Jump to ContentJump to Main Navigation
Byzantium and the Crusader States 1096–1204$

Ralph-Johannes Lilie

Print publication date: 1994

Print ISBN-13: 9780198204077

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780198204077.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 30 March 2017

(p.259) Appendix I Laodicea, Antioch, and Byzantium between ad 1098 and 1105

(p.259) Appendix I Laodicea, Antioch, and Byzantium between ad 1098 and 1105

Byzantium and the Crusader States 1096–1204
Oxford University Press

We learn about the events in North Syria in the years between the conquest of Antioch in the First Crusade and Bohemond's journey to the West mainly from the works of Anna Comnena, Radulph of Caen, and Albert of Abe. Unfortunately, there are discrepancies within each author as well as between all three which make the presentation of a reasonable chronological sequence almost impossible and which, so far, cannot be satisfactorily resolved from the secondary literature. Since by far the most comprehensive source, Anna Comnena, also contains the greatest number of obscurities, we will give here a summary of the series of events in her account.

Book 11 Chapter 4

The crusader army marches through Asia Minor to Antioch and begins the siege.

Kerbogha's relieving army approaches the city.

Chapter 5

Conquest of western Asia Minor by Byzantium.

Chapter 6

Philomelium; discovery of the Holy Lance; victory over Kerbogha; march to Jerusalem; coronation of Godfrey.

Chapter 7

War between Egypt and the Kingdom of Jerusalem; battle at Ramleh; capture of many Frankish nobles and of Godfrey; Alexius secures the liberation of the prisoners (this also recounted in book 12 chapter 1); Godfrey is sent to his brother Baldwin in Edessa.

The Emperor's message to Raymond of Saint-Gilles to hand over Laodicea to Andronicus Tzintzilukes and the districts of Valania and Maraclea to the men of Eumathius Philocales (Governor of Cyprus) and himself to conquer additional places; Raymond marches with 400 men against Tortosa which he takes easily, (p.260) wins a battle against the Atabeg of Damascus, besieges Tripoli, and, with Byzantine help, builds the fortress of Pilgrim's Mountain; on receiving news of the handing-over of Laodicea, Bohemond sends Tancred with an army to besiege the city. A relief attempt by Raymond is unsuccessful. Tzintzilukes decides to hand over the city.

Chapter 8

Death of Godfrey; successor is his brother Baldwin since Raymond refuses the offer and, a little later, goes to Constantinople; description of the crusade of 1101; Raymond returns to his troops in Tripoli. He dies, naming William Jordan as his successor.

Chapter 9

News of the occupation of Laodicea by Tancred reaches Constantinople; Alexius sends an unsuccessful embassy to Bohemond; thereupon the held commander, Butumites, is sent with troops to Cilicia; finally Cilicia and Germanicea (Marash) are conquered and placed under Monastras.

Chapter 10

A Pisan fleet goes east and plunders the Byzantine coast on the way. A Byzantine squadron under Taticius (just returned from Antioch) and Landulph is sent against them, defeats the Pisans, and pursues them to Syria; an embassy to Bohemond is unsuccessful whereupon the Byzantines retire to Constantinople, the Pisans having united with Bohemond; Alexius orders the occupation of Corycus and Seleucia in Cilicia Trachea.

Chapter 11

A year later, a Genoese fleet goes east, followed by the Byzantine field commander Cantacuzenus. Laodicea and its surroundings are besieged by the Byzantines and captured, except for the citadel; Bohemond together with the troops of Tancred and Raymond (!) makes an attempt at relief but is able only to supply the citadel with fresh provisions; Monastras conquers Cilicia.

Chapter 12

Bohemond goes to Italy.

Book 12 Chapter 1

News of the battle at Ramleh (see above, 11. 7) reaches Constantinople; Alexius arranges the release of the prisoners and demonstrates in this way the injustice of Bohemond's accusations.

(p.261) Chapter 2

Cantacuzenus and Monas tras recalled from Laodicea and Cilicia to oppose Bohemond and replaced by other commanders.

Chapter 3

Tancred conquers Cilicia.

It is clear that Anna's chronology cannot be correct. The Pisan fleet must certainly be placed in 1099 and not, as Anna leads us to suppose, in 1103–4. None the less, there are a few clues which give a framework for a chronological arrangement for the events recounted by Anna. The expedition to Antioch must clearly be dated to 1097–8, and the further advance to Jerusalem 1099. Anna then recounts further events concerning Jerusalem, namely the great defeat of the Franks at Ramleh, the capture of Godfrey and of many princes, and the Emperor's efforts to get them released. She could not possibly mean the battle at Ascalon in 1099, for this was a great victory. Godfrey took part in no further great battle, for in 1100 he died. Ramleh was the setting for a whole series of battles between 1101 and 1105.1 For our purposes, the most likely was the battle of 1102. Here, indeed, the Franks suffered a great defeat in which many fell and many, doubtless, were taken prisoner. For a time the King was presumed dead and the situation was stabilized only with difficulty.2 It is quite possible that news of this kind reached Constantinople and moved the Emperor to negotiate for a ransom of the prisoners. Anna makes the double mistake of telling not only that the King was captured, but also that he was Godfrey and not Baldwin. Probably in the confusion of the battles of Ascalon and Ramleh she lost the overall view and set the defeat of Ramleh in the chronological place of the victory of Ascalon—which she does not mention.

The Emperor's letter to Raymond of Saint-Gilles is considered further below. But first, to the conquest of Tortosa. The city had been captured once already, during the First Crusade, but then given up again. From different Western sources it appears that Raymond captured it only after the crusade of 1101, that is, in 1102.3 The battle against the army of Damascus can also be dated. According to Ibn al Athir, this falls in AH 495—which corresponds to the time between 26 October 1101 and 15 October 1102. According to Anna, Raymond marched against Tortosa with 400 men; Ibn al Athir gives him 300.4 In the same year, Bohemond is said to have been freed.5 That is not quite correct since Bohemond was set free only in 1103 (that is, AH 496), but it is none the less quite close.6 Radulph of Caen also reports the march against Tripoli and gives the army 4,000 men.7 If we subtract a zero (perhaps the figure of 400 men (p.262) seemed incredible to the chronicler and he simply increased it, unless it is a question of a mistake in copying) then we arrive at Anna's figure. Raymond's expedition may thus be placed in 1102.

