
of King Leon V (1320-41), was murdered, perhaps because the king felt he
had gained too much power, or because he believed that this nobleman
had been responsible for the deaths of his predecessor King Oschin (1307-20)
and the latter's sister.^ By this point religious disputes were also escalating,
for in 1309-10 King Oschin had been forced to break up a large demonstration at Sis against a plan to accept papal authority and introduce Catholic
practices within the church.29 This plan would have attracted more western
support against the Muslims, but the hostility it aroused indicates that it
was risky for any ruler to foster too close links with the Latins. Leon V
married Constance of Aragon, widow of Henry II of Cyprus (1285-1324),
in an effort to gain help from this quarter, but he was apparently murdered
as a consequence. A similar fate befell Henry II's nephew Guy of Lusignan,
who succeeded Leon V as Constantine IV (1342-44). Even though Guy's
mother was a member of the royal Armenian family, he was quickly assassinated and had 'an ephemeral reign, because the troops rebelled'.30 These
actions were short-sighted, for when Constantine VI (1367-73) eventually
succeeded in establishing an alliance with Peter I of Cyprus (1359-69), the
latter was able to halt an attack on Corycus by the nearby Turkish emirate
of Karamania.31 Instead, the chronicler Jean Dardel claimed that the middle
decades of the fourteenth century witnessed such a constant succession of
internal feuds that some barons were still plotting to overthrow the last
king, Leon VI (1374-5), even as the Mamluks were conducting the final
siege of Sis.32
Although Jean Dardel was doubtless prone to exaggeration, the Armenian
rulers generally failed to overcome the religious and political rifts which
divided their kingdom. Leon II's attempts to try to reduce the power of
noble families like the Hethoumids had only achieved relative success for
as long as Leon himself was alive. The anarchy which followed his death
and the ease with which Constantine was able to place first Philip and then
Hethoum on the throne suggest that Leon's authority had only been based
on his own dynamic personality. Indeed, Constantine's ruthless treatment
of Isabelle and Philip shows that Leon had not instilled the barons with a
sense of respect for the institution of monarchy, and that he had failed to
prevent the Hethoumids from continuing to act as though they were above
the law. While these problems were kept in check during the long and
relatively stable reign of Hethoum I (1226-69), they reappeared once regular
Muslim attacks began to sap the economic and military power of Hethoum's
1/4
17
l/O
UNKNOWN C R U S A D E R CASTLES
had belonged to an individual baron in 1198 but was given to the Order by
Hethoum I and Queen Isabella in 1236.5 Once again, the timing of this grant
probably had a political dimension to it, for Hethoum had been supported
by the Teutonic Knights during the succession dispute of the 12203 and
wished to remain on friendly terms with the imperial forces of Frederick
II active in the Holy Land at that time.6 Architecturally, Harunia was smaller
than Amouda and amounted to little more than an elongated keep with a
cramped central courtyard, two floors of shooting galleries and a large
rounded tower in the north-west corner. The only possible traces of Latin
occupation are to be found in this tower, which may have been repaired
by the Teutonic Knights and perhaps even used by them as a chapel.7 In
addition, the Order probably held even smaller fortified houses or towers
in the Cilician countryside, for both Amouda and Harunia had been granted
to the brothers along with considerable surrounding estates. In 1212, for
example, Willbrand of Oldenburg noted that the Teutonic Knights were in
possession of Cumbethefort, a village situated somewhere between Misis
and Tarsus which had been formally handed over by Leon II at the same
time as Amouda.8 According to a document dating from 1271, the Order also
possessed a toll station in the vicinity of Servantikar known as the Black
Tower.9 This minor stronghold has recently been identified as Hasanbeyli,
a medieval tower located near the eastern approaches to the Amanus Gates.
This identification implies that the Teutonic Knights held properties quite
far beyond the Cilician Plain and cut off from estates around Harunia itself
by the Amanus mountains and by lands belonging to the lords of Servantikar. It may therefore be that the Black Tower lay to the west of Servantikar
nearer Harunia in a spot which has not yet been identified.10
This brief description of properties held by the Teutonic Knights gives
us some idea of what functions their Armenian castles were expected to
fulfil. It is clear that on one level they held no great military significance
in their own right, being simply a means of maintaining good relations
with the German emperors and thereby legitimising first the coronation of
Leon II (1198) and then the reign of Hethoum I, who only acquired the
throne by forcing Leon's daughter to marry him (1226).n This applies most
of all to Amouda, which lay too far away from any frontiers for its garrison
to take rapid action against hostile invaders. It may therefore have been
granted to the Teutonic Knights in return for more general political or
military help rather than any local strategic support. On the other hand,
Harunia's location relatively close to the Amanus Gates indicates that this
castle did have an immediate military role to play, something which
Hethoum I no doubt had in mind when he referred to the Knights' constant
struggle with 'the enemies of Christ' in his donation charter.12 Seljuk troops
177
had already used the Amanus Gates to attack the region in the past,13 and
during the 12205 and 12305 there were more raids which put the Armenians
under severe pressure.14 Later attacks through the Amanus Gates by the
Mamluks were to show just how common it was for Muslim armies to use
this route when approaching from the south east. Presumably Hethoum I
hoped that the Teutonic Knights would help him to stop hostile incursions
of this kind either by drawing the Muslims into a lengthy siege of Harunia
itself, or by sending troops to defend the Amanus Gates in the same way
that he himself did against Baybars in 1266.15
If the intention had been to secure the eastern frontier of Cilicia against
future Muslim attacks, Hethoum and his successors must have been sorely
disappointed. As we have already seen, in 1266,1275 and 1298 large Mamluk
armies could not be prevented from passing through the Amanus Gates
and then ravaging the Cuidan Plain. Indeed, in 1266 Amouda itself was
taken and sacked by the Mamluks. Bearing in mind that the fortifications
of Harunia were no stronger, these castles must have been too small to halt
attackers by forcing them to stop and undertake lengthy sieges.16 To some
extent this problem could not have been foreseen by Leon II and Hethoum I,
for the size and professionalism of Muslim armies grew steadily in the
r/8
course of the thirteenth century. The steady increase in Muslim power also
had wider implications, for the Teutonic Knights had castles to defend in
the Holy Land, which lay closer to Egypt and presumably took precedence
in terms of troops and resources. They may therefore have been prevented
from constructing more fortifications in Cilicia or sending more soldiers
there. The loss of Acre in 1291 ultimately persuaded them to concentrate
on their Baltic activities.17
These problems should not obscure the fact that Amouda and Harunia
were important in other ways. First, it is clear that these strongholds gave
the Teutonic Knights a certain amount of influence over Armenian politics,
for Leon II and his successors did not wish to offend an Order which gave
them a link with the German emperors and was so closely associated with
their elevation to royal status. The Teutonic Knights received lands and
enjoyed royal protection under both Leon II and Hethoum I, even though
the latter ruler gained the throne against the wishes of Leon's designated
heiress Isabelle. Hethoum appears to have been far more concerned about
remaining on friendly terms with the Teutonic Knights than the Hospitallers,
who were forced to deliver the castle of Silifke to him because they supported
Isabelle during the upheavals which followed the death of Leon II.18
Amouda, Harunia and their surrounding estates also provided the Order
with a major source of income. Harunia, for example, was granted along
with its own small town, numerous surrounding villages and abbeys, plus
many estates, orchards, woods and waterways.19 Amouda also received a
steady supply of fish from the adjacent Ceyhan river, whilst the claim that
over 2OOO people were sheltering there during the Mamluk raid of 1266
implies that this was a well-populated area providing plenty of scope for
agriculture.20 Indeed, when the Teutonic Knights were given Amouda in
1212, they had only been in existence as a Military Order since the 11905,
and during their early years it must consequently have represented one of
their most lucrative possessions. The value of their Armenian properties
was also increased by the fact that the Teutonic Knights were exempt from
paying taxes when buying or selling various goods, even though they themselves could impose tolls on others travelling through their lands.21 Amouda
may have been used for this purpose because of its close proximity to the
Ceyhan river and its location between Sis and the Amanus Gates.22 At the toll
station known as the Black Tower the Knights who garrisoned Harunia
made a good income from people who had emerged from the Amanus
Gates and were travelling further north.23 As time went by the Armenian
fortifications belonging to the Teutonic Knights doubtless had more significance as centres of local farming, trade and defence, and as a way of
maintaining good relations with the German emperors, than they did as a
THE M I L I T A R Y
ORDERS
1J9
l8o
UNKNOWN CRUSADER
CASTLES
The history of Silifke during this period can be used to illustrate the
various ways in which Leon hoped that closer links with the Military Orders
would benefit him. It has already been shown that until 1207 this castle,
along with Goumardias and Norpert, belonged to the Armenian baron Henry
and his three sons Constantine, Joscelin and Baldwin, but that they were
then deprived of their lands for plotting against the king.31 Leon may have
felt that granting these castles to the Hospitallers provided him with an
opportunity to present them to someone who did not belong to a strong
local clan, thereby reducing the threat of baronial unrest. However, Leon
was probably far more concerned about the Seljuks, and as early as 1207 he
presumably realised that Henry and his sons were not only disloyal but also
lacked the resources to deal with the growing Muslim threat. The fact that the
Hospitallers effectively rebuilt Silifke implies that under baronial control this
castle had been too weak to withstand a siege and was in desperate need of
improvement. Henry's rebellion may indeed not have been entirely unwelcome, for it gave Leon an excuse to transfer Silifke to a body which could
rectify the situation and also contribute urgently needed troops to the army.
Other transactions between Leon II and the Hospitallers during this
period confirm that the long regarded them as an important source of
military and financial support. In 1214 the Order provided Leon with 20,000
Saracen bezants in return for the right to collect revenues from several
other estates, including the port of Canamella, which lay along the coast
between Alexandretta and Misis. These estates were to be handed back to
Leon, provided he could repay his loan within two years; otherwise the
Order could keep them.32 At this time the Hospitallers also paid a sizeable
amount of cash for the lordship of Vaner, whose original Armenian lord,
Vasil, had died without an heir. It has been suggested that these estates
were centred around the fortress of Gkvelioglu, an impressive ruin situated
on an outcrop south of Misis, but this cannot be verified as the relevant
documents do not mention fortifications.33
The scale of Leon IFs dealings with the Hospitallers suggests that he was
experiencing financial difficulties. Partly caused by his struggle with the
Seljuks, they were compounded by Leon's long and costly campaign to gain
control over Antioch. Until 1201 this city had normally been controlled by
Bohemond III, but his death in that year caused a succession dispute to
break out between his son Bohemond IV and his grandson RaymondRoupen, who was the product of a marriage between Bohemond Ill's
deceased older son Raymond and Leon II's niece Alice. Inevitably Leon took
the side of Raymond Roupen, who as a small boy became little more than
a pawn in his great uncle's plans to extend Armenian power into Syria.
Bohemond IV, on the other hand, was supported by the Templars, which
THE M I L I T A R Y O R D E R S
l8l
doubtless accounts for Leon II's growing involvement with the Hospitallers.34
Originally they had supported Bohemond IV, but in 1204 they changed
sides.35 The need to preserve their support must have influenced Leon's
decision to grant them Silifke six years later. In 1207 Raymond Roupen also
promised them Jbala, a coastal town within the principality of Antioch
which was actually held by the Muslims at that time.36 As a result the
Hospitallers eventually occupied half of Jbala, although there is no evidence
that they ever recaptured Bikisrail, another ex-crusader stronghold to the
north east of Margat granted to them by Raymond Roupen in 1210.37 These
grants indicate that Raymond Roupen saw himself as the legitimate ruler of
the principality, but he was unable to take control over Antioch itself until
1216, when the Hospitallers helped him to overcome local opposition by
garrisoning the ramparts and the citadel. Three years later anti-Armenian
factions within the city joined forces with Bohemond IV and drove Raymond
Roupen and his Hospitaller allies back to Ciucia, thereby ending any hopes
of incorporating Antioch into Leon II's kingdom.38
Leon II's motives for fostering friendly relations with the Hospitallers
were therefore similar to those which drove him and his successor to
establish links with the Teutonic Knights. Both these Orders proved to be
useful political allies, for one helped bring Leon a royal crown from the
German empire, whilst the other gave him a strong local ally in his struggles
with Bohemond IV. Like their German counterparts at Harunia, the Hospitallers were also granted Silifke and its adjoining estates in order to defend
a vulnerable frontier of the Cilician Plain against the Seljuks. The impressive
archaeological remains of Silifke, its successful defence in 1216, and the
Hospitallers' offer of 400 horsemen to the royal army all suggest that they
initially fulfilled this objective with more success than the Teutonic Knights.
This was partly because they had greater economic and military resources
to draw on, and partly because they had much wider experience of managing
such territories from the Holy Land, where Margat and Crac des Chevaliers
both formed centres of vast frontier lordships which the Hospitallers used
to wage war against neighbouring Muslims and to protect Prankish lands
nearer the coast.39 It is clear that after 1210 the new castle of Silifke began
to take on a similar role, for it became the residence of the local preceptor.40
Like the estates around Amouda and Harunia, the lands which were attached
to Silifke must have generated very considerable income, thereby offsetting
the cost of rebuilding the fortress and providing troops for the royal army.
Although there were clear benefits both for the Armenians and for the
Hospitallers, Leon's death in 1219 quickly caused all these arrangements to
unravel. In that year Raymond Roupen's control of Antioch ended, and his
subsequent attempt to seize the Armenian throne, carried out 'with the
l82
UNKNOWN CRUSADER
CASTLES
THE M I L I T A R Y O R D E R S
183
between the masonry of the lower bailey and numerous other Templar sites,
including Darbsak. Similarly, the various defensive and residential structures
of the upper bailey are arranged in a compact, typically Armenian fashion,
but a more detailed study of individual architectural features within these
buildings reveal certain similarities with the Templar citadel of Tortosa.48 It
would be wrong, therefore, to argue that Baghras is largely an Armenian
construction simply because it is a mountain castle situated in a remote area.
Indeed, it has been claimed that 'the Armenian presence here is no more than a
flirtation' and that the importance of the site to Leon II has been exaggerated.49
This conclusion does not, however, concur with the historical evidence.
The Arab historian Ibn al-Athir, for example, wrote that Leon 'marched on
this place which was near his territory. He rebuilt Baghras carefully and
stationed a garrison there to carry out raids on the surroundings'.50 According to Willbrand of Oldenburg, Baghras was 'a very powerful castle,
with three strong walls and towers around it, situated in the last mountains
of Armenia. It carefully guards the entrances to that land, whose ruler, the
king of Armenia, holds it'.51 These descriptions suggest that Leon's presence
at Baghras from 1191 onwards was more significant than a 'flirtation', and
that contemporaries considered the castle to be strategically very important.
This is confirmed by Leon's own efforts to hold on to Baghras for as
long as possible rather than to return it to its original Templar owners.
From the 11905 onwards Leon was often prepared to discuss the issue of
Baghras but frequently used stalling tactics, such as asking the pope to
arbitrate in the matter, rather than dealing with the Templars directly. In
1199, for example, Leon tried to persuade Innocent III that Baghras was
his by right of conquest, but the pope quickly rejected this.52 During the
ensuing years Innocent's representatives were kept almost constantly busy
trying to solve both this problem and the parallel Antiochene succession
dispute between Bohemond IV and Raymond-Roupen.53 In 1203 the situation worsened when Leon tried to capture Antioch itself, but he was
pushed back by a combined force of Templars defending their headquarters
and supporters of Bohemond IV.54 Subsequent fighting provoked Leon into
taking temporary possession of Roche Roussel and Roche Guillaume, the
remaining Templar castles in the Amanus march, and then justifying his
actions by portraying the Order's efforts to retrieve Baghras as unprovoked
aggression against the Armenians.55 Two years later Leon also made a failed
attempt to recapture Darbsak from Aleppine forces, whose friendship with
Bohemond IV at this time provides us with an interesting example of a
Franco-Muslim alliance against a fellow Christian.56 Needless to say, these
activities did not please the Templars, and in 1211 another period of failed
negotiations and sporadic warfare culminated in a major Templar attack
184
Baghras. Ground-plans of the castle, scale 1:10,000. A. Lower fortress and lower floor
of upper fortress. B. Upper floor of upper fortress (ground-floor premises stippled)
with the earlier bunding phase (Byzantine?) in black and the portions dating from the
twelfth to thirteenth centuires cross-hatched, i. Lower fortress. 2. Forecourt and
gatehouse. 3. Lower gallery. 4. Palace premises. 5. Great Tower. 6. Magazine chambers.
on the area, including perhaps Baghras itself.57 By this point the pressure
applied by the papacy, the Templars and Bohemond IV, combined with
the Seljuk threat in the west and the financial shortages implied by Leon's
transactions with the Hospitallers, all conspired to bring about a change
in Armenian policy. In 1212 Leon agreed to restore all Templar properties,
and in 1213 the excommunication which had previously been imposed upon
him was lifted by Innocent III. Even then Leon still managed to stall the
return of Baghras until 1216, twenty-eight years after Saladin's invasion.58
Leon's determination to hold on to Baghras for such a long time, and
his ultimate decision to give it back to the Templars, reveal, much about
THE M I L I T A R Y O R D E R S
185
the strategic importance of the castle. This stronghold's original twelfthcentury purpose had been to protect Prankish Syria against incursions from
the north, but Willbrand of Oldenburg noted that 'it overlooks Antioch
directly1.59 It could therefore be used to spot attacks aimed at Ciucia by
Bohemond IV. Alternatively, it could act as a starting point for Armenian
incursions toward Antioch, and it seems likely that Leon utilised Baghras
in this way during his failed attack on the city in 1203. Ibn al-Athir's claim
that the Armenian defenders of the fortress carried out frequent local raids
also suggests that Leon regarded Baghras as an important military asset in
his long campaign to gain control over Antioch.60 On the other hand, in
1226 the reinstalled Templar garrison successfully withstood an Aleppine
besieging army arriving from the east. Although this force had actually
been encouraged to attack Baghras by the Armenians and therefore posed
no threat to them, its failure to capture the fortress indicates that it could
also be used to defend the Cilician Plain against the Muslims.01 Depending
on who occupied it, Baghras could be used to protect or attack both
Antioch and Ciucia. It seems reasonable to conclude that whilst some
scholars have overestimated the extent of its Armenian remains, others
have underestimated its strategic importance.
Leon's actions also help us to understand the military role of other Templar
fortifications near Baghras. In the same way that Baghras guarded the Beln
Pass, so the Templar castle of Darbsak controlled the entrance to a more
northerly defile through the Amanus mountains during the twelfth century,
but it was never recaptured by the Order after 1188.62 Nevertheless Leon's
attempt to retake the fortress in 1205, another failed crusader attack in 1237
and Hethoum I's eagerness to reoccupy the site after the Mongol invasion
of 1260 all indicate that Darbsak had once been a key Templar stronghold
whose loss potentially threatened both Prankish and Armenian security.63
On the Cilician side of this valley, meanwhile, the Templars still retained
Roche Roussel (Hadjar Shoglan/Chilvan Kale) in the thirteenth century. This
castle was located on yet another steep outcrop which required few manmade defences, although some of the remains there, most notably a square
keep and a chapel, may date from the Templar occupation.64 It was from
here that the Templars launched their failed expedition against Darbsak in
1237, illustrating that the role of Roche Roussel had changed dramatically
since 1188. Before that date Rouche Roussel, Darbsak and Baghras had all
worked together to create a network protecting Antioch from the north, but
in the thirteenth century Roche Roussel found itself caught between the
Muslims in the east and the Armenians in the west. Presumably the unidentified neighbouring castle of Roche Guillaume suffered a similar fate
after 1188, when it continued to be guarded by the Templars.65
l86
THE M I L I T A R Y ORDERS
l8j
l88
U N K N O W N CRUSADER CASTLES
PART FOUR
Prankish Greece
1204-1380
18
Warfare
The Fourth Crusade (1202-04), consisting of a land army composed of
French and Italian troops and a powerful Venetian naval fleet, had originally
been planned as an offensive against Egypt. Through a combination of greed,
political intrigue and mutual distrust, the expedition ended up attacking the
Greeks instead.1 After the Crusade captured Constantinople in 1204, those
members of the expedition who chose to stay in Greece undertook the conquest of what remained of the Byzantine empire. This resulted in the creation
of several new Latin states around the Aegean. The Latin empire itself
encompassed Thrace and the northern fringes of Asia Minor, while Macedonia and parts of Thessaly formed the kingdom of Thessaloniki. Further
south, Byzantine Attica and Boeotia were turned into the Prankish duchy
of Athens, and below it the principality of Achaia covered the Peloponnese.
Meanwhile the Venetians established a number of important trading colonies
around the Aegean, particularly at Modon, Coron, Constantinople and Crete.
They also dominated numerous islands which lay either directly within the
Venetian sphere of influence or were colonised by individual Venetian
citizens. These included Corfu, Cephalonia and Euboea, as well as the
Cyclades, which formed the duchy of the Archipelago under the Saudo
dukes of Naxos. Eventually other islands came to be controlled by rival
powers, most notably the Genoese, who held Chios and several neighbouring
trading posts between the fourteenth and sixteenth centuries. In addition,
from 1306 onwards the Hospitallers subjugated Rhodes.
Leading Greeks who had fled from Constantinople during the Fourth
Crusade also created three new states situated on the fringes of the Byzantine
world. The capital cities of these states lay at Trebizond along the Black
Sea, at Nicaea in Asia Minor and at Arta in Epiros. All three of these
territories, and in particular the latter two, saw themselves as the natural
heirs to the Byzantine empire and consequently often clashed both with
each other and with their unwelcome new Latin neighbours. To the north
of Constantinople, the Franks were also threatened by the Bulgars, who
had been opponents of the Byzantine empire before 1204 and now continued
to launch frequent attacks into northern Greece in conjunction with an
aggressive tribe of horsemen known as the Cumans.2
192
WARFARE
193
through the Byzantine empire, pillaging and looting as they went, before
they eventually found themselves in the pay of Duke Gautier. However,
after some initial successes an argument broke out between the duke and
the volatile Catalans over pay, and as a result they took to the field against
Gautier and his Achaian allies. This decisive encounter, traditionally
known as the battle of Cephissus but actually fought at Halmyros (southern
Thessaly), took place in March 1311 and led to the death of the unfortunate
Gautier along with virtually all of his Athenian knights. In its wake Gautier's
lands were conquered by the Catalans, who then colonised and ruled the
duchy of Athens until the late fourteenth century. During this period
the Catalans acknowledged the overlordship of the kings of Aragon, who
held the ducal title and ruled Athens through representatives whom they
regularly sent out to Greece.3
The dramatic advances made by the Greeks against the Latins had an
important effect on the internal history of Achaia. Originally the Prankish
rulers of this crusader state had recognised the Latin emperors as their
overlords, but after Michael VIII recaptured Constantinople in 1261, and
gained control over Mistra, Old Mania and Monemvasia the following year,
it became essential for the Villehardouin princes of Achaia to find a powerful
new suzerain who could aid them against the Greeks. During the 12605,
therefore, William II of Villehardouin (1246-78) allied himself with Charles
of Anjou, brother of Louis IX of France and ruler of Naples (1266-85). As
part of this alliance it was agreed that, if William did not produce a male
heir, Achaia would eventually pass under Charles's control, which was
exactly what happened following William's death in 1278. These events
turned out to be a mixed blessing for the Peoloponnese, for on the one
hand Charles and his Angevin descendants did provide sporadic military
support for the Peloponnese (as William had hoped), but on the other
relatively few of them ever visited the region, preferring instead to send
royal representatives from Italy or to grant parts of Achaia to their own
followers. Perhaps the most important such figure was the wealthy Florentine
lord Niccolo Acciajuoli, who acquired large parts of northern Morea, including the castellany of Athens, between the 13308 and 13505. Toward the
end of the fourteenth century Niccolo's descendants also made significant
advances against the Catalans of Athens, who were detested in France and
Italy because of their close links with the Aragonese, arch-enemies of the
Angevins and the Avignon papacy.
The fact that the Angevin rulers of Achaia resided in Naples rather than
Greece gradually had the effect of weakening central authority in the Peloponnese, so that the region's late fourteenth- and early fifteenth-century
history was reduced to a series of internal clashes between various local
194
UNKNOWN CRUSADER
CASTLES
factions. In the north the Florentine lords of Corinth, allied with a new
company of Navarrese mercenaries who had arrived in the 13808, fought
against the Catalans. In the south east the Greeks of Mistra continued to
advance, whilst Achaia itself sporadically found itself disputed between rival
Angevin claimants. As the Turkish threat grew, the Venetians and the
Hospitallers also became more involved in the politics of Latin Greece,
sometimes placing entire areas under their own protection. Indeed, from
the 13905 onwards, local Christian squabbles became little more than an
irrelevance as the Ottoman Turks began to overrun the entire region regardless of whether it was controlled by Greeks, Italians, Franks or Iberians.
By 1460 the Byzantine empire had disappeared and most of the Greek
mainland had been incorporated into the Ottoman empire. During the next
two centuries all remaining Christian islands in the eastern Mediterranean
also fell to the Turks.4
During this period the military strength of the Latins and their opponents
varied considerably. The relatively short life span of the kingdom of Thessaly
and the Latin empire, for example, indicates that here western settlers found
themselves under extreme pressure from the very beginning. It has been
estimated that the total number of cavalry, including both knights and
mounted sergeants, available to the first Latin emperor, Baldwin of Flanders
(1204-5), only came to between 500 and 1000 horsemen.5 Contemporary
accounts bear this out, for in late 1204 a mere 120 knights left Constantinople,
crossed the Bosphorus and began to conquer as many lands as they could
from the Greeks of Nicaea.6 Similarly, Baldwin's successor Henry (1206-16)
campaigned against the Bulgars with a total of 400 knights in 1206, and
was accompanied by 260 knights when he advanced against the Nicaean
Greeks in mi.7 Initially these figures do not seem small when compared
with Greek forces, for the largest reliable total given for a contemporary
Nicaean army is that of 2000 cavalry (including 800 Latin mercenaries)
at the battle of Antioch on the Meander, an encounter with the Seljuk
Turks which took place in i2ii.8 It should also be noted that the advances
of 1204 and Henry's campaign of 1211 both ended in Prankish success,
suggesting that Latin and Nicaean forces were evenly balanced at this
time, particularly when the Greeks were also struggling with their Seljuk
neighbours to the east.
To some extent the same may have applied to the Cuman and Bulgar
troops who threatened the northern border of the Latin empire and the
kingdom of Thessaly. In 1208, for example, a mere 2000 Latins allegedly
confronted 33,000 Bulgars in northern Greece, but the fact that this encounter ended in a Prankish victory certainly casts doubt on the latter figure
and suggests that the Bulgars could not possibly have enjoyed such an
196
U N K N O W N CRUSADER CASTLES
WARFARE
197
the principality of Achaia and the Venetian colonies were all smaller in size
and were protected by seas, mountains or other natural features such as
the isthmus of Corinth. Moreover, the further Latin conquerors travelled
from Constantinople, the more they benefited from the increasing regionalism which had been apparent within the Byzantine empire since before the
Fourth Crusade. When Boniface of Montferrat, the man who established
the kingdom of Thessaly, undertook his conquest of northern Greece in
1204, he received a warm welcome from some locals who regarded Constantinople as a distant and corrupt absorber of taxes, and consequently
felt that it was best simply to come to terms with the Latins.14 Similarly,
when the Franks reached the Peloponnese local Greek lords ('archons')
were often allowed to keep their incomes provided that they recognised
their new Prankish masters. By contrast, the volatile political situation in
Constantinople itself before and during the Fourth Crusade, followed by
the flight of many leading Greeks to Nicaea or elsewhere, left little room
for a quiet transition of power in northern Greece.15
The relative stability which existed in south-western Greece during the
first half of the thirteenth century ensured that the rulers of Achaia were
more prosperous and enjoyed greater military power than their Latin neighbours in Thessaloniki and Constantinople. The total number of knights
settled in the region has been estimated at between five and six hundred,
and according to one rather romanticised source Prince Geoffrey II of
Villehardouin (1228-46) kept eighty knights just at his own court.16 These
figures are confirmed by the Chronicle of Morea, if we can believe its claim
that, at the battle of Tagliacozzo (1268), Geoffrey's successor William II
(1246-78) fought alongside his overlord Charles of Anjou with 400 cavalry
brought from Achaia.17 This implies that, even though it was smaller, the
military strength of Achaia was almost equal to that of the original Latin
empire, and that its rulers had enough resources to intervene in conflicts
beyond the Peloponnese. Indeed, in 1236 Geoffrey II of Villehardouin sent
help to Constantinople whilst it was being besieged by the Bulgarians and
the Greeks of Nicaea.18 In 1249 Geoffrey's brother William II joined Louis
IX's crusade against Egypt with a fleet of twenty-four Achaian ships. This
episode took place at roughly the same time as the Latin empress of
Constantinople made a desperate plea to Louis for 300 knights to help her
and her husband defend their capital, and therefore highlights the contrast
in military strength between the Franks of Achaia and those of northern
Greece.1?
The military strength of the principality of Achaia should not be overestimated, however, for it still took the Villehardouins many decades to
subjugate the Peloponnese fully, and they did not capture the impregnable
198
WARFARE
199
of the fighting are still impossible to come by. Thus in 1292 the Chronicle of
Morea stated that the Epirote city of loannina was besieged by a Byzantine
(formerly Nicaean) force of 30,000 infantry and 14,000 cavalry, but that
this entire expedition retreated in panic when Florent of Hainault, prince
of Achaia (1289-97) and ally of Epiros, approached with an army which
probably only contained four or five hundred men!25 Once again, the
Byzantine figures should clearly be disregarded, for the number of western
knights living in this region suggests that most clashes between Greeks and
Latins cannot have involved more than a few hundred horsemen, and
perhaps only a couple of thousand men in total. Indeed, it has been suggested
that during the entire late Byzantine period most Greek campaigns 'probably
involved hundreds rather than thousands of troops'.26
Whilst the Nicaean Greeks eventually succeeded in conquering Achaia,
we have seen that it was the Catalans who finished off the Prankish duchy
of Athens. The ruthlessness with which they did so suggests that they were
unusually well-trained and aggressive soldiers. According to the chronicler
Muntaner the Catalan company consisted of 2500 cavalry plus 4000 Almogavers (light infantry) and 1000 other footsoldiers27 These figures seem
high, although the total of around 2500 horsemen was also given by another
source.28 On the other hand, Muntaner claimed that, when they defeated
Duke Gautier I of Athens in March 1311, the Catalans overcame a Prankish
army containing 30,000 infantry, plus 700 knights of whom all but two
were killed. Even if all the lords of Athens and Achaia had turned up, these
are impossibly high totals and unfortunately cast doubt on the other troop
numbers given by this chronicler.29 The decisiveness of the Prankish defeat
does at least suggest that Gautier had underestimated both the fighting skill
and the size of the Catalan foe, even if the total number of combatants
involved in this encounter was greatly exaggerated.
The Catalans also participated in another form of warfare which affected
much of the Aegean area during the thirteenth and fourteenth centuries:
piracy and raiding. Even before they arrived in the duchy of Athens the
Catalans had a terrible reputation for such activities in Byzantine Greece,
where 'there was not a town or city that was not looted and burnt' by
them.30 From the mid fourteenth century onwards the Turks became an
even greater threat, causing so much destruction that by the 13805 some
Aegean islands had been completely abandoned by their original inhabitants.
Even as early as 1358 a document recording the transfer of Corinth castle
to the Florentine lord Niccolo Acciajuoli makes it clear that the Turks were
having a devastating effect on the region: one of Niccolo's primary duties
was to persuade people to return to the area around Corinth, which had
been overrun so many times by Greek, Catalan or Turkish raiders that
200
U N K N O W N CRUSADER CASTLES
WARFARE
201
John III Vatatzes of Nicaea (1222-54), the Venetians were able to break the
naval blockades around the city and effectively save the Latin empire from
destruction.39 Apart from the Italian city states, other Latin powers also had
warships available to them. In 1236 many vessels from Achaia assisted in
the relief of Constantinople and William II of Villehardouin later contributed
a fleet of twenty-four ships to the crusade of St Louis. At the beginning of
the fourteenth century the Catalans also brought their own fleet to the
Aegean, using it in countless raids along the coasts of northern Greece. In
1305, for instance, after the Catalans had fallen out with their former
employer Andronikos II, they sent five galleys out from their temporary
headquarters at Gallipoli to attack neighbouring Byzantine targets.40
The financial, administrative and political problems of the late Byzantine
state often prevented the Greeks from raising adequate naval forces to deal
with such aggressors. During the 12808, for example, Andronikos II reduced
the size of the Byzantine fleet in order to cut costs.41 However, it would
be misleading to assume that the Greeks were entirely overawed at sea, for
at times, and in particular during the reign of Michael VIII Palaiologos
(1259-82), the Byzantine authorities were able to build up naval forces by
relying on mercenaries. Thus in the 12705 Michael VIII employed the Latin
pirate Giovanni de lo Cavo to act as his admiral in the Aegean.42 Latin
seamen also did not have matters entirely their own way because of the
growing threat of the Turks. Even the ruthless Catalans had to be aware of
this danger, for during the winter of 1303-4 they sent their fleet to a secure
winter anchorage at Chios 'because the Turks, with barques, ravage all [the
Aegean] islands'.43
This summary of thirteenth- and fourteenth-century warfare around the
Aegean makes it clear that there were many motives for the Latins who
lived there to construct or occupy castles. Although exact troop numbers
are almost impossible to establish, most of the Greek or Latin armies active
in the region only contained between one and five hundred horsemen. Even
when additional foot soldiers accompanied them, campaigns were rarely
undertaken by more than one or two thousand men. Consequently the
Latins who settled in Greece after the Fourth Crusade were not as heavily
outnumbered by individual opponents as some of the contemporary chroniclers would lead us to believe. There were, however, ultimately too few
Latins living in northern Greece to halt the combined attacks of many
different Nicaean, Epirote, Cuman and Bulgarian armies. The fall of the
duchy of Athens in 1311 suggests that Prankish settlement was so fragile that
a single defeat in a pitched battle could seal the fate of an entire crusader
state. Eventually the Franks proved equally incapable of halting the gradual
loss of Achaia after the Greek acquisition of Mistra in 1262. In addition,
202
the Aegean was almost constantly affected by some form of localised raiding,
rebellion or piracy. While this did not necessarily threaten national security
in the short term, it ultimately ground down the economic and military
strength of states such as Achaia. In short, this was an extremely insecure
world, and in order to protect themselves against it Latins needed fortifications to make up for their lack of troops and to defend their property
against the constant threat of enemy attack.
19
Military Architecture
The sheer quantity of medieval ruins still standing in Greece and Turkey
bears witness to the turbulent history of the region. This should make the
task of identifying and describing crusader fortifications relatively simple,
but this is not the case. There are a number of problems which make the
study of Prankish castles around the Aegean fraught with difficulties, not
least the lack of historical sources. Bearing in mind that there are certain
periods, particularly in the fourteenth century, when 'we cannot reconstruct
completely' the history of southern and central Greece in general, it is
hardly surprising that the fate of individual castles frequently remains obscure for decades at a time.1 Moreover, when the Latins who settled in this
area constructed or repaired fortifications, they adopted the same tactic of
recycling older masonry as their Byzantine predecessors. As a result, it is
often impossible to tell whether a wall composed of classical masonry blocks
robbed from a much older structure is the work of Greeks or Latins. Trying
to establish if a strongpoint is Prankish, Catalan, Navarrese or attributable
to another western dynasty (such as the Acciajuoli lords of Corinth) is even
harder.
Inevitably, these problems have led to arguments about dating and origin
amongst the few archaeologists and historians who have studied the subject.
Such arguments can perhaps best be summed up by looking at the castle
of Androusa, located at the heart of the principality of Achaia in the
south-western Peloponnese. The present remains of Androusa include sections of a single curtain wall, flanked by several rounded, square and
polygonal salients, as well as a large tower whose design and unusually thick
walls suggest that it acted as a keep. The castle's masonry is typical for
medieval Greece: a mixture of small, uncut stones quarried locally, shards
of pottery and other recycled fragments found at the site, and larger,
presumably classical, masonry blocks used to strengthen corners, doorways
and other weak spots.2
In his monumental work, on the castles of Morea, Antoine Bon dates
this stronghold to the mid-thirteenth century, with possible additions in
the fourteenth century. The Aragonese version of the Chronicle of Morea
appears to confirm this when it states that Androusa was built by William
204
II of Villehardouin, who reigned from 1245 until 12/8.3 But there are problems with this conclusion. A number of features at Androusa, and in
particular the use of decorative brickwork and the presence of a pentagonal
open gorge tower along the north curtain, are typically Byzantine, but
extremely rare in Prankish military architecture.4 Bon accounts for these
anomalies by suggesting that the crusaders used local craftsmen, which may
well be true, but he never seems to consider the possibility that some of
the defences at Androusa were constructed either before 1205 or during the
brief period of Greek domination in the area between the 14205 and 14505.5
Likewise, when discussing the date of the oldest fortifications at Mistra,
Bon again lets the historical evidence take precedence, basing his conclusion
that the castle must have been constructed 'in its entirety' by William II
on a statement to that effect in the Chronicle of Morea.6
In a clear reference to Bon's work, the Byzantine scholar David Winfield
challenges these findings on archaeological and architectural grounds,
asserting that at Mistra 'there is nothing in the hill top citadel and not
very much on the site as a whole to indicate that Villehardouin did more
than reoccupy a typical Byzantine hill town'. Winfield then goes on to
claim that in Greece 'a number of fortifications that are probably Byzantine
have been assigned without question to the Franks'.7 Both scholars therefore
take somewhat extreme views, with Bon preferring to believe the Chronicle
of Morea even if this contradicts the archaeological evidence, and Winfield
taking a more Byzantine stance which immediately assumes that the written
sources must be wrong. In trying to resolve this argument we immediately
stumble upon two further problems. First, it should be noted that many
of the key fortifications occupied by the Latins between 1204 and 1380
were subsequently taken over, repaired or even totally rebuilt by the
Venetians. By 1715 the medieval defences of sites like Corinth, Nauplia,
Monemvasia and Patras had all been partially or totally hidden below new
ramparts and artillery bastions.8 From around 1500 onwards the Turks
likewise demolished some older crusader defences, as appears to have been
the case at Boudonitza,9 or carried out their own alterations and improvements, as can be seen at Modon.10 Secondly, and perhaps as a result of all
these other difficulties, subsurface archaeological exploration of medieval
strongpoints in Greece has barely even begun. It cannot yet compete with
the kind of detailed studies which have been carried out on Crac des
Chevaliers and other sites in the Holy Land. As a result it would be
pointless, and in most cases virtually impossible, to describe the appearance
and history of every stronghold in Prankish Greece. It is more useful to
give specific examples of medieval fortifications in order to shed light on
their various functions.
MILITARY
ARCHITECTURE
205
All crusader fortifications in Greece can be placed in three general categories: those which the Latins built from scratch on sites showing very little,
if any, evidence of previous occupation; those which were constructed
incorporating older Byzantine or classical structures; and, finally, those
which were already in good condition at the time of the Fourth Crusade,
and were simply reoccupied by the Franks. In Thrace, Macedonia, Asia
Minor and eastern Thessaly most strongholds fall into this latter category,
mainly because the crusaders were too poor or in control too briefly to
carry out their own repairs and building programmes. Thus Appolonia,
located on the furthest outskirts of Prankish territory facing Nicaea, was
referred to as 'one of the strongest and most imposing castles to be found'
as early as 1204, and clearly needed few improvements before the crusaders
could garrison it with their own troops.u North of Apollonia, a series of
equally well defended coastal strongholds formed the backbone of Prankish
power in Asia Minor. Perhaps the most important of these was Spiga (Pigae),
one of the last Latin outposts to fall to the Nicaean Greeks in 1225. Located
on a promontory forming a good natural harbour, Spiga was protected by
a powerful Byzantine rampart flanked by a series of closely set pentagonal
towers, which must have been maintained by the Franks and are still largely
preserved to this day.12
On the European side of the Bosphorus, the crusaders inherited some
even more spectacular Byzantine fortifications, particularly at Constantinople
itself. This city's landward side was protected by the famous double walls
of Theodosius II (408-50), as well as a number of later structures, including
the rampart of Manuel Komnenos (1143-80), which was located near the
2O6
Blachernae palace and bore the brunt of the Prankish land offensive in i2O3.13
That these defences had been kept in good condition is attested to by the
events of the Fourth Crusade, when the Latins failed to take Constantinople
from the west, being forced to concede that 'never was a city so well
fortified'.14 Interestingly, the Venetians had far more success when they
launched a naval attack on the seaward defences and managed to capture
twenty-five towers along the Golden Horn by lowering wooden platforms
from their ships onto the ramparts.15 In preparation for the second Prankish
siege of Constantinople in April 1204, the Greeks were forced to heighten
these towers 'with two or three wooden storeys', but were still unable to
prevent the Venetians from gaining access to the city in almost exactly the
same spot. This suggests that the single sea wall may have been lower, weaker
and possibly in a far worse state of repair than the land defences. It also
illustrates the huge advantage which Venetian naval power gave the crusaders
over their Greek opponents.16
If the thirteenth-century sea defences of Constantinople were not in a
particularly good state, the Franks were fortunate that the Venetians could
patrol the Bosphorus for them, and that most of their opponents attacked
the city by land. Baldwin of Flanders and his successors were all so lacking
in troops and resources that 'no repairs are attested for the period of the
Latin empire', and it seems that until 1261 the sheer size of the Theodosian
walls alone saved the city from early capture by the Greeks or the Bulgars.17
These observations are confirmed by the fact that Michael VIII Palaiologos
carried out extensive repairs on the ramparts between the 12605 and 12805.18
Contemporary accounts of the Prankish conquest of territories to the west
of Constantinople suggest that many Byzantine settlements in Thrace, Macedonia and Thessaly were similarly well protected at the beginning of the
thirteenth century. Demotika, for example, was described as 'a very fine,
strong and wealthy castle' at this time, while the fortress of Christopoli
(Kavala) was, according to Villehardouin, 'one of the strongest in the
world'.19 Descriptions of the Thracian rebellion of 1205 and Kalojan's subsequent campaigns in the area also reveal that Arcadiopolis, Stenimaka,
Philippopolis, Rousion and Rodosto all had urban fortifications of some
kind. Meanwhile Adrianople was so well defended that Prankish attempts
to capture the city during the initial Latin conquest of Greece failed despite
a sustained siege involving ladders, catapults and sapping.20
Special reference should be made to Thessaloniki, whose vast defences,
begun in the fourth and fifth centuries, formed a triangle around the city,
with an impregnable hill top citadel at its apex.21 The history of these
defences is obscure during the crusader period, although the few facts
available to us suggest that they were in good repair in 1204, and that they
MILITARY ARCHITECTURE
2O/
208
it is likely that one of the few sites which remained untouched by the
newcomers was Monemvasia, located on the east coast of the Laconian
peninsula. Its situation on a vast, sheer-sided rock separated from the
mainland by a narrow stretch of water made it so inaccessible that it was
not captured by the Franks until 1249, following a bitter three-year siege.
By 1262, however, the castle had already been handed back to the Greeks
as part of William II's ransom following the battle of Pelagonia (1259).26 As
a result, the Franks barely had time to integrate Monemvasia into the
principality of Achaia, and it is extremely unlikely that they ever built any
new fortifications there. This is confirmed by the archaeological evidence,
which suggests that the Byzantine citadel is the only pre-Venetian structure
on the site.27
The second category of castles and fortifications covers the vast number
of strongpoints which incorporated a combination of both Prankish and
older Byzantine or classical defences. Structures of this type varied enormously. As one would expect, they occurred most commonly in Achaia and
the duchy of Athens, where Latin rule lasted much longer, but there were
also a few such fortifications further east. At Nicomedia, for example, the
Franks fortified a large Byzantine church and surrounded it with outer
defences in a simple but effective arrangement which withstood at least one
Nicaean siege.28 In 1206 the crusaders also reached Cyzicus, a narrow headland not far from Spiga, and found that 'there had been in ancient times
a fortress with walls, towers and ditches; and they were nearly in ruins.
And the army of the Franks entered it, and Peter of Bracieux, to whom
the land had been assigned, began to rebuild it, and to construct two castles
and two entrances'.2' Other castles in Asia Minor which were built or
repaired using similar methods included Charax, Civetot and Panormos,
all of which lay along the coast.30
Within the European half of the Latin empire, the chronicler Henry of
Valenciennes wrote that in 1208 the emperor Henry decided to reconstruct
the ruined castle of Pamphilon in Thrace. Having defeated a vast Bulgar
and Curtan invasion force, Henry went to this spot, where he and his
marshal swore not to leave 'until the walls had been rebuilt and repaired'.
Local labourers were subsequently recruited and the work was completed
with such speed and determination that a new Frankish garrison had
been installed before the onset of winter. Although this castle must have
been situated somewhere along the main road between Constantinople and
Philippopolis, its exact location has never been established.31 Its successful
reconstruction in such a determined fashion suggests that the reign of Henry
(1206-16) represented the high point of the Latin empire. However, when
reading about the fortification of places like Cyzicus and Pamphilon, one
MILITARY ARCHITECTURE
2O9
is still struck by the impression of great haste, and the lack of money, troops
and resources. Clearly the ephemeral nature of Prankish rule within the
Latin empire meant that it was not only convenient but essential to reoccupy
older Byzantine or classical ruins, where the immediate supply of ready-cut
stone facilitated the construction of new defences in a matter of weeks.
