warburg750021

The Arts in the Latin Kingdom of JerusalemAuthor(s): T. S. R. BoaseReviewed work(s):Source: Journal of the Warburg Institute, Vol. 2, No. 1 (Jul., 1938), pp. 1-21Published by: The Warburg InstituteStable URL: http://www.jstor.org/stable/750021 .Accessed: 22/04/2012 04:43

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

The Warburg Institute is collaborating with JSTOR to digitize, preserve and extend access to Journal of theWarburg Institute.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=warburghttp://www.jstor.org/stable/750021?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsp

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM

By T.S.R. Boase

In I883 Prutz published his lengthy Kulturgeschichte der Kreuzziige. Since Wilken's pioneer work in 1807, the history of the crusades and the medi- aeval Levant had been industriously rewritten. Michaud had added the full romantic flavour and further study of the Arab authorities, Heyd had defined the economic background in his Geschichte des Levantehandels. Prutz sought to add an exhaustive account of the thought and manners of the Latin kingdom, and his book is an interesting example of the methods of his time. The greater part of it is concerned with a comparative examination of Christianity and Mohammedanism as the background of the two civilizations, and the effect of such differing premises on the general structure of society. He saw clearly some of the points where crusader and Arab most misconceived one another, and he searched the written evidence skilfully for examples. The visual arts receive only a brief treatment. The scheme of the book had taken form when Prutz was attacking a fantastic attempt to find Barbarossa's bones in Tyre, and he was sceptical about archaeological results and the attribution of dates and authorship to Syrian ruins. The mind of man as reflected in his handicraft was still in 1883 a comparatively obscure subject of study, and one which a Kulturgeschichte could deal with in some twenty pages out of its quota of six hundred and forty two. De Vogifi had published in I86o his Eglises de la Terre Sainte and Rey in 1871 his Architecture Militaire des Croises en Syrie, books which mark the transition from travellers' accounts to systematic investigation; but photography had not as yet developed sufficiently to provide readily filed data; and much of the evidence was still inaccessible, forbidden to Christians as was the Haram El-Khalil at Hebron, or dangerously remote and brigand-infested. Even under these conditions much valuable work was done in the early years of this century,1 but it was not till the French and British mandates were established that new possibilities of research were opened up, and were by the terms of the mandate very definitely encouraged.2

The antiquities envisaged were in Palestine mainly biblical. The biblical interest has in fact in that country been at times somewhat exclusive, and has not readily admitted the claims of other subjects of research. But crusading monuments shared in the general facilities for new archaeological investigation, and in Syria, if the splendours of Baalbek rightly became

1 Particularly noteworthy are: Dussaud, "Voyage en Syrie," Revue Arche'ologique, 1896- 97; Clermont-Ganneau, Archaeological Resear- ches in Palestine, 1899; M. van Berchem, Voyages en Syrie, 1913-15.

In 1912 RR. PP. Vincent and Abel began their great work Jirusalem. (Paris 1912-26).

R. P. Meistermann, Guide de la Terre Sainte (Revised edition 1923) summarised the author's researches, published in detail in various pamphlets.

2 The British Mandate Article 21 provides

for "the execution of a law of antiquities" which includes protective measures, and requires qualifications before authorisations for excavation are granted. "The Admi- nistration of Palestine, however, shall not, in granting these authorisations, act in such a way as to exclude scholars of any nations without good grounds." Similar measures were laid down for Syria. Cf. F. Anus, "La protection des Monuments Historiques en Syrie et au Liban." Syria, XIII, 1932, pp. 293-99-

I

2 T. S. R. BOASE

the first charge for preservation, propaganda encouraged crusading researches as throwing light on the first Frankish settlement, of which the mandate was represented in some sense as a revival. "Sur les traces de nos soldats marchent toujours nos savants" : with such words was introduced the chief product of this period of expanded studies, Camille Enlart's two volumes on the religious and civil architecture of the crusading kingdom. Enlart was already an elderly man when he undertook the task, and his journeys in Syria are a striking tribute to his continued energy, but he did not live to see his book through the press. Shortly before his death he chose M. Paul Deschamps as his successor in the task of dealing with the military monuments of the crusaders, and Deschamps' detailed study of Crac des Chevaliers appeared in 1934. A second volume, a wider survey of the subject, is promised and said to be well advanced.

The ecclesiastical organisation of the crusading states was grouped under the two patriarchates of Antioch and Jerusalem. At its fullest development that of Jerusalem comprised four archbishoprics and thirteen suffragan sees, that of Antioch, where the ancient ecclesiastical organisation was never adapted to new needs, six archbishoprics and four bishoprics.2 Relative to their territorial basis, diocesan divisions were therefore numerous and tended to the multiplication of churches, a tendency which the frequency of sites of pilgrimage encouraged. The various monastic orders also sought to be represented in the Holy Land and added their buildings to the already crowded ecclesiastical map. The Italian city states, wherever they had settlements of any size, enjoyed diocesan immunities and built churches dependent on their home metropolitans. In the castles of the military orders, the chapels were large and rivalled in scale many of the churches. Acre was said in the thirteenth century to have had forty Latin churches. And as background were the places of worship of the earlier Christian inhabitants, with their bewildering diversity, Greeks, Armenians, Jacobites, Maronites.