Anna places the siege of Laodicea by Tancred before Baldwin was elected King of Jerusalem and before the crusade of 1101. That is obviously incorrect since Tancred went north only in the spring of 1101,8 after the capture of Bohemond in the summer of 1100, which is not mentioned by Anna. Thus he can have begun the siege at the earliest in the summer of 1101. Since, according to Radulph of Caen, this lasted a year and a half,9 but must have been ended before Bohemond's return in the summer of 1103, it could not have been begun later than the end of 1101 and, at latest, the beginning of 1102. The date of the early autumn of 1101 suggests itself, when the news of the defeat of the crusading army in Asia Minor reached Antioch.10 This brings us to the years 1101–2.

In what follows, Anna returns to 1100 (death of Godfrey, election of Baldwin, Raymond's journey to Constantinople) and concludes with an account of the crusade of 1101, and Raymond's return and death, which takes place in 1105.11 Thus once more she abandons the chronological order. The news of the occupation of Laodicea by Tancred and the consequent Byzantine embassy to Antioch should therefore be dated in 1102–3, more likely in 1103, since it was addressed to Bohemond.12 Thus the conquest of Cilicia would have taken place afterwards, in 1103 or 1104.13

The events concerning the Pisan fleet take us once more to 1099, which Anna also gives.14 According to this, the occupation of Corycus and Seleucia should be placed in 1099 or 1100. According to Anna, a Genoese fleet went east a year later, as a result of which the Byzantines were able to reconquer Laodicea and Cilicia. Since Laodicea fell to the Normans at the latest in 1103, we are here concerned with events of this, or of the following, year. A connection with the previous chapter 9 is suggested. In fact, at this time, 1104, there was a Genoese fleet in the East which, among other matters, supported Raymond at the siege of Tripoli and later co–operated in the battles for Acre.15 None the less, Anna is not completely wrong since, also in 1100, a Genoese fleet went to Syria and took part in the conquest of Caesarea among other activities.16 Anna Comnena, (p.263) however, mentions only one such fleet, so the conclusion seems to be that she mistakenly combined the news of the two fleets and placed both in 1100. In so doing, apparently, she simply transfers the events of 1104 and thus, probably quite unknowingly, combines 1099 with 1104, though at the expense of one of the two Genoese fleets. In chapter 1 of the twelfth book the Princess once more takes up the events at Ramleh (1102), which she had placed before 1101, here no doubt as propaganda against Bohemond, who imputed to the Emperor actions and sentiments inimical to the crusaders and for this reason had begun the crusade of 1106–8.17

If we examine the report as a whole, two chief mistakes appear which then give rise to others. The first is in wrongly placing the battle at Ramleh before the crusade of 1101 and, in consequence, wrongly placing Raymond's expedition against Tripoli and the beginning of the siege of Laodicea by Tancred. The reason could be that, after the report of the siege of Jerusalem, Anna simply continues the story of events there and in so doing confuses the different battles of Ramleh and Ascalon and so, quite suddenly, dates the whole sequence of events of Ramleh–Tripoli–Laodicea as 1099–1100 with Godfrey as King. Actually, she is recounting events of the years 1102 and 1103, with the surrender of Laodicea to Tancred. Consequently she is forced to start again. This happens in chapter 8, but she is in great difficulty with how to fit in Raymond, since he had taken part in the crusades in Asia Minor in 1101. Logically, he cannot return to Laodicea and proceed thence against Tortosa, since that had been related in chapter 7, but he must go direct to Tripoli, to his troops there. Thereby the sequence is indeed re-established but, unfortunately, at the cost of a well-nigh endless dispute in the secondary literature about the date of the surrender of Laodicea to Byzantium and the first battles at Tripoli.

The second mistake made by the Princess lies in the misplacing of the Pisa episode. Evidently, she herself saw that this must be dated earlier and she began again with an appropriate comment though evidently she could not get the connections right. The reason may be that she had correct information about the two Genoese fleets (1100 and 1104) but, perhaps because she had introduced the Pisan fleet too late, mistakenly put these two together as one and thus restored the connection with 1104, though at the cost of one fleet. The events concerning the fleet of 1104 she places in 1100 (that year's fleet effectively disappears) but at the same time the impression is thus given that the whole sequence of chapters 10 and 11 relates to 1103 and 1104, which has occasioned a great deal of discussion among the experts.

One of the first of these, Kugler, attempted to put the misplaced chronology into order and, by a comparison with Western sources, arrived at dating the whole sequence of events from the Emperor's embassy to Bohemond to the Genoese fleet as from 1099–1100.18 To some extent he has been followed in (p.264) this.19 Kugler gives the following comparison, the Western sources, including the chronicle of Matthew of Edessa, corresponding with Anna's series:

Matthew and Western sources:

Anna Comnena:

Embassy to the crusader princes at Arqa

Embassy to Bohemond

Byzantine occupation of Marash

Byzantine conquest of Germanicea/Marash

Pisan fleet

Pisan fleet

Genoese fleet

Genoese fleet

Such a correspondence cannot be accidental. With Western sources and Matthew of Edessa in the majority, the Greek's sequence of events is accordingly to be dated 1099–1100. Chalandon has already raised doubts about the dating of the conquest of Germanicea20 and indeed we can record certain differences, whatever the admitted similarities in the two tables of events. The embassy, according to Western sources, goes to Arqa while, according to Anna Comnena, it goes to Bohemond and is concerned with Tancred's conquest of Laodicea.21 According to Matthew of Edessa, the Byzantines occupy Marash in 1099, since the city had been handed over to them in 1097 by the crusaders. That is entirely probable. We know that the crusading army on the way to Antioch went via Germanicea and why should it not have given back the city, as had been the case with the Cappadocian towns? Against this, according to Anna Comnena, Marash was conquered by the Byzantines. This, too, is probable since, in 1100, Bohemond had attacked the city without, however, being able to capture it, and had harried the neighbourhood.22 According to Radulph of Caen, Tancred, about 1101–2, reconquered Ciiicia, which the Byzantines, apparently from Corycus and Seleucia, had taken for themselves, and hence Marash may also have fallen to him once more.23 In fact, Anna gave the wrong date for Pisa24 and we saw that for Genoa we are really concerned with two events—a Genoese fleet was in the East in 1100 as well as in 1104. But Kugler's thesis fails here. For it is credible that Alexius, at the news of the fall of Laodicea, sent a last embassy to Antioch—immaterial whether to Tancred or to Bohemond25—and, after the failure of this, took the offensive. It is possible that the reason for Anna's mistake was that, while considering the news of the Genoese fleet of 1104, she suddenly remembered that there was something else (p.265) to report about Pisa. She tried, as it were, to take up the maritime aspect of the whole affair but in doing so got into great chronological difficulties which she solved, doubtless without any ill intention, by adding the second Genoese fleet to the Pisan and ignoring the fleet of 1104. But the whole sequence of events cannot thereby, as Kugler suggests, be dated to 1099–1100, but only chapter 10, the account of the Pisan undertaking and the Byzantine occupation of the harbours of Corycus and Seleucia in western Cilicia.