In south-western Greece and the Peloponnese, greater political stability
ensured that former Byzantine and classical sites were repaired, improved
or even rebuilt on a far larger scale. This did not necessarily mean that
crusader fortifications were constructed to a higher standard. At Patras, for
example, William Alemn, the first Prankish lord of the city, considerably
improved the defences of the lower bailey of the Byzantine citadel, but this
was done using masonry robbed from the neighbouring archbishop's
residence. Admittedly William was embroiled in a dispute with the archbishop at this time, and was therefore motivated by political as well as
practical concerns, but his desperate actions also reflect the poverty and
relative insecurity of many Prankish adventurers shortly after they arrived
in Greece32
At other sites, the amount of alteration undertaken by the crusaders
varied considerably, depending on whether these places were in ruins or
were still occupied at the time of the Prankish invasion. A good example
of the latter type of fortification is Kalamata, located on the coastal plain
of south-western Morea. In 1205 the western newcomer William of Champlitte only obtained its surrender by promising the defenders that he would
Patras: one of the many Byzantine strongholds occupied and repaired by the Latins
after 1204.
210
respect their land and property.33 This proves that the stout double baileys
of the fortress, which protected its more vulnerable eastern side, predate
the thirteenth century. It seems that the only structure at Kalamata built
by the crusaders was the keep, a huge, slightly rectangular building located
at the highest point of the inner bailey and constructed using the same
combination of classical ashlar blocks and small uncut stones as that already
described at Androusa.34
Another stronghold which was probably occupied almost continuously
from classical times up to the thirteenth century and beyond was Arcadia,
situated near the Ionian coast on the other side of the Messenian peninsula
from Kalamata. Like its neighbour, Arcadia also relied on its isolated position
at the top of a steep, narrow hill for much of its defensive strength. This
meant that the design of the castle was almost totally dictated by the shape
of the summit. It enabled the original builders to leave particularly inaccessible areas, such as the sheer north side of the hill, virtually devoid of
man-made defences, and to concentrate their efforts on the weaker southern
and eastern approaches. Here two successive baileys and curtain walls,
flanked by numerous towers and salients, defended the main access route
to the summit itself.35 Small sections of classical masonry, some of it still
in situ, suggest that this defensive arrangement is extremely old and was
merely improved upon by the Byzantines. When the crusaders first arrived
in 1205 they were immediately impressed by the sheer strength of Arcadia's
fortifications, and in particular the large tower at the summit of the castle,
which they believed had been built by 'giants'; a clear reference to the vast
classical blocks used in its construction. Indeed, Arcadia did not surrender
to the Franks until they had begun to bombard this tower with several
catapults.36 Many of the upper levels of Arcadia's walls, however, were built
using far smaller, uncut stones, whose appearance is so generic that it is
very difficult to date them accurately. It does nevertheless seem that the
large, round tower at the eastern corner of the site was constructed by
the Franks, like the keep at Kalamata.37
Numerous other fortifications in southern and central Greece can be
placed in the same group as Arcadia and Kalamata. These include Coron,
Patras, Athens and Thebes, taken by the crusaders in i205,38 and Corinth,
Nauplia and Argos, captured in 1210,1211-12 and 1212 respectively.39 At about
the same time the crusaders also occupied Neopatras, which they held until
its recapture by Theodore of Epiros in 1219.4 Almost all these places had
to be taken by force, a clear indication that they were fully functioning
military strongholds at the time of the Fourth Crusade. In terms of design
and location they also shared a number of important characteristics both
with each other and with Arcadia and Kalamata. Corinth, Argos, Neopatras
MILITARY ARCHITECTURE
211
and Thebes, for example, were all mountain castles, and were defended by
successive walls arranged around an ancient acropolis. All four sites had
either been occupied continuously since classical times or had been refortified
during the political upheavals of the eighth and ninth centuries.41 Although
Patras, Nauplia and Coron were located in less precipitous coastal areas, they
too were built on hilltops or promontories which had already acted as refuge
sites for many centuries.42 The great age of these sites can be seen from
their masonry, which still contains the same mixture of high quality ancient
blocks and more recent poorly cut stones as that used at Arcadia.43
From the early years of the thirteenth century onwards, the Latins carried
out their own repairs and improvements on these castles. It is remarkable
how often these alterations involved the addition of a strong, central tower
or keep. Such structures have already been mentioned at Kalamata and
Arcadia, and good examples have also survived at Corinth and Neopatras.44
It may have been the former structure that the Chronicle of Morea was
referring to when it claimed that repairs were carried out on 'the castle of
Corinth' by William II of Villehardouin (i246-78).45 Reference should also
be made to the famous medieval tower on the acropolis in Athens, which
was probably built by the Acciajuoli in the fourteenth century but was
unfortunately demolished in 1874. From photographs it seems that this
tower was unusually sturdy because it had been constructed almost entirely
from smooth marble slabs, without the usual filling of smaller stones or
rubbled mortar. Standing to a height of at least twenty-five metres, it must
have been one of the most impressive medieval structures anywhere in
Greece.46 It is also important to think of the acropolis itself not as a cluster
of classical ruins but as a formidable refuge point which was still being
used in a military capacity. As recently as 1205 it had successfully withstood
a major attack by Leon Sgouros, the Greek ruler of Corinth, during the
brief period of anarchy on the eve of the Prankish invasion.4?
Other improvements carried out by the Latins normally amounted to the
restoration or rebuilding of outer defences and curtain walls. Such work
was undertaken by William Alemn at Patras, whilst at Corinth the Acciajuoli
may have strengthened the ancient ramparts in the fourteenth century.48
One contemporary source also implies that the Venetians improved the
Byzantine defences of Coron considerably, although the proliferation of
post-medieval artillery defences makes this difficult to prove.4^ The Chronicle
of Morea also noted that in the thirteenth century there were no less than
two fortresses at Nauplia, implying that the lower slopes of the promontory
were just as well protected as the summit itself.so It is possible that a short
stretch of wall, a triangular bastion and two rounded towers visible beneath
a collapsed fifteenth-century talus represent a small section of these defences.
212
UNKNOWN CRUSADER
CASTLES
Certainly a postern gate incorporated into the wall and built using recycled
classical stone resembles later fourteenth-century Catalan work at Boudonitza and Salna (Amphissa), where doorways were framed with large,
antique masonry slabs.51 Similar building techniques were also used in the
construction of a tower at Thebes, the only remaining structure from a
larger castle erected by the Prankish lord of the city, Nicholas II of Saint
Omer, in 1287. According to the Chronicle of Morea this was a palatial
building in its heyday,52 but the recycled appearance of its surviving masonry,
combined with the fact that Lombard rebels in Thebes had withstood a
major attack by the emperor Henry in 1209, indicate that this was yet another
site whose defences were originally Byzantine and had then been extended
or repaired by Prankish newcomers rather than built from scratch.53
Many of these observations also apply to the ruined castles or acropolis
sites which the Latins reoccupied in southern and central Greece. Unlike
Arcadia, Corinth, Thebes and all the other strongholds just mentioned,
these places had not been maintained in recent times, and therefore lay
abandoned or were only inhabited as open settlements at the time of the
Fourth Crusade. Nevertheless, their natural strength and ready supply of
high quality stone made such ruins extremely attractive to the Latins, who
exploited them in the same way that the Prankish rebuilders of Cyzicus
had done in I2o6.54 Inevitably fortresses belonging to this group contained
far more new structures dating from the thirteenth and fourteenth centuries,
although their reliance on ancient ruins both for their location and for
their building materials meant that their overall appearance did not differ
very much from intact Byzantine strongholds such as Arcadia. At Boudonitza, an important castle on the northern frontiers of the Athenian
duchy, we find the familiar arrangement of successive baileys dominated
by a hill-top citadel, even though this stronghold was probably built almost
exclusively by the Franks and the Catalans. The masonry at Boudonitza
also confirms that the Latins, like their Byzantine predecessors, tended to
rely on rather low-quality uncut stones quarried locally if it was found that
there were not sufficient classical remains left on the site which could be
recycled.55 Other interesting examples of this type of fortification include
Salona and Zeitoun (Lamia) in central Greece, and Modon and Akova in
the principality of Achaia.5 At Modon the crusaders came across an open
settlement 'which had been without walls for a long time', even though
this city was apparently still inhabited by Greeks living inside the ruins of
much older urban defences. It seems that these remains were subsequently
used by the Latins when they refortified the site in 1205, and they may have
been recycled once more by the Venetians after they acquired Modon in
120p.57
MILITARY
ARCHITECTURE
213
Chlemoutsi: an
unusually well-built
fortress intended as
a royal residence for
the Villehardouin
rulers of Achaia.
At other sites whose ancient ramparts had been abandoned by 1204, the
Franks appear to have begun the process of restoration by constructing
large, isolated towers out of the surrounding ruins. This was certainly the
case at Akova, situated in the mountainous interior of the Peloponnese.
Here it seems that a robust rectangular keep dominated the castle, which
otherwise relied on a combination of sheer cliffs and thin, poorly constructed curtain walls for its defence.58 Similarly, the central donjon at
Boudonitza had walls two metres thick and an entrance positioned 2.2
metres above the ground, whereas the outer ramparts of the fortress were
barely half as thick and had flanking towers situated as much as 80 metres
apart At Salona, another medieval stronghold built on an ancient acropolis, a large tower located at the summit of the inner ward may belong
to the same category, although the fact that this structure was round, an
almost unique feature in central Greece, raises the possibility that it was
built by the Turks rather than the Franks or the Catalans.60 What is clear,
is that castles like Salona were generally dominated by central towers which
were far stronger than any surrounding fortifications. They are therefore
very similar to the more intact Byzantine strongholds already referred to
which had Latin donjons added to them, such as Corinth, Kalamata and
Neopatras.
Having looked at castles which were either intact in 1204, or were constructed on the ruins of far older fortifications, it is now time to discuss
214
MILITARY ARCHITECTURE
215
216
U N K N O W N CRUSADER CASTLES
Karytaina. This castle is similar to many other mountain castles bunt or occupied
by the Franks, including Mistra and Arcadia. Note the possible remains of an isolated
square keep in the upper courtyard.
MILITARY ARCHITECTURE
2l/
to that effect in the Chronicle of Morea, Bon has argued that the Achaian
fortresses of Androusa, Old Navarino, Mistra, Beaufort (Leutron), Old
Mania and Graki were also founded by the Franks after 12O4.69 But is this
apparently simple method of identifying new Prankish castles actually as
reliable as it seems? Architecturally, there are a number of problems with
this approach, for it has already been noted that Androusa may well have
existed in some form before 1204. Similarly, many of the features which
Bon regards as typically Prankish, such as low quality masonry or the use of
successive baileys to defend an isolated hilltop, have already been described
at numerous castles which were quite clearly Byzantine, including Kalamata,
Arcadia, Argos and Corinth. This suggests that even if Glarenza and Livadia
were built from scratch by the Latins, as appears to have been the case,
then their present remains are so similar to neighbouring Byzantine ruins
that it is pointless trying to isolate a typical 'Prankish style' from them.
Even within the last five years, archaeological exploration carried out at
Vardounia, a castle situated in the Taygetos range to the south of Mistra,
has only led to the conclusion that the site is of 'Prankish or Byzantine
origin' because of the close similarities between all medieval fortifications
in the Peloponnese.70
It has also been shown that similar uncertainties have led other scholars
to challenge Bon's argument that Mistra was founded by William II of
Villehardouin. The upper castle of this site was dominated by a large
rectangular keep, which, as we have seen, was an extremely common feature
in local Prankish military architecture. Most of the surrounding ramparts
and gateways, however, show strong Byzantine influence and have therefore
caused David Winfield to reject the traditional claims made by Bon and
others.71 These problems suggest that Bon has been too concerned to make
the archaeological evidence fit in with the Chronicle of Morea, rather than
allowing the medieval ruins to speak for themselves. At Karytaina, for
example, he acknowledges the presence of what appears to be much older
classical masonry incorporated into the keep, but the thirteenth-century
claims of the Chronicle seem to prevent him from seeing this as a possible
sign that parts of the castle already existed before 1204 and were therefore
Byzantine.72 Indeed, in the case of Kalavryta, another fortress in northern
Achaia, Bon has attributed the defences to the Latins even though the
Chronicle only listed this as an important Prankish lordship rather than
claiming that it had actually been built by the newcomers.73
Another difficulty with Bon's argument arises from the Chronicle of Morea
itself. This source survives in French, Greek and Aragonese versions, plus a
much shorter Italian translation dating from the sixteenth century. Although
these versions were all based on an older (probably French) original, they
2l8
UNKNOWN CRUSADER
CASTLES
only survive in manuscripts dating from the fourteenth century, and are
not always reliable when describing events which took place more than one
hundred years earlier.74 According to the Chronicle, for example, the Franks
who invaded Greece after the Fourth Crusade did not capture Corinth,
Argos and Nauplion until the 12408, even though the papal correspondence
of Innocent III indicates that these strongpoints actually fell between 1210
and 1212.75 In other cases it is clear that when the Chronicle speaks of French
settlers constructing new fortifications, it does not necessarily mean that
they were building castles from scratch on previously unoccupied sites. We
have already seen that the citadel of Thebes erected by Nicholas of Saint
Omer around 1287 cannot possibly have been anything other than an extension or reconstruction of a much older Byzantine strongpoint.76 Finally,
it is important to note that some versions of the Chronicle occasionally
attributed the construction of a single fortress to more than one man.
Karytaina, for example, was said to have been built by both Rugues de
Bruyere, the first lord of the castle during the early years of Prankish rule,
and by his later successor Geoffrey around the time of the fall of Monemvasia
(1249). Similarly, Gerald was attributed to two entirely different generations
of the same family.77 These anomalies suggest that we should not always
accept the Chronicle of Morea at face value when attempting to date Achaian
fortifications.
In a sense, we have come full circle and have returned to the architectural
and historical dilemmas outlined at the beginning of this chapter. Medieval
remains in Greece are often so similar in appearance that they can be
attributed to the Greeks, as Winfield has done, or to the Franks, as Bon
preferred, with equal validity. This situation probably sterns from the fact
that both Latins and Greeks used the same local materials and craftsmen
to build their fortifications. This certainly appears to have been done on
Cyprus, where the castle of Paphos has been dated to the early Prankish
period even though it contains a number of polygonal towers built in a
Byzantine rather than a Latin style.78 Returning to Greece, it is clear that
western settlers sometimes also built or repaired churches 'using Greek
craftsmen working in a Greek style'.79 In reality, therefore, it may not
actually be possible to identify a 'pure' thirteenth-century Prankish castle
simply by looking at its masonry and location, and then turning to the
Chronicle of Morea for confirmation.
One element of these castles which can probably be attributed to the
Franks with far more certainty is the keep. In other crusader states scholars
have long argued that square or rectangular stone towers of this kind
represented the earliest and most common structures erected by the first
Latin settlers. Speaking of the kingdom of Jerusalem, historians have
MILITARY ARCHITECTURE
21p
noted that 'the conquest was followed by the construction of large numbers
of keep-and-bailey castles',80 and that strongholds built by western lords
'almost always took the form of stone towers'.81 We have seen that in
the Holy Land this tradition of constructing towers continued into the
thirteenth century, both at small sites such as Tuklah and at much larger
castles such as Montfort. There are also examples from other areas of Latins
building such structures inside much older ruined or intact Byzantine
strongholds. Following Richard I's conquest of Cyprus, for example, it has
been argued that the first Lusignan rulers of the island built a series of
new towers where small Greek castles had once stood.82 In Cuidan Armenia
a Prankish keep was also added to the Roman and Byzantine acropolis of
Anavarza not long after the First Crusade.83 In Greece itself we have already
seen that the Franks built numerous similar towers at the intact Byzantine
castles of Athens, Corinth, Kalamata, Neopatras and Arcadia, as well as the
ruined fortresses of Akova, Boudonitza and Salona. Perhaps, therefore,
when the Chronicle of Morea spoke of Franks bunding new castles it was
actually referring to keep towers being added to much older Byzantine or
ancient defences. This theory would help to explain why supposedly 'pure'
crusader sites like Karytaina contain recycled ancient stone, and would
reconcile Bon's argument that many Achaian strongholds are Prankish
with Winfield's conclusion that they already existed before 1204. In other
words, it may be that Androusa, Mistra, Karytaina, Beaufort (Leutron),
Old Navarino and Gerald were not built from scratch by the Franks, as
Bon and others have assumed, but were older Greek sites which were
remodelled and had donjons added to them. This suggestion may also
explain the claims made by the Chronicle of Morea that Karytaina and
Geraki were built by more than one Prankish lord, for these statements
no longer seem illogical if they are interpreted as referring to continuing
repairs and alterations rather than the completion of entire new fortresses.
Hence at Karytaina, for example, some features which can probably be
attributed to the Franks with more certainty, such as the gothic hall or the
central keep, may have been completed by different people at different
times in the course of the thirteenth century.
The popularity of stone towers is also confirmed by the fact that such
structures made up the vast majority of smaller fortifications built by Latin
settlers. Strongholds of this kind were either left to stand in isolation or
were surrounded by an outer curtain wall. More than twenty towers of the
isolated variety have been identified in mainland Greece, and particularly
in the former duchy of Athens. Most of these structures are extremely
ruinous and are fast disappearing, but the tower of Markopoulo, located
about twenty-five kilometres south east of Athens, still stands to its original
220
U N K N O W N CRUSADER CASTLES
Haliartos:
a typical Latin
tower of the eastern
Mediterranean.
MILITARY ARCHITECTURE
221
222
U N K N O W N CRUSADER CASTLES
MILITARY ARCHITECTURE
223
Livadia. The almost total reliance of the Latins on local workmen and
existing fortifications also makes it virtually meaningless to try to identify
a distinct Prankish style of military architecture. It has been suggested that
the only exception to this general observation was the keep or isolated
tower, which was so popular with westerners both on the mainland and
on the Aegean islands that it can probably be attributed to Latin newcomers
with more confidence than certain types of masonry or particular stretches
of curtain wall.
A number of solutions can be put forward to explain the unimpressive
appearance of most medieval fortifications in Greece. One possibility is that
poverty prevented the completion of more elaborate defences; a point which
can be illustrated by taking a closer look at the history of some Achaian
fortresses. Akova, Graki, Karytaina, Patras and Kalavryta, for example, were
some of the most important lordships in Achaia, containing as many as
twenty-four fiefs each (Patras and Akova),101 and in certain cases possessing
special rights of jurisdiction.102 Most of these castles were held by the same
Prankish families until the late thirteenth or early fourteenth centuries,
when they were either lost to the Greeks or returned to the royal domain
because their owners had begun to die out.103 Others, such as the archbishop's fortress at Patras, were sold to or sought the protection of Venice,
which increasingly became the only western power strong enough to defend
Latin territories against the Greeks and the Turks.104 Yet even before this
decline in Latin power, which did not set in properly until the last third
of the thirteenth century, many Prankish castle owners do not seem to have
had the resources to improve the fortifications which they occupied. We
have already seen that at Karytaina some of the curtain walls were less than
one metre thick and so badly constructed that they have long since collapsed,
even though this lordship owed the service of twenty-two knights and was
therefore the third largest barony in all of Achaia.1Q5
The same contradictions can be found in central and northern Greece.
For many years during both the Prankish and the Catalan periods, the castle
of Boudonitza retained its virtual independence as a frontier lordship on
the northern outskirts of the duchy of Athens. It seems that it successfully
withstood the Epirote conquerors of Thessaly in the I22os,106 resisted the
aggressive Catalan conquerors of Athens,107 and even held up the Turks for
a while in the fifteenth century, yet its defences were relatively simple in
design, and so little money seems to have been spent on them that good
or even mediocre masonry was only used in the most vulnerable parts of
the castle.108 These observations also apply to other baronial castles in central
Greece, such as Salona, which became an important Prankish lordship
between 1205 and 1311.109
224
UNKNOWN CRUSADER
CASTLES
Even more surprising is the fact that many royal strongholds were just
as badly constructed. Androusa, for example, belonged to the princes of
Achaia throughout the medieval period until it was finally lost to the Greeks
in the 1420S, yet its walls were built using a slapdash mixture of broken
pottery, bricks and uncut stones.110 The castle of Livadia in the duchy of
Athens first belonged to the de la Roche dukes in the thirteenth century,
then rose further in importance under the Catalans, who made it the chief
residence and administrative headquarters of the vicar-general, the official
representative of the Aragonese dukes. As a result Livadia outshone Thebes,
and often rivalled both Athens and Neopatras, but one would hardly think
so by looking at the actual remains of the fortress.111 The most striking
example of this irony must surely be that of Constantinople itself, whose
rulers bore the imperial title yet lacked the money to repair the walls of
their own capital.112
Needless to say there were some exceptions, most notably Chlemoutsi,
which had been constructed with the help of a fortuitous windfall, and the
luxurious thirteenth-century castle of Saint Omer in Thebes.113 But, on the
whole, it seems that many medieval fortresses in Prankish Greece owed their
appearance to the poverty of their occupants. In the case of Baldwin of
Flanders and his successors, this can be explained in terms of the sheer
number of castles and amount of territory they were expected to govern.
As far as the nobility was concerned, it reflected the relatively humble
background of the settlers who came to Greece in the first place. William
Alemn, lord of Patras, used stone robbed from church properties to repair
his castle out of necessity as much as ruthlessness, even though on paper
his lordship was one of the most important in Greece, owing the service of
twenty-four knights. Even more significantly, William's ancestry in Europe
remains a mystery, and it seems that he was only one of many crusaders
who came to the east with little to lose and much to gain.U4 Architecturally
speaking, lack of adequate resources could also explain why so many of
William's fellow newcomers chose to add donjons or free-standing towers
to the former Byzantine castles which they took over. Structures of this kind
could be built relatively quickly and cheaply, whilst their compact shape
meant that they could be garrisoned by very few men even if older outer
defences were left in ruins or had to be abandoned in the face of a numerically
superior enemy. This situation can be contrasted with that of the other
crusader states, and in particular the Holy Land itself. Here the majority of
larger castles, and certainly those which were in royal hands or were held
by the greatest landholders, were constructed on a far grander scale by skilled
craftsmen using well-dressed stone. This difference in quality reflected the
contrast in wealth between men like William Alemn and powerful European
MILITARY ARCHITECTURE
225
crusaders such as Louis IX, who repaired the town walls of Caesarea, Jaffa,
Sidon, Haifa and Acre. Most of the other great thirteenth-century fortifications in the Holy Land, such as Crac des Chevaliers and 'Atlit, were built
by the Military Orders relying on large donations and land revenues collected
in western Europe.115
Another obvious reason why castles like Crac des Chevaliers were never
built in medieval Greece was that they were not actually needed. To some
extent the local terrain helps to account for this. The fortress of Karytaina,
for example, occupied a very steep hill top more than 150 metres above the
surrounding valleys. It must have been extremely difficult for any attacker
who was under fire to reach it safely, let alone deploy siege engines against
its ramparts. The Prankish lords of the fortress may well have considered
walls which were only ninety centimetres thick to be perfectly adequate, us Karytaina was only one of countless other castles in Thessaly, Boeotia
and the Peloponnese situated in mountainous areas, where massively thick
walls or perfectly dressed stones would often have been a waste of time
and money. We have already seen that mountain strongholds in Cyprus
or Cilician Armenia often relied on sheer cliffs rather than sophisticated
man-made defences to protect them against attackers.117
But this does not explain why more exposed strongpoints were sometimes
just as poorly defended. Notably, the crusader town of Glarentza was only
protected by a single ditch and curtain wall even though it was situated on
a relatively flat and open coastal plain.118 This was a far cry from the
elaborate double ramparts, overlapping towers and complex gateways which
protected coastal cities like Acre in the Holy Land. The obvious implication
is that warfare in medieval Greece, and in particular siege warfare, had not
yet reached the same level of sophistication as in the Holy Land. There late
thirteenth-century Mamluk armies regularly contained up to 12,000 (or
more) men, whilst major crusader castles needed to be capable of withstanding an array of Muslim sappers, catapults and other siege weapons.
By contrast, most of the warfare which took place in the Peloponnese, and
indeed the Aegean area in general, probably involved hundreds rather than
thousands of troops, and there was often a far greater risk of attacks by
raiders or pirates than by vast invasion forces fully equipped with siege
engines. Lack of adequate troops and siege engineers may also explain the
sporadic and indecisive nature of the fighting both in northern Greece and
in the Peloponnese.119 On at least one occasion the Greeks of Mistra failed
to conquer neighbouring Prankish lands because they had not brought
enough siege weapons with them.120 Some castles in this region, including
Karytaina, were also captured through bribery or stealth rather than a direct
assault; tactics which the Greeks may have fallen back on because they were
226
U N K N O W N CRUSADER CASTLES
20
228
U N K N O W N CRUSADER CASTLES
229
of Cyzicus came under siege in 1207, the emperor Henry organised a relieving
fleet which sailed across the Sea of Marmara, forcing Theodore Lascaris's
land and naval forces to retreat. If Cyzicus had been located inland, it is
extremely unlikely that Henry would have reached it in time or have had
enough troops at his disposal to face Lascaris in a land battle.13 These factors
explain why heavily fortified coastal strongholds, such as Modon and Coron
in the Peloponnese, were invariably the last western outposts to fall, sometimes outliving neighbouring inland castles by many decades.14
When it was no longer possible to defend even the most powerful fortifications, Latin garrisons could still use the sea as their final means of
escape. Many westerners, including the Latin emperor Baldwin II, managed
to flee from Constantinople in 1261 in much the same way that the defenders
of Acre, Sidon and other coastal outposts in the Holy Land were to do
exactly thirty years later.15 Many decades earlier Prankish troops had also
managed to make an orderly retreat from Kibotos in the vicinity of Nicomedia by sailing away across the Sea of Marmara to Constantinople (1207).
Interestingly, Kitobos had already withstood a Nicaean attack successfully,
but it was still decided to evacuate the castle because the Franks knew its
garrison was too weak to hold out for much longer. Kitobos provides us
with another example of how inadequate troops and resources rather than
a lack of military skill prevented the Franks from holding on to their
territories within the Latin empire.16
The fate of Kibotos merely confirms that the Latins could only hope to
sustain their fragile position in the former Byzantine empire by protecting
their seaborne links with each other and the West. In practice this meant
that they were heavily dependent upon the Venetians. It was Venice which
saved Constantinople in 1235 and 1236, and it was their desperate need for
Venetian assistance that forced the Franks to give up their claims to Modon
and Coron in 1209, even though these cities had originally been captured
by French rather than Italian crusaders.17 Other coastal fortresses would
never have surrendered in the first place if it had not been for the Venetians.
Nauplia and Monemvasia, for example, were both captured because Venetian ships enabled the Latins to blockade these strongholds by sea as well
as by land. In order to obtain Venetian support for his campaign against
Monemvasia, William II of Villehardouin confirmed the republic's possession of Coron and granted it certain privileges. In return, Venice provided
the Franks with four galleys for the campaign, and undertook to maintain
a further two galleys for the permanent defence of Achaia18 This agreement
illustrates how Venice dominated all naval activities in Prankish Greece, in
the same way that it had dominated the Fourth Crusade itself. This can be
contrasted with the situation in Cyprus and the Holy Land, where the
230
231
232
U N K N O W N C R U S A D E R CASTLES
strongholds which were situated near the sea. Similarly, the many isolated
towers constructed in harbours could protect people and revenues against
opportunistic raiders whilst at the same time helping to guide approaching
ships toward their destination. Towers of this kind have survived on Euboea
at Aliberi-Matsoukela and Rovies, and the latter may even date back to
the time of William of Villehardouin's campaigns on the island during his
civil war with Duke Guy I of Athens and the Venetians (1255-58)31
Many castles on the mainland also acted as refuge points, for Greek
coastal areas were just as likely to be attacked by pirates as the Aegean
islands. At Monemvasia, for example, the harbour was situated on a vulnerable strip of land facing the sea but was protected by the impregnable citadel
above it. Even though a devastating Catalan raid during the 12908 resulted
in the destruction of the town, its inhabitants escaped unhurt because
'they climbed up to the castle on the great rock which stands there'. In this
way an important anchorage and its civilian population were prevented
from falling into the hands of the dreaded Catalans.32 Another mainland
settlement which was often targeted by pirates was Corinth. From 1311
onwards this city and its immediate surroundings must have been particularly exposed to raids, for both the Catalans of Athens and the Greeks of
Mistra could easily reach it by land or sea. As a result, living conditions in
this area got so bad that by the middle of the fourteenth century large parts
of the countryside had been completely abandoned. The Angevin rulers of
Achaia responded by granting the castellany of Corinth to Niccolo Acciajuoli,
a powerful Florentine lord who was instructed to repair the citadel's defences
and was granted special privileges to help him meet the huge cost of
maintaining several smaller castles in the neighbourhood (1358). These
included that of St George, situated near the frontier with Mistra and
'valiantly' defended by its Latin garrison during the I35os.33 Several years
earlier Acciajuoli had also been granted Messenian lands in the southern
Peloponnese, equally devastated by years of Greek, Turkish and Catalan
pillaging. To halt this process, Acciajouli had used the same tactic as at
Corinth, erecting a new castle in the vicinity of Kalamata which could
protect the area against external raiders.34
Elsewhere fortifications situated in coastal regions played a more active
role in the fight against piracy, for they provided safe anchorages from which
to organise counter-strikes against seaborne raiders. By the mid fourteenth
century the most formidable intruders of this kind were the Turks. In order
to deal with this threat the papacy, the Venetians, the Cypriots and the
Hospitallers of Rhodes formed a powerful naval alliance known as the Holy
League. This was not simply regarded as a military force but as a continuation
of the crusading movement once land-based offensives against Jerusalem
233
itself had become unrealistic.35 The important contribution coastal fortifications made to the continuing struggle against Islam is shown by the
League's first and greatest victory: the capture of Smyrna, on the coast of
Asia Minor, in October 1344. This was achieved by a fleet of Cypriot vessels
sent from the fortified harbour of Famagusta, Hospitaller galleys stationed
at the Order's heavily defended base on Rhodes, and a large contingent from
the formidable Venetian navy. These forces all gathered at the Venetian port
of Negroponte on Euboea before attacking Smyrna itself.36 Although they
have long since disappeared, Negroponte's medieval defences at this time
were extensive and included a fortified bridge connecting Euboea with the
mainland. This structure was divided by a drawbridge 'no larger than to let
a galley pass through', so that Venetian ships anchored opposite the city
would be protected from Turkish, Genoese or Catalan raiders.37 The security
provided by sites like Negroponte, Rhodes and Famagusta enabled the Holy
League to launch its successful offensive on Smyrna and to maintain pressure
on the Turks throughout the eastern Mediterranean.38
Smyrna and other Christian outposts in Asia Minor, such as Cypriot-held
Corycus,39 were themselves further examples of the kind of coastal strongholds already mentioned at Cyzicus, Naxos and Monemvasia. All these sites
owed their very existence to a combination of good sea links and powerful
fortifications. Between them, these places created a network of safe anchorages which were not only used by major expeditions, such as the crusade
against Smyrna, but could also protect much smaller naval forces. During
the latter half of the thirteenth century, for example, the Venetians constantly
patrolled the waters around the Peloponnese with two of their own galleys
to reduce the general level of piracy rather than to deal with a specific
threat. These vessels no doubt operated between Venetian bases at Modon,
Coron, Negroponte and in the Cyclades, never straying too far from a
friendly port.40
Pirates themselves normally used coastal strongholds as bases from which
to launch their raids. One group which relied on this tactic were the Catalans.
After clashing with their former Byzantine employers, but before they
established themselves at Athens, they withdrew to Gallipoli for a few years
at the very beginning of the fourteenth century. They refortified this site
by digging huge ditches and erecting new wooden stockades, making it a
perfect headquarters from which to carry out naval attacks on neighbouring
harbours such as Panidos and Rodosto. On one occasion so many Catalans
left Gallipoli on a seaborne raiding party that the Greeks came close to
recapturing the town, which was almost empty.41 But the Catalans survived
this crisis, and Muntaner wrote that in the following months 'except the
cities of Constantinople and Adrianople and Christopoli and Salnica, there
234
was not a town or city that was not pillaged and burnt by us, nor any
place, unless it was a castle in the mountains'.42 This programme of systematic looting was only made possible by the strength of the defences at
Gallipoli itself, which enabled the Catalans to act without fear of reprisal
from the Byzantine authorities On the other hand, Muntaner's statement
indicates that the Greeks had been able to maintain the defences around
their larger cities, whilst his reference to the impregnability of mountain
castles explains why elevated strongholds like Monemvasia were so popular
with those who lived in exposed areas.43
Many other raiders and pirates were stationed on islands. Corfu, for
example, had to be cleared of Genoese pirates before the Venetians could
take control of the island in the wake of the Fourth Crusade.44 Elsewhere,
other Genoese forces, acting in closer cooperation with their native city
used similar tactics against their Venetian rivals. During the 13405 and 13505
Genoa tried to use its new colony on Chios as a springboard for further
expansion in the eastern Mediterranean, including the conquest of Euboea
from Venice. This project came to an abrupt end after the Genoese failed
to capture the port of Oreus along the north coast of the island in i35i.45
Almost 150 years earlier, the Genoese were equally unsuccessful in their
efforts to subjugate Crete and to defend it against the Venetians, who had
bought the island from Boniface of Montferrat shortly after the Fourth
Crusade. It was only by capturing the heavily fortified castle and harbour
of Palaiocastro, situated on the north coast of Crete, that Venice finally
ousted the Genoese and prevented them from establishing a base which
could have posed a serious threat to Venetian power in the Cyclades.46
The Genoese were very similar to their Venetian, Prankish and Catalan
rivals in that their military activities in the Mediterranean were almost
entirely centred around the construction, capture or defence of castles. Such
structures were often the only means by which the various Latin powers
operating in the eastern Mediterranean could hope to maintain control over
the islands and coastal regions of the former Byzantine empire. Offensively,
castles were an ideal way of creating bridgeheads which could be reinforced
by sea and used to make further conquests inland. In addition, they protected
important harbours which were needed for naval campaigns such as the
crusade against Smyrna. Defensively, powerful strongholds like Monemvasia
provided shelter for local people whilst preventing strategic anchorages from
falling into the wrong hands, for the activities of the Catalans at Gallipoli
proved that raiders could use such sites to wreak havoc over vast areas.
This last point also highlights the immense importance of coastal defences
as a means of protecting the sea routes between East and West, for fortified
settlements such as Glarentza were needed to maintain the steady flow of
235
arms, troops and supplies upon which Prankish Greece depended, Coastal
strongpoints were invariably the first to be captured by the Latins, and also
the last to fall, not only in the Aegean but throughout the crusader states.
21
THE O F F E N S I V E F U N C T I O N S OF I N L A N D F O R T I F I C A T I O N S 237
morale and the fighting strength of the Franks were beginning to wane.* A
comparable case of territorial conquest being based around castles dates
from the period of Catalan rule at Gallipoli. Once they had established
themselves here, the Catalans were not content to launch naval attacks
against neighbouring coastal settlements, but also expanded inland until
they dominated the entire surrounding peninsula. This meant having to
capture local Byzantine fortresses such as Maditos, situated to the south of
Gallipoli. Maditos only fell after an eight month siege, when the Catalans
managed to climb the walls during the garrison's afternoon siesta. Such a
prolonged campaign would not have been possible without the use of
Gallipoli as a base from which to bring up food, supplies and troop reinforcements, and it highlights the need for armies to have some kind of secure
starting point before they could take to the field.5
These considerations also applied to the many less ambitious campaigns
which were launched by Prankish forces to collect booty rather than gain
territory. As part of the Latin emperor Henry's efforts to re-establish
Prankish control in Thrace after the Bulgar invasion of 1205, no less than
120 western knights were stationed in the walled town of Rousion, to the
south of Demotika. During the winter of 1205-06, these troops carried out
a raid against hostile Greeks in neighbouring territories, killing many and
capturing forty horses. The purpose of this attack was not to retake lost
fortresses or land but rather to intimidate the local population and force
them to accept Prankish rule.6 The following summer Henry used Adrianople as a base for a similar incursion into Bulgar territories to the north,
even though the Franks clearly lacked the numbers to occupy such a large
area permanently. Instead, Henry simply hoped to gain some booty, inflict
economic damage on the region, and thereby deter the Bulgars from
attacking the Latin empire again. These tactics are virtually identical to
those used by the Hospitallers of Crac des Chevaliers and the Templars of
Saphet to keep their Muslim neighbours in the Holy Land at bay.7
In 1271 Michael VIII Palaiologos hoped to launch a major attack on the
Morea using troops he had assembled at Monemvasia. On hearing this
news, William II of Villehardouin immediately prepared for a pre-emptive
strike against the area to the north of Mistra, which was dominated by
Byzantine troops and rebellious local Greeks. William's forces therefore
marched to Glarentza, where they were joined by other troops sent by
Charles of Anjou, before moving east to Karytaina. Along the way, other
barons met them with further contingents and supplies for two months,
and at Karytaina itself the lords of both this castle and nearby Akova also
joined the expedition. Once everybody had arrived, William called a meeting
to discuss tactics, before setting off on a raid which not only halted Michael
238
VIII's own advance but also brought the Franks much booty and cattle.
During the preliminary meeting William remained camped along the river
below the castle of Karytaina, so that his army had a ready supply of food
and water but could still find shelter quickly in case of a Greek counterattack. The success of the entire campaign relied upon the food, supplies
and protection provided by Glarentza and Karytaina.8
In describing William's preparations for this expedition, the Chronicle
of Morea implies that those barons who joined the prince brought troop
contingents drawn from their own castle garrisons. The same source provides us with more evidence that this tactic was used by Greeks and Franks
alike. In 1259, on the eve of the battle of Pelagonia, the Franks decided to
confront their Nicaean opponents in open battle, thereby hoping to destroy
the Greek field army in one blow. 'Those who were most experienced in
warfare' argued that this would leave the Greek castles in Thrace and
Macedonia completely unprotected, and that their outnumbered defenders
would be forced either to flee or to surrender. Similar circumstances had
of course enabled Saladin to overrun the kingdom of Jerusalem in 1187, but
the crushing defeat at the battle of Pelagonia ended any Prankish hopes of
emulating this success in Greece.^ Nor do the westerners seem to have
learnt from this lesson, for in 1311 they lost so many troops at the battle of
Halmyros that the Catalans were able to do precisely what the Franks had
planned to do against the Nicaeans. At Livadia, for example, there were
not enough Latins left to prevent the local Greeks from rebelling and
opening the gates to the Catalans. Elsewhere the invaders simply occupied
former Prankish castles without a struggle, and some even married the
widows of those barons who had fallen only a few weeks earlier.10
The relatively small number of Latins settled in Greece made it risky to
commit too many garrison troops to major Prankish field armies. If such
forces were defeated, reinforcements were hard to come by and strongholds
were extremely vulnerable to enemy counter-attacks. In 1187 and 1311 these
factors led to the destruction of entire crusader states. It was more prudent
to avoid confrontations in the field altogether, and to rely on castles to
strengthen control over territory. In the Morea, William II of Villehardouin
used this strategy against the Melings, an aggressive Slavic tribe whose
homeland in the Mani peninsula was so mountainous and inhospitable that
it still refused to acknowledge Prankish overlordship as late as the 12403.
William realised that it would be useless to send a force of knights into
such an area, so he constructed (or reoccupied) the castles of Mistra and
Old Mania (Grande Maigne) at the northern and southern ends of the
peninsula. From here his troops could observe the Melings, launch punitive
raids against them and shelter local people from their depredations without
THE O F F E N S I V E F U N C T I O N S OF I N L A N D F O R T I F I C A T I O N S 239
having to carry out a systematic conquest of the entire region. For their
part, the Melings realised that 'pressured between these two castles, it was
impossible for them to resist the prince', and so they made peace with
him.11 William then built a third castle to the east called Beaufort (Leutron),
'all the better to contain [the local Slavs] and place them under his
rule'.12 The construction of this latter fortress suggests that Mistra and Old
Maina had been successful in suppressing the Melings, and that castlebuilding enabled William to keep his Slavic subjects in check without having
to occupy their homeland in its entirety.
By constructing castles, William was able to make up for the fact that he
lacked the troops to suppress the Melings through sheer weight of numbers.
At other times, however, the Franks did manage to defeat their opponents
in open battle, but still needed castles to make sure that the advantage
gained from a victory was not lost. This problem most affected the Latin
empire, where the emperor Henry had to keep campaigning almost continuously because 'he could not raise enough troops to defend his territories'.
As a result, the Franks were often obliged to rely on castles rather than
soldiers to maintain their borders.13 After victory over the Cumans in 1208,
for example, Henry immediately occupied the castle of Crucemont, about
thirty kilometres west of Philippopolis. He then rode south to Pamphilon,
where he constructed a new castle as quickly as possible before the onset
of winter. In so doing, Henry hoped to take full advantage of his victory,
secure his northern border, and prevent the Cumans and Bulgars from
simply reoccupying this region as soon as he had withdrawn to Constantinople.14 Similarly, during the 12603 Michael VIII Palaiologos enjoyed
considerable success against the Latins because, once he had harried Prankish
Greece with his land and sea forces, he 'occupied many places and built
powerful castles on mountains and in very strong passes'. Michael knew
that this would make it far harder for the Franks to regain this territory,
as they would only be able to do so by undertaking a series of lengthy
sieges.15
One way of responding to such tactics was to adopt a scorched earth
policy and to demolish all strongholds likely to be occupied by an advancing
enemy. In 1331, for example, the Catalans responded to news of Gautier II
of Brienne's imminent invasion by demolishing the fortress of Saint Omer
at Thebes 'to make sure that the duke of Athens would not take it in any
way and recover the duchy by using this castle'.16 In 1207 the emperor
Henry agreed to demolish both Cyzicus and the fortified church at Nicomedia in his peace treaty with Theodore Lascaris of Nicaea. By insisting
on the destruction of these two castles, Theodore hoped to prevent the
Franks from continuing to use them as bases for raids on the surrounding
240
U N K N O W N C R U S A D E R CASTLES
T H E O F F E N S I V E F U N C T I O N S O F I N L A N D F O R T I F I C A T I O N S 24!
forces in other areas. The two strongholds erected by the Franks at Corinth
were specifically intended to ensure that the besieged 'could not rush out
from the walls to collect water or any other supplies to sustain them'. If
they attempted to do so, the Franks could easily launch a counter-raid from
either of their forts in order to force the Greeks back inside the citadel.26
At Constantinople Villehardouin wrote that 'never had so many been besieged by so few in any city', and this explains why the crusaders established
their fortified camp opposite one particular stretch of the defences, rather
than trying to spread out around the whole city.2? It would have been
impossible for the Latins to surround vast sites such as Corinth or Constantinople entirely; but, by gathering all their forces together inside fortified
camps, they made the most of the troops available to them, and could still
blockade far larger castles for months or even years. Once again castles took
the place of troops in the field. These factors also explain why it took the
crusaders five years to capture Corinth: with so few soldiers available, an
all out assault on the walls was usually out of the question; if they could
not bombard a fortress into submission, the Franks simply had to wait until
hunger or low morale persuaded its garrison to surrender.28
The methods used by the crusaders at the siege of Corinth also reflect
their military fragility in general. Every Latin soldier, and in particular every
knight, was a precious asset for the Franks, who had enough trouble raising
adequate forces for their campaigns even at the best of times. Hence one
of the primary functions of all castles in Prankish Greece was to protect the
Latin field army, especially during a military crisis. In the spring of 1205,
for example, the first Latin emperor of Constantinople, Baldwin of Flanders,
was captured outside Adrianople, while his army suffered a heavy defeat at
the hands of the Greeks and the Bulgars. This disaster forced the crusaders
to flee south, past Pamphilon and toward the coast. It was not until they
reached the fortified city of Rodosto, situated along the Sea of Marmara,
that they finally found refuge from the pursuing forces of Kalojan.29 The
walls of Rodosto effectively saved the Prankish and Venetian army from
total annihilation, which may explain why the emperor Henry chose to
fortify Pamphilon three years later, by which date Adrianople had returned
to Prankish control. Pamphilon must have acted as a useful stopping off
point for Latin troops travelling from Rodosto or Constantinople toward
the northern frontier of the empire. Alternatively, Henry knew that, if
Adrianople were ever recaptured by his opponents, the new defences at
Pamphilon could protect westerners retreating south, who would otherwise
have to march all the way to Rodosto without any hope of shelter along the
route.30
William II of Villehardouin's campaign against the Greeks, undertaken
242
22
244
THE D E F E N S I V E F U N C T I O N S OF I N L A N D F O R T I F I C A T I O N S 245
the fortresses with all their supplies, so that our people could not find
anything to eat'. Those who tried to follow the locals were simply attacked
by Epirote troops hiding in the forests and valleys of the interior. As autumn
set in and the threat of starvation began to loom, Charles's forces had no
choice but to retreat. In this way the most powerful warriors of the Morea
and Cephalonia were defeated not by a rival army but by the strength of
the despotate's castles.e
Almost twenty years later the Catalans used the same strategy to thwart
an attempt by Gautier II of Brienne to recapture the duchy of Athens,
which his father had lost following the battle of Halmyros in 1311. In 1331
Gautier sailed from Brindisi with an army allegedly containing 800 French
knights which he had gathered together at great cost. During the ensuing
weeks Gautier led this force into Attica and waited for the Catalans to meet
him in battle, confident that they would be wiped out by his Prankish
warriors. But the Catalans 'did not want to come out and fight', and simply
waited inside their castles until Gautier, who had not come prepared for
siege warfare, ran out of money.7 The Catalans emerged victorious because
they understood that it would be more difficult for Gautier to hold his
army together in the field than it would be for them to wait inside their
strongholds. The invasions of 1205 and 1304 also collapsed through a combination of political and logistical problems long before any real fighting
had taken place. This 'wait and see' strategy meant that a region could be
defended effectively without the use of thousands of troops stationed along
its frontiers, for in 1304 Charles II failed to conquer Epiros even though
his forces had been allowed to swarm across its borders and reach Arta
totally unhindered. In the Holy Land, similar tactics were used during crises
such as the Khwarizmian invasion of 1244, showing that political boundaries
in the modern sense were meaningless during the crusader period.8 It was
castles and urban fortifications rather than lines on a map which defined
the limits of a ruler's authority.