There remains the question of when Raymond of Saint-Gilles surrendered Laodicea to Byzantium. Let us concentrate first on Anna Comnena, leaving aside the other sources. The whole sequence (chapter 7) is related by Anna under 1099–1100, in any event before the death of Godfrey and the crusade of 1101, but the actual events, as we have seen, must be dated to 1102 or 1101. It is therefore most probable that the Emperor's letter to Raymond belongs to this time since the Count went back to Syria only in the spring of 1102.26 It might be possible, if Ramleh is dated too early, for the subsequent events to be placed actually in 1100. But this is not correct, as we have seen, since the siege of Laodicea must be dated at the latest at the end of the year 1101, while the conquest of Tortosa clearly took place in 1102. The whole sequence of events is thus moved in time. But that leads to a double difficulty. Since Raymond captured Tortosa and began the siege of Tripoli only after 1101, while Anna recounts this as before 1101, she is forced to make the Count go direct to his troops before Tripoli after his return from Syria. This would not have been necessary if he could not have started the undertaking until after 1101. Moreover, he has to be required by letter to surrender Laodicea, Valania, and Maraclea, which clearly does not make sense since the Count stayed in the Eastern Empire from spring 1100 until spring 1102. It is equally improbable for the Emperor's letter to have been sent in 1102 since Raymond had just come from Constantinople and had certainly settled everything necessary with the Emperor there. In addition, Laodicea was under siege by Tancred in 1102 and one can hardly believe that, in this situation, the Emperor would have required the Count of Toulouse to hand over the besieged city and thus leave the way clear for the Normans. Thus, whether the fact of the Emperor's written communication to Raymond is an invention of Anna's or, as in the case of Pisa, the position of the letter in the time sequence is incorrect, there would arise the paradoxical situation that Anna Comnena gets the sequence of events Ramleh-Laodicea–Tripoli–siege of Laodicea wrong, but, by placing the surrender of Laodicea to the Emperor within the framework of the series, dates this event correctly by accident. This problem is also raised in the secondary literature but has not been satisfactorily settled. Four possibilities are presented:

  1. 1. Handing-over in the spring of 1098 in connection with the negotiations before Arqa.27

  2. 2. Handing-over in 1100, before the Count's journey to Constantinople.28

  3. (p.266) 3. Handing-over in the spring or summer of 1102 on the occasion of the Count's return from Constantinople.

  4. 4. Handing-over in the course of 1101 or at the beginning of 1102, when the Count was in Byzantium.

The third possibility has already been explored and rejected.29 Let us next consider the second possibility, which assigns an embassy of the Emperor to the end of 1099 or the beginning of 1100, to which the Count responded by travelling to Constantinople. It must be noted that, according to Anna Comnena, after receipt of the Emperor's embassy, Raymond first besieged Tripoli and only later, in any case after Godfrey's death, departed to Constantinople. That is evidently not true, since Godfrey died on 18 July 110030 while Raymond, in June of the same year, was no longer in Syria which, apparently, he had already left in February.31

The question is whether, in this short time, he was able to conquer the districts of Maraclea and Valania as well as attack Tripoli. As we recall, Raymond, together with Robert of Flanders and Robert of Normandy, went back in the autumn of 1099 from Jerusalem to Laodicea, where they prevented the conquest of the city by Bohemond.32 Subsequently the Count of Flanders and the Duke of Normandy went to Constantinople, while Raymond remained in Laodicea. When Bohemond and Baldwin set out for Jerusalem in November, and met before Valania, the whole coast was still in enemy hands.33 On their (p.267) return to the north, the two princes met Raymond in Laodicea at the end of January 1100.34 It is hardly likely that, in the interim, the Count would have proceeded against Tripoli, especially during Christmasùde. There remain, then, the weeks after 26 January. In this time, Raymond would have had to have conquered Valania and Maraclea and possibly undertaken an expedition against Tripoli. The news of this would have had to have reached Constantinople, the Emperor would have had to make the necessary decisions and send the order for Valania, Maraclea, and Laodicea to be surrendered to Byzantium and even for Raymond himself to make further conquests. But that is virtually impossible. Even if we calculate very narrowly and, following Hagenmeyer, accept the end of May as the latest possible date for departure, a period of just four months remains from which at least six weeks must be deducted for the ambassador's journey. If Raymond had already set out before the end of February, as Kugler with some reason supposes, it becomes completely impossible. So, all in all, it appears improbable that the surrender of Laodicea took place in this time, the more so since Raymond left his troops and his wife behind in Laodicea. Moreover, after all that the Count had done for the Emperor, it would be most unlikely that Alexius, without any more detailed negotiations, would have demanded that he leave his territory.

There remains, however, one more difficulty. Anna Comnena says that, before his journey to Constantinople, Raymond was fighting outside Tripoli. Radulph of Caen says the same, independently of the Greek,35 which in itself speaks for the correctness of the statement.36 But there is a difference. Anna Comnena indeed puts the expedition against Tripoli before the journey to Constantinople, but in a different context. According to her, the two events have nothing to do with each other: chapter 7, battle at Ramleh–Raymond's advance against Tripoli–siege of Laodicea by Tancred. Chapter 8, death of Godfrey–Raymond refuses the crown offered to him and travels to Constantinople–crusade of 1101–Raymond goes to Tripoli and dies. Since, in point of time, the whole sequence of events is placed too late, a comparison with Radulph of Caen is not correct. In the second sequence there is, in addition, an incorrect order since Raymond had already left Syria before Godfrey's death (see above).

The chronology is misplaced by Radulph of Caen also. The series is as follows:

  1. Chapter 141: capture of Bohemond.

  2. Chapter 142: death of Godfrey.

  3. Chapter 143: Baldwin becomes the new King.

  4. Chapter 144: Tancred begins the siege of Laodicea.

  5. Chapter 145: Raymond's attempt at relief is beaten back. In this year, Tancred conquers the Turks, Egyptians, and Raymond of Toulouse. Raymond attacks (p.268) Tripoli with 4,000 men and fortifies the Pilgrim's Mountain. Since his strength is insufficient, he goes to Constantinople, where he confers with Alexius about war against Tancred; on the return journey, however, he is captured in Tarsus and eventually set free on oath.