The sheer amount of damage which invading armies could inflict on an
area still made it desirable to stop such forces as quickly as possible, and
this was the primary function of frontier castles. One of the most important
such strongholds was Corinth, which acted as a kind of buffer against any
invasion forces from the north because of its location at the entrance to
the Peloponnese. In 1205, for example, it brought Boniface of Montferrat's
advance through Byzantine Greece to a halt. Although the subsequent
campaign in the Morea by William of Champlitte and Geoffrey of Villehardouin showed that this castle could be circumvented, the crusaders knew
that it was dangerous to leave such a powerful citadel in the hands of the
enemy. For as long as it remained uncaptured, its garrison could cut off
246
U N K N O W N CRUSADER CASTLES
the Franks' retreat and leave them stranded in the Peloponnese. This also
explains why Corinth retained its strategic importance after the Catalans
settled in Athens, and again during the period of Turkish expansion, when
the Acciajuoli spent vast sums on maintaining its defences. Corinth did not
finally fall to the Turks until 1458, by which time it had been retaken by
the Greeks.9
Far to the north of Corinth, other Latin-held castles guarded the strategic
mountain passes which connected northern Greece with the duchy of
Athens. Perhaps the most famous of these was Boudonitza, near Thermopylae, where Leon Sgouros tried to prevent Boniface of Montferrat from
entering Boeotia in 1205. It is possible that some of Boudonitza's oldest
medieval fortifications date from this period and from the 12205, when
Honorius III demanded that the castle should be strengthened in response
to Theodore Komnenos's Epirote invasion of Latin Thessaly. Like Corinth,
Boudonitza could not in itself prevent a hostile force from moving south;
but anyone wishing to control the area permanently would sooner or later
have to return and capture the castle, or risk being cut off.10 The importance
of frontier castles as a means of delaying invaders and preventing them
from making permanent conquests also gave these fortifications a special
standing in the feudal structure of the principality of Achaia. This is made
clear in the Assizes of Romania, the law code of Prankish Greece, which
stated that any such strongholds held by the prince could not be destroyed
or handed over to the enemy without the consent of his most powerful
vassals. This rule was strictly adhered to in 1262, when the noble ladies of
Achaia, along with the few local lords who had not been killed or captured
at the battle of Pelagonia, held lengthy debates on whether to hand over
Monemvasia, Mistra and Old Mania as a ransom for Prince William. In
the end they agreed to do so, although those who had argued that this
would enable the Greeks to 'throw us out of the country' were ultimately
proved right.11
Frontier strongholds could only delay or intimidate an invader rather
than stopping him in his tracks, but there were a few examples of medieval
fortifications in Greece which were comparable to more elaborate continuous
barriers such as Hadrian's Wall or the Great Wall of China. Most famously,
the ancient Hexamilion Wall across the Isthmus of Corinth was rebuilt
several times from 1415 onwards in an effort to halt Turkish attacks on the
Morea; something which was never achieved because there were not enough
troops or resources available to defend the wall properly.12 Less well known
were the efforts made by Andronikos II in about 1307 to prevent the Catalans,
who had by that time left Gallipoli and moved further west toward Thessaloniki, from returning to Constantinople. As a result Andronikos blocked
THE D E F E N S I V E F U N C T I O N S OF I N L A N D F O R T I F I C A T I O N S 247
off the narrow valley situated between the steep hills rising up from the sea
and the small fortified promontory occupied by the city of Christopoli.
Without access to this vital coastal route the Catalans would have found it
almost impossible to lead their horses back into Thrace through the woods
and mountains of the interior. Thus Andronikos may unwittingly have
contributed to the downfall of Prankish Athens by encouraging the Catalans
to keep moving west rather than returning to Byzantine territory.13
Decisive events such as the Catalan conquest of Athens or the Greek
acquisition of Mistra were in reality rare. For most of this period strongholds
were simply expected to protect people against localised raids rather than
to defend the frontiers of entire Christian states. These raids could be
perpetrated by a whole variety of Greek, Catalan, Turkish or even Prankish
aggressors, and they ensured that most communities were situated within
easy reach of some form of fortification. At Monemvasia, for example,
people living in the town managed to survive a Catalan raid by climbing
up to the citadel above.14 Likewise, in 1262-63 a large Greek force from
Mistra entered the principality of Achaia along the same route which William
II later used in 1271, past the castle of Karytaina and towards the city of
Andreville. Along the way the Greeks stopped at the fortress of Veligosti,
where 'they destroyed the market and left the castle intact'. This suggests
that local people could at least rely on this fortress to save their lives, even
if their houses were still exposed to the Greek attackers.15 Thirty-five years
later Isabelle of Villehardouin (1289-1307) constructed another castle in this
area which served a similar purpose, for it provided a place of refuge for
rural communities whilst at the same time ensuring that local peasants paid
their taxes to the Franks rather than the neighbouring Greek garrison of
Gardiki.16 Similar tactics were still being used in the fourteenth century,
when Niccolo Acciajuoli constructed a new Messenian stronghold to the
south of Veligosti to 'provide safety for the province of Kalamata'.17
There are many Prankish towers dotted across southern Greece and the
Aegean islands. Recent field work carried out in south-western Boeotia, an
area which belonged to the duchy of Athens during the thirteenth and
fourteenth centuries, indicates that minor Latin lords built such structures
close to or inside small rural communities so that they could provide
emergency shelter for Latin settlers and a place to store precious belongings.
The construction of Haliartos, an unusually well-preserved tower, can be
explained by the fact that a neighbouring 'large village or small medieval
town' was thriving during the Prankish period. Not far away the droughts
of 1989 and 1990 lowered the levels of a lake and revealed the remains of
another tower and a number of surrounding structures whose outer walls
formed a kind of 'protective outer facade for defensive purposes'.18 While
248
towers and buildings of this kind could never have withstood a prolonged
siege, they must have been an ideal way of protecting small numbers of
people, along with their crops and cattle, from robbers or pirates.
The presence of so many towers in certain parts of Greece also raises
some interesting questions about the nature of rural settlement and the
state of the local economy. On the one hand, these structures suggest that
areas like Boeotia were extremely insecure and affected by constant localised
warfare, yet it can also be argued that it was both expensive and timeconsuming to build stone towers of this kind, and that they could only
have been erected during relatively long periods of peace and prosperity.
Indeed, the field work carried out in south-western Boeotia suggests that
the population of this region still continued to grow between 1300 and
1460 despite the political upheavals of the period.19 It is also worth noting
that some of the better preserved Latin structures on the island of Euboea
contain decorative elements such as the circular brick feature still visible
on a tower at Phylla. Embellishments of this kind would surely not have
been added to defences constructed at haste and with few resources in the
face of overwhelming Turkish raids.20 Whilst it is true that the towers of
Prankish Greece were primarily intended for defensive purposes, they were
not necessarily built during the most intensive periods of warfare. It follows
that the more devastated by fighting a region became the fewer strongpoints
it contained, because there were not enough people or resources available
locally to facilitate the construction of castles in the first place. As was
noted in Chapter 4, a similar conclusion has been made in recent years
with regard to the Holy Land.2i
Personal touches of the kind mentioned at Phylla indicate that for many
smaller western landholders the construction of a new tower represented a
major financial commitment. Once it was completed it therefore became
a source of pride and a status symbol within the community. It is likely
that towers were intended to be inhabited continuously rather than left
empty during peace time and only occupied temporarily whenever raiders
appeared. If this is the case, Latin settlers may have spent far more time in
the Greek countryside than one might have thought. It is often argued that
Latins lived almost exclusively in cities like Glarentza, and that they rarely
visited the estates and villages which they held in the countryside.22 Yet it
seems odd for so many western newcomers to have constructed expensive
(and sometimes even personalised) stone towers if these buildings were only
used for fleeting visits, handed over to local Greek representatives or simply
allowed to stand empty for most of the year. Recent research carried out
in the Holy Land suggests that, far from being reluctant to leave their castles
and walled cities, the Franks often lived and worked in the countryside
250
U N K N O W N CRUSADER CASTLES
the lower bourg at Naxos, whose outer curtain wall could still be seen early
this century. The dukes themselves, along with their Venetian followers,
probably lived at a higher level inside the actual castled In this way the
Saudos were able to minimise the threat of violence between the two
peoples, and also perhaps stress the symbolic overlordship of the Venetians
over the locals. That settlements of this kind were protected both by an
outer wall and by an inner citadel must also have made them attractive
places to live, for their outer fortifications acted as a deterrent against pirates
and minor raiders, whilst their central defences could still halt the progress
of more determined invaders.
A similar arrangement protected most Latin-held cities on the mainland.
When Kalojan attacked the Macedonian stronghold of Serres in 1205, the
defenders eventually had to abandon the outer town but could still retreat
into the castle, 'which was very strong'.31 In the end, the citadel of Serres
could not hold out against Kalojan indefinitely, and its garrison suffered
the same fate as hundreds of other Macedonians who were robbed or killed
by the Cumans and the Bulgars.32 During the Prankish conquest of Achaia
the spirit of the local Greeks was finally broken because they ran out of
castles in which to take shelter. Like the defenders of Serres, they realised
that further resistance was useless once 'they could not get help from
anywhere else'. The fate of these people illustrates the importance of strongholds as refuge sites in times of war. Without such places, the inhabitants
of bourgs, towns and villages could not escape being robbed, enslaved or
killed.33 As in the Holy Land, the population relied on fortified houses,
castles and cities to protect itself against anything ranging from full-scale
invasions right down to common thieves and burglars.34
Castles could not always be expected to withstand enemy attacks in
isolation. No fortress was powerful enough to resist a determined besieging
army indefinitely. It was just a matter of time before Serres had to capitulate
once it became clear that there were no Prankish troops on their way to
relieve the castle. This realisation must have had a negative effect on the
morale of the garrison, thereby bringing the fall of the citadel even closer.
In a comparable incident which took place during the summer of 1292, the
same fate nearly befell the Epirote strongpoint of loannina when it came
under attack from Byzantine (formerly Nicaean) forces hoping to conquer
the despotate of Epiros. When they heard that Florent of Hainault, ruler
of Achaia, count Richard of Cephalonia and Thomas the despot were all
marching toward them, the Byzantine Greeks decided, however, to raise
the siege rather than 'wait for battle and be defeated and flee dishonourably'.
In this way loannina, and ultimately perhaps the entire despotate, were
saved from a Byzantine invasion.35
THE D E F E N S I V E F U N C T I O N S OF I N L A N D F O R T I F I C A T I O N S
25!
252
UNKNOWN CRUSADER
CASTLES
23
254
The historical evidence also suggests that the Greeks, lacking the equipment
needed to besiege castles and noting that others might already have come
to terms with the Franks, tended to stop fighting as soon as local strongholds
were occupied by the crusaders. In Messenia, for example, the Franks
initially experienced problems imposing their authority over the local
Greeks, but once they captured the strategic fortress of Kalamata this
situation changed dramatically. 'Afterwards', wrote Villehardouin, 'more
Greeks from the country submitted to them than ever before.'5
Kalamata became one of the many Byzantine castles strengthened by the
construction of new Latin fortifications, and a similar process was going
on all over the Morea during the early years of the thirteenth century. The
men responsible for these changes were figures like William Alemn of
Patras, newly created barons who began 'to change their surnames and take
the names of the fortresses they were building'.6 In so doing, they were
actively encouraged by the rulers of Achaia, who specifically stated in the
Assizes of Romania that the twelve most powerful vassals of the principality
could build their own strongholds unhindered. This policy ensured that
the countryside was subjugated as quickly as possible: the more Prankish
castles there were, the more pointless it became for the Greeks to rebel. It
may also have been intended to encourage Prankish nobles to stay in Greece
and consolidate their new baronies, thereby stabilising the political situation
between fellow newcomers, not just between Greeks and Franks.7
The need to create strong and viable new lordships as quickly as possible
meant that other westerners relied on the same tactics as those used by the
Achaian Franks. At Coron, the ever expanding thirteenth-century fortifications of the citadel enabled the Venetians to dominate much of the
surrounding countryside, not just the city itself.8 During the first three
decades of Venetian rule on Crete, castles and towers were also used to
protect Italian settlers against the locals, who resented having to share their
land with the newcomers and were less well treated than the Greeks in the
Morea.9 The mountainous areas of western Crete proved particularly
troublesome in this respect, so the Venetians eventually built the fortress
of Suda on the north-west coast of the island to try to bring order to this
region. While this gradually wore down the resistance of the Greeks, it is
interesting to note that, by placing Suda by the sea, the Venetians created
an escape route for themselves in case it ever became necessary to withdraw
from the area completely. Suda's location also suggests that the Venetians
were happy to contain rather than conquer western Crete, in the same way
that William II of Villehardouin decided to surround the Melings with a
ring of frontier castles.10
Despite such tactics the Greeks still rebelled openly from time to time,
F O R T I F I C A T I O N S AND I N T E R N A L SECURITY
255
particularly in areas where Latin rule was already relatively weak. After
Michael VIII Palaiologos acquired Maina, Mistra and Monemvasia in 1262,
for example, a combination of local grievances and encouragement from
the Byzantine authorities often led to uprisings in the south-eastern Peloponnese. In 1296, for example, an obscure argument between a Greek lord
and a Prankish knight was used as the excuse for a massive insurrection in
northern Laconia. This in turn resulted in the fall of St George, an important
Latin frontier castle, after a traitor within the ramparts lowered a ladder to
Greeks waiting below. In response, Florent of Hainault organised a lengthy
campaign against St George, which incidentally involved the use of small
forts similar to those built by the Franks at Corinth between 1205 and mo.11
This must have been successful, for according to the Chronicle of Morea
St George belonged to the Latins again about ten years later.12
Most other rebellions mentioned by contemporary sources were caused
by more specific grievances. Heavy taxation, combined with a general feeling
that the Latins were not maintaining established Byzantine customs, seem
to have sparked off a second uprising around St George in 1302.13 Similar
complaints probably led to such a large number of disturbances on Crete.14
On Chios heavy-handed tactics by the Genoese later led to so much local
resentment that between 1329 and 1346 Genoa was forced to abandon the
island altogether. No doubt the Genoese were encouraged to do so because
of the strong political support, and perhaps even direct military assistance,
given to the inhabitants of Chios by Andronikos II.15 Like the Muslim
peasants who rebelled after the battle of Hattin, the Greeks also took their
chances whenever the Latins had been defeated by an external foe and were
in serious military trouble.16 The Thracian rebellion of 1205 was clearly
timed to coincide with Kalojan's invasion of the area, whilst in 1207 the
Greek capture of the local Prankish lord precipitated a more spontaneous
uprising near Nicomedia.17 Insurrections also broke out in the wake of both
the battles of Halmyros (1311) and Pelagonia (1259), devastating encounters
which inflicted such heavy losses on the Franks that it proved impossible
for them to maintain law and order.18
Even if castles sometimes failed to prevent rebellions, such structures
could still minimise their impact. This point is perhaps best illustrated by
returning to the uprising of 1302, supposedly caused by the heavy taxes
imposed on the Greeks by the Achaian prince Philip of Savoy, but clearly
also backed by the Byzantine authorities at Mistra. This rebellion started
well for the Greeks, who captured and destroyed the castles of Ste Hlne
and Crevecoeur located near Karytaina, before the Franks even had time
to react. But by the time the rebels reached Beaufort (another fortress near
Karytaina, which should not be confused with its namesake in the Mani
256
U N K N O W N C R U S A D E R CASTLES
peninsula),19 they had already lost the element of surprise and realised that
it would be impossible to storm the castle without suffering heavy losses
at the hands of the Prankish crossbowmen stationed along the ramparts.
They therefore changed their minds and marched to St George, only to
find that they lacked the equipment to besiege it. To solve the problem,
they asked for a trebuchet to be brought from Monemvasia, but by now
they had lost so much time that the initial impetus of the rebellion was
slipping away. Meanwhile, Philip of Savoy and his barons were able to raise
an army and march south toward St George, forcing the rebels to retreat
into the mountains or back to Mistra. As a result, all resistance crumbled,
and Philip was able to rebuild or strengthen those castles which had been
attacked, collect any outstanding taxes and deal with the leaders of the
uprising.20 Despite managing to destroy two entire castles, the Greek rebels
had failed because they lacked the resources to undertake a more extensive
campaign involving lengthy sieges. They needed the element of surprise to
succeed. Once this had been taken away from them, they did not have their
own strongholds nearby in which to regroup or take shelter. The Prankish
defenders of St George and Beaufort, on the other hand, knew that they
had time on their side, and were happy to adopt the same tactic of wait
and see as that used by the Catalans against Gautier II of Brienne in 1331.
In doing so they prevented the rebellion from spreading any further, and
ultimately caused it to collapse entirely.
By retreating inside their castles rather than trying to confront rebels in
the field, the Franks could also keep their casualties to a minimum. In 1302
this was only a temporary measure until Philip of Savoy turned up, but on
other occasions it was less clear if and when a relieving army would arrive.
This was certainly the case for the unfortunate Renier of Trit, lord of
Philippopolis and Stenimaka, a town and castle situated on the northern
fringes of the Latin empire. When Kalojan invaded this area and the local
Greeks rose up in arms against the Latins (1205), Renier found himself cut
off at Philippopolis with 120 knights. During the ensuing weeks, small groups
from this force tried to leave the city and make the dangerous nine-day
journey to Constantinople despite Renier's efforts to persuade them to stay.
Most of them probably suffered the same fate as Renier's son and brother,
who were captured and beheaded along with at least thirty other knights
soon after departing from Philippopolis. As if this were not bad enough,
Renier also began to hear rumours that the inhabitants of Philippopolis
itself were planning to rebel and deliver the city to Kalojan. He therefore
retreated to the castle of Stenimaka with his fifteen remaining knights and
stayed there, cut off from the outside world, until a relieving force from
Constantinople finally reached him in June 1206. By this point Renier had
F O R T I F I C A T I O N S AND I N T E R N A L SECURITY
257
been stranded for thirteen months, but by refusing to leave his stronghold
he had avoided the fate of most of his companions, and had survived one
of the worst rebellions of the entire crusader period.21
Somewhat less spectacular tactics enabled the Genoese to withstand an
uprising on Chios in 1347. This rebellion had been organised by a local
nobleman, who had gathered together a force of Greeks and mercenaries
which was so large that it may well have outnumbered the total number
of westerners on the island. Rather than trying to confront them, the
Genoese responded by withdrawing inside the powerful citadel of Chios.
Here they were besieged for a while, until the arrival of reinforcements
caused the entire insurrection to fizzle out. Once again, fortifications rather
than superior troop numbers had saved both the lives and the territorial
claims of the Latin newcomers, enabling them to wait in safety until further
assistance arrived from elsewhere.22
For the actual rulers of crusader states there was always the added threat
that their Latin as well as their Greek vassals would rebel. The obvious way
to prevent this was to make sure that the most powerful castles in any given
area belonged to the local ruler rather than his barons. In the Catalan duchy
of Athens, the key centres Thebes, Livadia, Siderokastron, Neopatras and
Athens all belonged to the royal domain of the Aragonese dukes.23 These
sites were guarded by inaccessible hill top citadels whose capture would have
been beyond the means of any local baron. In 1275 not even the Byzantine
army of Michael VIII proved capable of taking Neopatras.24 In Achaia the
strategic fortresses of Corinth, Kalamata and Androusa, as well as Chlemoutsi
and Glarentza, were all held by the rulers of the principality. Between 1249
and 1262 these were joined briefly by Mistra and Monemvasia.25 Further
east, the fortified cities of Constantinople and Thessaloniki were of course
also royal properties, although Constantinople in particular contained a very
large Venetian quarter.26
Having gained possession of key strongholds like these, it was essential
for individual rulers to maintain their defences. When Kalojan sacked the
royal city of Serres during the Bulgar invasion of 1205, he deprived Boniface
of Montferrat of an important economic and military centre and therefore
undermined the latter's political domination over the surrounding area. As
soon as Kalujan retreated, Boniface refortified the site, thereby re-establishing his own authority over the local Greeks and Latins, and at the same
time protecting these people against future Bulgar invasions.27 Even if ramparts were kept in perfect condition, this was only good enough if the
troops who manned them could be trusted. As a result, garrisons were often
changed on the accession of new sovereigns. In 1301 Philip of Savoy, ruler
of Achaia by virtue of his marriage with Isabelle of Villehardouin, 'had the
258
U N K N O W N C R U S A D E R CASTLES
castellans and constables and some of the sergeants changed in all the castles
of his principality of Morea, and placed in them some of the people he
had brought from Piedmont and Savoy'.28 The purpose of this policy was
to avoid treachery and foster loyalty toward individual rulers, but it did
not always work. The traitor who let Greek rebels inside St George in 1296,
for example, had the suspiciously Latin sounding name of Boniface.29
Likewise, individual Franks at Karytaina and neighbouring castles were
prepared to hand these places over to the Greeks in return for personal
financial gain.30
Most rebellions or acts of treachery which involved fellow westerners
were caused by wider political disagreements. During the earliest days of
Latin rule at Constantinople arguments arose between Boniface of Montferrat and his overlord, Baldwin of Flanders, regarding the kingdom of
Thessaly. Thessaloniki in particular became a sore point between the two
men. Boniface hoped to make it his new capital, whereas Baldwin was well
aware of the city's traditional imperial status.31 Matters deteriorated even
further in the autumn of 1204, when Baldwin managed to seize Thessaloniki
before Boniface could get to it. As a result, the latter decided to rebel openly
against his lord and hastily laid siege to Adrianople, which had also been
captured recently by Baldwin. The emperor's followers inside Adrianople
therefore appealed to Constantinople for help, and a relieving army was
dispatched from the capital which forced Boniface to raise his siege and
come to terms. Boniface was subsequently allowed to keep Thessaloniki in
exchange for recognising the emperor's overlordship and territorial claims
in Thrace.32 Boniface had been prevented from conquering Thrace, and
perhaps even overthrowing Baldwin, because he had failed to capture Adrianople, whose role in the affair can be compared with that of Beaufort or
St George during the Greek rebellion of 1302.
Sometimes, fortifications tended to erode rather than consolidate central
authority. This is hardly surprising when one considers that some Latin
nobles had inherited sprawling classical or Byzantine strongholds originally
intended to withstand far larger besieging armies than anything the Franks
could muster. If vassals who held such sites were able to resist external
invasion forces numbering thousands of men, it must have been tempting
at times to defy a royal army composed of just a few dozen knights. This
certainly seems to have been the case during the winter of 1208-9, when a
powerful group of Lombard nobles who had settled in northern Greece
rebelled against the Latin emperor Henry. This rebellion was centred around
a plot to replace Demetrius, the infant son of Boniface of Montferrat (died
1207), with a rival claimant to the kingdom of Thessaly.33 In order to protect
Demetrius Henry was obliged to lead an expedition into the kingdom of
F O R T I F I C A T I O N S AND I N T E R N A L SECURITY
259
2O
claimant succeeded in taking the other man's castles. Philip only emerged
victorious after Hugh promised to leave Greece in return for a substantial
pay-off.38
Incidents like these illustrate how castles could prevent lords from
imposing their will on troublesome vassals or rival claimants. But the
chronology and frequency of rebellions can also be used to shed more light
on the decline of central authority in general. Many of the rebellions which
have been mentioned occurred in the latter half of the fourteenth century,
when the rulers of Latin states in Greece were increasingly absent in western
Europe. By contrast, the fact that there were relatively few disturbances in
Achaia before the loss of Mistra, Monemvasia and Old Mania (1262) suggests
that the principality enjoyed better internal security during this period. To
some extent, this is confirmed by contemporary descriptions of Andravida
(Andreville), the thirteenth-century capital of Achaia, which was situated
in the middle of an open plain, 'without any walls or a citadel'. Instead
of trying to rectify this situation, the earliest rulers of Achaia preferred
to reside at Chlemoutsi, which provided them with shelter but still lay
conveniently close to their centre of government.39
This situation would have been unthinkable in most other crusader states.
If Constantinople, Salonika, Acre, Antioch or Tripoli had not been protected
by major urban fortifications these cities, along with all the territories
around them, would have been overrun in a matter of weeks. Indeed, the
only Prankish settlement in the entire eastern Mediterranean which was at
all comparable with Andravida was Nicosia, the capital of Cyprus. Before
1291, and possibly until as late as the 13505, Nicosia was only defended by
a citadel and probably did not have any urban fortifications to protect it
against a major invasion force. Cyprus was also similar to the early thirteenth-century principality of Achaia in that both were geographically
isolated and therefore less exposed to external invaders than Christian
territories which shared long borders with hostile Greek, Bulgar or Muslim
states. As a result Cyprus and pre-i26os Achaia frequently enjoyed greater
internal stability than their neighbours. This may explain why the Franks
did not make an effort to fortify Andravida until the Greeks had taken
over Mistra and were beginning to attack the principality regularly. These
attacks encouraged local people to rebel, thereby increasing the need for
urban defences even further.40 A similar process took place on Cyprus,
whose capital may not have had adequate urban defences around it until
improvements were carried out following the Genoese occupation of
Famagusta in the i37os.41
Returning to the Aegean, it would be a mistake to conclude that Latin
settlers and Greek natives all hated each other, and that the two communities
F O R T I F I C A T I O N S AND I N T E R N A L S E C U R I T Y
201
24
THE M I L I T A R Y
ORDERS
263
defending the castle of St George. This fortress was situated not far from
Karytaina, which fell to Mistra at about the same time and was located so
far inland that by this point the Franks clearly needed all the help they
could get to stem the Greek advance.6 During the 13305 the German traveller
Ludolph of Sudheim also reported that 'in Achaia, or Morea, there are
brothers of the Teutonic house guarding very strong castles [and] constantly
fighting against the dukes of Athens and the Greeks'.7 This indicates that
as the fourteenth century progressed the Order was kept even busier by
Catalan as well as Byzantine incursions against the remaining Prankish lands
in the north and west of the Peloponnese. Presumably the Teutonic Knights
continued to participate in this struggle until the very end, for the last
known references to Mostenitsa date from the early fifteenth century, by
which time most of Messenia already belonged to the Greeks of Mistra.8
Beyond the Peloponnese it is impossible even to gather this amount of
evidence about the Knights. If they did hold any strongholds in the Latin
empire before 1261, or the in the duchy of Athens until 1311, all historical
and archaeological traces of their existence have long since disappeared.9
Due to their arrest and disappearance at the beginning of the fourteenth
century, the evidence concerning the Templars is even sketchier. This
problem can be illustrated by looking at the history of Palaiopolis, a small
village situated in central Elis. In 1210 this and two other settlements were
granted to the Templars by leading members of the Prankish invasion force,
including William of Champlitte.10 After the fall of the Templars, the site
was taken over by the Hospitallers, who either inherited or constructed
the medieval tower situated on a hill near the village. The present remains
of this structure indicate that it was rectangular, incorporated reused
classical masonry and had been equipped with a large vaulted cellar or
cistern. In short, this was a typical Prankish tower of the kind erected all
over Greece, although the presence of a cistern suggests that it was slightly
more substantial than the majority of such fortifications.11 Sixty years
after the dissolution of the Templars, 'the castle of Palaiopolis' still belonged
to the Hospitallers.12 At this time the Order also held a castle, or, more
likely a fortified tower, at Laffustan (Phostena) in northern Achaia, another
site originally granted to the Templars. Presumably, Laffustan evolved from
a Templar village into a Hospitaller stronghold in the same way that
Palaiopolis did.13
Beyond the Peloponnese, papal documents dating from 1210 reveal that the
Templars had held Lamia (Zeitoun) and the neighbouring (but unidentified)
Byzantine stronghold of Ravennika until the previous year. These sites had
been granted to the Order by Boniface of Montferrat and other Latin
newcomers shortly after their initial arrival in Thessaly. Thereafter the Order
264
UNKNOWN CRUSADER
CASTLES
THE M I L I T A R Y O R D E R S
205
206
THE M I L I T A R Y O R D E R S
267
Military Orders fulfilled two basic functions, one military and the other
economic. Turning first to the military role, it is clear that Ravennika and
Lamia had originally been granted to the Templars to protect the duchy of
Athens and the eastern fringes of the kingdom of Thessaly from attacks by
the Greeks of Epiros. These castles were needed 'to defend the land' and
presumably the Hospitallers occupied neighbouring Gardiki for the same
reasons.30 The emperor Henry's confiscation of Lamia and Ravennika from
the Templars may have been foolish, for by the 12203 the Epirote Greeks
had overrun this area and were even threatening Thebes.31
In broader terms, the Templars' links with these castles so soon after the
Fourth Crusade indicate that the Order was involved with the initial Latin
conquest of Greece. In 1206 the Latin emperor granted the Byzantine settlements of Satalia and Pergamum, both of which lay in areas of Asia Minor
which had not yet been conquered by the Franks, to the Templars and the
Hospitallers respectively.3^ This suggests that the first western settlers in
Greece were attempting to pursue the same policies as their Christian
neighbours in Cuidan Armenian and the Holy Land, where frontier strongholds were granted to the Military Orders in order to encourage expansion,
spread the financial burden of defence and protect vulnerable territories
against external aggressors. In the former Byzantine empire, these policies
ultimately failed through a combination of internal conflict with the Latin
emperor Henry and overwhelming external pressure from the Greeks of
Epiros and Nicaea. These two neighbours had already overrun virtually all
Latin territories in northern Greece and Asia Minor by the mid 12205.
Clearly a similar process took place in southern Morea about a century
later, when the Teutonic Knights were called upon to garrison castles like
St George in a valiant but ultimately failed attempt to stop the Greeks of
Mistra from advancing into the principality of Achaia.
The Hospitaller garrison of Sykaminon seems to have performed a similar
role by the time Niccolo da Martoni visited it in 1395, but this cannot have
been the castle's original function before the arrival of the Ottoman Turks,
when it had still been located deep inside Latin territory. The presence of
other Hospitaller properties between this area and Athens, combined with
Niccolo da Martoni's reference to 'the port of Sykaminon', suggests that
the site had originally been intended as an economic and administrative
centre which could communicate with Rhodes or western Europe via the
sea.33 Similar observations can be made about Achaian strongholds such as
Mostenitsa and Palaiopolis, mere towers which were not located anywhere
near important frontiers before the loss of Mistra in 1262. On the other
hand, Mostenitsa lay very close to other German estates in Messenia which
became the subject of a land dispute with nearby Modon during the early
268
UNKNOWN CRUSADER
CASTLES
fifteenth century. The fact that they were later granted Castiello Nuevo, and
had also acquired a house inside the walls of Chlemoutsi in 1237, implies
that the Teutonic Knights gradually became relatively powerful landowners
in the area. It seems that Mostenitsa must have acted as the focal point for
a large agricultural domain, and that it was primarily designed to protect
produce, cattle and farm revenues rather than any major settlements or
strategic frontiers. The castle's fortifications only needed to be strong enough
to deter local criminals, rebels and pirates.34 This arrangement had its
parallels further east, where towers such as Hospitaller Kolossi on Cyprus
were often used to house regional commanders and safeguard local farming
against thieves or pirates.35
The changing role of the Military Orders in Prankish Greece, and in
particular the Hospitallers, who started off as fairly minor landowners in
the years immediately after the Fourth Crusade but eventually controlled
all of Achaia between 1376 and 1381, reflected the wider political history of
the region. The insignificance of the Orders in thirteenth-century Achaia
implies that the principality enjoyed relatively good internal stability at this
time, for no ruler would actively encourage the Hospitallers, Templars and
Teutonic Knights to build up vast castellanies and estates unless it was
absolutely necessary. Muslim pressure had already created a situation in
Palestine and Syria where the three Orders held virtually independent
lordships whose presence safeguarded Christian territories but eroded central authority.36 Similarly, the appearance of the Hospitallers and Templars
at Gardiki, Lamia and Ravennika in the early thirteenth century, followed
by the much later activities of the Hospitallers and Teutonic Knights in
Achaia, indicate that the power of the Military Orders in Greece rose
whenever the military situation deteriorated. During more peaceful periods,
their local castles and estates could still provide the Orders with revenues
which helped to fund military activities in other areas.
PART FIVE
25
Residences
It would be misleading to assume that crusader fortifications were constantly
involved in some form of warfare. Although the Christian strongpoints built
or occupied in the eastern Mediterranean, like all medieval fortifications,
were primarily military structures, it does not follow that their occupants
were always engaged in fighting. Clearly there were periods of intense
warfare, such as the Third Crusade (1189-92) or the sieges undertaken by
Baybars (1260-77), but we should not allow these famous episodes to overshadow the fact that most castle garrisons enjoyed years or even decades
of tranquillity between relatively short bursts of conflict. During these more
peaceful times, castles had only to cope with the threat of minor raids,
piratical attacks or the activities of local criminals. In this sense, crusader
castles were no different from similar strongholds all over medieval Europe.
Recent research indicates that the more stable and prosperous the twelfthcentury kingdom of Jerusalem became, the more castles, towers and fortified
residences were built there by its Latin inhabitants.1 This is clear evidence
that strongpoints were not simply built to withstand large-scale invasion
forces but were intended to act as agricultural and administrative centres
whose owners could protect themselves from the dangers of medieval
society. The same argument can be applied to areas beyond the kingdom
of Jerusalem and to the period after 1187. For example, in Syria, Cyprus,
Cilician Armenia and Greece, many of the castles held by the Military
Orders were little more than fortified farmhouses used to administer local
estates. We will now take a closer look at day-to-day activities of this kind.
For the sake of convenience these have been described as 'non-military'
functions, although it will quickly become apparent that they were virtually
inseparable from the more overt military activities which have already been
discussed.
Thanks to the security provided by such structures, castles frequently
acted as the permanent homes of local lords. They were therefore provided
with numerous features intended to make daily Ufe as pleasant and comfortable as possible. Inevitably these were most common within strongholds
belonging to the actual rulers of Latin states in the East, such as the princes
of Achaia, whose residence at Chlemoutsi in the north-western Peloponnese
272
UNKNOWN CRUSADER
CASTLES
contained an unusually large number of state rooms equipped with fireplaces, latrines and cisterns, and was built on a far grander scale than any
other Prankish castle in Greece.2 To the north, the Prankish, Catalan and
Florentine lords of Athens also converted the classical ruins of the acropolis
into a palace which was described as 'a great and wondrous work' by a
European traveller who saw it in the fourteenth century.3 Similarly, the
kings of Cyprus enjoyed palatial accommodation inside the citadels of
Kyrenia and Nicosia, whilst in the summer they retreated to the much
cooler and healthier mountain fortress of St Hilarin, whose upper baileys
housed extensive royal apartments.4
St Hilarion's role as a summer retreat also highlights the need for its
living quarters to provide as much respite from the hot climate as possible,
for during the crusader period a large gothic vault was added to this fortress,
situated at the cliff's edge and deliberately left open to the elements. The
sole purpose of this vault appears to have been to give the Lusignan kings
a shaded spot where they could sit and enjoy the view, s Similar needs no
doubt help explain why some castles were provided with gardens, which
supposedly existed on top of the acropolis in Athens and possibly also at
Montfort.6 These were no doubt used for relaxation and recreation as well
as the cultivation of herbs and vegetables. Finally, it should be noted that
Turkish baths, remains of which have been unearthed at Paphos, Belvoir
and 'Atlit, were also adopted by the Latins,7 whilst at Beirut, one room
within the citadel even contained a marble fountain in the shape of a
dragon. This would have provided a decorative centrepiece to the room
and at the same time cooled the surrounding air.8
According to Willbrand of Oldenburg, the residential rooms at Beirut
also enjoyed extensive views over the Mediterranean and the surrounding
countryside, and had marble floors and ceilings designed to represent
gently lapping waves and the symbols of the Zodiac.9 In Greece, the mid
thirteenth-century castle of Thebes was likewise famous for its magnificent
murals depicting the Latin conquest of Syria.10 Although these have long
since disappeared, traces of contemporary frescoes have been discovered at
the Hospitaller strongholds of Margat and Crac des Chevaliers, as well as
the transjordan castle of Kerak.11 These remains suggest that most of the
religious or communal rooms of larger crusader fortifications were in fact
decorated in this way.
References to Latin lords actually living in such rooms on a daily basis
are rare and usually coincidental. It only becomes clear, for example, that
leading members of the Teutonic Knights used Montfort as a regular
residence from incidents such as the resignation of the Master there in 1244
following a lengthy dispute within the Order.12 Similarly, a round chamber
RESIDENCES
2/3
within one of the highest inner towers at Crac des Chevaliers, which was
supported by a beautifully decorated gothic vault and lit by two relatively
large windows, was probably reserved for the Master of the Hospitallers,
even though there are no specific references to it being used for this
purpose.13 Such buildings were usually only referred to on special occasions,
for castles did not merely act as the residences of those permanently settled
in the East, but were also used to accommodate important guests and
crusaders. During Louis IX's crusade to Egypt (1249-50), for example, his
wife Queen Margaret spent much of her time at 'Atlit, which, according to
Oliver of Paderborn, contained an entire 'palace' within its inner bailey.14
Presumably it was here that Queen Margaret stayed, and even gave birth
to a new son.15 Following his return from Egypt, Louis himself also visited
several key strongholds in the kingdom of Jerusalem, including Jaffa and
Sidon, whose outer defences were repaired at this time.16
There are numerous other thirteenth-century examples of important
travellers being put up in castles. At the time of the Fifth Crusade, Andrew
II of Hungary visited Margat and Crac des Chevaliers, and was so impressed
by these strongholds that he gave the Hospitallers certain estates in his
homeland to express his gratitude for their generosity.17 About eighty years
later, during negotiations over a possible marriage alliance between the
principality of Achaia and the duchy of Athens, Isabella of Villehardouin
deliberately chose to meet Duke Guy de la Roche and his entourage in an
area of Achaia where there were enough castles to accommodate all the
various nobles attending this meeting.18 Elsewhere it seems that particularly
significant fortresses, including Montfort, whose role as the headquarters
of an international Military Order probably guaranteed a steady flow of
visitors from Acre and beyond, even had specific accommodation set aside
for guests. In order to house these visitors comfortably, the Teutonic Knights
constructed a guest hall immediately below the castle at some point between
1229 and 1260. This structure measured approximately forty metres by ten
metres, and its vaulting, windows and doorways were clearly the work of
highly skilled craftsmen following a typically gothic style. It was impressive
and spacious enough to accommodate an entire household within easy reach
of Montfort itself.19
The arrival of important visitors and other special occasions such as weddings and coronation ceremonies were often accompanied by much feasting
and celebrating, and large, rectangular halls similar to that which has survived at Montfort provided a fitting backdrop for these events. Extensive
remains of such halls can still be seen at numerous castles in the eastern
Mediterranean, including St Hilarin and Kyrenia on Cyprus, Karytaina
and Chlemoutsi in Greece, and Beaufort, Sidon and Crac des Chevaliers in
274
U N K N O W N C R U S A D E R CASTLES
the Holy Land. These generally date from the middle years of the thirteenth
century and were all built in a gothic style similar to that used for such
structures in the West.20
It is possible to gain some idea of how these halls were used from
contemporary descriptions of celebrations following the coronation of Henry
II as king of Jerusalem in 1286. These took place in the auberge, a hall
owned by the Hospitallers in Acre's suburb of Montmusard, and included
re-enactments of Arthurian legends and other stories over a period of fifteen
days.21 No doubt the castle of Corinth witnessed similar scenes about twenty
years later, when the Achaian ruler Philip of Savoy (1301-6) organised a
famous tournament there which lasted for about three weeks and involved
virtually all the lords and knights of Prankish Greece.22 Philip clearly enjoyed
this form of entertainment, for in 1303 he also met the duke of Athens at
the neighbouring castle of Vostitza, so that the latter could do homage to
him before both men spent eight days 'feasting and celebrating with their
barons and the knights who were with them'.23
Thirteenth- and fourteenth-century castles were increasingly designed to
enhance the romantic and heroic image of knights, and formed part of the
growing interest in chivalry/The magnificent state rooms which have already
been mentioned at Chlemoutsi, for example, were intended to impress
visitors and probably helped to establish the reputation of the castle's builder,
Geoffrey II of Villehardouin (1228-46), as the epitome of knightly virtue.
According to a later source Geoffrey 'constantly maintained eighty knights
with golden spurs' at his court, 'whom he gave all that they required besides
RESIDENCES
2J5
their pay'.24 Similarly, in 1252 Joinville wrote that on hearing of Louis IX's
imminent arrival, the count of Jaffa adorned his castle with a pennon and
a shield bearing his arms at each of the five hundred openings in the
battlements, in order to present the king and his French followers with an
almost idealised image of a rich and powerful fortress.25 Its castles contributed to the reputation which the eastern Mediterranean enjoyed in the
West as an exotic and romantic destination for knights wishing to prove
their courage as well as their piety.
At other times, castles provided convenient meeting points for more
solemn occasions. In 1262, for example, the castle of Nildi, whose location
somewhere in Achaia has not yet been identified, played host to the famous
Parlement des Dames. At this gathering the wives of lords captured by the
Greeks at the battle of Pelagonia (1259), including Prince William II of
Villehardouin (1246-78), discussed the ransom terms demanded by the
Byzantines for the release of their husbands.26 In 1209 the Latin emperor
Henry (1206-16) called a similar assembly of Prankish and Lombard lords
at the citadel of Thessaloniki, in order to negotiate a settlement to the
succession dispute affecting that city.27 Once again, the large size and
impressive architecture of castle halls made them ideally suited to meetings
of this kind.
Spacious halls and communal rooms were limited to the largest and most
important strongholds, but further down the social scale many smaller
castles and isolated towers were clearly also regarded as permanent homes.
At the Red Tower, a relatively simple twelfth-century building in the vicinity
of Caesarea, traces of mosaics and painted plaster have been found in the
upper parts of the structure.28 Similar features are virtually unheard of in
the isolated towers of Prankish Greece, most of which seem to have been
far more primitive in terms of their internal appearance. However, at least
one such building is known to have contained a number of residential
rooms arranged over several floors, and to have had an oven and a wine
press attached to it. Clearly this was a fortified farm house rather than a
mere refuge site, and it has justifiably been compared with the Hospitaller
complex at Kolossi, another isolated keep from which the Order administered its sugar plantations on Cyprus.29 Returning to Greece, we have seen
that many similar towers there were built in open, fertile countryside,
implying that they belonged to, and sometimes acted as the day-to-day
residences of, Latin settlers. They were therefore situated near rural estates
rather than strategic hill tops or lines of communication. Recent field work
in south-western Boeotia has confirmed this, and has shown that local
Prankish towers 'previously believed to exist in isolation from contemporary
settlement' did in fact stand within or close to small farming communities.30
2/6
UNKNOWN CRUSADER
CASTLES
In recent years it has also been suggested that these towers were status
symbols, and that their height did not merely give their occupants greater
protection, but reflected the wealth of the men who built them.31 The best
examples of such structures are to be found on the island of Euboea, where
it has been shown that well over fifty towers were constructed by wealthy
Lombard, Greek and (perhaps) Venetian landholders, partly in response to
late fourteenth- and early fifteenth-century Turkish incursions, but partly
also to impress neighbours.32 One of these towers situated in the village of
Politika measures approximately six metres square at its base, contains three
floors with a crenellated parapet above, and stands to a height of fifteen
metres. Its beautifully constructed first-floor entrance, round-headed windows and plaque above the doorway all reflect the aspirations of a wealthy
local landholder.33
Politika should also be compared with other medieval towers built in an
urban rather than a rural setting. Acre's skyline appears to have been
dominated by similar buildings which belonged to different Orders and
trading nations constantly trying to construct slightly taller towers than
their rivals.34 This suggests that the destruction of Genoa's tower following
her defeat by the Venetians and the Pisans in 1258 was intended as an insult
to civic pride just as much as a military precaution against future Genoese
aggression.35 Beyond the Holy Land itself, reference can also be made to
the tower which originally stood on the acropolis in Athens, as well as the
sole remaining urban tower in Negroponte, both of which were probably
once very similar to surviving structures of this kind in Tuscany.36
Returning to Politika itself, one opening in the upper parts of the tower
may originally have led to a wooden balcony.37 This feature may simply
have been intended for light and ventilation, but it could also have been
used for lordly display, giving physical reality to a landholder's economic
and social superiority over his dependants. On a far grander scale, the great
hall at Kyrenia had large balconies overlooking the central courtyard of the
fortress, where the kings of Cyprus may have presented themselves to their
vassals below.3 If this was the case, Politika and Kyrenia may represent
eastern Mediterranean equivalents of similar openings in English castles
such as Newark and Sherborne, whose role has been compared to that of
'the balcony at Buckingham Palace'.39 The presence of such balconies,
particularly in the exterior wall of the tower at Politika, does at any rate
suggest that fashion, symbolism and comfort could at times take precedence
over military considerations. Hence the strongpoints of the Latin east were
often intended to display the wealth and power of local lords, not only to
those invited inside castles but to those who lived and worked around them
every day.
26
Prisons
Strong walls and deep ditches meant that it could be just as difficult to get
out of fortresses as it was to break into them, making them ideal for use
as prisons. 'Atlit, for example, appears to have been the principal gaol in
the East for the entire Templar Order, and the Rule of the Templars records
several cases of violent or dishonest brothers being locked up there.1 The
Hospitallers likewise used part of their heavily fortified headquarters at Acre
for this purpose,2 whilst at Beirut Willbrand of Oldenburg wrote that
troublesome citizens were placed in the castle moat.3 This is hardly surprising
when one considers that ditches of this kind could be well over ten metres
deep, and would therefore have been impossible to climb out of.4 It seems
unlikely that prisoners were allowed to roam around freely. They were
probably either shackled or locked up in cells similar to that cut into the
side of the famous rock-hewn ditch at Saone in Syria.5
At other times, those kept in castles were political prisoners rather than
common criminals. Hence the long internal conflict between Bohemond
VII of Tripoli and Guy II, lord of Gibelet, came to a dramatic end in 1282
when the latter was finally captured and supposedly left to starve to death
in the moat at Nephin near Tripoli.6 Good examples of similar incidents
abound from the period following the restoration of King Henry II of
Cyprus (1285-1324), who was deposed briefly by his rebellious brother
Amaury between 1306 and 1310. After Amaury himself had been murdered,
many of his followers were held in the dungeons at Kyrenia, where they
were only fed a small amount of bread and water each day, and were forced
to share cells barely two metres across. They eventually starved to death.7
Such episodes may have been unusually grim, but the fact that Henry's
successors Hugh IV (1324-59) and Peter I (1359-69) both imprisoned their
enemies at Kyrenia suggests that this was in fact the most important gaol
on Cyprus during the crusader period.8 At other times, neighbouring Buffavento, whose location on a rugged mountain top made it ideal for such
a role, was also used to house political opponents, including those followers
of Amaury of Tyre who had been lucky enough not to get sent to Kyrenia.9
During the 13805 a knight imprisoned here by King James I (1382-98) even
managed to escape by resorting to the classic trick of using a sheet as a
2/8
UNKNOWN CRUSADER
CASTLES
makeshift rope.10 Some years earlier, it seems that Peter I of Cyprus had
also intended his new citadel at Nicosia, known as the Margarita Tower,
to replace Kyrenia as the largest royal prison in the kingdom. Toward the
end of his reign, Peter even forced one of his disobedient nobles to work
alongside the slaves excavating the moat of this tower, but this so enraged
his other barons that it may well have contributed to the king's subsequent
murder. It seems that the Margarita Tower soon came to symbolise Peter's
oppressive rule, implying that he had built it to intimidate his Nicosian
vassals as much as to defend his capital against the Mamluks and the
Genoese.11 Shortly after these events, Leontios Makhairas also referred to
one of Hugh IV's enemies being 'cast into the prison of Famagusta', although
it is unclear whether this was actually inside the royal citadel or not.12
In Cuidan Armenia, castles were occasionally used to imprison religious
rather than political dissenters, for there were frequent arguments over the
merits of forging closer links between the Catholic and Armenian Churches
in exchange for more western aid against the Muslims. In 1307 or 1308,
following the Armenian Catholicos Constantine's decision to accept papal
supremacy, riots broke out in Sis orchestrated by 'monks and religious men,
priests and deacons, as well as doctors and bishops and many people, both
men and women, who refused to accept the use of water in the chalice at
mass, and other innovations. The king Oschin, with the agreement of the
patriarch and the barons, arrested all of them, imprisoned the doctors in
the citadel, and put to death many men and women, and some religious
men ,..'13
Apart from criminals or political and religious opponents, prisoners of
war were frequently held in castles. In 1262, for example, William II of
Villehardouin inflicted such a heavy defeat on Greeks attacking central
Achaia that they later had to be distributed to several different neighbouring
strongholds.14 After the fall of Acre in 1191, all 500 Christians who had
been held in the city were released, while the most important Muslim
captives were locked up in one of the towers.15 This tower may have been
situated in or near Acre's citadel, for according to Ibn al-Furat this was
where Muslims were later imprisoned during the reign of Sultan Baybars
(1260-77).16
Prisoners of this kind were subsequently often reduced to the status of
slaves, being obliged to work in order to survive. From 1240 onwards
about looo Muslim captives helped to reconstruct Saphet in Galilee,17
whilst in the late fourteenth century the Cypriots were happy to employ
Genoese prisoners during the fortification of Nicosia, even though they
were fellow Christians.18 This suggests that the Latins relied on captured
soldiers to make up for their desperate shortage in manpower, although
PRISONS
2/9
27
l8l
back into line and regain a vital source of income which had been lost to
the Greeks of Mistra. This example has a close parallel in the Holy Land,
for the rebuilding of Saphet in the early 12405 quickly brought the Templar
garrison of this castle considerable amounts of money from the 10,000
people living in the surrounding countryside. Many of these people had
previously fallen under the control of Damascus.6
Elsewhere in the Holy Land, similar links between castles, internal stability
and taxation are revealed in a Hospitaller document dating from 1263, which
stated that a Muslim settlement in lower Galilee was refusing to pay its
taxes to the Order, because of waning Prankish control in the area. This
example can be used to illustrate the fragility of a system which relied on
isolated strongpoints rather than superior numbers to suppress a hostile
population. However, it also implies that until the reign of Baybars, who
systematically pushed the Latins out of the region during the 12605, local
people had been paying up on time for decades without ever voicing
complaints.7 There are numerous cases of castles in the Holy Land being
used to impose order over local people and to collect revenues from them
dating back to the twelfth century, such as the fortress built on Mons
Glavianus above Beirut during the reign of Baldwin II (1118-31) .8
Thanks to their international structure and far-reaching commitments,
the Military Orders provide us with several examples of even greater amounts
of money being stored in castles. Most famously, the Templars safeguarded
their many financial activities by keeping their main treasury inside their
powerful headquarters at Acre, whilst the European wing of their banking
system was centred around an equally impregnable tower at the Temple in
Paris.9 Judging by John of Joinville's account of Louis IX's stay at Acre in
the summer of 1250, the former building was used almost like a modern
bank by crusaders who could deposit or withdraw money there as required.10
The funds held in Acre were also relied on by the Order itself, for the Rule
of the Templars stipulated that 'when the Master wants to go to Tripoli or
Antioch, he can take 3000 bezants or more from the treasury if necessary,
to aid the houses there'.11 More locally, the Rule also indicates that other
Templar castles and houses had their own stores of money for various
military and domestic expenses.12 Similar arrangements clearly existed within
major castles belonging to other Orders, such as the Teutonic Knights'
headquarters at Montfort.13
Apart from fines, taxes and stores of money, farm revenues and important
administrative documents were kept inside castles. At Saphet, for example,
the seven towers of the inner bailey housed 'numerous offices for all
necessary requirements' and it was presumably from here that the Templars
administered neighbouring estates and organised the daily running of their
282
UNKNOWN C R U S A D E R CASTLES
castle.14 Beyond major strongholds such as Saphet, the cultivation of farmland also centred around smaller towers and fortified structures. During
the twelfth century, and possibly again after 1192, the tower of Qaqun in
the lordship of Caesarea was 'used to enforce the lord of Caesarea's authority
over his seigneur/, even though he himself probably rarely visited this
site.15 Qaqun was only one of countless minor fortifications which had been
built between Acre and the southern fringes of the kingdom of Jerusalem
during the twelfth century in order to manage local estates and farmlands.16
Many smaller strongholds beyond the Holy Land itself, such as Hospitaller
Kolossi on Cyprus or the numerous Prankish towers in Greece, fulfilled the
same function.
On a far larger scale, the administrative institutions of entire crusader
states could be protected by fortifications. The principal mint of Achaia
was situated inside the walls of Glarentza, close to the Villehardouin centres
of government at Chlemoutsi and Andravida in the north-western Peloponnese.17 On Cyprus, the secrete, an archive recording royal debts, privileges,
rents and other earnings, was incorporated into the partially fortified palace
at Nicosia. During the 13905 this important office was moved into the new
castle built by King James I, making it far more secure against potential
Genoese or Mamluk incursions, and ensuring that James's citadel protected
the infrastructure of his kingdom as well as the inhabitants of his capital.18 During the thirteenth century, strongholds in the east belonging to
the Military Orders performed the same function, for they were used to
administer vast estates both in Europe and the Holy Land. Between 1204
and 1206 Margat played host to a General Chapter of the entire Hospitaller
Order, no doubt making use of the spacious chapel and large communal
rooms of the fortress.19
28
284
a collection point for surrounding farmlands, whilst on Cyprus the Hospitallers may have stored sugar cane in a fortified building next to the tower
of Kolossi.7 Another important industry which was often organised around
fortifications was the extraction of salt. Salt mined near 'Atlit, for example,
was probably brought inside the walls of this fortress for storage, export or
consumption.8 Fourteenth-century salt works owned by the Venetians on
Corfu and the Hospitallers on Castellorizzo were supervised from nearby
towers, where both the salt and the men who extracted it could be sheltered
from the ravages of Turks and pirates.9 This arrangement can be compared
with that of Recordane, a fortified Hospitaller mill situated on the plain of
Acre, defended by a sturdy two-storey tower.10
Whenever a conflict erupted, strongpoints could shelter many other valuable belongings which would normally have remained in the countryside.
During one of several mid thirteenth-century clashes between the Greeks
of Mistra and the Franks of central Achaia, local peasants took their cattle,
produce and anything else they could carry with them inside the nearest
stronghold.11 This incident illustrates why frontier castles and strategic
lookout posts were so important, for if farmers did not receive adequate
warning of an imminent attack they were forced to leave many of their
possessions behind. Hence the Cuman and Bulgar invasion of Latin-held
northern Greece in 1205 caused so much devastation because the attackers
did not simply overrun the area, but 'took the cattle in the countryside'
with them when they left.12 Likewise, in January 1374 Genoese invasion
forces on Cyprus reached the key northern stronghold of Kyrenia so quickly
that they managed to capture the cattle grazing in nearby fields before they
could be brought inside the fortress.13 If peasants were not alerted early
enough, or were subsequently unable to find shelter for their livestock, the
impact on the local economy could be disastrous, and the fact that farmers
had themselves escaped death or enslavement became immaterial if their
only source of food and income had been destroyed.
Castle garrisons were also keen to protect and participate in agricultural
activities as they relied on food produced locally just as much as farmers
and peasants did. The Hospitallers of Margat, for example, collected more
than 500 wagon loads of crops annually from the fertile slopes below the
castle, whilst at Saphet fresh fish was delivered daily from the River Jordan
and the Sea of Galilee.14 Surplus crops which were not needed by the
garrison itself were sold off at market; a policy which provided the castle
of Areas near Tripoli with 'considerable revenues, the annual income of its
lands coming from imposts, cane and cultivated fields, and amounting to
a large sum'.15
Other food which was not needed immediately could be stored for use
A G R I C U L T U R E AND D A I L Y L I F E
285
286
28/
undercrofts and other buildings which provided shelter and flanking fire
in wartime but which were normally used for storage and accommodation.39
There are also numerous cases of Baybars keeping prefabricated catapults
inside fortresses, which could be brought out, along with specially trained
siege engineers and sappers, whenever the sultan intended to attack a
Prankish stronghold.40 For most of their existence, therefore, fortified
structures were actually used as arsenals and storage points or for other
mundane domestic activities, rather than the waging of war.
If months or even years could pass between incidents of warfare, it follows
that many of those who lived inside strongholds rarely, if ever, got involved
in any fighting. Figures which have survived for the castle of Saphet tend
to confirm this, for they record that the peace time garrison of approximately
1700 people consisted of 430 knights, sergeants, turcopoles and crossbowmen,
plus another 820 staff and 400 slaves who were needed 'for labour and other
offices' within the castle.41 In other words, more than 70 per cent of the
garrison was primarily employed to carry out repairs, and to staff the
kitchens, workshops and various other amenities which have already been
referred to, even if they could doubtless assist in the defence of Saphet in
an emergency. This would not of course have applied to the slaves living at
the castle, some of whom may well have been the same Muslim prisoners
of war used in its original construction.42 The rest of the ancillary staff may
have been native Christians, for judging by contemporary accounts of the
Muslim capture of Saphet in 1266, they also made up most, if not all, of
the sergeants and crossbowmen.43 Bearing in mind that virtually all the
turcopoles may likewise have been native horsemen employed as mercenaries, it becomes evident that, even if we include the full-time soldiers
stationed there, Saphet's garrison only contained a tiny minority who actually
came from western Europe.44
Saphet provides us with incomparably detailed figures regarding its garrison, but it seems that a similar situation prevailed inside many other Latin
castles. In 1263, for example, the Hospitallers and the Templars rejected a
proposed prisoner exchange between the Latins and the Mamluks because
'their [Muslim] slaves were very useful to them, for they were all craftsmen
and it would have been very expensive to hire other craftsmen'.45 Although
the sources do not explicitly say that these unfortunate people were kept
in castles, it seems probable that Saphet was not the only fortress in the
Latin east which was constructed and repaired using slave labour.
Plenty of circumstantial evidence also exists for the presence of paid staff
in castles, who, like their colleagues at Saphet, were probably only expected
to fight if they came under direct attack. Indeed, it has been estimated that
Crac des Chevaliers, Margat and Mount Tabor had total garrisons of 2000,
288
U N K N O W N CRUSADER CASTLES
looo and 1300 people respectively, but only housed 520, 260 and 340 fighting
soldiers each.46 When in 1302 the Muslims captured the island of Rouad,
which lay opposite Tortosa and had been occupied by the Templars as a
forward base for attacks on the mainland, they also found 120 knights, 500
archers and 400 'men and women of the population'.4? Unlike Saphet, the
servants did not, therefore, outnumber the full-time troops, but they were
nevertheless extremely numerous for such a small and perhaps otherwise
uninhabited island. They clearly also included a large number of women,
disproving the impression given by some military historians that castle
garrisons were completely male dominated.
Similar arrangements appear to have been made in Greece, where the
number of western settlers was often as low, if not lower, than in the Holy
Land. In the Peloponnese, for example, the fortified town of Nauplion was
said to have contained two castles, one of which was granted to the Greeks,
whilst the other was occupied by the Franks. This implies that western
settlers garrisoned the upper citadel, but relied on locals to man the lower
defences of the city.48 Moreover, the Assizes of Romania suggests that Greeks
did, at times at least, perform military service to Latin lords.49 When the
Greeks of Epiros overran Thessaly in the early i22os, their blockade of
Thessaloniki appears to have lasted for several weeks or months. Unless
this only applied to the citadel, it seems impossible that Latins living there
could have defended the city's vast ramparts for so long without the aid of
native citizens or mercenaries.so Incidents of this kind suggest that, like
their neighbours further east, many Latins living around the Aegean often
had to boost their castle garrisons with locals who acted as servants or
part-time soldiers.
It is clear that peasants, farmers and castle garrisons were all closely
dependant upon each other. Those who lived and worked inside strongpoints
relied on neighbouring estates for food, supplies and revenues, but these
goods could only be provided if peasants in the countryside felt safe enough
to go about their work. This interdependence between peasants and garrisons
was extremely important to the Latins; if it broke down, their control over
any given region could collapse remarkably quickly. During Sultan Baybars'
raid on the county of Tripoli in 1270, his soldiers' horses 'grazed on the
meadows and crops of Hisn al-Akrad (Crac des Chevaliers), and this was
one of the reasons why it was captured, since its only provision came from
its crops and these were all used for pasture by the Muslim troops at this
time'.51 Moreover, previous attacks carried out by Baybars during the 12608
had clearly already taken a heavy toll on the area, for in 1268 the Hospitaller
Master complained that a lack of troops and resources was making the
strain of maintaining this and surrounding fortresses almost unbearable for
289
the Ordert Another document dating from 1263 also noted that the region
around Rafania, which lay just to the north east of Crac des Chevaliers,
'had once been the seat of a bishopric, but is now desolate and abandoned
by the Christians'.53 These reports imply that Crac des Chevaliers's storerooms were virtually empty when Baybars returned in 1271 and took the
fortress in a mere three weeks.34 Further south, similar circumstances led
to the fall of Montfort, whose territories were so eroded by enemy incursions
that by 1268 all but ten of its surrounding villages had been lost to the
Muslims. As a result, in 1270 Montfort's defenders made a temporary
arrangement with the Hospitallers, whereby they would be allowed to grow
crops for the coming year on land belonging to this latter Order. Before
the year was through, however, Montfort fell to Baybars, who had effectively
starved its garrison into submission without even needing to undertake a
lengthy sieged5
Beyond the Holy Land, constant raids by the Greeks of Nicaea, who from
1261 onwards were able to use Mistra as a starting point for attacks on
Prankish Achaia, left large parts of the principality in a similar state. As a
result, William II of Villehardouin made himself a vassal of Charles of
Anjou, in the hope that the latter could bring military aid from southern
Italy.56 Charles responded by sending arms and supplies from Brindisi,
including grain, to various castles in Albania and Greece.57 Clearly, enemy
incursions had damaged the economy to such an extent that Latin garrisons
could no longer rely on a steady supply of local food, and were therefore
incapable of defending themselves properly. About eighty years later further
raids by Greeks, Turks and Catalans persuaded the rulers of Naples to grant
the castellany of Corinth to Niccolo Acciajuoli, who was likewise expected
to repair local castles and bring desperately needed external aid to this
highly exposed area of the Peloponnese.58
29
Trade
The problems experienced by Crac des Chevaliers, Montfort and several
royal castles in Greece help to explain why Latin fortifications with direct
access to the sea survived far longer than their neighbours further inland.
Thanks to the naval strength of the Italian city states, the Latins remained
unrivalled at sea for most of the thirteenth and fourteenth centuries, but
on land they found themselves hopelessly outnumbered by the Greeks and
the Muslims. Hence inland strongpoints which relied solely on locally grown
produce for their income and food supply suffered far more economic
damage from land-based enemy incursions than coastal sites such as Acre
and Tripoli, whose defences protected lucrative seaborne trade routes, not
just vulnerable rural estates.
This can be illustrated further by looking at the growth of Famagusta,
a Cypriot port which quickly replaced Acre as the most important Latin
trading centre in the eastern Mediterranean after the latter city fell to the
Muslims in 1291. Famagusta acquired its first real urban defences during
this period. According to the sixteenth-century historian Estienne de
Lusignan, Henry II of Cyprus 'fortified it and made it resemble Acre in
every way, creating a market there, where all the strangers of the Orient
arrived, and traded in all sorts of merchandise'.1 This view has been
criticised in recent years on the grounds that Henry was simply responding
to the growing Mamluk threat, and was not consciously trying to create
a replacement for Acre. There seems to be no reason to think, however,
that Henry did not have both these concerns in mind, for Famagusta's
walls prevented the city from being captured until 1374 but clearly also
protected a booming economic centred This is given further credence by
the fact that Famagusta's post-1291 expansion occurred at exactly the
same time as the decline of Limassol, situated along the southern coast of
Cyprus. During the thirteenth century these two cities had not differed
much in terms of commercial importance, for Limassol had been used as
a staging post for the crusades of Frederick II (1228-29) and Louis IX
(1248-49).3 After the fall of Acre, however, when Cyprus became
the new front line between Christianity and Islam, Limassol remained
unfortified. As a result it was considered far less secure, making it
TRADE
291
292
U N K N O W N CRUSADER CASTLES
Cyprus remained the most peaceful of all the crusader states throughout
the thirteenth and fourteenth centuries.11
Large, heavily defended quarters in close proximity to each other, which
had only emerged at Acre thanks to its commercial importance and inadequate central government, were not constructed elsewhere. Such bases were
either actively discouraged by local rulers, as was the case on Cyprus, or
were simply not needed in ports where the volume of trade, and the
competition for that trade, was far less intense. In lesser ports, the Italians
normally only held much smaller towers or fortified houses, which were
simply used to store goods, revenues and administrative records. In 1294,
for example, one chronicle referred to a Genoese tower at Limassol, which
incidentally confirms that until the late thirteenth century this city still
played host to a fair proportion of Cypriot trade.12 Further west, contemporary records indicate that, during the following century, the Venetians
(or those living under Venetian protection) held many similar towers on
Aegean islands and the coasts of Prankish Greece.13 A well-preserved example
of such a tower, which has three floors, roof top battlements and stands
to a height of fifteen metres, can be seen near the Euboean coastal settlement
of Aliberi, w
In Cilician Armenia maritime trade during this period was almost all
limited to Corycus and Ayas, whose economic importance rose dramatically
during the second half of the thirteenth century, partly because of their
trade links with the Mongol Il-Khanate of Persia, and partly because of
their status as the only mainland ports still in Christian hands after the fall
of Acre. The numerous customs dues and tolls imposed on Italian merchants
using Ayas were collected by a representative of the Armenian kings, the
captain of customs, whose administration may well have been located in
the land castle of this city. The fact that this stronghold was pillaged by
Venetian sailors in 1307 also suggests that tolls were stored in it and that
it played a major role in the running of the port.15
At some sites, fortifications were also designed in such a way that they
controlled the arrival and departure of individual merchants and vessels.
During the 12603 Venice and Genoa regularly fought over the Tower of the
Flies: its position on a reef in Acre's harbour meant that it dominated the
85 metre wide entrance to the port, and whoever occupied it could therefore
control much of the economic life of the city.16 Similarly, the fortified bridge
connecting the Aegean island of Euboea with the Greek mainland was
divided by a drawbridge, which the Venetians of Negroponte no doubt
used to impose tolls on certain vessels, whilst at the same time preventing
Genoese or Turkish ships from getting through.17 These observations also
apply to land-based trade, for tolls could easily be collected from merchants
TRADE
293
Margat: tower used as a toll station and situated between the fortress and the
Mediterranean sea.
as they passed through city gates. In 1266 John of Ibelin, lord of Beirut,
decided to give the Order of St Lazarus ten bezants a year from the customs
dues imposed on tradesmen entering or leaving his city.18 At Acre access
to the royal Court of Chain, which was the main customs house, was
likewise guarded by an iron gate regulating the flow of goods between the
port and the rest of the city. Rulers of Acre jealously guarded their monopoly
over this lucrative area, whilst other trading nations competed fiercely for
access to the harbour front. The walls built in and around cities were
sometimes intended to keep trading rivals out just as much as political
opponents.19
Once merchants had left the safety of the fortified cities and were travelling
across the countryside, smaller Prankish strongholds offered them protection
against bandits and highwaymen. Possibly the best example of such a structure is that of Destroit, a watch tower in the vicinity of 'Atlit. Destroit was
built at a point where the rocky terrain forced the main coastal path into
a narrow defile, which could easily be used to ambush travellers.20 The
restricted nature of the site made it an ideal place for levying tolls from
tradesmen, and the Templars may well have done so in the same way that
the Hospitallers did near Margat. Here a wall had been constructed running
from the fortress itself 'toward the sea near the port of Margat', so that
people travelling between the county of Tripoli and the principality of
Antioch were obliged to pass through a small gate and pay a fee before they
could continue their journey. This operation was supervised from a tower
near the gate, whose occupants carried out the dual task of protecting
travellers against robbers, and at the same time raising revenues for the
294
30
296
U N K N O W N CRUSADER CASTLES
the neighbouring castle of Margat. The chapel of this fortress may have
become the bishop's new cathedral, serving the needs of the surrounding
diocese.6 Alternatively, it is possible that services conducted for the benefit
of those who did not belong to the actual garrison took place in another
building located in the castle town, for at Crac des Chevaliers it appears
that local peasants used a small church situated just outside the main
entrance rather than the much more grandiose chapel of the fortress itself.7
In Syria, the Latin patriarchs of Antioch also managed to outlive the
Mamluk conquest of the principality in 1268 by taking shelter inside the
fortress of Cursat, which was situated in the rugged interior and was not
finally lost to Baybars until 1275.8 This castle had presumably been chosen
as the principal residence of the patriarchs, and the depository of their
treasure, because of its strength and inaccessible location. In the very same
year that Cursat fell, Mamluk pressure forced the catholicos, or head of
the Armenian Church, to retreat from the exposed Cuidan Plain to the
impregnable stronghold of Vagha, located high in the Anti-Taurus mountains and eventually the home of all the Armenians' most important
relics.9 In the neighbouring Taurus chain, the equally inaccessible fortress
of Lampron became an important centre for Armenian miniaturist painting
in the medieval period, and therefore helped to preserve local culture
despite the deteriorating military climate.10 By taking such precautions,
both the Franks and the Armenians hoped that their holiest possessions
would not share the same fate as the undefended church of Nazareth,
which was demolished by Baybars in 1263 in an effort to undermine
Christian morale.11
Several Prankish castles in the Holy Land were themselves thought to be
extremely important in the fight against Islam. By garrisoning Mount Tabor,
the Hospitallers were not merely occupying a strategic vantage point in
central Galilee but were also defending the supposed scene of Christ's
Transfiguration.12 At Saphet it was considered equally important that the
new Templar castle stood on the ruins of a mosque and a synagogue.13
Gregory IX was referring to the religious as well as the military importance
of Montfort when he spoke of its proximity to the Muslims and its vital
contribution to the defence of the Holy Land. Crusader fortifications could
take on great spiritual significance in this way as the furthest outposts of
Christendom, particularly when they were guarded by the Hospitallers,
Templars or Teutonic Knights.14 The members of these Orders were all
monks as well as warriors, and their fortresses acted as monasteries, not
just places of war. When 'Atlit was built at the time of the Fifth Crusade,
for example, it was seen as a perfect retreat for the Templars from the
'sinful' city of Acre. It must have been regarded as a most suitable base
297
for an Order which had originally been heavily influenced by the reclusive
Cistercians.15
The monastic role of castles belonging to the Orders was also reflected
in their architecture, for they were normally provided with beautiful chapels
which the brethren could use for their daily services. Well-preserved
examples of such buildings can be found at Margat, Crac des Chevaliers
and Chastel Blanc, where the thirty metre long chapel of the Templars
formed the lower floor of the keep itself.16 At 'Atlit, and possibly Saphet,
there were also round churches of the type normally associated with the
Templars, although these are likelier to have been inspired by the Constantinian rotunda within the Holy Sepulchre rather than the Temple of the
Lord, as was once thought.17 At Crac des Chevaliers the Hospitallers also
added an intricate gothic arcade to the central hall of the castle to act as a
monastic cloister. It has even been suggested that by designing several of
their earlier castles, including Belmont in Judaea and Belvoir near Lake
Tiberias, around a cloister-type central courtyard, the Hospitallers contributed to the development of concentric fortifications almost by accident, for
an isolated central keep clearly obstructed the traditional monastic layout
which this Order hoped to achieve.18 Extensive archaeological research
carried out at 'Atlit has confirmed that the many vaulted halls built around
the inner bailey of this stronghold also corresponded 'to the dormitory,
refectory, infirmary, and guest apartments of a monastery'.19 After the
Templars acquired Beaufort and Sidon, they constructed large, communal
halls at both sites, so that these formerly baronial castles were better suited
to the daily life of monastic communities.20
By building their own chapels within their strongholds, the Military Orders
also excluded themselves from the authority of the local clergy. The papacy
encouraged this process by granting them a series of privileges, including
the right to appoint chaplain brothers for the knights' spiritual needs, and
sweeping exemptions from the payment of tithes. As a result, their castles
gave the Hospitallers, Templars and Teutonic Knights ecclesiastical as well
as military independence, for in theory at least they were answerable only
to the pope. It is hardly surprising that Latin churchmen in the East, whose
income from tithes could drop sharply as a consequence of the growth in
estates belonging to the Military Orders, were vocal in calling for their
privileges to be withdrawn.21 This must have been a complex issue, as it
has already been shown that some members of the clergy, such as the
bishops of Banyas, relied on fortifications held by the Orders to protect
them against the Muslims. By the late thirteenth century this bishopric had
become almost entirely dominated by the Hospitallers, who effectively
treated it as part of the lordship of Margat.22 A few miles down the coast,
298
U N K N O W N C R U S A D E R CASTLES
the same dilemma confronted the bishops of Tortosa, for vast Templar and
Hospitaller estates around Crac des Chevaliers, Chastel Blanc and Tortosa
itself had left them in charge of a mere handful of local churches, yet their
very own cathedral lay behind ramparts garrisoned by the Order of the
Temple.23
Although the Hospitallers, Templars and Teutonic Knights were envied
for their wealth and power, it was clear to everybody that their castles were
sorely needed to defend Christian territories. Such structures were so important that it became popular for crusaders who lacked the troops and
resources to attack the Muslims to spend their time in the East constructing
and financing new Latin fortifications. The thirteenth-century sea castle at
Sidon was built in the late 12205 by European crusaders waiting for Frederick
II to arrive from the west, whilst 'Atlit was also known as Pilgrims' Castle
because members of the Fifth Crusade largely paid for its construction.24
Both Louis IX and Richard I also took part in castle building, as it was
another way of expressing their piety once further campaigns against the
Muslims had become impracticable. In 1252 Louis IX specifically helped his
men to construct a new citadel at Jaffa 'to earn his indulgence'.25 Ultimately,
Latin fortifications became linked with the religious ideas which underpinned the entire crusading movement, for contemporaries did not
distinguish between the more down-to-earth military functions of such
buildings, and their highly symbolic role as fortified monasteries and outposts of Christianity.
3l
Conclusion
We have seen that each of the four regions which have been considered all
had certain architectural characteristics. In Cilician Armenia, the horseshoe
towers and sinuous curtain walls of mountain castles like Servantikar and
Yilan represented a distinct Armenian style, which was closely copied by
the Hospitallers at Silifke, Margat and Crac des Chevaliers, but which differed
from that of most other Prankish castles in certain respects. In Greece the
ready availability of earlier Byzantine or classical fortifications meant that
there were few truly Latin strongpoints; and even, when western settlers
did build new fortifications, these tended to be of a low quality and difficult
to distinguish from pre-1204 structures. A similar situation appears to have
prevailed on Cyprus, although it has been shown that major new defences
which were probably built to a much higher standard were erected at
Famagusta, Kyrenia and Nicosia after 1291. Meanwhile, in the Holy Land
the most substantial and sophisticated Latin fortifications of the entire
period were built at sites like Acre, Tyre, Tortosa and 'Atlit.
From this it might seem that the Franks who lived in the Holy Land
were technologically superior to Latins in other regions. Indeed, as castles
like Crac des Chevaliers grew in complexity, it is tempting to suggest that
a steady chronological advancement took place, whereby older, simpler
designs were gradually rejected in favour of more sophisticated ones which
were better able to withstand Muslim siege weapons. This argument can
be applied to major Syrian strongholds such as Crac des Chevaliers, but to
view military architecture in general as a precise science which unswervingly
moved forward because of new technological discoveries tends to oversimplify the evidence. For example, it has been shown that even though the
castrum-type castle consisting of a straightforward rectangular enclosure was
an extremely simple design, it was not abandoned after the twelfth century,
as some have suggested, but turned up again in the late fourteenth century,
when lames I of Cyprus built such a structure at Sigouri. Similarly, it would
be wrong to assume that, after they started to build massive concentric
fortifications such as 'Atlit, the Military Orders no longer constructed castles
with an isolated central keep, for the Teutonic Knights did just that when
they erected their new headquarters at Montfort during the late 12205. The
300
U N K N O W N CRUSADER CASTLES
CONCLUSION
3O1
did not even need complete walls around them because they were situated
on precipitous hill tops, whereas coastal sites such as Tortosa required huge
ramparts to protect them from land attackers. Once again these considerations make it dangerous to try to suggest that military architecture
progressed in a straight chronological line. For example, some scholars have
implied that the castle of Montfort, completed in the late 12205, was not
as sophisticated as Saphet, rebuilt in the early 12405. Whilst this may be
true on purely architectural grounds, it does not prove that the Franks had
become technologically more advanced because fifteen years had passed
since the completion of Montfort. In fact, Montfort may have been stronger
that Saphet in some ways, and simply required fewer man-made defences
because it was situated on a remote mountain spur. Thus the military
architecture of the eastern Mediterranean could vary considerably according
to local conditions and to local needs.
The Latins and the Armenians often found themselves heavily outnumbered by their opponents. In the Holy Land, internal squabbles amongst
the Muslims and the peace treaties with the Christians which resulted from
them meant that this did not become a serious problem until the 12405,
and even more so after 1260. Between that date and 1291 the crusader states
were wiped out by Mamluk armies which could be more than ten times
as large as those of the Franks. Until 1375 a similar process took place in
Cilician Armenia, where the Armenians had managed to hold their own
against the Seljuks but were far too weak to stop the Mamluks (and sometimes the Mongols). After 1291 the meagre Christian forces on Cyprus also
had to be on the alert for a Mamluk invasion; a threat which had not really
existed before the fall of Acre. Meanwhile around the Aegean none of the
Greek or Bulgar armies which confronted the Latins seem to have been as
large or as skilled in siege warfare as those of the Egyptians, but they could
still outnumber the western newcomers if they took up arms at the same
time. The kingdom of Thessaly and the Latin empire were quickly wiped
out by the simultaneous efforts of the Greeks in Nicaea and Epiros, plus
the Bulgars and the Cumans. In south-western Greece the protagonists were
more evenly matched, but here too the Latins were eventually ground down
first by their Greek enemies and then more decisively by the Ottoman
Turks.
Close comparisons can be made between Cyprus, Greece, Cilician Armenia
and the Holy Land. In this last area we have seen that major strongpoints
such as Acre and Saphet were so powerful that even though they only had
garrisons of one or two thousand Christians, they could successfully withstand tens of thousands of Muslim attackers. Although the defenders of
such sites had to cope with the destruction of the surrounding countryside,
3O2
UNKNOWN CRUSADER
CASTLES
they could wait behind their defences until inadequate supplies, lack of
progress or bad weather forced the Muslims to retreat. In this way fortifications made up for the lack of soldiers in the field and enabled the heavily
outnumbered Christians to remain in the Holy Land for most of the
thirteenth century. Exactly the same tactics were used in many other parts
of the eastern Mediterranean. For example, even though the number of
combatants involved was far smaller, the Catalans successfully defended the
duchy of Athens against Gautier II of Brienne's invasion of 1331 by retreating
inside their castles. Strong fortifications (especially around Constantinople)
also explain why the Latin empire survived for as long as it did. Elsewhere
they enabled the Armenians to retain their independence until 1375 and
prevented the Genoese from conquering any Cypriot territories beyond
Famagusta. Famous episodes such as the successful Christian defence of
Tyre against Saladin in 1187 had many less well-known parallels throughout
the thirteenth and fourteenth centuries.
The written sources also reveal two other important facts regarding the
role of Christian fortifications during hostile incursions. First, it is clear
that there were no nationwide networks of castles which could all communicate with each other using fire or smoke signals. Other scholars have
rejected this argument when discussing the Latin towers built in Greece,
and it should also be rejected with regard to the county of Tripoli and
Cilician Armenia. This does not rule out the possibility that much smaller
groups of strongpoints were linked in this way, as was clearly the case on
Rhodes and on Cyprus. But in other regions, and particularly Cilician
Armenia, a combination of political, logistical and architectural factors
would have made it impossible for an early warning system of this kind to
have extended over many hundreds of miles. Secondly, and also regarding
Cilician Armenia, it has been shown that the Armenians were not radically
different from the Latins when it came to constructing urban fortifications.
Far from turning their backs on city life and preferring to dwell in or near
mountain castles, they continued to inhabit a number of urban settlements
which were defended by walls or citadels. Throughout the eastern Mediterranean, therefore, there were many Christian communities which relied on
city ramparts to protect them against external attackers, regardless of
whether they were Latin or Armenian.
Similarities can also be identified between the Holy Land and other areas
when it comes to the attacking role of fortifications. This is most apparent
in Prankish Greece, where we have seen that the Latins used strongpoints
to consolidate new territorial gains, to launch punitive raids against opponents such as the Bulgars and to blockade Greek strongholds like Corinth.
These activities were very similar to those undertaken by the Franks in the
CONCLUSION
3O3
Holy Land, where Saphet, Crac des Chevaliers and numerous similar fortresses enabled the outnumbered Christians to extend their authority over
large tracts of the surrounding countryside. In both these regions the Latins'
lack of troops meant that there was also a close interdependence between
castles and field armies. Castles needed external assistance if they came
under siege, but at the same time Christian soldiers operating in the field
relied on neighbouring strongholds for food, water and shelter.
During the last one hundred years scholars' thoughts on the defensive
role of crusader fortifications in the Holy Land have changed dramatically.
Once it was argued that virtually every local stronghold, ranging in size
from Crac des Chevaliers to isolated watch towers, formed part of a vast
network of castles intended to protect Christian lands from external invaders.
But more recent research has shown that we need to distinguish between
very large fortresses, like Belvoir or Saphet, and the vast majority of smaller
Prankish castles, towers or fortified farmsteads. These latter structures were
simply intended to protect the local agricultural activities of individual
Christian lords. They had nothing to do with external security on a nationwide scale and were no different from the thousands of strongholds erected
all over Europe during the middle ages. Most current scholarship has focused
on the kingdom of Jerusalem in the twelfth century, but in this book it has
been suggested that the same arguments can be applied to the other crusader
states in the Holy Land and to the period after 1187. We have seen, for
example, that the fortress of Goliath in the county of Tripoli was sporadically
destroyed by major Muslim incursions, but that for most of its thirteenthcentury existence it lay in a relatively peaceful area where it was only
expected to safeguard neighbouring Hospitaller estates against much smaller
Muslim raids or the activities of local criminals.
For certain parts of their history it is clear that many fortifications beyond
the Holy Land were equally unaffected by full-scale warfare. Before 1291
Cyprus was so untroubled by the threat of an external invasion that its
Prankish rulers did not bother to construct any large new castles. Nor was
there any need for them to allow their leading vassals to do so. After 1291
fear of a Mamluk invasion led to a change in policy, but as it turned out
no major attackers turned up until the Genoese invasion of 1374. Consequently, with the exception of sporadic internal clashes such as the civil war
of 1229-33, Cypriot fortifications hardly ever faced anything more serious
than occasional raids by pirates. Furthermore, if we apply the argument that
more castles were actually built in peace time to Prankish Greece, it becomes
apparent that this region also enjoyed certain periods of calm and prosperity.
Western settlers would not have had the time or the money to erect the
numerous towers which have been mentioned on Euboea, in the duchy of
3O4
Athens and elsewhere if these areas had been affected by constant warfare.
Although the evidence is not so compelling for Cilician Armenia, certain
features, such as the local castles, estates and toll stations belonging to the
Teutonic Knights, suggest that this was another prosperous agricultural
region before the devastating Mamluk incursions of the late thirteenth and
fourteenth centuries. Indeed, in an effort to dispel the misconception that
castles were only ever used for fighting or that the crusader states were in
a permanent state of war, the last part of this book was devoted in its entirety
to the many non-military functions of fortifications.
Having looked at strongpoints throughout the eastern Mediterranean, it
is also possible to identify certain common strands regarding their role in
suppressing local populations. Greek or Muslim rebels normally lacked the
strength to capture even relatively small Latin strongholds. Structures of
this kind enabled western settlers to survive numerous native uprisings,
including the Cypriot rebellion of 1192 and the failed Muslim attack on
the Tower of David in 1229. As a result, the mere presence of Latin
fortifications often prevented the outbreak of violence. Insurrections were
usually limited to periods when the Latins had been defeated in battle or
when rebels were able to gain external assistance. Local Muslims took up
arms after the battle of Hattin, and the Greeks living around Mistra were
frequently in a state of war with the Franks of Achaia after this castle had
passed under Byzantine control in 1262. When the Latins were stronger,
however, the threat of major rebellions receded. It should again be stressed
that some crusader states were not nearly as violent as one might think:
the notion that Franks living in the Holy Land hardly ever dared to venture
outside their fortified strongpoints for fear of being attacked by Muslim
peasants has been exaggerated in the past.
Although they helped to maintain law and order, it is clear that fortifications also played a prominent role in the many internal clashes which
took place between fellow westerners. The problems which Frederick II
had in trying (and failing) to get the Templars of 'Atlit to submit to his
rule indicate that large Christian strongholds could be used by their owners
to defy the will of local rulers. Beyond the Holy Land comparable incidents
took place on Cyprus during the civil war of 1229-33 and in fourteenthcentury Greece. The most chronic difficulties of this kind affected the
Armenian kings of Ciucia, who never really succeeded in imposing their
full authority over the lords of impregnable castles like Lampron. On
the other hand, the fact that most rulers possessed far more lands and
castles than any of their vassals could also strengthen their rule, a situation
which was most apparent on Cyprus. Indeed, the relatively calm history
of Cyprus, which was of course physically isolated, suggests that central
CONCLUSION
3O5
authority was weakest in those areas which were most exposed to external
attackers. In the Holy Land Christian rulers had to allow the barons and
the Military Orders to build massive castles so that they could protect
themselves against the Muslims, despite the threat that this posed to royal
power. Similarly, by the late fourteenth century Achaia was under so much
pressure from Greek, Catalan or Turkish aggression that royal government
disappeared completely for a while and the entire region was leased to the
Hospitallers. Fortifications could either enhance or undermine internal
stability, depending on local conditions.
In Cyprus, Greece and the Holy Land there was a very strong interdependence between coastal fortifications and Christian shipping. In the
same way that field armies needed friendly castles to act as refuge points,
so fortified ports like Acre, Famagusta and Glarentza provided Latin vessels
with safe anchorages where they could conduct their trade and replenish
their stocks of food and water. It has been written that in the Holy Land
maritime links with western Europe were 'absolutely essential for the continued survival of the crusader states' because they provided the constant
stream of troops and resources needed to sustain the war with the Muslims.1
Clearly this was also the case for Latin territories in Greece and on Cyprus,
whilst the Armenians' growing links with the West were equally dependent
upon the sea. Cyprus acted as a safe base for numerous crusades of the
period, including those of Frederick II and Louis IX. In Greece we have
seen that Glarentza provided a life line between the Peloponnese and Brindisi, and that for much of its existence the Latin empire was effectively
reduced to Constantinople, a fortified haven which could only be reached
safely by sea. Throughout the eastern Mediterranean virtually all Latin
territories were eventually reduced to coastal outposts of this kind, and
after the fall of Acre in 1291 it was only by fortifying ports like Smyrna that
the crusading movement could sustain any kind of presence at all on the
Muslim mainland. As far as the Christians were concerned, the only negative
aspect of this naval activity was that rival maritime powers such as the
Genoese, the Venetians and the Catalans relied on heavily defended ports
like Tyre and Gallipoli to wreak havoc or wage war against each other.
In the period before 1291, all three Military Orders were most active in
the Holy Land as an inevitable consequence of their spiritual goals. They
had the resources needed to build and guard fortifications which were large
enough to keep up with the growing Muslim threat. From the 12505 onwards
they took over more and more former baronial strongpoints, such as Sidon
and Arsuf. Castles which were situated on the frontiers and were therefore
most exposed to Muslim attacks were likeliest to pass to the Hospitallers,
Templars or Teutonic Knights. These included Ascalon, Saphet, Beaufort,
3O6
U N K N O W N C R U S A D E R CASTLES
Crac des Chevaliers and Baghras. During the desperate final years of Christian rule coastal strongholds such as Margat, Tortosa and 'Atlit effectively
also became frontier outposts. Even though this high level of castle ownership could sometimes put the Orders beyond the reach of local rulers and
therefore undermine central authority, it is clear that their contribution
considerably lengthened the life span of the crusader states.
Despite the huge commitment which the Orders made to the Holy Land,
they were already prepared to undertake a limited number of similar projects
elsewhere in the Mediterranean before 1291. In Cilician Armenia the castles
of Silifke and Harunia, held by the Hospitallers and the Teutonic Knights
respectively, were both frontier strongholds intended to bolster this region's
defences against the Seljuk Turks. Silifke was particularly interesting, because
at one time this little-known fortress must have rivalled the much more
famous Hospitaller castle of Crac des Chevaliers, both in terms of its
appearance and its local status. However, a combination of financial, military
and political pressures forced the Order to give up Silifke, whilst in the end
Harunia probably contributed very little to the struggle with the Turks. In
Greece the Templar frontier castles of Lamia and Ravennika, along with
the neighbouring Hospitaller-occupied fortress of Gardiki, had a very
ephemeral existence before they were swallowed up by the Greeks of Epiros.
Furthermore, it has been argued in this book that the Military Orders were
not granted any Cypriot castles after 1192 which were specifically intended
to guard the island's frontiers. It has been suggested that the castle of Paphos
belonged to the king rather than to the Hospitallers, and that the Templars
simply used Gastria to administer their local estates rather than to protect
eastern Cyprus against a threatened Byzantine invasion.
The year 1291 clearly marked a dramatic turning point for all three Orders.
Cyprus became a base for the late thirteenth- and early fourteenth-century
efforts by the Hospitallers and the Templars to regain a toe hold in the
Holy Land. During the next hundred years the Hospitallers were involved
in the sporadic but ultimately failed projects to use Cilician Armenia as a
starting point for the reconquest of Syria. The fall of Acre also changed the
Military Orders' attitude toward the Peloponnese. Their presence there had
been minimal during the previous eighty years, but in the fourteenth century
both the Teutonic Knights and the Hospitallers participated in the doomed
struggle against the Greeks, the Catalans and the Ottoman Turks. During
these clashes the Orders' local castles, which had originally been situated
far away from any exposed frontiers, suddenly found themselves in the
firing line and therefore took on a similar role to that of earlier frontier
castles like Crac des Chevaliers, Harunia or Lamia. The Hospitallers' acquisition of the entire principality of Achaia between 1376 and 1381 can be
CONCLUSION
307
Abbreviations
ABSA
BZ
CHSB
EHR
HC
MGH SS
PEQ
PG
PL
QDAP
RHC
RHCArm
RHCLois
RHCOc
RHCOr
RHGF
RIS
RIS n.s.
ROL
RRH
RS
Notes
Notes to Introduction
1. R. C. Smail, Crusading Warfare, 1097-1193, second edition with an introduction
by C. Marshall (Cambridge, 1995).