  6. Chapter 146: after one and a half years' siege, Laodicea is captured.

  7. Chapter 147: the crusade of 1101; Bohemond is ransomed out of captivity.

  8. Chapter 148: siege of Edessa by the Muhammadans.

Here Radulph has Tancred besiege Laodicea before Raymond's march on Tripoli and his journey to Constantinople are reported. Similarly, the conquest of Laodicea is placed before the crusade of 1101, which, as we have seen, is impossible. Tancred came to Antioch in the spring of 1101, while Raymond had left Syria in the first half of 1100 and the conquest of Laodicea must be dated at 1102 or 1103, that is, after the crusade of 1101. Apart from that, Radulph reports quite precisely on the fortification of the Pilgrim's Mountain that, similarly, takes place in the period after Raymond's return from Byzantium. The chapters are at least in the wrong order. Moreover, according to Radulph, Raymond had nothing at all to do with the undertaking of 1101. The chronicler does not say, in as many words, that, before the beginning of the siege of Laodicea, Raymond marched against Tripoli, but he gives that impression. The explanation for Radulph's account may be as follows. Tancred besieged Laodicea. The city was Christian, however, and practically speaking had belonged to Raymond of Toulouse since the autumn of 1099. For all the conflicts between Raymond and Bohemond/Tancred it would have been morally disreputable to besiege the property of another crusader and it may be that the chronicler wished to erase this impression. But he could do that only if the Count of Toulouse raised no further claim on Laodicea. Even a journey of Raymond to Constantinople could hardly be so presented, since (in Radulph's view) it would have been even more immoral to attack the city while its lord was absent even if this absence was to conquer another district for himself. And this is just what happened with the undertaking against Tripoli. Here may be the explanation for Radulph's account, a wish to present in the best light the behaviour of his hero, who might well have been totally indifferent to this saving of his honour. It is clear from the events themselves that Raymond did not place a permanent garrison on the Pilgrim's Mountain, which he had fortified before his journey to Constantinople, as Radulph tells us. The Count's wife remained in Laodicea and consequently at least a part of the troops of Toulouse also.37 Tortosa was not conquered until 1102. If Raymond actually had left a body of troops before Tripoli as Anna Comnena also relates, this, given the situation, could have been only very small and as long as Tortosa remained Muhammadan could not have been reached by land except by a greater armed force, which at that time Raymond no longer had at his disposal. Thus such an action also would have been most improbable. In the few months of 1099#1100 that he remained in Laodicea, Raymond could not in fact have undertaken any action against Tripoli. Admittedly it is highly probable that he planned such an operation and went to Constantinople to seek permission and (p.269) help from Alexius. But this precludes the possibility of the Emperor's embassy mentioned by Anna. As far as this is historical it could only have been in the spring of 1098 or the year 1100.

The decisive question which will have to be settled is whether Laodicea came into the Count's possession at all during or before the siege of Arqa. Chalandon and Yewdale suppose so, and the Hills suggest it.38 During the siege of Arqa, as already explained in Chapter 1, Raymond had become lord of Tortosa and Maraclea at least.39 There is no dispute about Valania. There is only a general supposition that the city fell into the hands of the crusaders during the siege of Antioch. But this assumption cannot be proved and is not likely to be correct. Radulph of Caen, who here is our only witness, reports that the princes had withdrawn to the neighbourhood of Antioch, that is, Robert of Normandy to Laodicea,40 Robert of Flanders to the so-called ‘Vallis Comitis’ with the towns of Balena, Bathemolin, Corsebel, Barsoldan, and others, Godfrey to Iskanderun, Tancred to Harenc, Bohemond into the ‘Vallis Doxae’ (Daphne etc.), and Raymond to Rubea etc. The ‘Vallis Comitis’ to which Robert of Flanders retreated lies not to the south of Antioch, as the editors of the ‘Gesta Tancredi'’assume, identifying the town of Balena as Valania, but, on the contrary, to the north of Antioch. In consequence, Balena cannot be identified as Valania but as Balanèe—also known as Balaine/Bellanè.41

There remains the question of Laodicea. If Raymond held the city in the spring of 1099 and Tortosa and Maraclea as well, Valania may have fallen to him. But did he occupy Laodicea, as Albert of Aix states, or not? The sources are contradictory; even in the secondary literature there is no agreement.42 Let us examine the statements of the different sources and groups of sources.

The most comprehensive is Albert of Aix, who speaks of Laodicea in two places, the first during the siege of Antioch and the second occasion in relation to the siege of Laodicea by Bohemond in the autumn of 1099. The accounts are contradictory. In the first we are told how the pirate Guynemer of Boulogne (p.270) conquered and held Laodicea without informing the crusaders outside Antioch or sending them help. Eventually, it is said, he was attacked and taken prisoner by Turcopoles and soldiers of the Emperor. He was only set free from this captivity by Godfrey.43

The second account is more extensive. Guynemer is said to have assembled a fleet of pirate ships from many places, including Friesland, Flanders, and Provence, Raymond's lordship. Thereupon, it is said, he came with this fleet to Laodicea, conquered the city, and surrendered it to Count Raymond after the siege of Antioch. Guynemer is said to have then been captured by Turcopoles and soldiers of the Emperor and freed by Godfrey only much later. Raymond, it is said, as the army set off once more, for Jerusalem, surrendered the city to the Emperor in accordance with the oath taken in Constantinople. When the princes learned, on their return from Jerusalem, that Bohemond was besieging Laodicea and thus wronging the Emperor as well as the Count, they compelled him to desist.44

Albert's accounts contradict each other. The first is not concerned with Raymond. The crusaders before Antioch do not know of Guynemer's conquest of Laodicea. They learn of this only when Godfrey arranges Guynemer's liberation. Since Godfrey, according to all the other sources also, had nothing to do with Laodicea, which in fact concerns only Guynemer, Robert of Normandy, and perhaps, on account of the alleged conquest of Valania, Robert of Flanders, he cannot have arranged this release until the spring of 1099, when he went via Laodicea to the south, to the besieging army before Arqa.45 In contradiction to this, the crusaders in the second account were wholly informed of the events in Laodicea and the city was even surrendered to one of them, to Raymond himself. But even that is contradictory.

And the second account shows internal contradictions. Guynemer surrendered Laodicea to Raymond after the siege of Antioch, at the earliest, therefore, in June 1098. The report then suggests that Guynemer was captured by the Greeks and only set free by Godfrey. Raymond is said to have surrendered the city when the march on Jerusalem was resumed. Chronologically this would mean that Raymond took over Laodicea at the earliest in June 1098. In November 1098 Raymond set out from Antioch to the south and is stated to have surrendered the city to the Byzantines and subsequently evacuated his troops. Since it appears improbable that Guynemer was captured while Raymond's men, who owed the city to him, occupied Laodicea, he can have been captured only in December, to be set free by Godfrey at the beginning of March. But one can scarcely describe this period of three months as ‘a very long time’ (‘post plurimum tempus’). However if, as Kugler does,46 Albert of Aix is correlated with Anna Comnena and the handing-over placed at the return from Arqa, that is, in April 1099, there remains no time whatever for the capture of Guynemer by the Byzantines and his subsequent liberation by Godfrey, unless we suppose that the Provencals watched this affair without taking any action. (p.271) There is a further consideration which makes the occupation of Laodicea by Raymond at this time appear incredible. When Godfrey of Bouillon and Robert of Flanders marched south in the spring of 1099, they besieged Gibel (Jabala), a city about twenty kilometres beyond Laodicea, for a week.47 If, as Kugler supposes, Raymond had been at this time in possession of Laodicea and Valania, as well as of Tortosa and Maraclea in the south, it would have been unlikely that Jabala alone could have held out.

David is most probably correct when he supposes that Albert of Abe, in his second account, was looking for a reason why Raymond should attack Bohemond for the benefit of the Byzantines.48 Since the chronicler takes no account of the part played by the Duke of Normandy in Laodicea, one may even suppose that there is here a confusion between the two princes, the more so since Albert may have been misled by later events involving Raymond and Laodicea into backdating the beginning of this connection to the time of the siege of Antioch.49

We have not made a comprehensive examination of the other sources since, for our problem, they are relatively unimportant. The best summary, which, with a few deviations, is followed here, is that of David.50 The ‘Gesta Tancredi’ of Radulph of Caen is the most important source.51 According to this, Englishmen in the Emperor's pay occupied Laodicea. Aware of their weakness, they called on the crusaders for help, though not on Raymond, but on Robert of Normandy. He answered the call, spent some time in Laodicea (winter 1097–8), but finally listened to the complaints of his comrades and went back to Antioch.52 A certain problem arises here, namely that Edgar Aetheling, the leader of these Englishmen, can be shown to be in Scotland in March 1097 although in the winter of this year he must already have been in Laodicea, which presupposes an extraordinarily rapid journey, since he also entered the Emperor's service.53 (p.272) The difficulty might be solved if, though Edgar Aetheling in person did not, indeed, leave Scotland until the spring of 1097, his fleet had preceded him. The negotiations in Constantinople would not have required much time since Englishmen had served in the Emperor's army for some years so that Edgar would not have found the idea strange.54

It is most probable that Guynemer of Boulogne first took Laodicea but was then overcome by the Englishmen in the service of the Emperor, perhaps on his orders. They then called on the Duke of Normandy, whom they knew, for help, since they felt themselves to be too weak on their own. He came, about the turn of the year 1097–8, but then went back to Antioch and the Englishmen and the Byzantines took over the city once more. This does not exclude the possibility that Robert returned once or twice in 1098. Since he had no further interests in the East, there were no difficulties over the handing back.55 In this way the imprisonment of Guynemer may be explained. The Duke, who had been called on not by him but by the Byzantines, was not interested in him. Godfrey arranged his release in the spring of 1099.56 The city remained with the Byzantines while the remaining coastal towns, that is, only Maraclea and Tortosa, were surrendered at the advance of the crusaders and only on Raymond's return were they captured and held for any length of time.57 That Raymond had nothing whatever to do with the events appears from the fact that he did not take the coast road, which would have been the nearest if he had maintained a garrison in Laodicea, but marched to the east of the mountains and reached the coast again only at Arqa.58

Thus, of the possibilities considered, only the fourth remains, that is, the surrender to Byzantium during the course of the year 1101, or at the beginning of 1102, when Raymond was in the Eastern Empire. At first, Anna Comnena's statement that Alexius, in writing, required the Count to surrender Laodicea, Vaiania, and Maraclea to him appears to argue against this hypothesis. In 1101–2 this is not possible, since the Count was then in Byzantium. But this objection disappears when we reflect that Anna transfers the whole sequence of events concerning Ramleh–Tripoli–Laodicea from 1102 to before 1101. Since she—correctly—connects the march against Tripoli with the surrender of Laodicea, she is in fact compelled to let the Count receive the relevant written embassage in order to account for his actions. If the whole episode had been correctly dated, a letter would not have been necessary. Perhaps, even, she found in her notes, or whatever she used as a draft (the Alexiad being much too (p.273) detailed and exact to be a simple work of memory, in spite of the little mistakes the Princess made from time to time), an indication of the existence of such a letter which, however, had not been addressed to Raymond of Saint-Gilles, but to Eumathius Philocales, the Governor of Cyprus, and to Andronicus Tzintzilukes. It might well be possible that Anna, who, in any case, was in difficulties on account of her mistaken chronology, simply decided that there must have been such a letter to the Count of Toulouse.

When could such a surrender have taken place? The exact point in time can no longer be established; none the less, the second half of 1101 or even the beginning of 1102 suggest themselves. After the defeat in Asia Minor of the crusaders led by Raymond, and his return to Constantinople, the Count found himself in an unenviable situation. For a short time at least, he was out of favour with the Emperor.59 He had to rely on Byzantine support if he wanted to proceed against Tripoli, since his forces were insufficient to sustain a war for any length of time. Apart from that, he had hardly enough troops to hold Laodicea and make war against Tripoli; Alexius for his part was interested in having an ally dependent on him in Tripoli as well as having a base in Laodicea against Bohemond or Tancred. So, in the time suggested, they may both have come to an agreement, whereupon the order was sent to Philocales and Tzintzilukes to take over the city or at least to get everything ready for this to be done, while Raymond himself followed more slowly with the participants in the operation of 1101 and eventually was imprisoned in Tarsus by Tancred and released again on oath. One might argue that Alexius could hardly have dispensed with the presence of the Provençal troops as long as Laodicea was besieged by Tancred. Yet we do not know exactly when Tancred began this siege. Possibly this was later. Apart from that, it seems, in any case, incredible that Tancred actually conducted such a close siege for a year and a half, as his eulogist Radulph of Caen claims. The Antiochene army was strong indeed, but hardly strong enough to besiege Laodicea and, simultaneously to fight all the rest of the battles in 1101 and 1102 which Tancred's activities had occasioned in practically every quarter of his region and beyond.60 They are astonishing enough as it is. Moreover it is not very likely that Tancred attacked Laodicea energetically while the Provençal army and Raymond' wife were still within the walls. On the other hand, a loose blockade would be possible, drawn closer only when Raymond and his men had left the district in the direction of Tortosa. The beginning of Tancred's action, as Kugler has suggested, may well be placed with the arrival of the news of the disaster of the crusade of 1101 in Asia Minor, which excluded any possible threat to Antioch from that side.61