2. C. Marshall, Warfare in the Latin East, 1192-1291 (Cambridge, 1992).
3. D. Pringle, Secular Buildings in the Crusader Kingdom of Jerusalem: An Archaeological Gazetteer (Cambridge, 1997).
4. Cyprus: C. Enlart, L'art gothique et de la Renaissance en Chypre, 2 vols (Paris,
1899). Cilician Armenia: R. W. Edwards, The Fortifications of Armenian Cilicio,
Dumbarton Oaks Studies, 23 (Washington, DC, 1987); H. Hellenkemper, Burgen
der Kreuzritterzeit in der Grafschaft Edessa und im Knigreich Kleinarmenien
(Bonn, 1976). Greece: A.Bon, La More franque: recherches historiques, topographiques et archologiques sur la principante d'Acae, 1204-1430, with a volume
of plans and plates (Paris, 1969). See also W. Mller-Wiener, Castles of the
Crusaders, trans. J. Brownjohn (London, 1966).
5. See below, pp. 161-67.
310
p. 204; Abu Shama, 'Deux jardins', RHCOr, vol. 5, pp. 116, 152; Ibn al-Athir,
'Kamel Altevarykh', RHCOr, vol. 2, pp. 84-86; Abu'1-Fida, 'Annales', p. 74.
6. See below, pp. 13-14, 69.
7. 'Eracles', pp. 369-74; Matthew Paris, Chronica malora, ed. H. R. Luard, RS, 57,
7 vols (London, 1872-83), vol. 3, pp. 159-61,172-77; Abu Shama, 'Deux jardins',
RHCOr, vol. 5, p. 186; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 17576; al-'Ayni, 'Le collier de perles', in RHCOr, vol. 2, pp. 187-94; al-Makin Ibn
al-'Amid, Chronique des Ayyubides, 602-658/1205-6-1259-60, trans. A-M. Edd
and F. Micheau (Paris, 1994), p. 41.
8. 'Eracles', pp. 416-19; 'Continuation de Guillaume de Tyr de 1229 a 1261, dite
du manuscrit de Rothelin', in RHC Oc, vol. 2, p. 554 (hereafter 'Rothelin');
Matthew Paris, Chronica malora, vol. 4, pp. 64-65, 140-43; Abu'1-Fida,
'Annales', pp. 120, 122; Abu Shama, 'Deux jardins', RHCOr, vol. 5, p. 193
(mentions Beaufort). See also P. Thorau, The Lion of Egypt: Sultan Baybars I
and the Near East in the Thirteenth Century, trans. P. M. Holt (London, 1992),
pp. 18-19.
9. For an outline of these crusades and the rebuilding of Ascalon, see P. Jackson,
'The Crusades of 123941 and their Aftermath', Bulletin of the School of Oriental
and African Studies, 50 (1987), pp. 32-60. See also 'Eracles', pp. 413-22; 'Rothelin', pp. 526-56; 'Les gestes des Chiprois', in RHCArm, vol. 2, pp. 72526
(hereafter 'Gestes'); Matthew Paris, Chronica malora, vol. 4, pp. 25-30, 71,
78-80, 138-45, 166. Jerusalem: Abu'1-Fida, 'Annales', pp. 117-18; al-Ayni, 'Le
collier de perles', pp. 196-97; 'Rothelin', pp. 529-31.
10. Chronica de Mallros, ed. J. Stevenson (Edinburgh, 1835), pp. 156-62; Matthew
Paris, Chronica malora, vol. 4, pp. 299-311, 337-40; 'Rothelin', pp. 561-63; 'Eracles', pp. 42728; al-'Ayni, 'Le collier de perles', p. 198.
11. Matthew Paris, Chronica malora, vol. 4, pp. 34143; 'Rothelin', pp. 564-65;
'Eracles', pp. 429-33; 'Gestes', p. 740; John of Joinville, Hlstoire de Saint Louis,
ed. N. de Wailly (Paris, 1874), pp. 288-90; Abu Shama, 'Deux jardins', RHCOr,
vol. 5, pp. 193-94; Abu'1-Fida, 'Annales', pp. 122-25; Ibn al-Furat, 'Selections
from the Tarikh al-Duwad wa'1-Muluk', in Ayyublds, Mameluks and Crusaders,
ed. and trans. U. and M. C. Lyons, with notes and an introduction by
J. S. C. Riley-Smith, 2 vols (Cambridge, 1971), vol. 2, pp. 4-8, 10-11, 46.
12. For an outline of Louis IX's Crusade, see J. Richard, Saint Louis: Crusader
King of France, ed. S. Lloyd and trans. J. Birrell (Cambridge, 1992), pp. 85-152.
13. For an outline of Baybars' career, see Thorau, The Lion of Egypt, especially at
pp. 142-210.
14. For an outline of these events, see Mayer, The Crusades, pp. 269-74; J- RileySmith, The Crusades: A Short History (London, 1987), pp. 203-7; P- M. Holt,
The Age of the Crusades: The Near East from the Eleventh Century to 1517
(London, 1986), pp. 99-106; R. Irwin, The Middle East In the Middle Ages: The
Early Years of the Mamluk Sultanate, 1250-1382 (London, 1986), pp. 6284.
15. For an outline of Muslim divisions during the first half of the thirteenth
century, see R. S. Humphreys, From Saladin to the Mongols: The Ayyublds of
311
Damascus, 1193-1260 (New York, 1977), pp. 41-307; Holt, The Age of the Crusades, pp. 60-67.
16. For an outline of the fall of the Ayyubids and the rise of the Mamluks in
Egypt, the Mongol invasion of Muslim Syria in 1260 and the accession of
Baybars, see Humphreys, From Saladin to the Mongols, pp. 309-63; Holt, The
Age of the Crusades, pp. 82-89; Thorau, The Lion of Egypt, pp. 27-88. For the
Mongols, see D. Morgan, The Mongols (Oxford, 1986), pp. 147-58.
17. R. S. Humphreys, 'The Emergence of the Mamluk Army', Studia Islamica, 45
(1977), PP- 74-76.
18. Oliver of Paderborn, 'Historia Damiatina', ed. H. Hoogeweg, in Bibliothek des
Litterarischen Vereins in Stuttgart, vol. 202 (1894), P- 24519. For references, see above, note 11.
20. Lyons and Jackson, Saladin, pp. 279-361; R. C. Smail, Crusading Warfare, second
edition with an introduction by C. Marshall (Cambridge, 1995), pp. 64-75.
21. A. S. Ehrenkreutz, ' The Place of Saladin in the Naval History of the Mediterranean Sea in the Middle Ages', Journal of the American Oriental Society, 75
(i955), PP-100-16.
22. 1271: 'Annales de Terre Sainte', ed. R. Rhricht and G. Raynaud, Archives de
rOrient Latin, 2 (1884), pp. 427-61, at p. 455 (text B); 'Gestes', p. 778; 'Eracles',
p. 460; al-Makrizi, Histoire des sultans mamlouks de l'Egypte, trans. M. E. Quatremre, 2 vols in 4 parts (Paris, 1845), vol. i(b), p. 87; Ibn al-Furat, 'Selections',
pp. 152-54; Marino Saudo, 'Liber secretorum fidelium crucis', in Gesta Dei
per Francos, ed. J. Bongars (Hanover, 1611), p. 224. See also C. Hillenbrand, The
Crusades: Islamic Perspectives (Edinburgh, 1999), pp. 571-72; J. H. Pryor, Geography, Technology and War: Studies in the Maritime History of the
Mediterranean, 649-1572 (Cambridge, 1988), pp. 112-34.
23. C. Marshall, Warfare in the Latin East (Cambridge, 1992), pp. 32-37; D. Ayalon,
'Studies on the Structure of the Mamluk Army', Bulletin of Oriental and African
Studies, 15 (1953), pp. 203-28, 448-76 at p. 223 and 16 (1954), pp. 57-90 at
pp. 67-70; Thorau, The Lion of Egypt, pp. 98-100, 175, 196.
24. 'Gestes', pp. 804, 806-7.
25. Ayalon,'The Mamluk Army', particularly at 15 (1953), p. 222,16 (1954), pp. 70-71;
Marshall, Warfare in the Latin East, pp. 32-33.
26. al-Makrizi, Histoire des sultans, vol. i(b), pp. 27-31; Ibn al-Furat, 'Selections',
pp. 88-96. See also 'Gestes', pp. 764-66; 'Eracles', pp. 454-55; Marino Saudo,
'Liber secretorum', pp. 222-23.
27. Ayalon, 'The Mamluk Army', 15 (1953), pp. 222-28. See also Marshall, Warfare
in the Latin East, pp. 33-34, and for more details on the composition, equipment
and training of thirteenth-century Muslim armies in general, see Hillenbrand,
The Crusades: Islamic Perspectives, pp. 439-67.
28. al-Makrizi, Histoire des sultans, vol. i(b), p. 78.
29. al-Makrizi, Histoire des sultans, vol. 2(a), pp. 120-31; Abu'1-Fida, 'Annales',
pp. 163-64; 'Gestes', pp. 805-18. See also Francesco Amadi in Chroniques d'Amadi et de Strambaldi, ed. R. de Mas Latrie (Paris, 1891), pp. 219-26; Florio
312
NOTES TO PAGES
8-11
313
States: The Role of the Templars', in B.Z.Kedar (ed.), The Horns of Hattin
(Jerusalem and London, 1992), pp.314-26.
46. 'Gestes', p. 767. For a brief outline of the Aragonese crusade, see Thorau, The
Lion of Egypt, pp. 199-201.
47. 'Gestes', pp. 767-68; Ibn al-Furat, 'Selections', pp. 137-39; al-Makrizi, Histoire
des sultans, vol. i(b), p. 77. The figure of 15,000 Muslims was reported by a
Prankish knight held prisoner at Saphet, who later escaped to Acre. See 'Gestes',
p. 768.
Notes to Chapter 2, The Holy Land: Military Architecture
1. 1186: Cartulaire general de l'orare des Hospitaliers de St Jean de Jerusalem,
1100-1310, ed. J. Delaville Le Roulx, 4 vols (Paris, 1894-1906), vol. i, no. 783,
pp. 491-96, no. 809, p. 505. Description based on P. Deschamps, Les chateaux
des croiss en Terre Sainte, vol. 3, La defense du comt de Tripoli (Paris, 1973),
pp. 272-84; H. Kennedy, Crusader Castles (Cambridge, 1994), pp. 16779;
E. G. Rey, Etude sur les monuments de ^architecture militaire des croiss en Syri
et dans l'ile de Chypre (Paris, 1871), pp. 19-38; W. Mller-Wiener, Castles of the
Crusaders, trans. J.M.Brownjohn (London, 1966), pp.57-58; T.E.Lawrence,
Crusader Castles, ed. D. Pringle (Oxford, 1988), p. 88; C. Cahen, La Syrie du
Nord a l'poque des croisades et la principaut franque d'Antioche (Paris, 1940),
pp. 171-72; A. Boas, Crusader Archaeology: The Material Culture of the Latin
East (London, 1999), pp. 113-15.
2. Cartulaire, vol. i, no. 144, pp. 116-18, no. 391, pp. 266-68.
3. P. Deschamps, Les chateaux des croiss en Terre Sainte, vol i, Le Crac des
Chevaliers (Paris, 1934), pp. 142-305; Kennedy, Crusader Castles, pp. 150-63;
Rey, Etude, pp. 39-67; Mller-Wiener, Castles of the Crusaders, pp. 59-62; Lawrence, Crusader Castles, pp. 7788; Boas, Crusader Archaeology, pp. 11213.
4. Deschamps, Le Crac des Chevaliers, pp. 279-83; idem, La defense du comt de
Tripoli, pp. 283-84. Earthquakes: Ibn al-Athir, 'Kamel Altevarykh', RHCOr,
vol. i, p. 572 (1170); Abu Shama, 'Deux jardins', RHCOr, vol. 4, p. 154 (1170);
Ibn al-Athir, 'History of the Atabegs of Mosul', hi RHCOr, vol. 2, pp. 261-62
(1170); Abu'1-Fida, 'Annales', pp. 40 (1170), 79 (1200/01), 83 (1202/3).
5. Abu Shama, 'Deux jardins', RHCOr, vol. 4, p. 154.
6. Silifke: see below, pp. 179-80.
7. 'The Templars and the Castle of Tortosa in Syria: An Unknown Document
Concerning the Acquisition of the Fortress', ed. J. Riley-Smith, EHR, 84 (1969),
pp. 278-88, at pp. 284-86.
8. M. Braune, 'Die mittelalterlichen Befestigungen der Stadt Tortosa/Tartus', Damaszener Mitteilungen, 2 (1985), pp. 44-54; Kennedy, Crusader Castles,
pp. 134-44; Rey, Etude, pp. 69-83, 211-14; Deschamps, La defense du comt de
Tripoli, pp. 289-91.
9. Rey, Etude, pp. 85-92; Deschamps, La defense du comt de Tripoli, pp. 252-58;
Mller-Wiener, Castles of the Crusaders, pp. 51-52; Kennedy, Crusader Castles,
314
NOTES TO PAGES
14-17
pp. 138-41; Lawrence, Crusader Castles, pp. 138-41; Boas, Crusader Archaeology,
p. 96.
10. 1218: Oliver of Paderborn, 'Historia Damiatina', p. 235.
11. Deschamps, La defense du comt de Tripoli, pp. 313-16; Barber, The New Knighthood, pp. 81-82, 85; Kennedy, Crusader Castles, pp. 68-73; Mller-Wiener,
Castles of the Crusaders, p. 53.
12. al-Makrizi, Histoire des sultans, vol. i(b), pp. 84-85; Ibn al-Furat, 'Selections',
pp. 143-49; 'Gestes', pp. 768, 777; 'Eracles', p. 460; Marino Saudo, 'Liber secretorum', p. 224; 'Annales de Terre Sainte', pp. 454 (text A), 455 (text B).
13. al-Makrizi, Histoire des sultans, vol. 2(a), p. 8o; 'Gestes', p. 792; Marino Saudo,
'Liber secretorum', p. 229.
14. al-Makrizi, Histoire des sultans, vol. 2(a), p. 126; Marino Saudo, 'Liber secretorum', p. 232.
15. RRH, no. 1447, p. 377; al-Makrizi, Histoire des sultans, vol. 2(a), p. 222.
16. Oliver of Paderborn, 'Historia Damiatina', pp. 169-71. See also Jacques de
Vitry, Lettres de Jacques de Vitry, ed. R. B. C. Huygens (Leiden, 1960), pp. 99100; Matthew Paris, Chronica maiora, vol. 3, p. 14; Burchard of Mount Sion,
'Descriptie Terrae Sanctae', in Peregrinatores medii aevi quatuor, ed.
J. C. M. Laurent (Leipzig, 1873), p. 83; C. N. Johns, 'Excavations at Pilgrims'
Castle ('Atlit): The Ancient Tell and the Outer Defences of the Castle', QDAP,
3 (1933-34) > PP-145-64, at pp. 152-64; C. N. Johns, Guide to 'Atlit: The Crusader
Castle, Town and Surroundings (Jerusalem, 1947), pp.36-67. See also M. Benvenisti, The Crusaders in the Holy Land ( Jerusalem, 1970), pp. 179-82; Kennedy,
Crusader Castles, pp. 124-27; D. Pringle, Secular Buildings in the Crusader Kingdom of Jerusalem: An Archaeological Gazetteer (Cambridge, 1997), nos 21, 22,
pp. 22-24; D. Pringle, 'Town Defences in the Crusader Kingdom of Jerusalem',
in I. A. Corns and M. Wolfe (eds), The Medieval City under Siege (Woodbridge,
i995)> PP- 69-121, at pp. 91-92; P. Deschamps, Les chateaux des croiss en Terre
Sainte, vol. 2, La defense du royaume de Jerusalem (Paris, 1939), pp. 32-33; Rey,
Etude, pp. 95-100; D. Pringle, The Churches of the Crusader Kingdom of Jerusalem, 3 vols (Cambridge, 1993 onwards), vol. i, pp. 69-71; Boas, Crusader
Archaeology, pp. 110-12.
17. 'Gestes', p. 818; Marino Saudo, 'Liber secretorum', p. 232.
18. 1168: RRH, no. 447, p. 116.1240: 'Eracles', pp. 417-18, 435; 'Rothelin', pp. 551-53.
19. 'Un noveau traite du texte De constructione castri Saphet', ed. R. B. C. Huygens,
Studi Medievali, series 3, vol. 4, part i (1965), pp. 335-87, at pp. 383-84, lines
160-93; D. Pringle, 'Review Article: Reconstructing the Castle of Safad', PEQ,
117 (1985), pp. 139-49. Another description of Saphet is given by Huygens in
his introduction to De constructione castri Saphet (pp. 370-77), but Pringle
(pp. 141, 142, 145) disagrees with many of his conclusions. See also Pringle,
Gazetteer, no. 191, pp. 91-92; Deschamps, La defense du royanme de Jerusalem,
pp. 140-42; Benvenisti, The Crusaders in the Holy Land, pp. 199-201; Kennedy,
Crusader Castles, pp. 128-29; Pringle, Churches, vol. 2 (1998), pp. 206-9; Boas,
Crusader Archaeology, p. 118.
315
20. Ihn al-Furat, 'Selections', p. 87; al-Makrizi, Histoire des sultans, vol. i(b), p. 27;
'Annales de Terre Sainte', p. 452 (text B).
21. 1189: Ihn al-Athir, 'Kamel Altevarykh', RHCOr, vol. i, pp. 712-13, 737; Imad
ad-Din al-Isfahani, Conqute de la Syri et de la Palestine par Saladin, trans.
H. Masse (Paris, 1972), pp. 150-54; Cont. WT, p. 75. For a description of elvoir,
see J. Prawer, The Latin Kingdom of Jerusalem (London, 1972), pp. 300-7;
Pringle, Gazetteer, no. 46, pp. 32-33; Pringle, Churches, vol. i, pp. 120-21; Boas,
Crusader Archaeology, pp. 106-8.
22. Cartulaire, vol. 2, no. 2726, p. 777, no. 2811, pp. 815-17.
23. Abu'1-Fida, 'Annales', pp. 86-87; 'Eracles', p. 317; Marino Saudo, 'Liber secretorum', p. 206; Benvenisti, The Crusaders in the Holy Land, pp. 360-61;
Pringle, Churches, vol. 2, pp. 66-67.
24. Oliver of Paderborn, 'Historia Damiatina', pp. 165-67; 'Eracles', p. 324; Jacques
de Vitry, Lettres, p. 98; Abu Shama, 'Deux jardins', RHCOr, vol. 5, pp. 163-64.
25. Oliver of Paderborn, 'Historia Damiatina', pp. 171-72; Jacques de Vitry, Lettres,
p. 108; 'Eracles', pp. 330-31; Matthew Paris, Chronica maiora, vol. 3, p. 14; Abu
Shama, 'Deux jardins', RHCOr, vol. 5, pp. 165-66.
26. Benvenisti, The Crusaders in the Holy Land, pp. 361-62; Pringle, Churches, vol.
2, p. 68. 1263: Ibn al-Furat, 'Selections', p. 56 n. i.
27. 1190: Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 395-400,441; Imad ad-Din,
Conqute de la Syri, pp. 159-62, 210; Ibn al-Athir, 'Kamel Altevarykh', RHCOr,
vol. i, pp. 738-39; Abu'1-Fida, 'Annales', p. 61; Cont. WT, pp. 79-81. 1268: Ibn
al-Furat, 'Selections', pp. 108-12; al-Makrizi, Histoire des sultans, vol. i(b), p. 51;
'Gestes', p. 771; 'Eracles', p. 456. Descriptions: Deschamps, La defense du
royanme de Jerusalem, pp. 198-208; Pringle, Gazetteer, no. 44, p. 31; MllerWiener, Castles of the Crusaders, pp. 62-63; Rey, Etude, pp. 127-32; Boas,
Crusader Archaeology, p. no.
28. Tabulae ordinis Theutonid, ed. E. Strehlke (Berlin, 1869), no. 63, pp. 51-53;
RRH, no. 1002, p. 263.
29. D. Pringle, 'A Thirteenth-Century Hall at Montfort Castle in Western Galilee',
Antiquaries Journal, 66 (1986), pp. 52-82, at pp. 54-56; Pringle, Gazetteer, no.
156, pp. 73-75; Benvenisti, The Crusaders in the Holy Land, pp. 335-37; Rey,
Etude, pp. 146-51; Marshall, Warfare in the Latin East, pp. 108-11; Kennedy,
Crusader Castles, pp. 129-31; Pringle, Churches, vol. 2, pp. 40-42; Boas, Crusader
Archaeology, pp. 109-10.
30. Marshall, Warfare in the Latin East, pp. non.
31. Deschamps, La defense du comt de Tripoli, p. 309; Kennedy, Crusader Castles,
pp. 68-70.
32. Deschamps, La defense du royanme de Jerusalem, pp. 211-13, 217-20; Pringle,
Gazetteer, no. 225, p. 103; Pringle, Churches, vol. 2, p. 372.
33. Akkar: Ibn al-Furat, 'Selections', pp. 147-49 (1271); al-Makrizi, Histoire des
sultans, vol. i(b), p. 85; 'Gestes', p. 777. Cave de Tyron: Pringle, Churches, vol.
2, p. 372; Pringle, Gazetteer, no. 225, p. 103.
34. Saladin captured Beirut in 1187 and demolished the town walls in 1190. See
3l6
Cont. WT, pp. 78, 98. 1197: Arnold of Lbeck, 'Chronica Slavorum', pp. 205-6.
Ibelins: 'Gestes', pp. 678-79.
35. Willbrand of Oldenburg, 'Itinerarium Terrae Sanctae', ed. S. de Sandoli, in
Itinera Hierosolymitana Crucesignatorum, 4 vols (Jerusalem, 1978-84), vol. 3,
p. 204.
36. R. du Mesnil du Buisson, 'Les anciennes defenses de Beyrouth', Syria, 2 (1921),
pp. 235-57, SV-V; I- France, Western Warfare in the Age of the Crusades, 10001300 (London, 1999), p. 268 n. 33. See also Pringle, 'Town Defences', pp. 87-88;
Pringle, Gazetteer, no. 45, p. 32; Pringle, Churches, vol. i, pp. 111-12; Boas,
Crusader Archaeology, pp. 45-46.
37. For a full list of these and other castles whose destruction was ordered by
Saladin, see Itinerarium peregrinorum et gesta Regis Ricardi, ed. W. Stubbs, RS,
38, 2 vols (London, 1864-65), vol i, pp. 280-82; Ambroise, L'estoire de la Guerre
Samte, ed. G. Paris (Paris, 1897), lines 6840-69, col. 183; Imad ad-Din, Conqute
de la Syri, pp. 231, 345-46. The extent to which Saladin's orders were actually
carried out seems to have varied.
38. Caesarea: Itinerarium, pp. 254-56; Benvenisti, The Crusaders in the Holy Land,
p. 138. Arsuf: Itinerarium, p. 282; Matthew Paris, Chronica maiora, vol. 2, p. 376.
Jaffa: Itinerarium, p. 284; Cont. WT, p. 135; Ambroise, L'estoire, lines 6941-7053,
cols 185-88. Ascalon: ibid., lines 7778-8086, cols 208-16; Itinerarium, pp. 315-17;
D. Pringle, 'King Richard I and the Walls of Ascalon', PEQ, 116 (1984), pp. 13347, at pp. 136-42. See also Pringle, 'Town Defences', pp. 84-85 (Ascalon), 89-90
(Caesarea), 93 (Jaffa), 94 (Arsuf).
39. Itinerarium, pp. 396-423; Matthew Paris, Chronica maiora, vol. 2, pp. 387-90;
Cont. WT, pp. 145-47; Imad ad-Din, Conqute de la Syri, pp. 384-87; Abu
Shama, 'Deux jardins', RHCOr, vol. 5, pp. 67-71; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 64-65.
40. Rey, Etude, pp. 205-10; Benvenisti, The Crusaders in the Holy Land, pp. 121-25;
P. Deschamps, 'Les entrees des chateaux des croiss en Syri et leurs defenses',
Syria, 13 (1932), pp. 369-87, at p. 386; Boas, Crusader Archaeology, pp. 44-45.
41. Ascalon: Itinerarium, pp. 428-29; Cont. WT, p. 153. Jaffa: Cont. WT, pp. 191-93;
Arnold of Lbeck, 'Chronica Slavorum', p. 204; Abu Shama, 'Deux jardins',
RHCOr, vol. 5, pp. 116.152; Abu'1-Fida, 'Annales', p. 74; Ibn al-Athir, 'Kamel
Altevarykh', RHCOr, vol. 2, pp. 84-86.
42. Caesarea: Oliver of Paderborn, 'Historia Damiatina', pp. 168-70; 'Eracles',
p. 325. Arsuf: Willbrand of Oldenburg, 'Itinerarium', p. 232.
43. 'Rothelin', pp. 531-32, 553, 555-56; 'Eracles', pp. 413-14, 421; Matthew Paris,
Chronica maiora, vol. 4, p. 143; Pringle, 'King Richard I and the Walls of
Ascalon', pp. 143-46; Benvenisti, The Crusaders in the Holy Land, pp. 119-20,
125-26. Pringle and Benvenisti both argue that no attempts were made to
rebuild Ascalon's town walls after 1192, but Pringle challenges Benvenisti's
earlier view on the exact location of the citadel. See also Pringle, Gazetteer,
no. 20, p. 21; Pringle, 'Town Defences', p. 85.
44. Cartulaire, vol. 2, no. 2320, p. 615. Egyptians: 'Rothelin', p. 565; 'Eracles', p. 741;
3V
Matthew Paris, Chronica malora, vol. 4, p. 343; 'Annales de Terre Sainte', p. 442;
Ihn al-Furat, 'Selections', pp. lo-ii; Abu Shama, 'Deux jardins', RHCOr, vol.
5, p. 194; Abu'1-Fida, 'Annales', p. 125.
45. 'Eracles', p. 334; Oliver of Paderborn, 'Historia Damiatina', p. 244; Jacques de
Vitiy, Lettres, pp. 101-2.
46. Chronique d'Ernoul et de Bernard le Trsorier, ed. L. de Mas Latrie (Paris, 1871),
pp. 458-61; 'Eracles', p. 373; Matthew Paris, Chronica malora, vol. 3, p. 175.
47. Jaffa: Ibn al-Furat, 'Selections', p. 107; Joinville, Histoire de Saint Louis, pp. 282
84, 306-8. Caesarea: ibid., pp. 256, 258, 336; Benvenisti, The Crusaders in the
Holy Land, p. 141.
48. 'Gestes', p. 728; 'Annales de Terre Sainte', p. 440. Quote: Joinville, Histoire de
Saint Louis, p. 308.
49. Cartulaire, vol. 3, no. 2972, p. i, no. 2985, p. 6; RRH, no. 1302, p. 341, no. 1313,
P- 343! 'Gestes', pp. 758-59; J. S. C. Riley-Smith, The Knights of St John in Jerusalem and Cyprus, 1050-1310 (London, 1967), pp. 133-34.
50. Arsuf and Caesarea: 'Gestes', p. 758; 'Eracles', p. 450; Marino Saudo, 'Liber
secretorum', p. 222; 'Annales de Terre Sainte', pp. 451-52; al-Makrizi, Histoire
des sultans, vol. i(b), pp. 6-15; Ibn al-Furat, 'Selections', pp. 70-71, 73-82. Jaffa:
'Gestes', p. 771; 'Eracles', p. 456; Marino Saudo, 'Liber secretorum', p. 223;
al-Makrizi, Histoire des sultans, vol. i(b), p. 51; Ibn al-Furat, 'Selections', pp. 1068.
51. Pringle, 'King Richard I and the Walls of Ascalon', p. 145; Pringle, 'Town
Defences', p. 85; Pringle, Gazetteer, no, 20, p. 21; Pringle, Churches, vol. i, no.
18, pp. 66-67.
52. I. Roll, 'Medieval Apollonia-Arsuf: A Fortified Coastal Town in the Levant of
the Early Muslim and Crusader Periods', in M. Baiard (ed.), Autour de la
Premire Croisade (Paris, 1996), pp. 595-606, at pp. 602-6; Benvenisti, The
Crusaders in the Holy Land, pp. 13435; Pringle, Churches, vol. i, pp. 5961;
Pringle, Gazetteer, no. 19, pp. 20-21; Pringle, 'Town Defences', p. 94; Boas,
Crusader Archaeology, pp. 4344.
53. Pringle, Churches, vol. \, pp. 264-67, and see no. 109, pp. 267-68, nos 111-12,
p. 270; Pringle, Gazetteer, no. no, p. 52; Marshall, "Warfare in the Latin East,
pp. 139-44, especially p. 141; Pringle, 'Town Defences', pp. 93-94; Boas, Crusader
Archaeology, pp. 49-50.
54. 'Eracles', pp. 432-33; Joinville, Histoire de Saint Louis, pp. 288, 290.1241: Matthew
Paris, Chronica malora, vol. 4, p. 142.1247: see above, p. 3. Description: Z. Razi
and E. Braun, 'The Lost Crusader Castle of Tiberias', in Kedar, The Horns of
Hattin, pp. 216-27, especially at pp. 221-27; Pringle, Gazetteer, no. 222, pp. 101-2;
Pringle, 'Town Defences', pp. 94-95. See also Pringle, Churches, vol. 2, pp. 351-53
and no. 267, p. 365; Boas, Crusader Archaeology, p. 57.
55. K. G. Holum, R. L. Hohlfelder, R. J. Bull and A. Raban, King Herod's Dream:
Caesarea on the Sea (New Haven and London, 1988), pp. 231-32; Benvenisti,
The Crusaders in the Holy Land, pp. 143-44; Pringle, Gazetteer, no. 76, p. 43;
Pringle, 'Town Defences', pp. 89-90.
3l8
56. Quote: Burchard of Mount Sion, 'Descriptie', p. 28, and see pp. 27-28; Willbrand of Oldenburg, 'Itinerarium', p. 206; Deschamps, La defense du comt de
Trpoli, pp. 300-1.
57. 1188: Imad ad-Din, Conqute de la Syri, p. 125; Ibn al-Athir, 'Kamel Altevarykh',
RHCOr, vol. i, p. 718.1271: According to Muslim chroniclers, Baybars held
Maraclea by this date. See Ibn al-Furat, 'Selections', pp. 150, 166; al-Makrizi,
Histoire des sultans, vol. i(b), p. 100. For a general history of Nephin and
Maraclea during the crusader period, see Deschamps, La defense du comt de
Tripoli, pp. 297-300, 323-26.
58. Ibn 'Abd al-Rahim, 'Vie de Kalavun', extract cited in J. F. Michaud, Bibliothque
des Croisades, vol. 4 (Paris, 1829), pp. 551-52; Deschamps, La defense du comt
de Tripoli, pp. 325-26. The exact history of Maraclea during the 12705 and
1280S remains unclear. See R. Irwin, 'The Mamluk Conquest of Tripoli', in
P. Edbury (ed.), Crusade and Settlement (Cardiff, 1984), pp. 246-49, at pp. 2484959. 1227-28: Chronique d'Ernoul et de Bernard le Trsorier, pp. 458-59; 'Bracks',
p. 365; 'Gestes', p. 676.1291: 'Gestes', p. 817; al-Makrizi, Histoire des sultans, vol.
2(a), pp. 130-31. Descriptions: H. Kalayan, 'The Sea Castle of Sidon', Bulletin
du Muse de Beyrouth, 26 (1973), pp. 81-89 and plates 1-10; Pringle, Churches,
vol. 2, no. 242, pp. 323-28; Pringle, Gazetteer, no. 201, pp. 94-95; Deschamps,
La defense du royanme de Jerusalem, pp. 229-33; Rey, Etude, pp. 154-59; Kennedy, Crusader Castles, pp. 122-24; Pringle, 'Town Defences', p. 88;
Mller-Wiener, Castles of the Crusaders, p. 70.
60. Deschamps, La defense du comt de Trpoli, pp. 208-15; Mller-Wiener, Castles
of the Crusaders, pp. 64-65; Kennedy, Crusader Castles, pp. 65-66.
61. 1188-97: 'Eracles', pp. 72, 227, 228; Cont WT, pp. 58, 98. It is possible that the
Embriaco family was allowed to cultivate lands in this region beyond the fall
of Tripoli in 1289 and Acre in 1291. See Irwin, 'The Mamluk Conquest of the
County of Tripoli', p. 249, and for a general history of Gibelet see Deschamps,
La defense du comt de Trpoli, pp. 203-8.
62. Joinville, Histoire de Saint Louis, pp. 302, 336; Pringle, Gazetteer, no. 201,
pp. 94-95; Deschamps, La defense du royaume de Jerusalem, pp. 227-29; Kennedy, Crusader Castles, p. 121; Mller-Wiener, Castles of the Crusaders, p. 70;
D. Pringle, 'A Castle in the Sand: Mottes in the Crusader East', in Chateau
Gaillard, 18 (Caen, 1998), pp. 187-91, at p. 188.
63. See below, p. 103.
64. Charles de Terre Sainte provenant de l'abbaye de Josaphat, ed. H. F. Delaborde
(Paris, 1880), no. 46, pp. 9596 (quote p. 96). See also Pringle, 'Town Defences',
pp. 86-87.
65. E. G. Rey, 'Etude sur la topographic de la ville d'Acre au XlIIe sicle', Mmoires
de la socit nationale des antiquaries de France, 39 (1874), pp. 115-45, at
p. 131; Benvenisti, The Crusaders in the Holy Land, p. 95. 1291: See below,
pp. 50-51.
66. The pilgrim abbot Daniel, cited in C. N. Johns, 'The Citadel, Jerusalem: A
319
Summary of Work Since 1934', QDAP, 14 (1950), pp. 121-90, at p. 164. See also
ibid., pp. 140-44, 165.
67. Johns, 'The Citadel, Jerusalem', pp. 169-88; A. J. Boas, Crusader Archaeology,
pp. 17-19; Pringle, Gazetteer, no. 115, pp. 54-55; Benvenisti, The Crusaders in
the Holy Land, pp. 52-53.
68. See below, pp. 42-43.
69. William of Tyre, Chronicon, ed. R. B. C. Huygens, 2 vols (Turnhout, 1986), vol.
i, bk. io, c. 26, p. 486, lines 1-18. For descriptions, see Deschamps, La defense
du comt de Tripoli, pp. 293-95; Mller-Wiener, Castles of the Crusaders, pp. 4243; Kennedy, Crusader Castles, p. 63. The Mamluks rebuilt the citadel early in
the fourteenth century; see al-Makrizi, Histoire des sultans, vol. 2(b), p. 281.
70. T. A. Sinclair, Eastern Turkey: An Architectural and Archaeological Survey, 4
vols (London, 1987-90), vol. 4 (1990), pp. 244-48 and plate 52; Cahen, La Syrie
du Nord, p. 129; Rey, Etude, p. 190.
71. Abu'1-Fida, Annales, pp. 162-63; al-Makrizi, Histoire des sultans, vol. 2(a),
pp. 102-3; 'Annales de Terre Sainte', p. 460 (text B).
72. Ibn al-Furat, 'Selections', p. 122.
73. Ludolph of Sudheim, 'De itinere Terrae Sanctae', ed. F. Deycks, in Bibliothek
des Litterarischen Vereins in Stuttgart, vol. 25 (Stuttgart, 1851), pp. 39-40.
74. Willbrand of Oldenburg, Ttinerarium', p. 200; Burchard of Mount Sion,
'Descriptio', p. 23. The famous fourteenth-century Genoese map of Acre is
reproduced at the back of Marino Saudo, 'Liber secretorum', and also in
D. Jacoby, 'Crusader Acre in the Thirteenth Century: Urban Layout and
Topography', in D. Jacoby, Studies on the Crusader States and on Venetian
Expansion (Northampton, 1989), c. 5, pp. 1-45, at figs 1-4. For more details of
these and other maps and the known topography of crusader Acre, see
A. Kesten, The Old City of Acre: Re-Examination Report, 1993 (Acre, 1993). For
other modern studies on Acre's defences, see Rey, 'Etude sur la topographic',
pp. 118-32; R. Frankel, 'The North-Western Corner of Crusader Acre', Israel
Exploration Journal, 37, part 4 (1987), pp. 256-61; Boas, Crusader Archaeology,
pp. 34-35; Pringle, 'Town Defences', pp. 81-84; Pringle, Gazetteer, no. 5, pp. 15
16; Mller-Wiener, Castles of the Crusaders, pp. 72-74; Benvenisti, The Crusaders
in the Holy Land, pp. 93-95.
75. Joinvie, Histoire de Saint Louis, pp. 334, 336; Marino Saudo, 'Liber secretorum', p. 219. It is clear that Montmusard was already fortified in some way
by the 12505. For more details, see D. Jacoby, 'Montmusard, Suburb of Crusader
Acre', in Jacoby, Studies on the Crusader States, c. 6, pp. 205-17, particularly
at p. 213; Pringle, 'Town Defences', pp. 83-84.
76. Sinclair, Eastern Turkey, vol. 4, pp. 242-48 and plates 47, 51-54; Rey, Etude,
pp. 185-204; Cahen, La Syrie du Nord, pp. 127-33; Deschamps, La defense du
comt de Tripoli, pp. 46-47; Boas, Crusader Archaeology, p. 42.
77. Sinclair, Eastern Turkey, vol. 4, pp. 243-48.
78. Tripoli: Willbrand of Oldenburg, 'Itinerarium', p. 210. Tyre: William of Tyre,
Chronicon, vol. i, bk. 13, c. 5, pp. 591-92, lines 1-17; Willbrand of Oldenburg,
32O
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
no.
111.
112.
113.
321
Rey, Etude, p. 40. Margat: see above, p. 13. Beaufort: see above, p. 18, Saphet:
see above, pp. 1617 and De constructione castri Saphet, line 255, p. 386.
Deschamps, La defense du comt de Trpoli, pp. 311-12; Kennedy, Crusader
Castles, p. 78; Boas, Crusader Archaeology, p. 102.
Boas, Crusader Archaeology, p, 101; Marshall, Warfare in the Latin East, p. 100;
Pringle, Gazetteer, p. 10.
D. Pringle, 'Aqua Bella: The Interpretation of a Crusader Courtyard Building',
in Kedar, The Horns of Hattin, pp. 147-67, at p. 147.
D. Pringle, The Red Tower (London, 1986), pp. 15, 18-19, 73-75 (and for more
examples of similar structures see p. 20); S. Tibbie, Monarchy and Lordships
in the Latin Kingdom of Jerusalem, 1099-1291 (Oxford, 1989), p. 139; Pringle,
Gazetteer, no. 229, p. 104.
Pringle, The Red Tower, pp. 15, 63-68, 70; Pringle, Gazetteer, no. 168, pp. 83-84;
Benvenisti, The Crusaders in the Holy Land, pp. 198-99.
Pringle, The Red Tower, pp. 16-18; Deschamps, La defense du comt de Trpoli,
pp. 317-19; Kennedy, Crusader Castles, pp. 73-75; Mller-Wiener, Castles of the
Crusaders, p. 52; Boas, Crusader Archaeology, p. 95.
Kennedy, Crusader Castles, pp. 75-77; Rey, Etude, pp. 101-2. For further
examples, see Boas, Crusader Archaeology, pp. 93-100; D. Pringle, 'Towers in
Crusader Palestine', in Chateau Gaillard, 16 (Caen, 1994), pp. 335-50.
See above, pp. 8-9.
Kennedy, Crusader Castles, p. 175.
Oliver of Paderborn, 'Historia Damiatina', p. 170.
See, for example, Kennedy, Crusader Castles, p. 134 and plate 46; Braune, 'Die
mittelalterlichen Befestigungen der Stadt Tortosa', plates 17, 21, 21.
Beirut: Willbrand of Oldenburg, 'Itinerarium', p. 204. Sidon: Deschamps, La
defense du royaume de Jerusalem, p. 232. 'Atlit: Johns, 'The Faubourg and its
Defences', p. 123. Maraclea: Deschamps, La defense du comt de Trpoli, p. 326.
Ascalon: Matthew Paris, Chronica maiora, vol. 4, p. 243. Tiberias: Razi and
Braun, 'The Lost Crusader Castle of Tiberias', p. 224.
al-Makrizi, Histoire des sultans, vol. i(b), p. 7.
Caesarea: Benvenisti, The Crusaders in the Holy Land, p. 141. Crac des Chevaliers: Deschamps, Le Crac des Chevaliers, pp. 189-90. We have seen that this
was one of many Syrian castles damaged by earthquakes in the crusader period;
hence sloping revetments were also intended to strengthen walls against this
threat.
Extract from the anonymous 'Tashrif', trans, and ed. F. Gabrielli in Arab
Historians of the Crusades, trans, from Italian by E. J. Costello (London, 1969),
p. 336, and see pp. 334-38.
Oliver of Paderborn, 'Historia Damiatina', pp. 169-70.
See above, pp. 13-14, 16-17. Saphet may have had a combination of rounded
and rectangular towers; see Pringle, Gazetteer, no, 191, p. 91.
See below, pp. 145-47.
Kennedy, Crusader Castles, pp. 134, 137.
322
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
323
141. Tabulae ordinis Theutonici, no. 64, p. 53, no. 66, p. 54, no. 72, pp. 56-57; Prawer,
The Latin Kingdom, pp. 308-9; Pringle, 'A Thirteenth-Century Hall', p. 53;
Kennedy, Crusader Castles-, pp. 129-31.
142. Montfort withstood an attack in 1266 before falling to Baybars in 1271. See
Ibn al-Furat, 'Selections', pp. 87 (1266), 106-12 (1271); al-Makrizi, Histoire des
sultans, vol. i(b), pp. 27 (1266), 87 (1271); 'Annales de Terre Sainte', p. 452 (text
B) (1266); 'Eracles', p. 460 (1271); 'Gestes', p. 778 (1271).
143. Quote: Ibn al-Furat, 'Selections', p. 148, and see pp. 143-49; al-Makrizi, Histoire
des sultans, vol. i(b), pp. 84-85; 'Gestes', pp. 768, 777; Marino Saudo, 'Liber
secretorum', p. 224; 'Annales de Terre Sainte', pp. 454 (text A), 455 (text B).
144. William of Tyre, Chronicon, vol. i, bk. 11, c. 26, p. 535, lines 26-27; Boas,
Crusader Archaeology, p. 117.
145. Smail, Crusading Warfare, pp. 232-33; Marshall, Warfare in the Latin East,
p. 100; Pringle, Gazetteer, p. 10.
146. Chevedden, 'The Trebuchet', p. 59.
147. France, Western Warfare in the Age of the Crusades, p. 3. For the walls of
Constantinople, see below, pp. 205-6.
Notes to Chapter 3, The Holy Land: Fortifications and External Security
1. R. Ellenblum, 'Three Generations of Prankish Castle-Building in the Latin
Kingdom of Jerusalem', in Balard, Autour de la Premire Croisade, pp. 517-51,
at pp. 520-29. For Belvoir, see also Prawer, The Latin Kingdom of Jerusalem,
pp. 300-7; Pringle, Gazetteer, no. 46, pp. 32-33.
2. Riley-Smith, 'The Templars and the Castle of Tortosa', pp. 278-79, 284-85.
3. Imad ad-Din, Conqute de la Syrie, pp. 121-22,124-26; Baha'-al-Din, 'Anecdotes
et beaux traits de la vie de sultan Youssof', in RHCOr, vol. 3, pp. 106-7; Abu
Shama, 'Deux jardins, RHCOr, vol. 4, pp. 349-63; Cont. WT, pp. 86-87; 'Eracles', pp. 119-22.
4. Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 346, 381-86 (Saphet, Kerak,
Montreal), 344-49, 386-89 (Belvoir), 395-400, 441 (Beaufort); Baha'-al-Din,
'Anecdotes', pp. 104-5,117-20 (Belvoir), 118-20,122 (Kerak, Montreal, Saphet),
121-22,132,151 (Beaufort); Imad ad-Din, Conqute de la Syri, pp. 104-7, i47~5
(Kerak and Saphet), 104,152-53 (Belvoir), 159-62, 210 (Beaufort); Ibn al-Athir,
'Kamel Altevarykh', RHCOr, vol. i, pp. 712-13, 734-35, 737; Cont. WT, pp. 75-76
(Belvoir and Saphet), 76, 86 (Montreal and Kerak), 79-81 (Beaufort).
5. Tyre survived a Muslim siege in the winter of 1187-88. Saladin avoided direct
attacks against both Tripoli and Antioch, preferring to loot and occupy castles
around them. See Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 341-45 (Tyre),
349-51 (Tripoli), 364-81 (Antioch); Baha'-al-Din, 'Anecdotes', pp. 98-99,102-3
(Tyre), 108-18 (Antioch); Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. i,
pp. 694-96 (Tyre); Imad ad-Din, Conqute de la Syri, pp. 63-80 (Tyre), 121-22
(Tripoli), 144-45 (Antioch); Cont. WT, pp. 76-79 (Tyre), 85-86 (Tripoli), 87
(Antioch).
324
NOTES TO PAGES
39-41
325
26. 'Gestes', p. 786. See also 'Annales de Terre Sainte', p. 457 (text A); Marino
Saudo, 'Liber secretorum', p. 228.
27. Ibn al-Furat, 'Selections', p. 122. See also al-Makrizi, Histoire des sultans, vol.
i(b), p. 5328. Sinclair, Eastern Turkey, vol. 4, pp. 242-48; Rey, Etude, pp. 185-204; Cahen, La
Syrie du Nord, pp. 127-33.
29. Ibn al-Furat, 'Selections', p. 122. See also pp. 121-26; al-Makrizi, Histoire des
sultans, vol. i(b), pp. 52-54.