With this timing, the surrender of Valania and Maraclea to Byzantium is (p.274) explained. In 1099 a conquest would hardly have been possible, on account of time, but why should not Raymond's followers, during his two years' absence, have brought the district around Laodicea under their control? They did not, on the other hand, risk any greater undertakings, as is shown by the example of Tortosa, which was taken only in 1102, by Raymond himself. In any case, both towns did not long remain with Byzantium, but were lost once more in the continuing battles with Antioch, not, however, to the Normans, but to the Muhammadans from whom in 1109 Tancred finally seized them.62

With this we can establish an approximate series of events in North Syria, which appears thus.63


Surrender of Germanicea/Marash to the Byzantines (Matthew of Edessa 50).


Siege of Antioch (Anna Comnena 11. 4–6).


Siege of Arqa.


Byzantine embassy to the crusader princes before Arqa (Raymond d'Aguilers 125 f.; William of Tyre 7. 20, 307 f.).

Capture of Jerusalem (Anna Comnena 11. 6).

Arrival of the Pisan fleet before Laodicea (Anna Comnena 11. 10).

Unsuccessful Byzantine embassy to Bohemond (ibid.).

Siege of Laodicea by Bohemond (ibid.).

Relief of Laodicea by the princes returning from Jerusalem (cf. Hagenmeyer, Chronologie, no. 430).

Byzantium occupies Corycus and Seleucia (Anna Comnena 11. 10).


Bohemond and Baldwin go to Jerusalem and back (cf. Hagenmeyer, Chronologie, nos. 433, 436).

(p.275) 1100

Byzantium reconquers Cilicia (time uncertain, end of 1099 first half of 1100, Radulph of Caen 706).

Raymond goes to Constantinople (Anna Comnena 11. 8; Radulph of Caen 708).

Bohemond attacks Marash unsuccessfully and harries the neighbourhood (Matthew of Edessa 50 f.).

Bohemond captured before Melitene (Matthew of Edessa 51 f.; Radulph of Caen 704, cf. Kugler, Boemund und Tankred, 16).

Arrival of a Genoese fleet in Syria (‘Liberatio Orientis’ 117).

Death of Godfrey; Baldwin becomes King of Jerusalem (Anna Comnena 11. 8).


Tancred arrives in Antioch to take over the regency (cf. Kugler, Boemund und Tankred, 20).

Unsuccessful crusade to eastern Asia Minor (Anna Comnena 11. 8; Radulph of Caen 709).

Byzantium takes over Laodicea, Valania, and Maraclea from Raymond of Saint-Gilles (Anna Comnena 11. 7).

Tancred begins the siege of Laodicea (Anna Comnena 11.7; Radulph of Caen 706–9).

Tancred conquers Cilicia and Marash (time uncertain: 1101 or 1102, Radulph of Caen 706).


Raymond, on the return journey from Constantinople to Laodicea, is captured by Tancred in Tarsus and released on oath (Radulph of Caen 708; Albert of Aix 583; cf. Hill and Hill, Raymond IV of Saint-Gilles, 133–5).

Raymond captures Tortosa and begins the siege of Tripoli (Anna Comnena 11. 7–8; Radulph of Caen 707 f.; Ibn ai Athir ‘Kamel-Altevarykh’, 211 f.).

Battle of Ramleh (Anna Comnena 11. 7).

Alexius tries to secure the release of the Franks captured in the battle of Ramleh (time uncertain, between 1102 and 1105, Anna Comnena 11. 7; 12. 1).

Unsuccessful attempt by Raymond to relieve Laodicea (1102 or 1103) (Anna Comnena 11. 7; Radulph of Caen 707, might refer to capture of Raymond on the return from Constantinople).


Capture of Laodicea by Tancred (Anna Comnena 11. 7).

Bohemond returns from captivity (Radulph of Caen 703; cf. Kugler, Boemund und Tankred, 24).

Unsuccessful Byzantine embassy to Bohemond (time uncertain; 1103 or 1104, Anna Comnena 11. 9).

(p.276) 1104

Arrival of a Genoese fleet in Syria (Anna Comnena 11.7).

Cantacuzenus recaptures Laodicea, but not the fortress, which the Normans continue to hold (Anna Comnena 11. 11).

Butumites and Monastras conquer Cilicia and Germanicea/Marash (Anna Comnena 11. 9–11).

Marash is recaptured by Jocelin of Courtenay (Matthew of Edessa 75).

Defeat of the armies of Antioch and Edessa by the Seljuks at Harran (cf. Runciman, Crusades, ii. 40 ff.).

Unsuccessful attempt by Bohemond to retake Laodicea (Anna Comnena 11. 11).


Bohemond goes to Italy (Anna Comnena 11. 12).


Death of Raymond of Saint-Gilles (Anna Comnena 11. 8).

After 1105

Cantacuzenus and Monastras are recalled from Laodicea and Cilicia (Anna Comnena 12. 2).

Tancred reconquers Laodicea and Cilicia (Anna Comnena 12. 2).

In general there are only a few spectacular changes. None the less, it appears, which is important, that Bohemond began the hostilities against Byzantium while the first Byzantine countermeasures were taken only in 1100 and continued in 1104. After a short episode of about two years from 1097 to 1099, Byzantium retook Laodicea only in 1101–2 and only at this time made binding agreements with the Count of Toulouse. In the same way, in 1099 as well as in 1103–4, a peaceful agreement with Bohemond was sought and only after that was military activity developed. The moderation of the Emperor's policy in the years 1099 to 1105 is just as obvious as the aggressive nature of the policy of the Normans of Antioch.


(1) Cf. Röhricht, Jerusalem, chs. 1–3; Runciman, Crusades, ii. 71 ff.