30. 'Eracles', pp. 384-85.
31. al-'Ayni, 'Le collier de perles', p. 196.
32. 'Rothelin', p. 529.
33. Annales prioratus de Dunstaplia, ed. H. R. Luard, in Annales monastiti, vol. 3,
RS, 36 (London, 1866), p. 150. This siege was recorded in few of the sources,
and remains something of a mystery. See Marshall, Warfare in the Latin East,
pp. 243-45; Jackson, 'The Crusades of 1239-41 and their Aftermath', p. 38.
34. 'Rothelin', pp. 529-30; Ibn al-Furat, 'Selections', p. 62; al-'Ayni, 'Le collier de
perles', pp. 196-97; Abu'1-Fida, 'Annales', pp. 117-18.
35. Johns, 'The Citadel, Jerusalem', p. 169. 1244: Matthew Paris, Chronica maiora,
vol. 4, pp. 299-311, 337-40; 'Rothelin', pp. 561-63; 'Eracles', pp. 427-28; al-'Ayni,
'Le collier de perles', p. 198.
36. See above, p. 22.
37. Ibn al-Furat, 'Selections', p. 122; al-Makrizi, Histoire des sultans, vol. i(b), p. 53.
38. Bar Hebraeus, Chronography, p. 448.
39. 'Gestes', p. 807. Benvenisti, The Crusaders in the Holy Land, p. 27, suggests
30,000 for Acre and Tyre.
40. See above, pp. 23-24.
41. See below, pp. 8384.
42. Rey, 'Etude sur la topographic', p. 132; Marshall, Warfare in the Latin East,
p. 67.
43. See maps of Acre at the back of Marino Saudo, 'Liber secretorum'; in Jacoby,
'Crusader Acre', figs 1-4; Mller-Wiener, Castles of the Crusaders, p. 73.
44. al-Makrizi, Histoire des sultans, vol. i(b), p. 8; Ibn al-Furat, 'Selections', p. 72.
45. 1188: Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 353-55; Cant. WT, p. 87.
1232: 'Gestes', p. 701.
46. Oliver of Paderborn, 'Historia Damiatina', p. 255.
47. Itinerarium, pp. 396-423; Matthew Paris, Chronica maiora, vol. 2, pp. 387-90;
Cont. WT, pp. 145-47; Imad ad-Din, Conqute de la Syrie, pp. 384-87; Baha'al-Din, 'Anecdotes', pp. 323-33; Abu Shama, 'Deux jardins', RHCOr, vol. 5,
pp. 67-71; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 64-65.
48. 'Gestes', pp. 704-8; Amadi, Chronique, p. 155; Bustron, Chronique, p. 86.
49. 'Gestes', p. 758.
50. Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 349-51; 'Eracles', pp. 119-21.
51. 'Eracles', p. 334; Oliver of Paderborn, 'Historia Damiatina', p. 244; Jacques de
Vitry, Lettres, pp. 101-2.
326
52. 'Gestes', p. 817; al-Makrizi, Histoire des sultans, vol. 2,(a), pp. 130-31.
53. Cont. WT, pp. 44-45, 56.
54. Ibn al-Furat, 'Selections', pp. 72, 75, 77; al-Makrizi, Histoire des sultans, vol.
i(b), pp. 8, 10; 'Gestes', p. 758; 'Eracles', p. 450 (mentions Arsuf only).
55. Oliver of Paderborn, 'Historia Damiatina', p. 171. See also Matthew Paris,
Chronica maiora, vol. 3, p. 14; Johns, Guide to 'Atlit, pp. 49-50.
56. Deschamps, La defense du royanme de Jerusalem, pp. 229-31; Rey, Etude, p. 157.
57. The northern breakwater of this harbour was built out of Roman columns;
another example of crusader recycling. See Rey, Etude, pp. 222-23; Benvenisti,
The Crusaders in the Holy Land, pp. 143-44; Holum, Hohlfelder, Bull and
Raban, King Herod's Dream, pp. 231-34.
58. Rey, Etude, pp. 173-74; Du Mesnil du Buisson, 'Les anciennes defenses de
Beyrouth', p. 244.
59. Tyre: In 1242 Ibelin opponents of Frederick II lowered the harbour chain at
Tyre in order to let their Venetian allies into the city. See 'Gestes', pp. 732-35.
See also Deschamps, La defense du royanme de Jerusalem-, p. 136; Rey, Etude,
pp. 167-69. Acre: Jacoby, 'Crusader Acre in the Thirteenth Century', pp. 8-10;
R. Gertwagen, 'The Crusader Port of Acre: Layout and Problems of Maintenance', in Balard, Autour de la Premire Croisade, pp. 553-82, at pp. 559-60.
Gertwagen argues that the layout of Acre's port was quite different from that
envisaged by many previous scholars.
60. 1262: 'Gestes', p. 755; 'Annales de Terre Sainte', p. 450 (text B); Marino Saudo,
'Liber secretorum', p. 221; al-Makrizi, Histoire des sultans, vol. i(a), pp. 177-78;
Ibn al-Furat 'Selections', p. 50. 1229: 'Eracles', pp. 384-85.
61. 'Gestes', p. 741; 'Eracles', p. 433 says twenty-two galleys. For Egyptian naval
weakness, see above, pp. 6-7.
62. Ibn al-Furat, 'Selections', pp. 93-96 (Saphet), 122-26 (Antioch); al-Makrizi,
Histoire des sultans, vol. i(b), pp. 30 (Saphet), 52-54 (Antioch); 'Gestes', pp. 76466 (Saphet).
63. Joinville, Histoire de Saint Louis, p. 302.
64. Tabulae ordinis Theutonid, no. 72, pp. 56-57.
65. Ibn al-Furat, 'Selections', p. 122. See also al-Makrizi, Histoire des sultans, vol.
i(b), p. 5366. See above, p. 39.
67. Matthew Paris, Chronica maiora, vol. 4, p. 309.
68. al-Makrizi, Histoire des sultans, vol. i(b), p. 8; Ibn al-Furat, 'Selections', p. 72.
69. Joinville, Histoire de Saint Louis, pp. 302, 336.
70. 'Gestes', p. 752; 'Eracles', p. 444; Jackson, 'The Crisis', pp. 499-500.
71. De construction castri Saphet, lines 256-58, p. 386.
72. al-Makrizi, Histoire des sultans, vol. i(b), pp. 30 (Saphet), 51 (Beaufort), 84-85
(Chastel Blanc and Akkar); Ibn al-Furat, 'Selections', pp. 94-95 (Saphet), 112
(Beaufort), 143 (Chastel Blanc), 148-49 (Akkar).
73. S. B. Edgington, 'The Doves of War: The Part Played by Carrier Pigeons in
the Crusades', in Balard, Autour de la Premire Croisade, pp. 167-75.
327
328
93.
94.
95.
96.
329
33O
41. 'Gestes', pp. 814-16; Kesten, The Old City of Acre, pp. 69-81; Benvenisti, The
Crusaders in the Holy Land, pp. 104-9; Barber, The New Knighthood, pp. 241-43;
Riley-Smith, The Knights of St John, pp. 248-49; Rey, 'Etude sur la topographic',
pp. 13536, 139-40. The Hospitaller compound is in the process of being
excavated. For a summary of recent discoveries, see Boas, Crusader Archaeology,
pp. 37-41.
42. 'Gestes', p. 815; Kesten, The Old City of Acre, pp. 29-39 (Genoa), 40-45 (Venice),
59-67 (Pisa); Benvenisti, The Crusaders in the Holy Land, pp. 98100 (Pisa),
100-2 (Genoa), 102-4 (Venice); Rey, 'Etude sur la topographic', pp. 137-39;
Jacoby, 'Crusader Acre in the Thirteenth Century3, pp. 19-26 (Pisa), 26-30
(Genoa), 30-36 (Venice).
43. 'Gestes', pp. 742-48. See also 'Eracles', p. 443; 'Rothelin', pp. 633-35; 'Annales
de Terre Sainte', pp. 447-48; Marino Saudo, 'Liber secretorum', pp. 220-21;
Riley-Smith, The Feudal Nobility, pp. 215-17.
44. 1258: 'Gestes', pp. 742-50.1287-89: 'Gestes', pp. 800-2; 'Annales de Terre Sainte',
pp. 459-60.
45. 'Gestes', pp. 756-57, 768-69.
46. France, Western Warfare in the Age of the Crusades, p. 86.
331
6854, col. 183, lines 7177-214, col. 193. Description: D. Pringle, 'Templar Castles
between Jaffa and Jerusalem', in H. Nicholson (ed.), The Military Orders 2:
Welfare and Warfare (Aldershot, 1998), pp. 89-109, at pp. 92-94.
14. Pringle, 'Templar Castles', pp. 89109.
15. Oliver of Paderborn, 'Historia Damiatina', p. 171.
16. Matthew Paris, Chronica malora, vol. 4, p. 290; Pringle, 'Templar Castles',
pp. 94-102. 1244: Chronica de Mauros, pp. 156-62.
17. Joinville, Histoire de Saint Louis, p. 302; Pringle, 'Templar Castles', p. 96.
18. Coni. WT, pp. 76-79; Imad ad-Din, Conqute de la Syne, pp. 63-80; Baha'-alDin, 'Anecdotes', RHCOr, pp. 98-99, 102-3; Abu Shama, 'Deux jardins',
RHCOr, vol. 4, pp. 341-45; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. i,
pp. 694-96.
19. Itinerarium, pp. 27-28; Coni. WT, pp. 83, 85-86; Abu Shama, 'Deux jardins',
RHCOr, vol. 4, pp. 352-53> 356-5720. Battle of Arsuf: Itinerarium, pp. 262-80; Ambroise, L'estoire de la guerre sainte,
lines 6090-630, cols 163-77; Abu Shama, 'Deux jardins', RHCOr, vol. 5, pp. 3840; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 49-50; Baha'-al-Din,
'Anecdotes', pp. 258-61.
21. 'Eracles', pp. 224-26; Abu Shama, 'Deux jardins', RHCOr, vol. 5, pp. 116-17;
Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 86-87; Arnold of Lbeck,
'Chronica Slavorum', pp. 205-6.
22. Oliver of Paderborn, 'Historia Damiatina', pp. 165-67; 'Eracles', p. 324; Jacques
de Vitry, Lettres, p. 98; Abu Shama, 'Deux jardins', RHCOr, vol. 5, pp. 163-64.
23. 'Eracles', p. 461; 'Gestes', pp. 778-79; al-Makrizi, Histoire des sultans, vol. i(b),
p. 101; Ibn al-Furat, 'Selections', p. 155; al-'Ayni, 'Le collier de perles', p. 246.
See also Pringle, The Red Tower, pp. 60, 62; Marshall, Warfare in the Latin
East, p. 206.
24. Arnold of Lbeck, 'Chronica Slavorum', pp. 207-10; Cont. WT, pp. 195-97;
Abu'1-Fida, 'Annales', p. 74; Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2,
pp. 87-88. For a description of Toron, see Deschamps, La defense du royanme
de Jerusalem, pp. 117-18.
25. Baha'-al-Din, 'Anecdotes', p. 274.
26. Le trsor des diartes d'Armenie ou cartulaire de la chanclleme royale des Roupniens, ed. V. Langlois (Venice, 1863), no. u, pp. 130-31 ( Jbala), no. 12,
pp. 132-33 (Bikisrail); Cartulaire, vol. 2, nos 1262-63, PP-70-71 (Jbala), no.
1355, PP-122-23 (Bikisrail).
27. RRH, no. 949, pp. 251-52, no. 1043, p. 272.
28. Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 105-6.
29. 1237: Matthew Paris, Chronica malora, vol. 3, pp. 404-6; Abu'1-Fida, 'Annales',
pp. 112-13; Izz al-Din Ibn Saddad, Description de la Syrie du Nord, trans. A-M.
Edd-Terrasse (Damascus, 1984), pp. 265-66. The importance of Templar
castles north of Antioch will be discussed in more detail below. See pp. 182-86.
30. Ibn al-Furat, 'Selections', p. 115.
31. Ibn al-Furat, 'Selections', pp. 115,126. See also Jackson ('The Crisis', pp. 494-96),
332
who argues that Bohemond made these conquests through sheer force rather
than with the assistance of the Mongols. 1188: Imad ad-Din, Conqute de la
Syne, pp. 133-40; Baha'-al-Din, 'Anecdotes', pp. 112-115; Abu Shama, 'Deux
jardins', RHCOr, vol. 4, pp. 368-74.
32. Darbsak: Cahen, La Syrie du Nord, p. 705. Darbsak is listed as one of the
castles Baybars acquired from the Armenians. See Abu'1-Fida, 'rmales', p. 152;
Ibn al-Furat, 'Selections', p. 166. 1268: ibid., p. 126.
33. 'Gestes', p. 778; Ibn al-Furat, 'Selections', p. 155.
34. Matthew Paris, Chronica malora, vol. 3, pp. 10-11; 'Eracles', pp. 323-25; Jacques
de Vitry, Lettres, p. 99; Oliver of Paderborn, 'Historia Damiatina', pp. 167-68.
35. Joinville, Histoire de Saint Louis, pp. 310-18.
36. 'Gestes', p. 752; 'Eracles', p. 444; Jackson, 'The Crisis', pp. 499-500.
37. 'Gestes', p. 767.
38. al-Makrizi, 'Histoire d'Egypte', pp. 126-28, 126 n. 3-4, 128 n. i; Abu'1-Fida,
'rmales', p. 81.
39. Homs and Hama: al-Makrizi, 'Histoire d'Egypte', p. 135; Ibn al-Athir, 'Kamel
Altevarykh', RHCOr, vol. 2, p. 105; Kamal-ad-Din, 'L'histoire d'Alep', pp. 44-45.
Jbala: al-Makrizi, 'Histoire d'Egypte', pp. 135, 127m; 'Eracles', pp. 247-48.
40. Ibn al-Furat, 'Selections', pp. 83-84.
41. 'Gestes', p. 784; Marino Saudo, 'Liber secretorum', p. 228; 'rmales de Terre
Sainte', p. 457 (text A). Hospitallers and Templars from Margat and surrounding strongholds may also have participated in an attack on Homs and Hama
undertaken by Prankish, Armenian, Mongol and Georgian troops in 1282. See
Hethoum the Historian, 'Table chronologique', p. 487.
42. France, Western Warfare in the Age of the Crusades, p. 218.
43. De constructione castri Saphet, line 251, p. 385.
44. Oliver of Paderborn, 'Historia Damiatina', pp. 171-72.
45. Oliver of Paderborn, 'Historia Damiatina', pp. 171-72; Jacques de Vitry, Lettres,
p. 108; 'Eracles', pp. 330-31; Matthew Paris, Chronica maiora, vol. 3, p. 14; Abu
Shama, 'Deux jardins', RHCOr, vol. 5, pp. 165-66.
46. Ibn al-Furat, 'Selections', p. 101.
47. Burchard of Mount Sion, 'Descriptio', pp. 83-84.
48. See above, p. 22.
49. Ibn al-Furat, 'Selections', p. 98; al-Makrizi, Histoire des sultans, vol. i(b), p. 32;
Willbrand of Oldenburg, 'Itinerarium', p. 210.
50. Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, p. 180.
51. al'-Ayni, 'Le collier de perles', p. 194; Deschamps, Le Crac des Chevaliers,
p. 128.
52. 'Eracles', pp. 403-5; Abu'1-Fida, 'Annales', p. 115.
53. Ibn al-Furat, 'Selections', p. 98; al-Makrizi, Histoire des sultans, vol. i(b), p. 32.
54. See below, pp. 280-81.
55. Joinville, Histoire de Saint Louis, pp. 310-18.
56. Deschamps, La defense du corate de Tripoli, pp. 283-84.
57. Ibn al-Furat, 'Selections', p. 148.
333
58. 'Gestes', p. 784 (1279), 786 (1281); Marino Saudo, 'Liber secretorum', p. 228;
'Annales de Terre Sainte', p. 457 (text A).
59. 'Rothelin', pp. 630-32; Marshall, Warfare in the Latin East, p. 143.
60. Bar Hebraeus, Chronography, pp. 492-93; Abu'1-Pida, 'Annales', p. 194.
61. La regle du Temple, cl. 168.
62. Matthew Paris, Chronica malora, vol. 4, p. 342.
63. Matthew Paris, Chronica maiora, vol. 3, pp. 404-6; Abu'1-Fida, 'Annales',
pp. 112-13.
64. 'Eracles', p. 455; Marino Saudo, 'Liber secretorum', p. 222.
65. For an outline of Saladin's extensive conquests in 1187 and 1188, see Lyons
and Jackson, Saladin, pp. 267-94.
66. Matthew Paris, Chronica maiora, vol. 4, p. 301; Riley-Smith in Ibn al-Furat,
'Selections', p. 173 n. 2.
67. 'Rothelin', p. 565; 'Eracles', p. 741; Matthew Paris, Chronica maiora, vol. 4,
p. 343; Ibn al-Furat, 'Selections', pp. 10-11; Abu Shama, 'Deux jardins', RHCOr,
vol. 5, p. 194; Abu'1-Fida, 'Annales', p. 125.
68. 'Eracles', p. 445; 'Gestes', pp. 752-53; Marino Saudo, 'Liber secretorum', p. 221;
Abu Shama, 'Deux jardins', RHCOr, vol. 5, p. 204.
69. De constructione castri Saphet, lines 204-9, P-384; Marshall, Warfare in the
Latin East, pp. 118-20.
70. The estimate of 900 knights comes from Abu Shama, 'Deux jardins', RHCOr,
vol. 5, p. 204. See also 'Eracles', p. 445; 'Annales de Terre Sainte', p. 449 (text
A).
71. 'Eracles', p. 441; 'Rothelin', p. 629; Marshall, Warfare in the Latin East, pp. 77-83;
Marshall, 'The French Regiment', pp. 3017.
72. Willbrand of Oldenburg, 'Itinerarium', pp. 208-10.
73. 'Eracles', pp. 403-5; Abu'1-Fida, 'Annales', p. 115.
74. Ibn al-Furat, 'Selections', p. 86.
75. Ibn al-Furat, 'Selections', p. 143; al-Makrizi, Histoire des sultans, vol. i(b), p. 84.
76. Cartulaire, vol. 4, no. 3308, pp. 291-93; Marshall, Warfare in the Latin East,
pp. 117-18.
77. De constructione castri Saphet, lines 204-9, P- 3^4; Cartulaire, vol. 2, no. 2727,
pp. 777-78; Marshall, Warfare in the Latin East, pp. 117-20.
78. 'Gestes', pp. 784, 786; Marino Saudo, 'Liber secretorum', p. 228; 'Annales de
Terre Sainte', p. 457 (text A).
79. 'Gestes', pp. 701, 704; Amadi, Chronique, pp. 149, 153; Bustron, Chronique,
pp. 81, 85.
80. Itinerarium, pp. 62, 73; Ambroise, L'estoire de la guerre sainte, lines 3060-76,
cols 82-83; 'Histoire des patriarches d'Alexandrie', extract cited in J. F. Michaud,
Bibliotheque des Croisades, vol. 4 (Paris, 1829), p. 257.
81. Sicily: Itinerarium, p. 168. Acre: Richard of Devizes, Cronicn de tempore Regis
Ricardi, ed. J. T. Appleby (London, 1963), p. 42.
82. See below, pp. 240-41.
83. Oliver of Paderborn, 'Historia Damiatina', p. 255.
334
NOTES TO PAGES
76-81
335
no. 783, pp. 491-96, no. 809, p. 505 (Margat), vol. 2, no. 2320, p. 615 (Ascalon),
no. 2726, p. 777, no. 2811, pp. 815-17 (Mount Tabor), vol. 3, no. 2972, p. i, no.
2985, p. 6 (Arsuf ).
11. Cartulaire, vol. i, no. 82, pp. 76-78 (Goliath), no. 519, pp. 353-54, no. 549,
pp. 371-72 (Chastel Rouge).
12. Ibn al-Furat, 'Selections', pp. 143, 147.
13. Cartulaire, vol. i, no. 783, pp. 491-96. For details of all Hospitaller properties
in the area, see Riley-Smith, The Knights of St John, pp. 93-95.
14. R. P. Harper and D. Pringle, 'Belmont Castle: A Historical Notice and Preliminary Report of Excavations in 1986', Levant, 20 (1988), pp. 101-18, at p. 102.
See also Pringle, Gazetteer, no. 207, p. 96.1191: Ambroise, L'estoire de la guerre
sainte, line 6859, col. 183.
15. Thus, for example, Marshall, Warfare in the Latin East (published 1992),
pp. 20-21, disagrees with Riley-Smith, The Knights of St John (published 1967),
pp. 415-16, 436-37. See also Pringle, Gazetteer, no. 46, p. 32 (Belvoir), no. 228,
p. 104 (Forbelet); Pringle, 'Survey of Castles in the Crusader Kingdom of
Jerusalem', p. 90 (Forbelet).
16. Pringle, Churches, vol. i, nos 31-32, p. 95; Pringle, Gazetteer, no. 32, p. 27.
17. La Tor de l'Opital: Riley-Smith, The Knights of St John, p. 135; Pringle, Gazetteer,
no. 69, p. 41. Turris Salinarum: Tibbie, Monarchy and Lordships, p. 150; Pringle,
Gazetteer, no. 67, p. 41.
18. RRH, no. 949, pp. 251-52, no. 1043, p. 272.
19. Ibn al-Furat, 'Selections', pp. 115 (Latakia), 128 (Jbala).
20. Willbrand of Oldenburg, 'Itinerarium', p. 212; Ibn al-Furat, 'Selections', p. 128
n. 5; Cahen, La Syrie du Nord, p. 171.
21. Cartulaire, vol. 2, no. 2117, pp. 486-87, no. 2120, p. 489; Riley-Smith, The Knights
of St John, pp. 446, 450. The Hospitallers had held Recordane since at least
1154. See Cartulaire, vol. i, no. 225, p. 173.
22. Cartulaire, vol. i, no. 879, p. 559; Pringle, The Red Tower, p. 59.
23. Cartulaire, vol. 2, no. 2141, p. 501, no. 2482, pp. 673-75.
24. Pringle, 'Templar Castles', pp. 89-109.
25. See D. Pringle, 'Templar Castles on the Road to the Jordan', in M. Barber
(ed.), The Military Orders: Fighting for the Faith and Caring for the Sick
(Aldershot, 1994), pp. 148-66.
26. 1187: Abu Shama, 'Deux jardins', RHCOr, vol. 4, p. 313.1192: Coni. WT, pp. 15152.
27. Matthew Paris, Chronica malora, vol. 4, pp. 288-91 (and see p. 339 for 1244);
Barber, The New Knighthood, pp. 143-44. In 1252 Gaza was formally recognised
as an Egyptian possession: RRH, no. 1199, p. 315.
28. RRH, no. 1318, p. 344. See also Barber, The New Knighthood, pp. 163-64.
29. RH, no. 1450, p. 379.
30. Tabulae ordinis Theutonid, no. 28-29, pp. 24-25.
31. Cartulaire, vol. i, no. 938, p. 594, no. 972, pp. 616-17. 1291: 'Gestes', pp. 808,
812-13.
336
32. Tabulae ordinis Theutonici, no. 44, pp. 35-36. This document refers to other
sections of the walls being guarded by the Templars. 1289: 'Gestes', p. 803.
33. Cartulaire, vol. i, no. 954, p.603 (Jaffa), vol. 2, no. 2160, p.310 (Sidon).
34. Tabulae ordinis Theutonici, no. 40, pp. 32-33. Jerusalem: see above, p. 26.
35. 'Rothelin', p. 636.
36. Beaufort: Rey, Etude, pp. 127-28; Deschamps, La defense du royaume de Jerusalem, pp.198, 208. Arsuf: Ibn al-Furat, 'Selections', p. 54; Benvenisti, The
Crusaders in the Holy Land, pp. 134-35.
37. De constructione castri Saphet, lines 203-4, P- 384.
38. La regle du Temple, cis 126-27. For more details on the income of the Orders,
see Forey, The Military Orders, pp. 98-132. Pilgrims and crusaders: see below,
p. 298.
39. Willbrand of Oldenburg, 'Itinerarium', p. 210; Deschamps, Le Crac des Chevaliers, pp. 203-4.
40. La regle du Temple, d. 228; Cartulaire, vol. 3, no. 3844, p. 453, cl. 12.
41. Ibn al-Furat, 'Selections', p. 122.
42. Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 364-67; Ibn al-Athir, 'Kamel
Altevarykh', RHCOr, vol. i, pp. 721-22; Baha'-al-Din, 'Anecdotes', pp. 111-12.
For a brief description and history of Saone, see Kennedy, Crusader Castles,
pp. 83-96.
43. Marshall, Warfare in the Latin East, pp. 141-42.
44. Joinville, Histoire de Saint Louis, pp. 306-8.
45. Tibbie, Monarchy and Lordships, p. 152, and see pp. 99-152, particularly
pp. 120-52.
46. Dramatically confirmed by the mocking letter Baybars sent to Bohemond VI
after he captured Akkar in 1271: Ibn al-Furat, 'Selections', p. 148.
47. 'Gestes', pp. 678-79. Saladin: Coni. WT, p. 98.
48. 'Eracles', pp. 43233; Joinville, Histoire de Saint Louis, pp. 288, 290.
49. Smail, 'Crusaders' Castles of the Twelfth Century', p. 147.
337
The Byzantine Empire, 1025-1204 (London, 1984), pp. 263-95. See also below,
pp. 191-92.
4. 'Gestes', pp. 820-22; Amadi, Chronique, p. 228; Bustron, Chronique, p. 128; Pope
Nicholas IV, Registres, ed. E. Langlois, 2 vols (Paris, 1886-1905), no. 6778. There
were no less than five Egyptian sultans between 1291 and 1310. See Holt, The
Age of the Crusades, pp. 103-13; Irwin, The Middle East in the Middle Ages,
pp. 79-102.
5. For a brief history of Egypt and its problems after 1291, see Holt, The Age of
the Crusades, pp. 107-206, particularly pp. 192-202. Naval weakness: Thorau,
The Lion of Egypt, pp. 101, 203, 206-8, and see above, pp. 6-7.
6. Edbury, The Kingdom of Cyprus, pp. 109-11,132-33,155-56,197-211; Hill, History
of Cyprus, vol. 2, pp. 379-417, vol. 3, pp. 589-91, and see below, pp. 117-22.
7. See below, pp. 108-10.
8. 'Eracles', pp. 360-62; 'Gestes', pp. 668-76; Amadi, Chronique, pp. 117-24; Edbury, The Kingdom of Cyprus, pp. 48-57; Riley-Smith, The Feudal Nobility,
pp. 159-66.
9. 'Gestes', pp. 676-82; 'Eracles', pp. 367-69, 376; Amadi, Chronique, pp. 124-34;
Bustron, Chronique, pp. 63-73; Edbury, The Kingdom of Cyprus, pp. 57-60.
10. 'Eracles', pp. 376-77; 'Gestes', pp. 684-94; Amadi, Chronique, pp. 136-46; Bustron, Chronique, pp. 74-79; Edbury, The Kingdom of Cyprus, pp. 60-61.
11. 'Gestes', pp. 699-724; 'Eracles', pp. 386-403; Amadi, Chronique, pp. 147-63; Bustron, Chronique, pp. 80-104; Edbury, The Kingdom of Cyprus, pp. 61-65.
12. 'Gestes', pp. 866-69, 871; Leontios Makhairas, Recital Concerning the Sweet
Land of Cyprus Entitled 'Chronicle', ed. R. M. Dawkins, 2 vols (Oxford, 1932),
vol. i, cc. 42-63, pp. 43-59; Amadi, Chronique, pp. 241-54, 259-62, 271-80,
298-391; Bustron, Chronique, pp. 135-41,148-62,176-243; Diomedes Strambaldi,
Chroniques d'Amadi et de Strambaldi, ed. R. de Mas Latrie (Paris, 1891), pp. 1825; Amauiy of Tyre's election charter of 1306 in 'Documents chypriotes du
debut du XlVe sicle', ed. C. Kohler, ROL, u (1905-8), pp. 440-52, at pp. 444-52;
Edbury, The Kingdom of Cyprus, pp. 109-31; Hill, History of Cyprus, vol. 2,
pp. 219-69.
13. Leontios Makhairas, Recital, cc. 90-281, pp. 81-269; Wiliam of Machaut, La
prise d'Alexandre ou chronique du ro Pierre Ier de Lusignan, ed. R. de Mas
Latrie (Geneva, 1877), pp. 19265; Amadi, Chronique, pp. 408-26; Bustron,
Chronique, pp. 259-76; Strambaldi, Chronique, pp. 35-114. These chroniclers
believed Peter was murdered more because of his tyrannical behaviour than
because of his expensive war policies. See J. Richard, 'La revolution de 1369
dans le royaume de Chypre', Btbliothque de l'Ecole des Chartes, 90 (1952),
pp. 108-23; P. Edbury, 'The Murder of King Peter I of Cyprus (1359-1369)',
Journal of Medieval History, 6 (1980), pp. 219-33; P. Edbury, 'The Crusading
Policy of King Peter I of Cyprus, 1359-69', in The Eastern Mediterranean Lands
in the Period of the Crusades, ed. P. M. Holt (Warminster, 1977), pp. 90-105;
Edbury, The Kingdom of Cyprus, pp. 175-79.
14. See above, pp. 7-9.
338
15. Giorgio and Giovanni Stella, 'Annales Genuenses', ed. G. Petti Balbi, RIS n.s.,
vol. 17, part 2 (Bologna, 1975), p. 167.
16. 'Gestes', pp. 806-7; Marino Saudo, 'Liber secretorum', p. 231; Edbury, The
Kingdom of Cyprus, p. 99.
17. See below, pp. 108-9.
18. 'Gestes', p. 712.
19. For the history of Cyprus between 1374 and 1571, see Hill, History of Cyprus,
vol. 2, pp. 404-96, vol. 3, pp. 497-1040.
339
340
341
60. Qaqun and Chastel Rouge: see above, p. 29. Kalamata, Arcadia, Corinth and
Neopatras: see below, pp. 209-11.
61. Megaw, 'A Castle in Cyprus', pp. 48-51, and see below, pp. 125-27. For Belvoir,
see Prawer, The Latin Kingdom of Jerusalem, pp. 300-7; Kennedy, Crusader
Castles, pp. 58-61, and see above, p. 17.
62. See above, pp. 13-19, 24-26.
63. Goliath: Cartulaire, vol. i, no. 82, pp. 76-78, and see above, pp. 28-29.
64. Marshall, Warfare in the Latin East, p. 100.
342.
18. 'Gestes', pp. 676-83; 'Eracles', pp. 367-69; Amadi, Chronique, pp. 124-32; Bustron, Chronique, pp. 63-70.
19. See above, pp. 91, m.
20. 'Gestes', p. 712, and see above, pp. 91, 94.
21. Amadi, Chronique, p. 250. See also Leontios Makhairas, Recital, c. 54, p. 53;
Bustron, Chronique, p. 139.
22. Amadi, Chronique, pp. 335-36, 343-44; Bustron, Chronique, pp. 199-201, 207.
23. See above, pp. 61-63, 79> 84-86.
24. Edbury, The Kingdom of Cyprus, p. 185. Military Orders: see below, chapter
twelve.
25. See above, pp. 59-61.
343
344
Crusades
1.
2.
3.
4.
5.
6.
7.
345
346
Hill, History of Cyprus, vol. 2, pp. 17074; Barber, The New Knighthood,
pp. 172-739. Bustron, Chronique, pp. 246-47. Gastria: Estienne de Lusignan, Description,
fol. 36.
10. Enlart, L'art gothique, vol. 2, p. 656.
11. Bustron, Chronique, pp. 16171; Enlart, L'art gothique, vol. 2, pp. 671-73.
12. 1279: 'Gestes', p. 784; 'Annales de Terre Sainte', p. 457 (text A); Amadi,
Chronique, p. 214; Bustron, Chronique, p. 116.1232: 'Gestes', p. 710.
13. 'A Register of the Cartulary of the Cathedral of Santa Sophia of Nicosia', ed.
J. L. LaMonte, Byzantion, 5 (1930), pp. 439-522, at no. 99, pp. 476-77.
14. For an outline of the Hospitallers' activities on Cyprus, see Coureas, The Latin
Church in Cypurs, pp. 155-72.
15. Bustron, Chronique, pp. 170-71. The fifteenth-century tower at Khirokitia is
referred to by Estienne de Lusignan, Description, fol. 35. See also Hill, History
of Cyprus, vol. 2, p. 23; Enlart, L'art gothique, vol. 2, pp. 671-73.
16. 'Gestes', p. 680.
17. 'Gestes', pp. 686-89; Amadi, Chronique, pp. 139-40.
18. Cartulaire, vol. 2, no. 1354, pp. 121-22; Bustron, Chronique, p. 171; Enlart, L'art
gothique, vol. 2, pp. 683-94.
19. Estienne de Lusignan, Description, fol. 18.
20. Tabulae ordinis Theutonici, no. 34, pp. 2728, no. 71, p. 56.
21. See J. Richard, Chypre sous les Lusignans: documents chypriotes des archives du
Vatican, XlVe et XVe sicles (Paris, 1962), p. 120. For a detailed discussion of
the Teutonic Knights' properties in Cyprus, see W. Hubatsch, 'Der deutsche
Orden und die Reichslehnschaft ber Cypern', Nachrichten der Akademie der
Wissenschaft im Gttingen: Philologisch-Historische Klasse (1955), pp. 245-306;
Coureas, The Latin Church in Cyprus, pp. 173-78.
22. 'Gestes', pp. 718-19; Amadi, Chronique, p. 173; Bustron, Chronique, p. 98. Acre:
'Eracles', pp. 373-74, and see above, p. 59.
23. See above, pp. 125-27.
24. Armenia: Acta Aragonensia, ed. H. Pinke, 3 vols (Berlin and Leipzig, 1908-22),
vol. 3, p. 146; Riley-Smith, The Knights of St John, pp. 199-200.
25. Riley-Smith, The Knights of St John, pp. 220-26.
26. Mas Latrie, Histoire, vol. 2, p. 90; Riley-Smith, The Knights of St John, p. 198.
27. Enlart, L'art gothique, vol. 2, pp. 684, 694-95. For more details on the sugar
industry, see M. -L. von Wartburg, 'The Medieval Cane Sugar Industry in
Cyprus: The Results of Recent Excavations', Antiquaries Journal (1983), pp. 298314.
28. Leontios Makhairas, Recital, c. 618, p. 609. For more details on the agricultural
uses of fortifications, see chapter twenty-eight.
347
348
rime', p. 519; Bar Hebraeus, Chronography, p. 418; Boase, 'The History of the
Kingdom', p. 25.
15. Cahen, La Syrie du Nord, p. 705. When Baybars conquered this region in 1268,
Darbsak was listed as one of the castles he took from the Armenians. See Ibn
al-Furat, 'Selections', pp. 126, 166; Abu'1-Fida, 'Annales', p. 152. For a brief
outline of the Mongol conquests, see Holt, The Age of the Crusades-, pp. 82-89;
Morgan, The Mongols, pp. 147-58.
16. Bar Hebraeus, Chronography, pp. 445-46.
17. Samuel of Ani, 'Chronographie', pp. 463, 466.
18. Samuel of Ani, 'Chronographie', p. 467.
19. For a discussion on the Seljuk army and its possible size, see A. Bombaci, 'The
Army of the Saljuqs of Rum', Istituto orientale di Napoli: Annali, new series,
38 (1978), pp. 343-69.
20. Simon de Saint-Quentin, Histoire des Tartares, ed. J. Richard (Paris, 1965),
p. 64.
21. Constable Sempad, Chronique, ed. Ddyan, p. 88; Vahram of Edessa,
'Chronique rime', p. 511.
22. Vahram of Edessa, 'Chronique rime', p. 513.
23. Oliver of Paderborn, 'Historia Damiatina', pp. 279-80. See also Constable
Sempad, 'Chronique', RHCArm, vol. i, pp. 644-46.
24. 'Gestes', p. 751; Jackson, 'The Crisis', p. 492. For a brief outline of the campaign
see Morgan, The Mongols, pp. 147-58.
25. Ayalon, 'The Mamluk Army', particularly at 15 (1953), p. 222 and 16 (1954),
pp. 70-71, and see above, p. 7.
26. Vahram of Edessa, 'Chronique rime', p. 522; Constable Sempad, Chronique,
ed. Ddyan, pp. 11718; Hethoum the Historian, 'Table chronologique', p. 487;
Samuel of Ani, 'Chronographie', p. 461; Bar Hebraeus, Chronography, p. 446.
27. Bar Hebraeus, Chronography, pp. 452-53.
28. al-Makrizi, Histoire des sultans, vol. 2(b), pp. 60-61.
29. Nerses Balientz, extract in Alishan, Sissouan, p. 469.
30. al-Makrizi, Histoire des sultans, vol. 2(b), pp. 60-61.
31. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 644.
32. al-Makrizi, Histoire des sultans, vol. 2(b), pp. 63-64.
33. See above, pp. 8-9.
34. Bar Hebraeus, Chronography, pp. 452-53.
35. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 666.
36. Constable Sempad, 'Chronique', RHCArm, vol. i, pp. 649-50.
37. Constable Sempad, Chronique, ed. Ddyan, p. 114.
38. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 668. The precision of
'twenty-one knights', as opposed to 'hundreds' or 'thousands' of troops, may
also make this one of Constable Sempad's few reliable figures.
39. M. Bartusis, The Late Byzantine Army: Arms and Society, 1204-1453 (Philadelphia, 1992), pp. 213-17.
40. Constable Sempad, Chronique, ed. Ddyan, pp. 94-95.
349
41. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 661, and see p. 659.
42. Constable Sempad, Chronique, ed. Ddyan, p. 115.
43. Constable Sempad, Chronique, ed. Ddyan, pp. 109-10.
44. For details of the royal domain, see below, pp. 159-60, 168.
45. Hethoum the Historian, 'Table chronologique', p. 484; Constable Sempad,
'Chronique', RHCArm, vol. i, pp. 644-45.
46. Constable Sempad, Chronique, ed. Ddyan, p. 115.
47. Constable Sempad, Chronique, ed. Ddyan, pp. 117-18; Hethoum the Historian,
'Table chronologique', p. 487; Vahram of Edessa, 'Chronique rime', p. 522;
Samuel of Ani, 'Chronographie', p. 461; Bar Hebraeus, Chronography, p. 446;
Boase, 'The History of the Kingdom', pp. 27-28.
48. Bar Hebraeus, Chronography, p. 454.
49. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 661.
50. William of Tyre, Chronicon, vol. i, bk. 13, c. 10, p. 598.
51. Fedden and Thomson, Crusader Castles, pp. 55-56.
35O
351
Fortifications,
352
353
354
355
356
357
358
41. Oliver of Paderborn, 'Historia Damiatina', p. 279. See also 'Eracles', p. 347.
42. See above, pp. 171-72, and Riley-Smith, The Knights of St John, pp. 158-60.
43. Constable Sempad, 'Chronique', RHCArm, vol. i, p. 648.
44. Riley-Smith, 'The Templars and the Teutonic Knights', pp. 92-97. See also
A. W. Lawrence, 'The Castle of Baghras', in Boase, The Cuidan Kingdom of
Armenia, pp. 34-83, at pp. 41-43.
45. Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 378-79; Ibn al-Athir, 'Kamel
Altevarykh', RHCOr, vol. i, pp. 731-32; 'Eracles', p. 136; Constable Sempad,
'Chronique', RHCArm, vol. i, p. 636; Vahram of Edessa, 'Chronique rime',
p.512; Bar Hebraeus, Chronography, pp.336-37; Lawrence, 'Baghras', pp.44,
45; Riley-Smith, 'The Templars and the Teutonic Knights', pp. 97-98.
46. For full descriptions of the site, see Lawrence, 'Baghras', pp. 49-83; R. W. Edwards, 'Bagras and Armenian Ciucia: A Reassessment', Revue des tudes
armniennes, 17 (1983), pp. 415-55, at pp. 419-32.
47. Lawrence, 'Baghras', pp. 55-56, 62-63.
48. Edwards, 'Bagras and Armenian Cilicia', pp. 420-23, 426, 429.
49. Edwards, 'Bagras and Armenian Cilicia', pp. 431-32.
50. Ibn al-Athir, 'Kamel Altevarykh', vol. i, p. 732.
51. Willbrand of Oldenburg, 'Itinerarium', p. 216.
52. Innocent III, 'Opera Omnia', ed. J. P. Migne, PL, vol. 214, lib. 2, no. 259, cols
819-20.
53. For full accounts of the arguments over Baghras and the Antiochene succession
dispute between the 11905 and 1216, see Cahen, La Syrie du Nord, pp. 596-623;
Riley-Smith, 'The Templars and the Teutonic Knights', pp. 97-107.
54. 'Gestes', p. 663; 'Annales de Terre Sainte', p. 435 (text B); Innocent III,
PL, vol. 215, lib. 8, no. 119, col. 689. See also Marino Saudo, 'Liber secretorum', p. 203; Hethoum the Historian, 'Table chronologique', p. 480; 'Eracles',
p. 257.
55. Innocent III, PL, vol. 215, lib. 7, no. 189, col. 504, lib. 8, no. 119, cols 689-90.
56. Ibn al-Athir, 'Kamel Atevarykh', RHCOr, vol. 2, pp. 317-18.
57. 'Eracles', pp. 317-18; Innocent III, PL, vol. 216, lib. 14, no. 64, cols 430-32.
58. Innocent III, PL, vol. 216, lib. 16, no. 7, cols 792-93; Riley-Smith, 'The Templars
and the Teutonic Knights', p. 107.
59. Willbrand of Oldenburg, 'Itinerarium', p. 216.
60. Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. i, p. 732.
61. Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 168-70; Constable Sempad,
'Chronique', RHCArm, vol. i, p. 648; Riley-Smith, 'The Templars and the
Teutonic Knights', p. 109.
62. 1188: Abu Shama, 'Deux jardins', RHCOr, vol. 4, pp. 376-77; Ibn al-Athir,
'Kamel Altevarykh', RHCOr, vol. i, pp. 730-31; Cahen, La Syrie du Nord,
pp. 144-4563. 1205: Ibn al-Athir, 'Kamel Altevarykh', RHCOr, vol. 2, pp. 317-18.1237: Matthew
Paris, Chronica maiora, vol. 3, pp. 404-6; Abu'1-Fida, 'Annales', pp. 112-13.1260:
Cahen, La Syrie du Nord, p. 705.
359
64. Cahen, La Syri du Nord, pp. 142-43; Sinclair, Eastern Turkey, vol. 4,
pp. 314-16.
65. I have followed Cahen's argument (La Syrie du Nord, pp. 143-45) that the
ruined fortress of Chilvan Kale should be identified as Roche Roussel rather
than Roche Guillaume. The latter castle may be another ruined stronghold
situated further east. See Boase, 'Gazetteer', pp. 159, 178.
66. Boase, 'Gazetteer', p. 177; Cahen, La Syrie du Nord, p. 141.
67. 'Eracles', pp. 405-6; Cahen, La Syrie du Nord, pp. 651-52.
68. Ihn al-Furat, 'Selections', p. 99. See also al-Makrizi, Histoire des sultans, vol.
i(b), p. 34.
69. Antioch: 'Eracles', p. 318; 'Annales de Terre Sainte', pp. 436-37; Hethoum the
Historian, 'Table chronologique', pp. 483-84; 'Gestes', p. 664; Constable Sempad, Chronique, ed. Ddyan, pp. 89-90.
70. Innocent III, PL, vol. 214, cols 1005-6.
71. Riley-Smith, 'The Templars and the Teutonic Knights', p. 108, from Innocent
III, PL, vol. 214, col. 1005, vol. 215, col. 504.
72. Riley-Smith, 'The Templars and the Teutonic Knights', p. no.
73. For example, in 1237 local Templars made a failed attempt to recapture Darbsak
from the Muslims. See Matthew Paris, Chronica maiora, vol. 3, pp. 404-6;
Abu'1-Fida, 'Annales', pp. 112-13.
74. Document trans, in Alishan, Sissouan, p. 239.
75. 'Gestes', p. 839.
76. Ihn al-Furat, 'Selections', p. 99; al-Makrizi, Histoire des sultans, vol. i(b), p. 34.
77. al-Makrizi, Histoire des sultans, vol. 2(b), p. 61; Riley-Smith, 'The Templars
and the Teutonic Knights', p. 115.
78. 'Gestes', p. 839.
79. Ibn al-Furat, 'Selections', pp. 121-26; al-Makrizi, Histoire des sultans, i(b),
pp. 52-54; 'Gestes', pp. 771-72; 'Annales de Terre Sainte', pp. 453-54; Marino
Saudo, 'Liber secretorum', p. 223; 'Eracles', p. 456; Bar Hebraeus, Chronography, p. 448.
80. Luttrell, 'The Hospitallers' Interventions', p. 122 n. 17.
81. J. Upton-Ward, 'The Surrender of Gaston and the Rule of the Templars', in
Barber, The Military Orders, pp. 179-88.
82. 1233: Le trsor des charles d'Armenie, no. 17, p. 140.
83. Ibn al-Furat, 'Selections', p. 99.
84. Hethoum the Historian, 'Table chronologique', p. 487. For more details on
the Hospitallers' involvement in this and other thirteenth-century Armenian
campaigns, see Luttrell, 'The Hospitallers' Interventions', pp. 119-22; RileySmith, The Knights of St John, p. 161.
85. Acta Aragonensia, vol. 3, p. 146.
86. Leontios Makhairas, Recital cc. 11628, pp. 103-13, cc. 132-33, pp. 117-19, c. 143,
p. 125, c. 150, pp. 131-33, and see above, pp. 127-28. For more details on the
Hospitallers' involvement in these and other fourteenth-century campaigns,
see Luttrell, 'The Hospitallers' Interventions', pp. 122-33.
360
87. For more details on these events, see Riley-Smith, The Knights of St John,
pp. 198-226; Luttrell, 'The Hospitallers' Interventions', pp. 124-28, and see
above, pp. 91, 112-13, 130.
88. Luttrell, 'The Hospitallers' Interventions', p. 128.
89. RHCArm, vol. i, p. xxxiv; Luttrell, 'The Hospitallers' Interventions',
pp. 128-29.
Notes to Chapter 18, Prankish Greece: Warfare
1. For an outline of the Fourth Crusade and its causes, see D. E. Queller and
T. F. Madden, The Fourth Crusade: The Conquest of Constantinople, second
edition (Philadelphia, 1997); C. M. Brand, Byzantium Confronts the West (Cambridge, Massachusetts, 1968), especially pp. 232-69, and see the introduction
to this book.
2. For an outline of these events, see P. Lock, The Franks in the Aegean, 12041500
(London, 1995), especially pp. 35-107, 135-60.
3. For an outline of these events, see Lock, The Franks, pp. 60-134; R- L. Wolff,
'The Latin Empire of Constantinople, 1204-61', in HC, vol. 2, pp. 187-233;
P. Topping, 'The Morea, 1311-1364', and 'The Morea, 1364-1460', in HC, vol.
3, pp. 104-40,141-66; K. M. Setton, 'The Catalans in Greece, 1311-1380', in HC,
vol. 3, pp. 167-224.
4. For an outline of these events, see Lock, The Franks, pp. 84-86, 92-95,127-34,
155-60; Topping, 'The Morea, 1311-1364', pp. 104-40 and 'The Morea, 13641460', pp. 141-66.
5. B. Hendrickx, 'A propos du nombre des troupes de la Quatrime Croisade et
de Fempereur Baudoin I', Byzantina, 3 (1971), pp. 29-41; Bartusis, The Late
Byzantine Army, p. 262.
6. Geoffrey of Villehardouin, La conqute de Constantinople, ed. N. de Wailly
(Paris, 1882), pp. 180, 184.
7. 1206: Villehardouin, La conqute, p. 256. 1211: Letter from Henry to the West,
dated January 1212, in RHGF, vol. 18, p. 533.
8. Bartusis, The Late Byzantine Army, p, 262.
9. Henry of Valenciennes, Histoire de l'empereur Henri de Constantinople, ed.
J. Longnon (Paris, 1948), pp. 46-47; letter from the emperor Henry to Pope
Innocent III, September 1208, in RHGF, vol. 19, p. 514.
10. Philip Mousket, 'Chronique rime', ed. baron de Reiffenberg, in Collection de
chroniques belges indites, 2 vols (Brussels, 1838), vol. 2, pp. 614-16; George
Akropolites, 'Annales', ed. I. P. Migne, in PG, vol. 140, cols 1058-59.
11. Villehardouin, La conqute, p. 202.
12. Villehardouin, La conqute, pp. 198-230, especially at p. 208. Quote: p. 250.
13. Villehardouin, La conqute, p. 200.
14. Niketas Choniates, 'Historia', ed. J. P. Migne, in PG, vol. 139, cols 983-87,
1023-27. For the conquest of northern Greece, see also Villehardouin, La
conqute, pp. 160-78; Robert of Clari, La conqute de Constantinople, ed.
361
P. Lauer (Paris, 1924), pp. 96-105; George Akropolites, 'rmales', cois 9991OO2.
302
part i (1938), pp. 280 (Crete), 283, 367 (Corfu); Marino Saudo, 'Vite de' duchi
di Venezia', ed. L. A. Muratori, in RIS, vol. 22 (1733), cols 533, 536 (Crete);
Andrea Navagerio, 'Storia Venezia', ed. L. A. Muratori, in RIS, vol. 23 (1733),
cols 984-89 (Crete and Corfu); Martin da Canal, 'La chronique des Veneciens',
ed. G. Galvani, in Archivio storico italiano, vol. 8 (1845), p. 351 (Crete). See also
J. K. Fotheringham, Marco Saudo, Conqueror of the Archipelago (Oxford, 1915),
pp. 81-87. After the Fourth Crusade, Crete was originally claimed by Boniface
of Montferrat, but then sold by him to the Venetians. See Urkunden zur lteren
Handels- und Staatsgeschichte der Republik Venedig mit besonderer Beziehung
auf Byzanz und die Levante, ed. G. Tafel and G. Thomas, 3 vols (Vienna,
1856-57), vol. i, pp. 512-15.
35. Livre de la conqueste, pp. 81-87; Libro de los fechos et conquistas, pp. 49-52;
Marino Saudo, 'Istoria', pp. 103-6; Andrea Dandolo, 'Chronicum Venetum',
p. 306. The final peace treaty ending this war was not signed until 1262; see
Tafel and Thomas, Urkunden, vol. 3, pp. 46-55. See also Longon, L'empire
latin, pp. 220-23; Lock, The Franks, pp. 88-92.
36. See below, pp. 253-57.
37. Luttrell, 'The Latins and Life on Smaller Aegean Islands', p. 146. Marino Saudo
('Istoria', p. 143) noted that Latins on other Aegean islands were also heavily
outnumbered.
38. Villehardouin, La conquete, pp. 92, 142.
39. 1235: Philip Mousket, 'Chronique rime', pp. 61416; Martin da Canal, 'La
chronique des Veneciens', pp. 363-67; Andrea Dandolo, 'Chronicon Venetum',
p. 295; Andrea Navagerio, 'Storia Venezia', col. 992; George Akropolites, 'Annales', cols 105859. 1236: Aubrey of Trois Fontaines, 'Chronica', pp. 938-39;
Philip Mousket, 'Chronique rime', p. 620.
40. Muntaner, Crnica, vol. 2, pp. 85-87. For further examples of the great skill
and ferocity of the Catalans and Aragonese in naval warfare, see J. H. Pryor,
'The Naval Battles of Roger of Lauria', in J. H. Pryor, Commerce, Shipping and
Naval Warfare in the Medieval Mediterranean (London, 1987), c. 6, pp. 179-216.
41. Bartusis, The Late Byzantine Army, p. 68.
42. Marino Saudo, 'Istoria', p. 132.
43. Muntaner, Crnica, vol. 2, p. 74, and see Pryor, Geography, Technology and
War, pp. 165-73.
363
6. Bon, La Mare franque, p. 639. The following scholars all agree with
Bon: K.Andrews, Castles of the Marea (Princeton, 1953), pp. 161, 173;
S. Runciman, Mistra (London, 1980), pp. 29-30, 95; Mller-Wiener, Castles
of the Crusaders, pp. 84-85. For a description and history of Mistra see
Andrews, Castles, pp. 161-82, and, in far greater detail, Runciman, Mistra,
pp. 9-146. See also To Chronikon tou Mareos, p. 200; Libro de los fechos et
conquistas, p. 49.
7. Foss and Winfield, Byzantine Fortifications, pp. 30, 34.
8. For more details, see Andrews, Castles, pp. 135,137-38,143-45 (Corinth); pp. 90,
91-92, 94-105 (Nauplia); pp. 19296, 198-99, 209-10 (Monemvasia); pp. 116,
117-19, 129 (Patras).
9. A. Bon, 'Fortresses medievales de la Grece centrale', Bulletin de correspondance
hellnique, 61 (1937), pp. 136-208, at p. 163.
10. Andrews, Castles, pp. 74-78, 81.
11. Villehardouin, La conqute, p. 190.
12. Foss and Winfield, Byzantine Fortifications, pp. 154-55. 1225: Nikephoros
Gregoras, 'Nicephori Gregorae Byzantina historia', ed. L. Schopen, in CSHB,
3 vols (Bonn, 1829-55), vol. i, p. 25; Aubrey of Trois Fontaines, 'Chronica',
p. 911; George Akropolites, 'Annales', cols 1038-39.
13. Foss and Winfield, Byzantine Fortifications, pp. 41-70; A. van Millingen,
Byzantine Constantinople: The Walls of the City and Adjoining Sites (London,
1899), pp. 51-58, 122-27, 164-7414. Villehardouin, La conqute, p. 134, and see ibid., pp. 90-102; Robert of Clari,
La conqute, pp. 44-51; Niketas Choniates, 'Historia', cois 923-27.
15. Niketas Choniates, 'Historia', col. 926; Villehardouin, La conqute, pp. 96-100;
Robert of Clari, La conqute, pp. 45, 47, 51.
16. Quote: Villehardouin, La conqute, p. 134. See also ibid., pp. 138-42; Robert of
Clari, La conqute, pp. 62, 69-77; Niketas Choniates, 'Historia', cois 947, 951-54.
For a description of Constantinople's sea walls, see van Millingen, Byzantine
Constantinople, pp. 178-267; Foss and Winfield, Byzantine Fortifications,
PP- 70-7317. Foss and Winfield, Byzantine Fortifications, p. 42.
18. George Pachymeres, 'Georgii Pachymeris de Michaela et Andronico Palaeologis', ed. I. Bekker, in CSHB, 2 vols (Bonn, 1835), vol. I, pp. 186, 187. The
Blachernae palace also had to be restored at this time, because it had been
left in such a bad state by the last Latin emperor, Baldwin II. See ibid., vol.
i, pp. 144, 161.
19. Villehardouin, La conqute, p. 166.
20. Villehardouin, La conqute, pp. 200, 204-6 (Arcadiopolis, Stenimaka, Philipoppolis); pp. 240-44 (Rousion); pp. 246-48 (Rodosto); pp. 206-16, 234-36
(Adrianople). For Adrianople, see also Niketas Choniates, 'Historia', col. 1002;
Robert of Clari, La conqute, pp. 105-6.
21. W. Miller, Essays on the Latin Orient (Cambridge, 1921), p. 279; Foss and
Winfield, Byzantine Fortifications, pp. 5, 8; M. Vickers, 'The Byzantine Sea Walls
364
365
pp. 116-19; Traquair, 'Mediaeval Fortresses', 279-80. Patras's close links with
Venice will be discussed again below, p. 230.
33. Livre de la conqueste, pp. 3738. See also Libro de los fechos et conquistas, p. 28.
34. Andrews, Castles, p. 34, and see pp. 30-35; Bon, La Mare franque, pp. 606-68;
Traquair, 'Mediaeval Fortresses', pp. 271-72.
35. Bon, La Moree franque, pp. 669-70; Andrews, Castles, pp. 85-89. As was noted
above (p. 100), Arcadia can be compared with the Greek mountain fortresses
occupied by the Latins on Cyprus.
36. Livre de la conqueste, p. 39. See also Libro de los fechos et conquistas, p. 28.
37. Andrews, Castles, p. 89; T. S. R. Boase, 'The Arts in Prankish Greece and
Rhodes', in HC, vol. 4, pp. 208-28, at p. 219.
38. Coron: Villehardouin, La conqute, p. 196; Livre de la conqueste, p. 37. Patras:
ibid., p. 30. Athens and Thebes: Niketas Choniates, 'Historia', col. 995.
39. The exact date of the fall of Corinth, Argos and Nauplia remains unclear.
They were initially besieged by Boniface of Montferrat in late 1204: Villehardouin, La conqute, p. 196; Niketas Choniates, 'Historia', cois 991, 998.
According to the Chronicle of Morea, Corinth and Nauplia were not captured
until the 12405: Livre de la conqueste, pp. 68-71; Libro de los fechos et conquistas,
p. 48; To Chromkon tou Mareos, pp. 188-97. However, papal documents from
the reign of Innocent III alluding to the fall of Corinth (PL, vol. 216, no. 6,
cols 201-2) and Argos (PL, vol. 216, no. 77, col. 598) suggest that the much
earlier dates given above are correct. For Argos, see also To Chronikon tou
Mareos, p. 104; Libro de los fechos et conquistas, p. 24. Libro de los fechos et
conquistas correctly attributes the capture of Corinth to Geoffrey I of Villehardouin (p. 43), but subsequently appears to make the same mistake as the
Livre de la conqueste by implying that Corinth fell much later in the thirteenth
century (p. 48). See also J. Longnon, 'Problmes de l'histoire de la principaut
de Moree', Journal des savants (1946), pp. 77-93,147-61, at pp. 156-57; Bon, La
Moree franque, p. 68.
40. Neopatras and surrounding territories had been granted to Boniface of Montferrat's followers in the partition treaty of 1204; Tafel and Thomas, Urkunden,
vol. i, pp. 486-88. It presumably fell to the crusaders during Boniface's campaign in the area; Villehardouin, La conqute, p. 178; Niketas Choniates,
'Historia', cois 986, 990-91; Nicol, The Despotate, pp. 35-36, 57-58. In 1319 the
Catalans captured Neopatras, and it became an important duchy under their
rule until the Turkish invasion of 1390. See Marino Saudo, 'Epistulae', ed.
J. Bongars, in Gesta Dei per Francos, vol. 2, ep. 3 (1325), p. 291; A. Rubio y
Lluch, 'Els Castells Catalans de la Grcia continental', Annuari de l'institut
d'estudis Catalans, z (1908), pp. 364-425, at pp. 399-413.
41. For general descriptions of these sites, see, Corinth: Andrews, Castles, pp. 13845. Argos: Bon, La Moree franque, pp. 674-76; Andrews, Castles, pp. 107-15.
Neopatras: Rubio y Lluch, 'Els Castells', pp. 399-400. Thebes: Bon, 'Forteresses
medievales', pp. 189-91.
42. For general descriptions of these sites, see, Patras: Bon, La Moree franque,
306
pp. 670-73; Andrews, Castles, pp. 119-29. Nauplia: ibid., pp. 92-105. Coron:
ibid., pp. 15-23.
43. Good examples of such masonry can still be seen at Argos and Patras. See
Andrews, Castles, p. 113, and figs 125-27 (Argos), p. 126 and fig. 144 (Patras);
Bon, La Mare franque, p. 673 (Patras).
44. Kalamata: Andrews, Castles, pp. 34-35. Arcadia: ibid., p. 89; Boase, 'The Arts
in Prankish Greece', p. 219. Neopatras: Rubio y Lluch, 'Els Castells', p. 400.
Corinth: Bon, La Mare franque, p. 674; Andrews, Castles, p. 140.
45. Libro de los fechos et conquistas, p. 49.
46. Miller, The Latins in the Levant, pp. 401-2; P. Lock, 'The Prankish Tower on
the Acropolis, Athens: The Photographs of William J. Stillman', ABSA, 82
(1987), pp. 131-33; P. Lock, 'The Prankish Towers of Central Greece', ABSA, 81
(1986), pp. 101-23, at pp. 107, 111-12.
47. Niketas Choniates, 'Historia', col. 991; Miller, The Latins in the Levant, pp. 31-32.
On the great strength of the acropolis in general, see K. M. Setton, Catalan
Domination of Athens, 13111388 (Cambridge, Massachusetts, 1948), pp. 188-89.
48. Corinth: Andrews, Castles, pp. 140-41.
49. Marino Saudo, 'Istoria', p. 106.
50. Livre de la conqueste, p. 71.
51. Nauplia: Bon, La Mare franque, p. 677; Andrews, Castles, pp. 93-94. Boudonitza: Bon, 'Fortresses medievales', p. 161. Salona: ibid., p. 177.
52. This was destroyed by the Catalans to prevent its recapture by the Franks.
However, the Catalans probably retained some urban defences at Thebes. See
Livre de la conqueste, pp. 220-21; To Chronikon tou Mareos, p. 524; Bon, 'Fortresses medievales', pp. 18791.
53. Henry of Valenciennes, Histoire de l'empereur, pp. 111-13.
54. Cyzicus: Villehardouin, La conqute, p. 272.
55. Bon, 'Fortresses medievales', pp. 148-63; Boase, 'The Arts in Prankish Greece',
pp. 214-15.
56. For general descriptions of these sites, see, Modon: Andrews, Castles, pp. 61-83.
Salona (Amphissa): Bon, 'Fortresses medievales', pp. 164-86; Rubio y Lluch,
'Els Castells', pp. 41325; Boase, 'The Arts in Prankish Greece', p. 215. Akova:
Bon, La Mare franque, pp. 634-35. Zeitoun (Lamia): Rubio y Lluch, 'Els
Castells', pp. 393-98; Boase, 'The Arts in Prankish Greece', p. 216.
57. 1205: Villehardouin, La conqute, p. 194. 1209: Martin da Canal, 'La chronique
des Veneciens', p. 349. Modon had originally been taken from the Greeks by
Prankish crusaders, but was recognised as Venetian in 1209 in return for
Venice's military support; see Andrea Dandolo, 'Chronicum Venetum',
pp. 283-84; Tafel and Thomas, Urkunden, vol. 2, pp. 97-100.
58. Bon, La Mare franque, pp. 634-35.
59. Bon, 'Fortresses medievales', pp. 152-61, particularly at p. 161.
60. Bon, 'Fortresses medievales', pp. 179-83; Rubio y Lluch, 'Els Castells', pp. 415i?.
61. To Chronikon tou Mareos, pp. 176-82; Libro de los fechos et conquistas, p. 49,
367
which incorrectly dates these events to c. 1256 and the reign of William II;
Honorius III, Regesta, vol. 2, nos 3162, 3163, pp. 516-17, no. 4480, p. 159. (Full
text of no. 4480 also in Innocent III, PL, vol. 216, cols 968-72.) See also Bon,
La Mare franque, pp. 94-97; R. L. Wolff, 'Politics in the Latin Patriarchate of
Constantinople', Dumbarton Oaks Papers, 8 (1954), pp. 225303, at p. 274.
62. Bon, La Mare franque, pp. 60829 (fireplaces, pp. 621-22); Andrews, Castles,
pp. 149-58; Traquair, 'Mediaeval Fortresses', pp. 272-79. Traquair should be
used with caution, as he incorrectly states that Chlemoutsi was built in the
14308. He is also used by Boase, 'The Arts in Prankish Greece', pp. 217-18.
63. Bon, La Mare franque, pp. 6027. See also Traquair, 'Mediaeval Fortresses',
pp. 272, 275-79, although this description is less reliable, and the attached
history of the site appears to confuse it with Chlemoutsi.
64. Bon, La Mare franque, pp. 324, 607.
65. Bon, 'Fortresses medievales', pp. 194-206; Rubio y Lluch, 'Els Castells', pp. 38387.
66. To Chronikon tou Mareos, pp. 130, 131, 210; Libro de los fechos et conquistas,
p. 29; Livre de la conqueste, p. 79.
67. Bon, La Mare franque, pp. 629-30, 631.
68. Bon, La Mare franque, pp. 63033.
69. Androusa: Libro de los fechos et conquistas, p. 49; Bon, La Mare franque,
pp. 637-39. Old Navarino (Port de Jone): Livre de la conqueste, p. 221; Libro
de los fechos et conquistas, p. 103; Bon, La Mare franque, pp. 668-69. For a
detailed description, see also Andrews, Castles, pp. 42-48. Mistra: Libro de los
fechos et conquistas, p. 49; To Chronikon tou Moveos, p. 200; Bon, La Mare
franque, pp. 639-42, and see also Andrews, Castles, pp. 168-74; Runciman,
Mistra, pp. 29-30. Beaufort (Leutron): Libro de los fechos et conquistas, p. 49;
To Chronikon tou Mareos, p. 202; Bon, La Mare franque, p. 504. Old Mania:
Livre de la conqueste, pp. 74-75; Libro de los fechos et conquistas, p. 49; Bon, La
Mare franque, pp. 503-4; R. Traquair, 'Laconia: I, Mediaeval Fortresses', ABSA,
12 (1905-6), pp. 259-76, at pp. 275-76. Gerald: To Chronikon tou Mareos, pp. 130,
210; Bon, pp. 642-45.
70. P. Burridge, 'The Castle of Vardounia and Defence in the Southern Taygetos',
in Lock and Sanders, The Archaeology of Medieval Greece, pp. 19-28, at p. 22.
71. Bon, La Mare franque, p. 639; Foss and Winfield, Byzantine Fortifications,
p. 30, and see above, pp. 204-5.
72. Bon, La Mare franque, p. 632.
73. To Chronikon tou Mareos, pp. 130, 131; Livre de la conqueste, p. 44; Bon, La
Mare franque, pp. 633-34, 636.
74. This book relies on the Greek ( To Chronikon tou Mareos), the Aragonese (Libro
de los fechos et conquistas) and the French (Livre de la conqueste) versions. The
Italian version is simply a shortened sixteenth-century translation of the Greek
version. For more details on these texts, see D. Jacoby, 'Quelques considerations
sur les versions de la chronique de More', Journal des Savants (1968), pp. 13389; Lock, The Franks, pp. 21-24.
368
3^9
37O
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
see Bon, 'Fortresses medievales', pp. 148-51. For variations in masonry, see
ibid., pp. 160-61.
Bon, 'Fortresses medievales', pp. 164-86; Rubio y Lluch, 'Els Castells', pp. 41325.
Bon, La Mare franque, pp. 637-39.
Othon de la Roche's ownership of Livadia is confirmed by documents dating
from 1214, issued during his dispute with the papacy. See L. A. Muratori,
ed., Antiquitates italicae medii aevi, 6 vols (Rome, 1738-42), vol. 5, cols 83336; Miller, The Latins in the Levant, pp. 69-70. Othon's arguments with
the church were similar to, and linked with, those of Geoffrey of Villehardouin; see above, pp. 213-14. For the subsequent history of Livadia, see
Bon, 'Fortresses medievales', pp. 191-206; Rubio y Lluch, 'Els Castells',
PP- 374-87.
Foss and Winfield, Byzantine Fortifications, p. 42. Urgent repairs were carried
out after Michael VIII Palaiologos regained the city in 1261; see George
Pachymeres, 'De Michaele et Andronico', vol. i, pp. 186, 187.
Thebes: see above, p. 212. Chlemoutsi: see above, pp. 213-14.
Libro de los fechos et conquistas, p. 28; Livre de la conqueste, pp. 43-44; Gerland,
Neue Quellen, p. 14; Bon, La Mare franque, p. 106 n. 3.
See above, Chapter 2.
Bon, La Mare franque, pp. 629-33.
Cyprus: see above, pp. 98-100. Cilician Armenia: see above, pp. 145-48.
Bon, La More franque, pp. 602-7.
Holy Land: see above, pp. 7-9, 24-26, 35. Greece: see above, pp. 194-202.
See below, pp. 255-56.
For example, in the 13205 the Greeks obtained Karytaina and other neighbouring castles by bribing their Latin garrisons; see Livre de la conqueste,
pp. 4045; Libro de los fechos et conquistas, pp. 140-43.
371
8. Buchn, Nouvelles recherches, vol. 2, no. 19, pp. 326-27, and see no. 24, pp. 33031, no. 25, p. 331, no. 27, pp. 332-33. See also Livre de la conqueste, pp. 178,
192-93.
9. Livre de la conqueste, p. 221; Libro de los fechos et conquistas, p. 103. For a
description, see Bon, La Mare franque, pp. 668-69; Andrews, Castles, pp. 404910. Livre de la conqueste, pp. 220-21; To Chronikon tou Mareos, p. 524; Lock, The
Franks, pp. 78-79.
11. Naxos: Sauger, 'Histoire nouvelle', in Fotheringham, Marco Saudo, p. 115, and
see pp. 70-71; Boase, 'The Arts in Prankish Greece', p. 222. Monemvasia: Andrews, Castles, pp. 192-210.
12. See above, pp. 196, 197.
13. VUlehardouin, La conqute, pp. 284-88; George Akropolites, 'Annales', col.
1042.
14. Modon and Coron did not fall until 1500; see Miller, The Latins in the Levant,
pp. 495-98; Andrews, Castles, pp. 15, 60.
15. George Akropolites, 'Annales', cols 1207-11; Nikephoros Gregoras, 'Byzantina
historia', vol. i, pp. 85-86; Andrea Dandolo, 'Chronicum Venetum', pp. 311,
369; George Pachymeres, 'De Michaele et Andronico', vol. i, pp. 140-48; Andrea
Navagerio, 'Storia Venezia', col. 199; Marino Saudo, 'Istoria', p. 115; Libro de
los fechos et conquistas, pp. 21, 51; Livre de la conqueste, pp. 26-27. Acre etc.:
see above, pp. 45-46, 50-51.
16. VUlehardouin, La conqute, pp. 278-82.
17. 1209: Andrea Dandolo, 'Chronicum Venetum', pp. 283-84; Tafel and Thomas,
Urkunden, vol. 2, pp. 97-100.
18. Livre de la conqueste, p. 68. See also Libro de los fechos et conquistas, p. 48. It
should be noted that the chronicle incorrectly dates the fan of Nauplia to this
period; it actually fell in 1212.
19. lacobo Zeno, 'Vita Caroli Zeni', ed. G. Zonta, RIS n.s., vol. 19, part 6 (Bologna,
1931-41), pp. 10-11.
20. In 1373 Venice sent two galleys to Patras. See Thiriet, Rgestes, vol. i, nos 520,
522, p. 130. For more details on the background to these events, see Bon, La
More franque, pp. 247-51; Miller, The Latins in the Levant, pp. 287-90; Gerland,
Neue Quellen, pp. 39-42.
21. Libro de los fechos et conquistas, pp. 121-24,127-37; Muntaner, Crnica, vol. 2,
pp. 159-60, 164-65. For more details on the background to this dispute, see
Miller, The Latins in the Levant, pp. 252-57.
22. Livre de la conqueste, pp. 3536.
23. Pryor, Geography, Technology and War, pp. 153-61, 165-75, and see above,
pp. 199-200.
24. See, for example, Marino Saudo, 'Istoria', pp. 132, 146.
25. See above, p. 222.
26. Niccolo da Martorii, 'Relation du plerinage', pp. 646-48; Luttrell, 'The Latins
and Life on the Smaller Aegean Islands', p. 151.
372
3/3
374
been one of the key leaders involved in the siege of Antioch (1097-98), which
was also conducted from small castles built by the crusaders to blockade the
city's gates. See Gesta Francorum, ed. and trans. R. Hill (Oxford, 1962), pp. 30,
39-40, 42, 4323. Livre de la conqueste, p. 33; Libro de los fechos et conquistas, p. 24.
24. Villehardouin, La conqute, pp. 92-94.
25. Livre de la conqueste, p. 30.
26. Livre de la conqueste, p. 69, and see Libro de los fechos et conquistas, p. 26.
27. Villehardouin, La conqute, p. 92.
28. These tactics can be compared with contemporary events in the Holy Land,
where we have seen that a number of sieges, most famously that of Acre
during the Third Crusade, were conducted from local makeshift fortifications.
See above, p. 76. Fall of Corinth: see above, p. 210 n. 39.
29. Villehardouin, La conqute, pp. 212-22; letter from the emperor Henry to
Innocent III, June 1205, RHGF, vol. 18, p. 527.
30. Henry of Valenciennes, Histoire de l'empereur, pp. 50-52.
31. 1271: Livre de la conqueste, pp. 178-80, and see also pp. 180-82; Libro de los
fechos, p. 84. 1302: Livre de la conqueste, p. 365.
32. Henry of Valenciennes, Histoire de l'empereur, pp. 3738.
33. Serres: Philip Mousket, 'Chronique rime', voi. 2, pp. 408-9. Poimanenon:
Nikephoros Gregoras, 'Byzantina historia', vol. i, p. 25; Aubrey of Trois Fontaines, 'Chronica', p. 911; George Akropolites, 'Annales', cols 1038-39.
375
vol. 9, no. 49924, p. 345. See also Buchn, Nouvelles recherches, vol. i, pp. 30-33;
Setton, Catalan Domination, pp. 38-41.
8. Holy Land: see above, pp. 3, 40.
9. Livre de la conqueste, pp. 32-34; Libro de los fechos et conquistas, pp. 23-26;
Niketas Choniates, 'Historia', cols 991, 998. For the history of Corinth in the
fifteenth century, see Andrews, Castles, p. 137.
10. 1205: Niketas Choniates, 'Historia', col. 991. 12205: Honorius III, Regesta, vol.
2, no. 4758, p. 207, no. 5464, p. 333.
11. Assises, article 19; Livre de la conqueste, pp. 120-23 (quote: p. 121); Libro de los
fechos et conquistas, pp. 65-66, 67-68.
12. Bartusis, The Late Byzantine Army, pp. 115-17.
13. Nikephoros Gregoras, 'Byzantina historia', vol. i, p. 246.
14. Livre de la conqueste, pp. 301-2, and see Muntaner, Crnica, vol. i, p. 186.
15. To Chronikon tou Moreos, p. 308.
16. Livre de la conqueste, pp. 328-29.
17. Buchn, Nouvelles recherches, vol. 2, no. 15, pp. 109-14, and see no. 29, pp. 15860.
18. Quotes: Bintiiff, 'The Prankish Countryside in Central Greece', p. 6.
19. Bintiiff, 'The Prankish Countryside in Central Greece', pp. 6-7.
20. Lock, 'The Towers of Euboea', no. 28, pp. 115-16.
21. Ellenblum, 'Three Generations of Prankish Castle-Building', pp. 520-23.
22. See, for example, Lock, The Franks, p. 293.
23. Holy Land: Ellenblum, Prankish Rural Settlement, especially pp. 3-6, 9, 12-19,
and see above, pp. 55-57.
24. Lock, 'The Medieval Towers of Greece', pp. 136-37; account of the Serbian
chronicler Daniel, in A. Soloviev, 'Histoire du monastre russe au Mont-Athos',
Byzantion, 8 (1933), pp. 213-38, at pp. 223-24.
25. Bartusis, The Late Byzantine Army, pp. 31617.
26. Innocent III, PL, vol. 216, no. 168, cols 956-57.
27. Fourteenth-century survey of royal rights and properties in the Morea, reproduced in Bon, La Mare franque, pp. 691-92, at p. 692. See also ibid., pp. 279,
646-48.
28. Bon, 'Forteresses medievales', pp. 162, 184.
29. Livre de la conqueste, p. 71.
30. Sauger, 'Histoire nouvelle', extract in Fotheringham, Marco Saudo, p. 115;
Fotheringham, Marco Saudo, p. 71. Generally, the inhabitants of Chios were
segregated in the same way, with the Genoese living in the actual citadel and
the Greeks inhabiting the surrounding bourg. See Balard, La Romanie gnoise,
vol. i, p. 226.
31. Villehardouin, La conqute, p. 232.
32. Villehardouin, La conqute, p. 250; Niketas Choniates, 'Historia', col. 1006.
33. Livre de la conqueste, p. 48.
34. See above, pp. 47-51, 54-59-
3/6
35. Livre de la conqueste, pp. 243-52 (quote: p. 251); Libro de los fechos et conquistas,
pp. loo-ioi.
36. Assises, article 26.
37. See above, pp. 3-4, 7-11, 192, 194-96.
38. Corinth: Bon, La Mare franque, p. 473; Andrews, Castles, pp. 136-37. Monemvasia: ibid., p. 207. Thebes and Salona: Bon, 'Fortresses medievales', pp. 164,
187.
39. Ludolph of Sudheim, 'De itinere Terrae Sanctae', p. 28.
40. Lock, 'The Prankish Tower on the Acropolis, Athens', p. 133; Lock, 'The Prankish Towers of Central Greece', p. 112. Buffavento: see above, p. 116.
41. Boeotia: Bon, 'Fortresses medievales', p. 148. Euboea: Lock, 'The Prankish
Towers of Central Greece', p. 102. Lock has since changed his views; see next
footnote.
42. Euboea: Lock, 'The Towers of Euboea', pp. 107-11, 118. See also Lock, 'The
Medieval Towers of Greece', pp. 139-40. Haliartos: Bintuff, 'The Prankish
Countryside in Central Greece', p. 6.
43. This was particularly the case in Cuidan Armenia; see above, pp. 157-61.
Notes to Chapter 23, Prankish Greece: Fortifications and Internal Security
1. Assises, article 70; Livre de la conqueste, p. 47; Libro de los fechos et conquistas,
p. 32.
2. Livre de la conqueste, pp. 34-35. See also Libro de los fechos et conquistas, p. 31.
3. Lock, The Franks, pp. 284-90. That Greeks in Achaia performed military service
is shown in Assises, article 71. The Greeks of Naxos are said to have helped
their Venetian conquerors combat pirates. See Andrea Dandolo, 'Chronicum
Venetum', p. 282; Fotheringham, Marco Saudo, p. 57.
4. See above, pp. 211, 213, 218-22, 224.
5. Villehardouin, La conqute, p. 196.
6. Livre de la conqueste, p. 79.
7. Assises, article 94.
8. Marino Saudo, 'Istoria', p. 106.
9. Andrea Dandolo, 'Chronicum Venetum', pp. 284-85, 288, 292 (see also p. 304);
Andrea Navagerio, 'Storia Venezia', cols 987-92; Marino Saudo, 'Vite de'
duchi', cols 545, 547, 549 (see also col. 557); Fotheringham, Marco Saudo,
pp. 87-103.
10. Andrea Navagerio, 'Storia Venezia', col. 991; Marino Saudo, 'Vite de' duchi',
col. 549; Fotheringham, Marco Saudo, p. loi.
11. Livre de la conqueste, pp. 318-27.
12. This is made clear by the fact that a Latin garrison defended St George during
another Greek rebellion in 1302 which will be discussed shortly. See Livre de
la conqueste, p. 367.
13. Livre de la conqueste, pp. 362-65, 373.
14. Fotheringham, Marco Saudo, pp. 89-92.
377
378
30. Livre de la conqueste, pp. 404-5; Libro de los fechos et conquistas, pp. 140-43.
31. Villehardouin, La conqute, pp. 164-66; Robert of Clari, La conqute,
PP- 97-98.
32. Villehardouin, La conqute, pp. 166-78; Robert of Clari, La conqute, pp. 97-100,
104-5; Longnon, L'empire latn, pp. 55-61.
33. Henry of Valenciennes, Histoire de l'empereur, pp. 55-118. See also Longnon,
L'empire latin, pp. 106-11; Miller, The Latins in the Levant, pp. 72-75; Lock,
The Franks, pp. 57-60.
34. Henry of Valenciennes, Histoire de l'empereur, pp. 59-61, 104-6.
35. Henry of Valenciennes, Histoire de l'empereur, pp. 79-118.
36. Frederick III had recognised Roger as vicar-general by August 1366. See Rubio
y Lluch, Diplomatari de I'Orient ctala, no. 271, p. 355. The pope also complained
about Roger's actions, and his use of Turkish mercenaries. See Lettres secretes
et curales se rapportant a la France, ed. G. Mollai (Paris 1955), nos 1047, 1050,
p. 163; Setton, 'The Catalans in Greece, 1311-1380', pp. 198-99, 202-4.
37. Livre de la conqueste, p. 85; Libro de los fechos et conquistas, pp. 50-51; Marino
Saudo, 'Istoria', p. 105.
38. Libro de los fechos et conquistas, pp. 152-55, and see above, p. 230.
39. Livre de la conqueste, p. 30. See also Bon, La More franque, pp. 318-20.
40. Livre de la conqueste, p. 137.
41. See above, pp. 104-5.
42. Innocent III, PL, vol. 216, no. 27, col. 564; Lock, The Franks, p. 212.
43. Holy Land: see above, pp. 54-59.
379
13. Survey of royal properties in Morea, 1371, in Bon, La Mare franque, p. 690;
Innocent III, PL, vol. 216, no. 150, col. 330.
14. Innocent III, PL, vol. 216, no. 136, cols 323-24; Rubio y Lluch, 'Els Castells',
PP- 393-9815. Henry of Valenciennes, Histoire de l'empereur p. no, no n. 4. See also P. Lock,
'The Military Orders in Mainland Greece', in Barber, The Military Orders,
PP- 333-339, at pp. 334-3516. Innocent III, PL, vol. 216, no. 144, col, 328; Lock, 'The Military Orders', p. 335.
17. Rubio y Lluch, Diplomatari de l'Orient ctala, no. 63, pp. 78-79.
18. Niccolo da Martoni, 'Relation du plerinage', pp. 655-56.
19. Innocent III, PL, vol. 216, lib. 13, no. 120, cols 3078, lib. 15, no. 69, cols 591-94.
20. Marino Saudo, 'Istoria', p. 130; Livre de la conqueste, pp. 216-17; Miller, The
Latins in the Levant, pp. 133-35.
21. Marino Saudo, 'Epistulae', ep. 3, p. 293.
22. To Chronikon ton Mareos, p. 214.
23. Riley Smith, The Knights of St John, pp. 215-16; A. Luttrell, 'The Hospitallers
at Rhodes, 1306-1421', in HC, vol. 3, pp. 278-313, at pp. 283-88.
24. Boase, 'The Arts in Prankish Greece', pp. 240-43; A. Luttrell, 'Lindos and the
Defence of Rhodes, 1306-1522', in A. Luttrell, The Hospitallers of Rhodes and
their Mediterranean World (Aldershot, 1992), c. 7, pp. 317-32.
25. Ludolph of Sudheim, 'De itinere Terrae Sanctae', pp. 27-28; Boase, 'The Arts in
Prankish Greece', pp. 231-40; A. Luttrell, 'The Military and Naval Organization
of the Hospitallers at Rhodes, 1310-1444', in Lutrell, The Hospitallers of Rhodes
and their Mediterranean World, c. 19, pp. 133-53, especially at pp. 136-44.
26. Luttrell, 'The Hospitallers at Rhodes', pp. 286-96; Luttrell, 'The Military
and Naval Organization of the Hospitallers', pp. 144-47; A. Luttrell, 'The Hospitallers of Rhodes Confront the Turks, 1306-1421', in Luttrell, The Hospitallers
of Rhodes and their Mediterranean World, c. 2, pp. 80116, and see above,
pp. 232-33.
27. Innocent VI, Lettres secretes et curiales, ed. P. Gasnault, M. H. Laurent and
N. Gotteri, 4 vols (Paris and Rome, 1959-76), vol. 4, no. 2133, pp. 75-76, no.
2134, p. 76; Luttrell, 'The Hospitallers at Rhodes', pp. 296-97.
28. Libro de los fechos et conquistas, pp. 159-60.
29. Luttrell, 'The Hospitallers at Rhodes', pp. 299-313; Lutrell, 'The Hospitallers
of Rhodes Confront the Turks', pp. 90-97; A. Luttrell, 'Intrigue, Schism and
Violence among the Hospitallers of Rhodes, 1377-1384', in A. Luttrell, The
Hospitallers in Cyprus, Rhodes, Greece and the West, 1291-1440 (Aldershot, 1979),
c. 23, pp. 30-48.
30. Innocent III, PL, vol. 216, no. 136, col. 323.
31. See above, pp. 192, 198.
32. 'Rgestes des empereurs latins de Constantinople, 1204-1261/72', ed. B. Hendrickx, Byzantina, 14 (1988), no. 41, p. 37, and see Lock, 'The Military Orders',
PP- 335-36.
33. Port: Niccolo da Martoni, 'Relation du plerinage', p. 655. See also A. Luttrell,
380
381
Sidon: Kalayan, 'The Sea Castle of Sidon', pp. 81-89; Deschampa, La defense
du royanme de Jerusalem, p. 232. Crac des Chevaliers: Deschamps, Le Crac des
Chevaliers, pp. 213-24.
21. 'Gestes', p. 793.
22. Livre de la conqueste, pp. 397-99.
23. Livre de la conqueste, p. 344.
24. Saudo, 'Istoria', p. loi. Chlemoutsi: Bon, La Mare franque, pp. 608-22; Andrews, Castles, pp. 154-58.
25. Joinville, Histoire de Saint Louis, pp. 282-84.
26. Livre de la conqueste, pp. 120-23; Libro de los fechos et conquistas, pp. 65-66,
67-68.
27. Henry of Valenciennes, Histoire de l'empereur, p. 62.
28. Pringle, The Red Tower, pp. 15, 105, 107, 124.
29. Documents sur le rgime des tenes dans la printipaut de More au XlVe sicle,
ed. J. Longnon and P. Topping (Paris, 1969), pp. 70-71; P. Lock, 'The Prankish
Towers of Central Greece', p. no. Kolossi: Enlart, L'art gothique, vol. 2, pp. 6839430. Bintliff, 'The Prankish Countryside in Central Greece', p. 6, and see pp. 1-18.
31. Lock, 'The Prankish Towers of Central Greece', p. 111; Lock, 'The Medieval
Towers of Prankish Greece' p. 138.
32. Lock, 'The Towers of Euboea', pp. 107-26.
33. Lock, 'The Towers of Euboea', p. 113, no. 13.
34. 'Gestes', pp. 814-15; Jacoby, 'Crusader Acre in the Thirteenth Century',
pp. 19-36; Rey, 'Etude sur la topographic', pp. 137-40; Pringle, 'Towers in
Crusader Palestine', pp. 337-38.
35. 'Gestes', p. 747; 'Annales de Terre Sainte', p. 448 (text B).
36. Athens: Lock, 'The Prankish Tower on the Acropolis, Athens', pp. 131-33; Lock,
'The Medieval Towers of Prankish Greece', p. 132. Negroponte: Lock, 'The
Towers of Euboea', p. 116, no. 31.
37. Lock, 'The Towers of Euboea', p. 113, no. 13.
38. Enlart, L'art gothique, vol. 2, p. 576.
39. D. Renn, 'Burhgeat and Gonfanon: Two Sidelights from the Bayeux Tapestry',
Anglo-Norman Studies, 16 (1994), pp. 177-98, at pp. 182-83, cited from C. Coulson, 'Cultural Realities and Reappraisals in English Castle-Study', Journal of
Medieval History, 22 (1996), pp. 171-207, at p. 175. I am very grateful to Dr
Coulson for providing me with this information.
382
22. Livre de la conqueste, pp. 245-47, 260. Chlemoutsi: Bon, La Mare franque,
pp. 608-22.
Notes to Chapter 27, The Non-Miltary Functions of Fortifications:
Taxation, Justice and Administration
1. Assises, articles 43, 177, and see article 9.
2. Assises, article 94. In 1275, for example, the high court gathered at Andravida
to hear an important land dispute. See Livre de la conqueste, pp. 197-211.
3. 'Nouvelles preuves de l'histoire de Chypre sous le regne des princes de la
mason de Lusignan', ed. L. de Mas Latrie, Bibliothque de l'Ecole des Chaes,
35 (Paris, 1874), pp. 120-21; Edbury, The Kingdom of Cyprus, p. 194.
4. 'Livre de Jean d'Ibelin', RHCLois, vol. i, p. 420.
5. Livre de la conqueste, pp. 328-29. See also Bon, La More franque, pp. 657-58.
6. De constructione castri Saphet, p. 386, lines 257-62.
7. Cartulaire, vol. 3, no. 3051, p. 64. See also Kedar, 'The Subjected Muslims',
pp. 160-74.
8. For this and similar twelfth-century examples, see Smail, Crusading Warfare,
p. 61.
383
384
10. Benvenisti, The Crusaders in the Holy Land, p. 251; Deschamps, La defense du
royanme de Jerusalem, p. 124; Pringle, 'Survey of Castles in the Crusader
Kingdom of Jerusalem', p. 89; Pringle, Gazetteer, no. 133, pp. 62-64.
11. Livre de la conqueste, p. 273.
12. Villehardouin, La conqute, p. 250.
13. Makhairas, Recital, c. 470, p. 453; Strambaldi, Chronique, p. 194.
14. Margat: Willbrand of Oldenburg, 'Itinerarium', p. 210. Saphet: De constructione
castri Saphet, p. 385, lines 229-33.
15. Ibn al-Furat, 'Selections', p. 85.
16. Margat: Willbrand of Oldenburg, 'Itinerarium', p. 210. Cave de Tyron:
Deschamps, La defense du royanme de Jerusalem, pp. 219-20.
17. 'Gestes', p. 686.
18. Marino Saudo, 'Lber secretorum', p. 229.
19. La regle du Temple, cl. 126, deals with supplies of this type being provided for
castles. Stables situated within the outer defences at 'Atlit could accommodate
a large number of cattle and horses. See C. N. Johns, 'Excavations at Pilgrims'
Castle ('Atlit): Stables at the South-West of the Suburb', QDAP, 5 (1935-36),
pp. 31-60.
20. Assises, article 159.
21. Deschamps, Le Crac des Chevaliers, p. 155; Lawrence, 'The Castle of Baghras',
pp. 58-59. See also Kennedy, Crusader Castles, p. 100.
22. Johns, 'The Faubourg and its Defences', p. 120.
23. Johns, Guide to 'Atlit, pp. 96-97.
24. William of Tyre, Chronicon, vol. 2, bk 22, c. 31, p. 1059.
25. Deschamps, Le Crac des Chevaliers, p. 189.
26. Beaufort: Deschamps, La defense du royanme de Jerusalem, p. 199. See also
Kennedy, Crusader Castles, pp. 99-100; Deschamps, La defense du comt de
Tripoli, p. 153.
27. Deschamps, Le Crac des Chevaliers, p. 152.
28. 'Gestes', p. 817; Bustron, Chronique, p. 127.
29. Paphos: Megaw, 'Supplementary Excavations', p. 324. Montfort: Pringle, 'A
Thirteenth-Century Hall', pp. 57-60, 68-71.
30. Deschamps, Le Crac des Chevaliers, pp. 203-4; Deschamps, La defense du comt
de Trpoli, pp. 153-54.
31. Crac des Chevaliers and Margat: Willbrand of Oldenburg, 'Itinerarium',
pp. 208-10. Saphet: De constructione castri Saphet, p. 384, lines 204-6.