(2) Cf. Röhricht, Jerusalem, 36 ff.; Runciman, Crusades, ii. 76 ff.

(3) Cf. Runciman, Crusades, ii. 58 ff.; Hill and Hill, Raymond IV de Saint-Gilles, 136 ff.

(4) Ibn al Athir, ‘Kamel-Altevarykh’, 211 f.

(5) Ibid. 212.

(6) Cf. Röhricht, Jerusalem, 45; Runciman, Crusades, ii. 39; Yewdale, Bohemond, 97 (only after 1104); Kugler, Boemund und Tankred, 24.

(7) ‘Gesta Tancredi’ 145, p. 707.

(8) Cf. Kugler, Boemund und Tankred, 19 f., 21; Röhricht, Jerusalem, 18.

(9) ‘Gesta Tancredi’ 146, p. 708.

(10) Cf. Kugler, Boemund und Tankred, 23; Chalandon, Comnène, i. 223.

(11) Cf. Runciman, Crusades, ii. 60 f.; Hill and Hill, Raymond IV de Saint–Gilles, 140.

(12) Kugler, Boemund und Tankred, 60 n. 7 dates it in 1099, but cf. in the following; Chalandon, Comnène, i. 223; the dating in the second half of 1103 is not completely certain although Bohemond is the addressee. Since Anna Comnena knows nothing of his imprisonment from 1100 to 1103 she must in any case send the embassy to Bohemond as she also attributes the second siege of Laodicea to him, which Tancred undertook on his orders.

(13) Thus Chalandon, Comnène, i. p. xviii.

(14) Anna Comnena 11. 10, pp. 41 f.; ‘As the Franks set out towards Jerusalem to conquer the cities of Syria…’ Cf. Kugler, Boemund und Tankred, 59 f. n. 7.

(15) ‘Liberatio Orientis’ 121; cf. Röhricht, Jerusalem, 46 f. (1103–4); Chalandon, Comnène, i, p. xviii dates this fleet in 1104 in contrast with Kugler, Boemund und Tankred, 60 n. 7; cf. also Favreau-Lilie, Italiener, 100 ff.

(16) ‘Liberatio Orientis’ 117; cf. also Favreau-Lilie, Italiener, 79 ff.

(17) Dölger, Regesten 1216 (after 25 May 1102) and 1220 (c. 1104) assumes two embassies, presumably because Anna gives two different names as envoys (Bardales, Nicetas Panukomites), but it seems to me that the second mention is too strongly determined by the propaganda against Bohemond to justify a separate new embassy. Perhaps the embassy was led by two men, which could explain the two different names.

(18) Kugler, Boemund und Tankred, 59 f. n. 7.

(19) Dölger, Regesten 1211 (Feb. 1099) and 1212 (c.Mar. 1099); Yewdale, Bohemond, 77 ff.; in part also Chalandon, Comnène, i, pp. xviii and 211 f.

(20) Chalandon, Comnène, i, pp. xviif.; Chalandon also gives general dating though without any more precise discussion of the sources. He recognizes correctly the identification of Ascalon-Ramleh, Tancred's siege of Laodicea, the Genoese fleet of 1104, and the incorrect placing of the Pisan fleet, but overlooks the fact that Anna in no way mixes the dates up ‘chaotically’ but keeps the correct order throughout her incorrect chronology which leads then inevitably to mistakes of fact.

(21) Kügler, Boemund und Tankred, 60 n. 7 explains this statement by the fact that Anna, among the different sieges of Laodicea, simply lost the overall view.

(22) Matthew of Edessa 50 f.

(23) ‘Gesta Tancredi’ 143, p. 706.

(24) Cf. Kugler, Boemund und Tankred, 60 n. 7; Chalandon, Comnène, i, p. xviii.

(25) Cf. above, n. 12.

(26) Cf. Kugler, Boemund und Tankred, 23; Hill and Hill, Raymond IV de Saint-Gilles, 133 f.

(27) Cf. Kugler, Boemund und Tankred, 59 f. n. 7; Yewdale, Bohemond, 78 f.; Chalandon, Comnène, i, pp. xiii and 212.

(28) David, Robert Curthose, 33; Hill and Hill, Raymond IV de Saint–Gilles, 130 and 127 solve the problem in the neatest fashion: on the one hand they support a handing-over at this time, on the other, however, they assume that Raymond gave Laodicea to Byzantium in 1098 or 1099 (before his advance to Jerusalem). They indeed refer to Albert of Aix 6. 55, 59, 60, pp. 501 ff., but this solution appears somewhat too elegant since they do not take sufficient account of the other sources (cf. below).

(29) On the first possibility cf. also above, Ch. 1, ‘The Evidence of the Sources: Bias and Objectivity’.

(30) Cf. Hagenmeyer, Chronologie, no. 482.

(31) In June 1100 the Count met the Venetian fleet at Cyprus, cf. Hagenmeyer, Chronologie, no. 464; on the basis of the reports of Matthew of Edessa and WUliam of Tyre, Kugler, Boemund und Tankred, 63 n. 18 assumes a departure before 25 February. The problem can no longer be solved on the basis of available information. In any case it is certain that the Count left Syria in the first half of 1100.

(32) Hagenmeyer, Chronologie, no. 430.

(33) Ibid., no. 433; that the coastal region at this time was still hostile is confirmed by William of Tyre 9. 14, p. 386. When Bohemond joined Baldwin outside Valania for the march on Jerusalem he camped, as Fulcher 1.33, p. 327 expressly states, not in, but outside the city; the ‘Gesta Tancredi’ 107, p. 681 similarly confirm that the coastal cities in 1098–9 were still Muhammadan; Raymond sent his chaplain Arnulf by ship from Arqa to Antioch in spite of the dangers: ‘propter navalia hostium, Tortuosae, Eracliae, Valoniae, Gibel, tandem Laodiciam pervenit, inde Antiocham per multa pericula, paucis comitantibus, excurrit.’ According to the ‘Liberatio Orientis’, at the time of the siege of Antioch Maraclea as well as Laodicea was Greek while Valania belonged to the Muslims. If this report is accurate we should have a further proof of the view expressed here albeit with the reservation that, at the time of the siege of Arqa, Maraclea had been given temporarily to the crusaders by the Greeks. In such a case, restoration to Byzantium, though reported by no source, would not be improbable. It seems incredible that the Greek garrison of Laodicea was strong enough to exercise effective control over Maraclea as well. In any case it did not remain Greek. At the latest during the siege of Laodicea by Bohemond it must have fallen once more to the Muhammadans. Then probably Raymond reoccupied it when he settled in Laodicea.