32. Benvenisti, The Crusaders in the Holy Land, p. 299, and see also pp. 37475 for
more details on crusader kitchens.
33. De constructione castri Saphet, p. 384, lines 210-15.
34. La regle du Temple, cl. 196.
35. La regle du Temple, cl. 662.
36. Megaw, 'Supplementary Excavations', pp. 333-34, 335, 337.
37. La regle du Temple, els 325 (masons), 604 (cobblers).
385
38. Saphet: De constmctione castri Saphet, p. 384, lines 185-90. Paphos: Megaw,
'Supplementary Excavations', p. 332.
39. Tabulae ordinis Theutonid, nos 27-29, pp. 23-25; Rey, 'Etude sur la topographic', p. 125.
40. See, for example, Baybars' preparations for his campaign against Saphet in
1266: al-Makrizi, Histoire des sultans, vol. i(b), p. 28.
41. De constmctione castri Saphet, p. 384, lines 204-10.
42. De constmctione castri Saphet, pp. 381-82, lines 115-24.
43. al-Makrizi, Histoire des sultans, vol. i(b), pp. 27-31; Ibn al-Furat, 'Selections'
pp. 88-96; 'Gestes', pp. 764-66; 'Eracles', pp. 454-55; Marino Saudo, 'Liber
secretorum', pp. 222-23.
44. For more details on turcopoles, see Marshall, Warfare in the Latin East,
pp. 58-60.
45. 'Gestes', p. 756. See also Kedar, 'The Subjected Muslims', pp. 152-54.
46. Marshall, Warfare in the Latin East, p. 119.
47. Amadi, Chronique, p. 238. See also 'Gestes', pp. 849-50; Bustron, Chronique,
P-13348. Livre de la conqueste, p. 71.
49. Assises, articles 70, 71.
50. Nikephoros Gregoras, 'Historia Byzantina', vol. i, pp. 25-28; George Akropolites, 'Annales', cols 1035-38; Sinogowitz, 'Zur Eroberung Thessalonikes im
Herbst 1224', p. 28.
51. Ibn al-Furat, 'Selections', p. 139.
52. Cartulaire, vol. 4, no. 3308, pp. 291-93.
53. RRH, no. 1329, p. 347.
54. Ibn al-Furat, 'Selections', pp. i43~49;'Gestes' pp. 768, 777; al-Makrizi, Histoire
des sultans, vol. i(b), pp. 84-85.
55. Ibn al-Furat, 'Selections', p. 130; al-Makrizi, Histoire des sultans, vol. i(b), p. 56;
Cartulaire, vol. 3, no. 3400, p 231. For the capture of Montfort in 1271, see Ibn
al-Furat, 'Selections', pp. 106-12; al-Makrizi, Histoire des sultans, i(b), p. 87;
'Gestes' p. 778.
56. Marino Saudo, 'Istoria', pp. 118-19; Livre de la conqueste, pp. 160,170-76; Libro
de los fechos et conquistas, pp. 91, 92, and see above, p. 193.
57. Buchn, Nouvelles recherches, vol. 2, nos 24, 25, 27, pp. 330-31, 332-33.
58. Buchn, Nouvelles recherches, vol. 2, nos 25-29, pp. 143-60.
386
4. Jacoby, 'The Rise of a New Emporium in the Eastern Mediterranean', pp. 14754; Enlart, L'art gothique, vol. 2, pp. 673-83.
5. The defences of Acre and Tyre protected the people and economic prosperity
of these cities on many occasions. See, for example, al-Makrizi, Histoire des
sultans, vol. i(a), pp. 199-200; Ibn al-Furat, 'Selections', pp. 57-59 (Mamluk
raid on Acre in 1263), and al-Makrizi, Histoire des sultans, vol. i(b), pp. 27-28;
Ibn al-Furat, 'Selections', p. 87 (Mamluk raid on Acre and Tyre in 1266). See
also above, p. 39.
6. Buchn, Nouvettes recherches, vol. 2, no. 9, pp. 98-103; Bon, La Mare franane,
pp. 320-22.
7. Lock, The Franks, pp. 154-55, 252-53; Miller, The Latins in the Levant, pp. 495,
498; Andrews, Castles, pp. 14 598. Patras: Gerland, Neue Quellen, pp. 89-92. Monemvasia: George Sphrantzes,
'Chronicon minus', ed. I. Bekker, CSHB (Bonn, 1838), pp. 399-404.
9. For detailed descriptions of these, see Rey, 'Etude sur la topographic', pp. 137
(Venice), 137-38 (Genoa), 138-39 (Pisa); Benvenisti, The Crusaders in the Holy
Land, pp. 98-100 (Pisa), 100-2 (Genoa), 102-4 (Venice); Jacoby, 'Crusader
Acre in the Thirteenth Century', pp. 19-26 (Pisa), 26-30 (Genoa), 30-36
(Venice); Kesten, The Old City of Acre, pp. 29-39 (Genoa), 40-45 (Venice),
59-67 (Pisa).
10. 'Gestes', pp. 742-48. For more details on this dispute, see Riley-Smith, The
Feudal Nobility, pp. 215-17.
11. Makhairas, Recital, c. 372, p. 353.
12. 'Gestes', p. 829.
13. See, for example, Thiriet, Rgestes, vol. i, no. 371, p. 97, and see Lock, 'The
Medieval Towers of Prankish Greece', p. 139; Lock, 'The Frankish Towers of
Central Greece', pp. 108-9; Lock, 'The Towers of Euboea', p. ill, nos 3, 4,
p. 116, no. 31, p. 117, no. 35. As we have seen, towers of this kind also often
acted as residences and status symbols. See chapter twenty-five.
14. Lock, 'The Towers of Euboea', p. 117, no. 35.
15. Le trsor des hartes d'Armenie, no. 23, pp. 170-75, and see pp. 35-38, 49-50.
For trade patterns in this area, see also Cahen, La Syrie du Nord, pp. 689-91;
Barber, The New Knighthood, p. 240.
16. 'Gestes', pp. 768-69; Jacoby, 'Crusader Acre in the Thirteenth Century', pp. 810; Gertwagen, 'The Crusader Port of Acre', pp. 559-60.
17. Andrews, Castles, pp. 187-91, particularly p. 187. For Venetian defence measures
regarding this bridge, see also Lock, 'The Towers of Euboea', p. 118.
18. RRH, no. 977, p. 257.
19. Jacoby, 'Crusader Acre in the Thirteenth Century', pp. 15-19. For a recent
reinterpretation of the layout of Acre's harbour, see Gertwagen, 'The Crusader
Port of Acre', pp. 554-68.
20. Oliver of Paderborn, 'Historia Damiatina', p. 169; Johns, Guide to 'Atlit, pp. 9498.
21. RRH, no. 1043, p. 272. The Templars were exempt from paying when using
387
this gate. See also Cartulaire, vol. 2, no. 2,058, pp. 455-57; Deschampa, La defense
du comt de Tripoli, pp. 28485.
22. Document translated and reproduced in Alishan, Sissouan, p. 239. See also
Edwards, Fortifications, pp. 147-49; Forstreuter, Der deutsche Orden, p. 65.
23. 'Gestes', p. 839.
388
Bibliography
PRIMARY SOURCES
Western Sources
Acta Aragonensia, ed. H. Pinke, 3 vols (Berlin and Leipzig, 190822).
Actes relatifs a la principante de Mare, 1289-1300, ed. C. Perrat and J. Longnon
(Paris, 1967).
Amadi, Francesco, Chroniques d'Amadi et de Stmtnbaldi, ed. R. de Mas Latrie (Paris,
1891).
Ambroise, L'estoire de la guerre sainte, ed. G. Paris (Paris, 1897).
'Annales Genuenses', ed. G. Petti Balbi, RIS n.s., vol. 17, part 2 (Bologna, 1975).
'Annales de Terre Sainte', ed. R. Rhricht and G. Raynaud, in Archives de l'Orient
latin, 2 (1884), pp. 429-61.
Antiquitates italicae medii aevi, ed. L. A. Muratori, 6 vols (Rome, 1738-42), vol. 5.
Arnold of Lbeck, 'Chronica Slavorum', ed. J. Lappenberg, in MGH SS, vol. 21 (1868).
Les Assises de Romanice, ed. G. Recoura (Paris, 1930).
Aubrey of Trois Fontaines, 'Chronica', ed. P. Scheffer-Boichorst, in MGH SS, vol.
23 (1874).
'Bans et ordonnances des rois de Chypre', in RHCLois, vol. 2.
Barbaro, Daniele, 'Cronica del Trivisano della citt di Venezia', short extract in
J. K. Fotheringham, Marco Saudo: Conqueror of the Archipelago (Oxford, 1915),
p. 106.
Benvenuto di San Giorgio, 'Historia Montisferrati', ed. L. A. Muratori, in RIS, vol.
23 (i733)Boniface VIII, Registres, ed. G. Digard et al., 4 vols (Paris, 1884-1939).
Burchard of Mount Sion, 'Descriptio Terrae Sanctae', ed. J. Laurent, in Peregrinatores
medii aevi quatuor (Leipzig, 1873).
Bustron, Florio, Chronique de l'ile de Chypre, ed. R. de Mas Latrie (Paris, 1886).
Cartulaire generai de l'orare des Hospitaliers de St Jean de Jerusalem, 1100-1310, ed.
J. Delaville Ie Roulx, 4 vols (Paris, 18941906).
Cartulaire general de Vordre du Temple, niy-iiso, ed. G. Albon (marquis d') (Paris,
1913)'The Cartulaire de Manosque: A Grant to the Templars in Latin Syria and a Charter
of King Hugh I of Cyprus', ed. P. W. Edbury, Bulletin of the Institute of Historical
Research, 51 (1978), pp. 174-81.
390
BIBLIOGRAPHY
3pl
392
UNKNOWN
CRUSADER CASTLES
Regesta Regni Hierosolymitani (1097-1291), ed. R. Rhricht (Innsbruck, 1893; Additamentum, 1904).
Rgestes des deliberations du snat de Venise concernant la Romanie, ed. F. Thiriet,
3 vols (Paris, 1958-61).
'Rgestes des empereurs latins de Constantinople, 1204-1261/72', ed. B. Hendrickx,
Byzantina, 14 (1988).
'A Register of the Cartulary of the Cathedral of Santa Sophia of Nicosia', ed.
J. L. LaMonte, Byzantion, 5 (1930), pp. 439-522.
La regle du Temple, ed. H. de Curzon (Paris, 1886).
Richard of Devizes, Chronicon de tempore Regis Ricardi Primi, ed. J. T. Appleby
(London, 1963).
Robert of Clari, La conqute de Constantinople, ed. P. Lauer (Paris, 1924).
Roger of Howden, Chronica, ed. W. Stubbs, RS, 51, 4 vols (London, 1868-71).
Le saint voyage de Jhrusalem du seigneur d'Anglure, ed. F. Bonnardot and A. Longnon (Paris, 1878).
Saudo, Marino (Torsello, or the Elder), 'Liber secretorum fidelium crucis', ed.
J. Bongars, in Gesta Dei per Francos, 2 vols (Hannau, 1611), voi. 2.
Saudo, Marino (Torsello, or the Elder), 'Epistulae', ed. J. Bongars, in Gesta Dei
per Francos, 2 vols (Hannau, 1611), voi. 2.
Saudo, Marino (Torsello, or the Elder), 'Istoria del regno di Romania sive regno
di Morea', ed. C. Hopf, in Chroniques grco-romanes indites ou peu connues
(Berlin, 1873).
Saudo, Marino (the Younger), 'Vite de' duchi di Venezia', ed. L. A. Muratori, RIS,
vol. 22 (1733).
Sauger, Robert, Histoire nouvelle des anciens dues de l'Archipel (Paris, 1699), extract
in J. K. Fotheringham, Marco Saudo: Conqueror of the Archipelago (Oxford, 1915),
pp. 113-22.
Simon de Saint-Quentin, Histoire des Tartares, ed. J. Richard (Paris, 1965).
Strambaldi, Diomedes, Chroniques d'Amadi et de Strambaldi, ed. R. de Mas Latrie
(Paris, 1891).
Supplementary Excerpts on Cyprus: Further Materials for a History of Cyprus, trans.
and ed. T. A. H. Mogabgab (Nicosia, 1941).
Surveys of royal properties in the Morea undertaken in the fourteenth century,
reproduced in A. Bon, La Mare franque: recherches historiques, topographiques
et archologiques sur la principaut d'Acae, 1205-1430 (Paris, 1969), pp. 689-92.
Tabulae ordinis Theutonici, ed. E. Strehlke (Berlin, 1869).
'The Templars and the Castle of Tortosa in Syria: An Unknown Document Concerning the Acquisition of the Fortress', ed. J. S. C. Riley-Smith, EHR, 84 (1969),
pp. 278-88.
Thomas Wykes, 'Chronicon', ed. H. R. Luard, in Annales Monastici, 5 vols, RS, 36
(London, 1864-69), vol. 4.
Urban V, Lettres secretes et curiales du pape Urbain V se rapportant a la France, ed.
G. Mollat (Paris, 1955).
Urkunden zur lteren Handels- und Staatsgeschichte der Republik Venedig mit
BIBLIOGRAPHY
393
besonderer Beziehung au/Byzanz und die Levante, ed. G. Tafel and G. Thomas, 2
vols (Vienna, 1856).
Villani, Johannis, 'Florentini historia universalis', ed. L. A. Muratori, RIS, vol. 13 (1728).
Willbrand of Oldenburg, 'Itinerarium Terrae Sanctae', ed. S. de Sandoli, in lunera
Hierosolymitana Crucesignatorum, 4 vols ( Jerusalem, 1978-84), vol. 3.
William of Machaut, La prise d'Alexandre, ou chronique du roi Pierre Ier de Lusignan,
ed. L. de Mas Latrie (Geneva, 1877).
William of Nangis, 'Gesta Ludovici IX', in RHGF, vol. 20 (1881).
William of Tyre, Chronicon, ed. R. B. C. Huygens, 2 vols (Turnhout, 1986).
Zeno, lacobo, 'Vita Caroli Zeni', ed. G. Zonta, RIS n.s., vol. 19, part 6 (1931-41).
Greek Sources
Akropolites, George, 'Annales', ed. J. P. Migne, in PG, vol. 140 (1887).
Choniates, Niketas, 'Historia', ed. J. P. Migne, in PG, vol. 139 (1894).
To Chroniken tou Mareos: The Chronicle of Marea, ed. J. Schmitt (London, 1904).
Gregoras, Nikephoros, 'Nicephori Gregorae Byzantina historia', ed. L. Schopen, in
CSHB, 3 vols (Bonn, 1829-55).
John VI Kantkouzenos, 'loannis Cantacuzeni eximperatoris historiarum libri IV',
ed. L. Schopen and B. Niehbuhr, in CHSB, 3 vols (Bonn, 1828-32).
Makhairas, Leontios, Recital Concerning the Sweet Land of Cyprus, Entitled 'Chronicle', ed. and trans. R. M. Dawkins, 2 vols (Oxford, 1932).
Pachymeres, George, 'Georgii Pachymeris de Michaele et Andronico Palaeologis',
ed. I. Bekker, in CHSB, 2 vols (Bonn, 1835).
394
Muslim Sources
Abu'1-Fida, 'Annales', in RHCOr, vol. i.
Abu Shama, 'Le livre des deux jardins', in RHCOr, vols 4-5.
al-'Ayni, 'Le collier de perles', in RHCOr, vol. 2.
al-Makin Ibn al'Amid, Chronique des Ayyoubides (602-658/1205-6 -1259-60), trans.
A.-M. Edd and F. Micheau (Paris, 1994).
al-Makrizi, 'Histoire d'Egypte', trans. E. Blochet, ROL, 9 (1902).
al-Makrizi, Histoire des sultans mamlouks de I'Egypte, trans. M. E. Quatremre, 2
vols in 4 parts (Paris, 1845).
Baha'-al-Din, 'Anecdotes et beaux traits de la vie de sultan Youssof', in RHCOr,
vol. 3.
'Histoire des patriarches d'Alexandrie', extracts in J. F. Michaud, Bibliothque des
Croisades, vol. 4 (Paris, 1829).
Ibn 'Abd al-Rahim, 'Vie de Kalavun', extracts in J. F. Michaud, Bibliothque des
Croisades, vol. 4 (Paris, 1829).
Ibn al-Athir, 'Kamel Altevarykh', in RHCOr, vols 1-2.
Ibn al-Furat, 'Selections from the Tarikh al-Duwal wa'1-Muluk', in Ayyubids, Mameluks and Crusaders, ed. and trans. U. and M. C. Lyons, with an introduction by
J. S. C. Riley-Smith, 2 vols (Cambridge, 1971).
Ibn Jubair, 'Extrait du voyage d'Ibn Djobeir', in RHCOr, vol. 3.
Imad ad-Din al-Isfahani, Conqute de la Syne et de la Palestine par Saladin, trans.
H. Masse (Paris, 1972).
'Izz al-Din Ibn Saddad, Description de la Syrie du Nord, trans. A-M Edd-Terrasse
(Damascus, 1984).
Kamal-ad-Din, 'L'histoire d'Alep', trans. E. Blochet, ROL, 5 (1897).
SECONDARY SOURCES
BIBLIOGRAPHY
395
Balard, M., 'L'activit commerciale en Chypre dans les annes 1300', in P. Edbury
(ed.), Crusade and Settlement (Cardiff, 1985), pp. 251-67.
Balard, M., La Romanie gnoise, Xlle- dbut du Xve sicle, 2 vols (Rome and Genoa,
1978).
Balard, M., 'The Genoese in the Aegean, 1204-1566', in B. Arbel, B. Hamilton and
D. Jacoby (eds), Latins and Greeks in the Eastern Mediterranean after 1204 (London,
1989), PP-158-74Barber, M., 'Supplying the Crusader States: The Role of the Templars', in B. Z. Kedar
(ed.), The Horns ofHattin (London and Jerusalem, 1992), pp. 314-26.
Barber, M., The New Knighthood: A History of the Order of the Temple. (Cambridge,
1994)Bartusis, M. C., The Late Byzantine Army: Arms and Society, 1204-1453 (Philadelphia,
1992).
Benvenisti, M., The Crusaders in the Holy Land (Jerusalem, 1970).
Bintiff, J., 'The Prankish Countryside in Central Greece: The Evidence from Archaeological Field Survey', in P. Lock and G. D. R. Sanders (eds), The Archaeology
of Medieval Greece (Oxford, 1996), pp. 1-18.
Boas, A., Crusader Archaeology: The Material Culture of the Latin East (London, 1999).
Boase, T. S. R., Kingdoms and Strongholds of the Crusaders (London, 1971).
Boase, T. S. R., 'Military Architecture in the Crusader States', in HC, vol. 4 (1977),
pp. 140-64.
Boase, T. S. R., 'The Arts in Prankish Greece and Rhodes', in HC, vol. 4 (1977),
pp. 208-50.
Boase, T. S. R., 'The History of the Kingdom', in idem (ed.), The Cilician Kingdom
of Armenia (Edinburgh, 1978), pp. 1-33.
Bombaci, A., 'The Army of the Saljuqs of Rum', Istituto orientale di Napoli: annali,
new series, 38 (1978), pp. 343-69Bon, A., 'Fortresses medievales de la Grece centrale', Bulletin de correspondence
hellnique, 61 (1937), pp. 136-208.
Bon, A., La Mare franque: recherches historiques, topographiques et archologiques sur
la principaut d'Achaie, 1204-1430, with a volume of plates and plans (Paris, 1969).
Brand, C. M., Byzantium Confronts the West (Cambridge, Massachusetts, 1968).
Braune, M., 'Die mittelalterlichen Befestigungen der Stadt Tortosa/Tartus', Damaszener Mitteilungen, 2 (1985), pp. 44-54.
Burridge, P., 'The Castle of Vardounia and Defence in the Southern Taygetos', in
P. Lock and G. D. R. Sanders (eds), The Archaeology of Medieval Greece (Oxford,
1996), pp. 19-^8.
Gaben, C., La Syrie du Nord a l'poque des croisades et la principaut franque
d'Antioche (Paris, 1940).
Canard, M., 'Le royaume d'Armnie-Cilicie et les mamelouks jusqu'au traite de
1285', Revue des tudes armniennes, 4 (1967), pp. 217-59.
Chevedden, P., 'Crusader Warfare Revisited: A Revisionist Look at R. C. Smail's
Thesis: Fortifications and Development of Defensive Planning', Journal of the
Rocky Mountain Medieval and Renaissance Association, 14 (1993), pp. 13-31.
396
UNKNOWN
CRUSADER
CASTLES
BIBLIOGRAPHY
397
Folda, }., 'The Crusader Frescoes at Crac des Chevaliers and Marqab Castle',
Dumbarton Oaks Papers, 36 (1982), pp. 177-210.
Forey, A. J., 'The Military Order of St Thomas of Acre', EHR, 92 (1977), pp. 481-503.
Forey, A. J., The Military Orders from the Twelfth to the Early Fourteenth Centuries
(London, 1992).
Forstreuter, K., Der deutsche Orden am Mittelmeer (Bonn, 1967).
Foss, C., and D. Winfield, Byzantine Fortifications: An Introduction (Pretoria, 1986).
Fotheringham, J. K., Marco Saudo: Conqueror of the Archipelago (Oxford, 1915).
France, J., Western Warfare in the Age of the Crusades, 10001300 (London, 1999).
Frankel, R., 'The North-Western Corner of Crusader Acre', Israeli Exploration
Journal, 37, part 4 (1987), pp. 256-61.
Frech, F., 'Die armenischen Burgen', Zeitschrift der Gesellschaft fr Erdkunde zu
Berlin, 9 (1915), pp. 576-80.
Gay, J., Le Pape Clement VI et les affaires d'Orient, 1342-1352 (Paris, 1904).
Geanakoplos, D. J., 'Greco-Latin Relations on the Eve of the Byzantine Restoration:
The Battle of Pelagonia, 1259', Dumbarton Oaks Papers, 7 (1953), pp. 101-41.
Geanakoplos, D. J., Emperor Michael VIII Palaeologus and the West, 1258-82: A Study
in Byzantine-Latin Relations (Cambridge, Massachusetts, 1959).
Gertwagen, R., 'The Crusader Port of Acre: Layout and Problems of Maintenance',
in M. Baiard (ed.), Autour de la Premire Croisade (Paris, 1996), pp. 553-82.
Gill, J., 'The Tribulations of the Greek Church in Cyprus, 1196-1280', Byzantinische
Forschungen, 5 (1977). PP- 73~93Godfrey, J., 1204: The Unholy Crusade (Oxford, 1980).
Gottwald, J., 'Burgen und Kirchen im mittleren Kilikien', BZ, 41 (1941), pp. 82-103.
Gottwald, J., 'Die Burg Til im sudstlichen Kilikien', BZ, 40 (1940), pp. 89104.
Gottwald, J., 'Die Kirche und das Schloss Paperon in Kilikisch-Armenien', BZ, 36
(1936), pp. 86-100.
Gough, M., 'Anazarbus (Anavarza)', Anatolian Studies, 2 (1952), pp. 85-150.
Grabois, A., 'La cit de Baniyas et le chateau de Subeibeh pendant les croisades',
Cahiers de civilisation medievale, 13 (1970), pp. 43-62.
Hamilton, B., The Latin Church in the Crusader States: The Secular Church (London,
1980).
Harper, R. and D. Pringle, 'Belmont Castle: A Historical Notice and Preliminary
Report of Excavation in 1986', Levant, 20 (1988), pp. 101-18.
Heffening, W., 'Eine Burgruine im Taurus', Reportium fr Kunstwissenschaft, 45
(1925), pp. 179-89Hellenkemper, H., Burgen der Kreuzritterzeit in der Grafschaft Edessa und im Knigreich Kleinarmenien (Bonn, 1976).
Hendrickx, B., 'A propos du nombre des troupes de la Quatrime Croisade et de
l'empereur Baudoin I', Byzantina, 3 (1971), pp. 29-41.
Hill, G., A History of Cyprus, 4 vols (Cambridge, 1940-52).
Hillenbrand, C., The Crusades: Islamic Perspectives (Edinburgh, 1999).
Holt, P., 'Baybars's Treaty with the Lady of Beirut in 667\i269', in P. Edbury (ed.),
Crusade and Settlement (Cardiff, 1985), pp. 242-48.
398
U N K N O W N CRUSADER CASTLES
Holt, P., The Age of the Crusades: The Near East from the Eleventh Century to 1517
(London, 1986).
Holt, P. (ed.), The Eastern Mediterranean Lands in the Period of the Crusades (Warminster, 1977).
Holum, K. G., R. L. Hohlfelder, R. J. Bull and A. Raban, King Herod's Dream: Caesarea on the Sea (New Haven and London, 1988).
Hubatsch, W., 'Der deutsche Orden und die Reichslehnschaft ber Cypern', Nachrichten der Akademie der Wissenschaftim Gttingen. Philologisch-Historische Klasse
(1955), PP- 245-306.
Humphreys, R. S., From Saladin to the Mongols: The Ayyubids of Damascus, 1193-1260
(New York, 1973).
Humphreys, R. S., 'The Emergence of the Mamluk Army', Studia Islamica, 45 (1977),
pp. 67-99.
I'Anson, E., 'Medieval and Other Buildings in the Island of Cyprus', Transactions
of the Royal Institute of British Architects, session 1882-83 (London, 1883), pp. 13-33.
Ilieva, A., 'The Suppression of the Templars in Cyprus According to the Chronicle
of Leontios Makhairas', in M. Barber (ed.), The Military Orders: Fighting for the
Faith and Caring for the Sick (Aldershot, 1994), p. 212-19.
Imber, C., The Ottoman Empire, 1300-1481 (Istanbul, 1990).
Inalcik, H., The Ottoman Empire: The Classical Age, 1300-1600, trans. N. Itzkowitz
and C. Imber (London, 1973).
Inalcik, H., 'The Ottoman Turks and the Crusades, 1329-1451' and 'The Ottoman
Turks and the Crusades, 1451-1522', both in HC, vol. 6 (1989), pp. 222-75, 3U-53Irwin, R., 'The Mamluk Conquest of the County of Tripoli', in P. Edbury (ed.),
Crusade and Settlement (Cardiff, 1985), pp. 246-49.
Irwin, R., The Middle East in the Middle Ages: The Early Years of the Mamluk
Sultanate, 1250-1382 (Beckenham, 1986).
Jacoby, D., Studies on the Crusader States and on Venetian Expansion (Northampton,
1989), containing 'Crusader Acre in the Thirteenth Century: Urban Layout and
Topography', ch. 5, pp. 1-45; 'Montmusard, Suburb of Crusader Acre: The First
Stage of its Development', ch. 6, pp. 205-17; 'The Rise of a New Emporium in
the Eastern Mediterranean: Famagusta in the Late Thirteenth Century', ch. 8,
pp. 145-79Jacoby, D., 'The Encounter of Two Societies: Western Conquerors and Byzantines
in the Peloponnese after the Fourth Crusade', American Historical Review, 78
(i973), PP- 873-906.
Jackson, P., 'The Crisis in the Holy Land in 1260', EHR, 95 (1980), pp. 481-513.
Jackson, P., 'The Crusades of 1239-41 and their Aftermath', Bulletin of the School
of Oriental and African Studies, 50 (1987), pp. 32-60.
Jackson, P., 'The End of Hohenstaufen Rule in Syria', Bulletin of the Institute of
Historical Research, 59 (1986), pp. 20-36.
Johns, C.N., A Guide to 'Atlit (Jerusalem, 1947).
Johns, C. N., 'Excavations at Pilgrims' Castle ('Atlit): The Faubourg and its Defences',
QDAP, i (1932), pp. 111-29; 'The South-Eastern Cemetary', QDAP, 2 (1932-33),
BIBLIOGRAPHY
399
pp. 41-104; 'The Ancient Tell and the Outer Defences of the Castle', QDAP, 3
(1933-34), PP-145-64; 'An Unfinished Church in the Suburbs', QDAP, 4 (1934-35),
pp. 122-37; 'Stables at the South West of the Suburbs', QDAP, 5 (1935-36), pp. 3160.
C. N. Johns, Pilgrims' Castle CAtlit), David's Tower (Jerusalem) and Qal'at ar-Rabad
('Ajlun), ed. D. Pringle with a foreword by A. Johns (Aldershot, 1997). Collected
studies containing all other listed works by C. N. Johns.
Johns, C. N., 'The Citadel, Jerusalem: A Summary of Work since 1934', QDAP, 14
(1950), pp. 121-90.
Kalayan, H., 'The Sea Castle of Sidon', Bulletin du Muse de Beyrouth, 26 (1973),
pp. 81-89.
Kedar, B. Z., 'The Subjected Muslims of the Prankish Levant', in J. M. Powell (ed.),
Muslims under Latin Rule, 1100-1300 (Princeton, 1990).
Kennedy, H., Crusader Castles (Cambridge, 1994).
Kesten, A., The Old City of Acre: Re-Examination Report, 1993 (Acre, 1993).
Lawrence, A. W., 'The Castle of Baghras', in T. S. R. Boase (ed.), The Cuidan Kingdom of Armenia (Edinburgh, 1978), pp. 34-84.
Lawrence, T. E., Crusader Castles, ed. D. Pringle (Oxford, 1988).
Lock, P., 'Castles and Seigneurial Influence in Latin Greece', in A. V. Murray (ed.),
From Clermont to Jerusalem: The Crusades and Crusader Societies, 1095-1500
(Turnhout, 1998), pp. 173-85.
Lock, P., 'The Prankish Tower on the Acropolis, Athens: The Photographs of
William J. Stillman', ABSA, 82 (1987), pp. 131-33.
Lock, P., 'The Prankish Towers of Central Greece', ABSA, 81 (1986), pp. 101-23.
Lock, P., The Franks in the Aegean, 1204-1500 (London, 1995).
Lock, P., 'The Medieval Towers of Prankish Greece: A Problem in Chronology and
Function', in B. Arbel, B. Hamilton and D. Jacoby (eds), Latins and Greeks in the
Eastern Mediterranean after 1204 (London, 1989), pp. 129-45.
Lock, P., 'The Towers of Euboea: Lombard or Venetian; Agrarian or Strategic', in
P. Lock and G. D. R. Sanders (eds), The Archaeology of Medieval Greece (Oxford,
1996). PP-107-26.
Longnon, J., L'empire latin de Constantinople et la principaut de Mare (Paris,
1949).
Longnon, J., Les compagnons de Villehardouin: recherches sur les croiss de la Quatrime Croisade (Geneva and Paris, 1978).
Longnon, J., 'Problmes de Fhistoire de la principaut de More', Journal des Savants
(1946), pp. 77-93, 147-61.
Longnon, J., 'The Prankish States in Greece, 1204-1311', in HC, vol. 2 (1962),
pp. 235-74.
Luttrell, A., Latin Greece, the Hospitallers and the Crusades, 1291-1440 (Aldershot,
1982), containing 'La corona de Aragon y la Grecia catalana, 1379-1394', eh. 11,
pp. 219-52; 'The Crusades in the Fourteenth Century", eh. 16, pp. 122-54; 'The
Hospitallers of Rhodes: Perspectives, Problems and Possibilities', eh. i, pp. 243-66.
Luttrell, A., The Hospitallers in Cyprus, Rhodes, Greece and the West, 1291-1440
40O
BIBLIOGRAPHY
401
Molin, K., 'Fortifications and Internal Security in the Kingdom of Cyprus, 11911426',
in A. V. Murray (ed.), From Clermont to Jerusalem: The Crusades and Crusader
Societies, 1095-1500 (Turnhout, 1998), pp. 187-99.
Molin, K., 'The Non-Military Functions of Crusader Fortifications, iiSj-c. 1380',
Journal of Medieval History, vol. 23, no. 4 (1997), pp. 367-88.
Morgan, D., The Mongols (Oxford, 1986).
Mller-Wiener, W., Castks of the Crusaders, trans. ]. Brownjohn (London, 1966).
Murray, A. V. (ed.), From Clermont to Jerusalem. The Crusades and Crusader Societies,
1095-1500 (Turnhout, 1998).
Nicol, D.M., The Despotate of Epirus (Oxford, 1957).
Nicol, D.M., The Last Centuries of Byzantium, 12611453 (London, 1972).
Pattenden, P., "The Byzantine Early Warning System', Byzantion, 53 (1983), pp. 258-99.
Powell, }. M., Anatomy of a Crusade, 1213-1221 (Philadelphia, 1986).
Prawer, }., Crusader Institutions (Oxford, 1980).
Prawer, }., The Latin Kingdom of Jerusalem (London, 1972).
Pringle, D., 'A Castle in the Sand: Mottes in the Crusader East', Chteau-Gaillard,
18 (Caen, 1998), pp. 187-91.
Pringle, D., 'Aqua Bella: The Interpretation of a Crusader Courtyard Building',
in B. Z. Kedar (ed.), The Horns of Hattin (London and Jerusalem, 1992),
pp. 146-67.
Pringle, D., 'A Thirteenth-Century Hall at Montfort Castle in Western Galilee',
Antiquaries Journal, 66 (1986), pp. 52-81.
Pringle, D., 'Crusader Castles: The First Generation', Fortress, i (1989), pp. 14-25.
Pringle, D., 'King Richard I and the Walls of Ascalon', PEQ, 116 (1984), pp. 133-47.
Pringle, D., 'Magna Mahumeria (al-Bira): The Archaeology of a Prankish New
Town in Palestine', in P. Edbury (ed.), Crusade and Settlement (Cardiff, 1985),
pp. 147-68.
Pringle, D., A. Petersen, M. Dow and C. Singer, 'Qal'at Jiddin: A Castle of the
Crusader and Ottoman Periods in Galilee', Levant, 26 (1994), pp. 135-66.
Pringle, D., 'Review Article: Reconstructing the Castle of Safad', PEQ, 117 (1985),
pp. 139-49Pringle, D., Secular Buildings in the Crusader Kingdom of Jerusalem: An Archaeological
Gazetteer (Cambridge, 1997).
Pringle, D., 'Survey of Castles in the Crusader Kingdom of Jerusalem, 1989: Preliminary Report', Levant, 23 (1991), pp. 87-91.
Pringle, D., 'Templar Castles between Jaffa and Jerusalem', in H. Nicholson (ed.),
The Military Orders 2: Welfare and Warfare (Aldershot, 1998), pp. 89-109.
Pringle, D., 'Templar Castles on the Road to the Jordan', in M. Barber (ed.), The
Military Orders: Fighting for the Faith and Caring for the Sick (Aldershot, 1994),
pp. 148-66.
Pringle, D., The Churches of the Crusader Kingdom of Jerusalem: A Corpus, vol i,
A-K (Excluding Acre and Jerusalem) (Cambridge, 1993); vol. 2, L-Z (Excluding
Tyre) (Cambridge, 1998).
Pringle, D., The Red Tower (London, 1986).
4O2
BIBLIOGRAPHY
403
Setton, K.M. (general ed.), A History of the Crusades, 6 vols (Madison, 1955-89).
Setton, K.M., Catalan Domination of Athens, 1311-1388 (Cambridge, Massachusetts,
1948).
Setton, K. M., 'The Catalans in Greece, 1311-1380', and 'The Catalans and Florentines
in Greece, 1380-1462', both in HC, vol. 3 (1975), pp. 167-224, 225-77.
Setton, K. M., The Papacy and the Levant, 1204-1571, vol. i, The Thirteenth and
Fourteenth Centuries (Philadelphia, 1976).
Sigonowitz, B., 'Zur Eroberung Thessalonikes im Herbst 1224', BZ, 45 (1952), p. 28.
Sinclair, T. A., Eastern Turkey: An Architectural and Archaeological Survey, 4 vols
(London, 1990).
Smail, R. C., 'Crusaders' Castles of the Twelfth Century', Cambridge Historical
Journal, 10 (1951), pp. 133-49Smail, R. C., Crusading Warfare, 1097-1193, second edition with an introduction by
C. Marshall (Cambridge, 1995).
Sterns, I., 'The Teutonic Knights in the Crusader States', in HC, vol. 5 (1985),
pp. 315-78.
Thomson, J., 'Castles in Cilicia', Geographical Magazine, 23 (April, 1951), pp. 569-77.
Thorau, P., The Lion of Egypt: Sultan Baybars I and the Near East in the Thirteenth
Century, trans. P. M. Holt (London, 1992).
Tibbie, S., Monarchy and Lordships in the Latin Kingdom of Jerusalem, 1099-1291
(Oxford, 1989).
Topping, P., 'The Morea, 1311-1364' and 'The Morea, 1364-1460', both in HC, vol.
3 (i975). PP-104-40, 141-66.
Traquair, R., 'Prankish Architecture in Greece', Journal of the Royal Institute of
British Architects, 31 (1923-24), pp. 33-48, 73-86.
Traquair, R., 'Laconia, I: Mediaeval Fortresses', ABSA, 12 (1905-6), pp. 259-76.
Traquair, R., 'Mediaeval Fortresses of the North-Western Peloponnesus', ABSA, 13
(19067), pp. 268-81.
Upton-Ward, J., 'The Surrender of Gaston and the Rule of the Templars', in
M. Barber (ed.), The Military Orders: Fighting for the Faith and Caring for the
Sick (Aldershot, 1994), pp. 179-88.
Van Milh'ngen, A., Byzantine Constantinople: The Walls of the City and Adjoining
Sites (London, 1899).
Vickers, M., 'The Byzantine Sea Walls of Thessaloniki', Balkan Studies, 2 (1970),
pp. 261-78.
Wartburg, M-L von, 'The Medieval Cane Sugar Industry in Cyprus: Results of
Recent Excavations', Antiquaries Journal (1983), pp. 298-314.
Wolff, R. L., 'Mortgage and Redemption of an Emperor's Son: Castile and the Latin
Empire of Constantinople', Speculum, 29 (1954), pp. 45-84.
Wolff, R. L, 'Politics in the Latin Patriarchate of Constantinople, 1204-1261', Dumbarton Oaks Papers, 8 (1954), pp. 225-303.
Wolff, R. L., 'The Latin Empire of Constantinople, 1204-1261', in HC, vol. 2 (1962),
pp. 187-233.
Youngs, G. R., 'Three Cilician Castles', Anatolian Studies, 15 (1965), pp. 11334.
Index
406
UNKNOWN CRUSADER
CASTLES
INDEX
407
408
UNKNOWN
CRUSADER
CASTLES
INDEX
409
41O
UNKNOWN CRUSADER
CASTLES
INPEX
411
412
INDEX
413
414
INDEX
415
4i6
U N K N O W N CRUSADER CASTLES
INDEX
417
4i8
Rafania 289
Ramlah 3
Ramon Muntaner, chronicler, 199,
233-34
Ravennika 263-64, 267, 268, 306
Raymond III, count of Tripoli, 45
Raymond, son of Bohemond III, 180
Raymond Roupen, claimant to
Antioch and Cilician Armenia, 60,
68, 82, 163, 171, 172, 180-81, 183,
186
Recordane, 54-55, 64, 82, 284, 328
n. 11
Red Tower, 82, 275
Renaud III, lord of Nephin, 60-61
Renier of Trit, lord of Philippopolis
256-57
Rey, E. G., archaeologist, 50, 51, 52
Rhodes 116, 127, 128, 130, 133, 188,
191, 232-33, 251, 267, 302;
description 265-66
Richard I, king of England, xvi, xvii,
xviii, xxi, 3, 6, 20, 40-41, 44, 53,
66, 67, 76-77, 82, 83, 89, 90,
95-96, 100-1, 106, 108, 110, 131,
219, 298, 324 n. 20
Richard, count of Cephalonia, 250
Richard of Cornwall's crusade
(1240^1) 3, 20, 22, 45, 53
Richard Filangieri, imperial marshal of
Frederick II, 10, 76, 91, 102, 111-12
Robert of Taranto, prince of Achaia,
230, 259-60
Roche Guillaume 182, 183, 185, 186,
187, 359 n. 65
Roche Roussel (Hadjar
Shoglan/Chilvan Kale) 68, 73, 182,
183, 186, 187, 359 n. 65;
description 185
Rodosto 206, 227, 233, 241
Roger de Lluria, vicar-general of the
duchy of Athens, 259
Rouad, island of, 127, 128, 129, 133,
288
INDEX
419
281-82, 283, 284, 286-87, 288, 295, Smail, R. C., historian, xviii, xx,
296, 297, 301, 303, 305, 313 n. 47;
51-52, 86
description 16-17, 314 n. 19, 321
Smyrna 188, 233, 234, 266, 305
n. Ill; 1266 siege 7, 9, 17, 36, 41, 48 Spiga (Pigae) 208; description 205
Satalia (Antalya) 128, 129, 138, 188, 267 Stenimaka 206, 256
Second Crusade (1147-49) xvi
Subeibe 49
Seleucia: see Silifke
Suda (Crete) 254
Seljuk Turks and Seljuk forces xiii,
Sykaminon 264, 267
137, 138, 140, 141, 153, 155, 168,
Syrian Gates: see Beln Pass
170, 174, 179- 80, 194, 301, 306
Selymbria 243-44
Tagliacozzo, battle of (1268), 197
Sempad, king of Cilician Armenia, 172 Tarsus 142, 143, 160, 162, 168, 169,
Sempad, Armenian Constable and
171, 176, 353 n. 73; description
chronicler, 142, 143, 163-64
162, 163-64
Serres 207, 242, 250, 251, 257, 259
Tchorlu 196, 236
Servantikar 148, 150, 151, 152, 153,
Telos 265
157, 159, 162, 166, 176, 179, 187,
Templars xvi, 3, 9, 14-17, 19, 32,
294, 299, 349 n. 4; description
33-34, 35, 38-39, 41, 43-44, 49,
145-47
51, 52, 54-55, 57, 59-61 62, 66-67,
Sherborne 276
68, 69, 70, 71, 72, 73, 74-76, 77,
Siderokastron 257
79-86, 89, 108, 113, 124, 126, 127,
Sidon 3, 18, 19, 26, 31, 45, 46, 47, 48,
130-34, 163, 171, 174, 175, 180,
51, 60, 65, 67, 69, 72, 77, 79, 80,
237, 278, 280, 281-82, 283, 285,
83, 85, 225, 229, 273, 285, 286,
286-88, 293, 295, 296-98, 304,
297, 298, 305, 330 n. 10;
305-7, 332 n.41; strongholds and
description 22-23, 28, 297
properties in the Holy Land 61,
siege weapons and engineers 8-9, 141,
80-81, 82-84; on Cyprus 131-32,
143-44, 147, 225, 286-87, 312
133; in Cilician Armenia 182-87; in
n. 35; catapults and trebuchets 8-9,
Prankish Greece 262, 263-65, 267,
268.
30-31, 32-34, 35, 37, 61, 93, 119,
141, 147, 256; mines 8, 31, 34, 141
Templar headquarters (Acre), 51, 59,
Sigouri, 107, 118, 122, 299;
61, 133, 281; (Paris) 281
description 105
Teutonic Knights xvi, 16, 18-19, 26,
Silifke (Seleucia) 14, 137, 138, 143,
36, 43-44, 61, 73, 74, 84, 132-33,
150, 152, 153, 155, 159, 170, 172,
161, 162, 170, 170-71, 172, 174,
178, 181-82, 187, 188, 299, 306;
182, 272, 273, 281, 283-84, 295,
description 179-80
296, 297-98, 299, 304, 305-7;
Sion Gate (Jerusalem) 26
strongholds and properties in the
Siquinum 188
Holy Land 61, 79-80, 83-84, 273,
Siros 231
286- 87; on Cyprus 132-33; in
Sis 138, 140, 141, 147, 148, 154, 155,
Cilician Armenia 161, 175-79, 181,
156, 157, 160, 161, 165, 168, 169,
187, 294; in Prankish Greece
173, 178, 278; description 164-65;
262-63, 267-68
1375 siege 164-65, 173
Thebes 198, 218, 224, 239, 251, 257,
42O
Trebizond 191
trebuchet: see siege weapons
Tripoli xiv, xvi, 3, 28, 39, 43, 44, 45,
49, 54, 60-62, 64, 67, 68, 77, 78,
80, 83, 260, 277, 281, 284, 285,
290, 323 n. 5; description 24,
25-26; 1289 siege 7, 24, 51, 83
Tukhlah 29, 50, 54, 82, 219
Tumlu 148, 158, 160, 168
Turris Salinarum 82
Tyre xiv, xvi, 3, 6, 24, 35, 39, 43, 45,
46, 47, 49, 54, 57, 62, 77, 78, 82,
84, 90, 91, 103, 107, 119, 144, 151,
291, 299, 305; description 23,
25-26; 1187 siege: 67, 302, 323 n. 5;
Ibelin attack on (1242) 10, 77, 326
n.59
Vagha 148, 150, 153, 155, 160, 163,
166, 168, 296; description 147-48
Vahram of Edessa, chronicler, 154,
158
Valania: see Banyas
Vardounia 217
Vasil, citizen of Tarsus, 163-64
Vasil, lord of Vaner, 170, 180
Veligosti 247
Venice xvi, 223, 228, 230, 234
Venetians xvii, xviii; in Holy Land 46,
61-62, 77, 276, 291-92, 305; on
Cyprus 90, 94, 95, 98, 104, 105,
109, 123; in Cilician Armenia 166,
292; during the Fourth Crusade
and in Prankish Greece, 191, 194,
200, 201, 204, 206, 211, 212, 22122, 228, 229-30, 231, 232-33, 234,
240, 241, 243, 250, 252, 254, 257,
259, 276, 284, 291, 292
Villehardouin: see Geoffrey, Isabelle,
William of Villehardouin
Vostitza 230, 274
William Alemn, lord of Patras, 209,
211, 214, 224, 254, 364 n. 32
INDEX
421