(34) Hagenmeyer, Chronologie, no. 445 (c.Jan. 1100).

(35) ‘Gesta Tancredi’ 145, pp. 717 ff., Ekkehard of Aura, ‘Chronicon’, 224, also cited in this context, makes no contribution to the question (cf. also Kugler, Boemund und Tankred, 67 n. 37).

(36) Thus too Röhricht, Jerusalem, 34; Chalandon, Comnène, i. 220; Hill and Hill, Raymond IV de Saint–Gilles, 128 f.; Kugler, Boemund und Tankred, 63 n. 18 on the other hand rejects such a supposition.

(37) Fulcher 1. 32, pp. 320 f.

(38) Chalandon, Comnène, i. 210; Yewdale, Bohemond, 78 f.; Hill and Hill, Raymond IV de Saint-Gilles, 127; David, Robert Curthose, 235 ff. rejects Albert of Aix's assertion.

(39) Cf. above, Ch. 1, ‘The Negotiations before Arqa’.

(40) ‘Gesta Tancredi’ 58, p. 649.

(41) On the march to Arqa, Robert of Randers and Godfrey of Bouillon together besieged Gibel (Jabala), situated between Laodicea and Valania, which was thus still Muhammadan. This also argues against an earlier conquest by Robert of Flanders; cf. Hagenmeyer, Chronologic no. 358; on the identification of Balena, named in the ‘Gesta Tancredi’, with Balanèe to the north of Antioch, cf. Cahen, Syrie du nord, index s.vv.; finally Deschamps, La Défense du comté de Tripoli et de la principauté d'Antioche, index s.vv. and vol. of Plates, General Map, Grid D 2/3.

(42) Supporting this: Chalandon, Comnène, i. 209 ff.; Hill and Hill, Raymond IV de Saint-Gilles, 127; Yewdale, Bohemond, 78 f.; against: Runciman, Crusades, i. 255 f., 272; David, Robert Curthose, 235 ff. Later (240 n. 41) David adds that there could have been a dose connection between Raymond and Laodicea, based on Raymond d'Aguilers 128, according to which the Count sent the brother of Adhemar of Le Puy to Laodicea to fetch thence the dead Bishop's cross and field chapel. This reference can scarcely be considered definite since we do not know how and when the things got there. Possibly they were taken back to Adhemar's home and were for that reason in Laodicea, then the crusaders' main post.

(43) Cf. David, Robert Curthose, 239 f.; Runciman, Crusades, ii. 255 f.; the account by Guibert of Nogent 7. 37, p. 254, which Runciman follows, sounds however very much like a legend and is contradicted by the Duke's later conduct.

(44) Ibid. 6. 55, pp. 500 f.

(45) Cf. Hagenmeyer, Chronologie, nos. 356–7.

(46) Kugler, Boemund und Tankred, 59 f. n. 7; similarly Chalandon, Comnène, i. 211; Yewdale, Bohemond, 78 f.

(47) Cf. Hagenmeyer, Chronologie, nos. 357–9.

(48) Thus David, Robert Curthose, 238.

(49) Cf. in addition above Ch. 1, ‘The Siege of Antioch’.

(50) David, Robert Curthose, 230–44; the questions are also discussed in Chalandon. Comnène, i. 208 f.; Runciman, Crusades, i. 255 f. n. 2; Hill and Hill, Raymond IV de Saint-Gilles, 127 do not go further into the question but follow Albert of Aix 6. 55–9, pp. 501–3.

(51) ‘Gesta Tancredi’ 58, p. 649.

(52) Cf. Hagenmeyer, Chronologie, no. 215 (c.mid-Dec. 1097).

(53) This led Runciman to place Robert's stay in the summer of 1098, as well as the fact that Radulph of Caen puts Robert's visit at the same time as the flight of Stephen of Blois; since none the less a few days later Robert took part in the battle against Kerbogha, his stay in Laodicea would be best dated in September 1098. In his reasoning Runciman overlooks two things: first that Radulph states that after three warnings the crusaders brought Robert back to Antioch under threat of excommunication, which would not have been possible in the autumn of 1098 when all the princes had left the city and, in addition, Robert was even occupying Laodicea as a base for the further advance to the south. During the famine of the winter of 1097–8 on the other hand it would have been quite probable. Secondly Radulph is here putting wrong dates to a whole series of events, putting all the various conquests of the princes in the environs of the city at this time, before the conquest of Antioch, when in fact they took place only after the victory over Kerbogha. It appears that the chronicler, in a kind of parenthesis, after the account of Robert's Laodicea episode, simply added the similar undertakings of the other princes without regard for the actual order of events.

(54) Cf. Runciman, Crusades, i. 228 n. 1.

(55) Cf. David, Robert Curthose, 239 f.; Runciman, Crusades, ii. 255 f.; the account by Guibert of Nogent 7. 37, p. 254, which Runciman follows, sounds however very much like a legend and is contradicted by the Duke's later conduct.

(56) Or was it to do with an intervention by Robert which Albert of Aix—either mistakenly or in order to glorify his hero—attributes to the Duke of Lorraine?

(57) See also above, Ch. 1 n. 198.

(58) That, at the time of the conquest of Antioch, Laodicea was Byzantine is confirmed by the ‘Liberatio Orientis’ 114 f., which states that at this time it was under the command of Eumathius Philocales (the Governor of Cyprus). That would be possible since Robert of Normandy had already given the city back to Byzantium at the beginning of the year; cf. David, Robert Curthose, 236, 240.

(59) Albert of Aix 8. 24, p. 574; Hill and Hill, Raymond IV de Saint-Gilles, 133 reject the idea of a disagreement between the two princes since the Emperor later once more supported Raymond. But that overlooks the difference between the—perhaps brief—personal disappointment of the Emperor and the demands of his policy in North Syria which made an accommodation with the Count of Toulouse imperative.

(60) Cf. Röhricht, Jerusalem, 41; somewhat more limited Kugler, Boemund und Tankred, 23 f. and 67 n. 37.

(61) Cf. Kugler, Boemund und Tankred, 23.

(62) Cf. ibid. 73 n. 24; Röhricht, Jerusalem, 80.

(63) The following source references are not intended to be complete but should give a quick general view of the sources dealt with in this chapter. Otherwise reference should be made to the corresponding sections in the first two chapters.