Of this vast scheme of church building little today remains. Much must have been work of a very minor nature, much of it, particularly in the North, adaptation of pre-existing edifices. The architectural history of the crusades is in fact one of destruction as well as creation, and they often pulled down to rebuild, with little regard for the merits of that from which they quarried. The preservation of the Basilica of Bethlehem in the form given it under Justinian has few parallels except in the case of small village churches. The famous proverb quoted by William of Tyre's translator "Chastel abatuz est demi refez" 3 was freely practised with regard to churches also, and if a Byzantine basilica was in need of repairs, as they mostly were, the crusaders preferred drastic remodelling along western lines.

1 Paul Leon, introduction to C. Enlart, Les Monuments des Croisis dans le Royaume de Je'rusalem. Architecture Religieuse et Civile, 2 vols. and two atlases, Paris, I928.

2The loss of Edessa and the re-occupation

of Tarsus and Mamistra by the Armenians early reduced these numbers.

3 Recueil des Historiens des Croisades. Hist. Occ., I, p. 697.

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 3

A similar treatment has been given to their own works : many were pulled down in the early days of the Mohammedan reconquest and were left like the church at Nazareth "quasi totam dirutam," 1 while their stones served other purposes, as at Lydda where Beibars built a bridge out of material from the church of St. George,2 whose cult had stood so high in crusading devotions. Those which survive do so for the greater part transformed into mosques. It is thus that Hebron, Ramleh, Beirut and Tarsus retain their cathedrals : Sebastyah, the great shrine of the Baptist, was for a time a mosque, but even so was allowed to fall into ruins, and the mosque was not rebuilt till a few years before the war, a small building which completed the ruin of the apses, but still left standing part of the walls of the nave to reveal the extreme excellence of their masonry. St. Anne at Jerusalem still carries on its fagade the inscription placed there by Saladin when he turned it into a school, and it was not till 1856 that it became a church once more, when the Sultan, grateful for French help in the Crimea, presented it to the Empress Eugenie : Giblet preserved its main structure and its beautiful adjoining baptistery as a Maronite Church: Notre Dame de Tortosa, the most perfect of them all, first a mosque, then a store-house, has now been cleared of surrounding obstructions, reinforced where the structure seemed in danger, and can show clearly its claim, with its foliage capitals and the marvellous cutting of its yellow stone, to be the most beautiful Latin church in Syria. The Church of the Holy Sepulchre, destroyed by riot and rebuilding, still stands with its faqade singularly unaltered, but it stands all too insecurely.3 These survive with some thirty or so minor churches, many of which are mainly Byzantine in construction, others mere ground plans, perceptible amongst later conglomerations of buildings. The rest are gone, destroyed by many different fates. Of the buildings of Antioch, the most complex civilization of the crusading settlement, hardly any traces are left : the great walls of Justinian, which resisted the first crusaders, were almost intact till 1835, when Ibrahim Pasha began to quarry stone from them, leaving only the foundations. Of its great cathedral of St. Peter, the site is not even known. Certainly here Byzantine influence was at its strongest, and it is here probably that the Franco-Byzantine synthesis was most fully worked out. Raoul of Caen keeps for us, with a surprising comparison, the splendours of this city as it appeared to the early crusaders :

Templa, quibus visis stupuisset pictor Achivus, Auri fusor Arabs, sculptores Scottus et Anglus.4

But Antioch is built on a hill-side in an earthquake zone, and what the

1 Ricold of Mont-Croix (1294) in Laurent, Peregrinatores medii aevi, p. Io7.

2 Clermont-Ganneau, "Le pont de Beibars a. Lydda." Rec. d'Arch. Orient., I, I888, p. 262.

3 As a result of the earthquake in i927 the crusaders' dome had to be removed. At present the building is in imminent danger, and large scale works are required.

4Gesta Tancredi, LXXVII. (Rec. Hist. Crois. Occid., III, p. 661). Typically this passage is commenting on destruction caused by the siege. Raoul tells us he was born at Caen: does his remark reflect Norman admiration for Anglo-Saxon sculpture? It would be a point of view thoroughly in accordance with recent research on the subject.

4 T. S. R. BOASE

infidel left standing, nature has overturned and buried.' In Jerusalem, in 1898, the Germans rebuilt St. Mary the Great as the Lutheran Church of the Redeemer and in 1906 the Greeks destroyed St. Mary of the Latins to build a bazaar : in 1916 the British shelled the cathedral at Gaza. So the tale goes on, yet even now this strip of coastal territory is so rich in remains of all periods that no government can hope to undertake the much needed preservation that will save them from further decay. All that can be done, and the French have led the way admirably in doing it, is that they should be systematically described and photographed.

In plan the crusading churches are remarkably uniform. The more important buildings consist of a nave and two aisles, the smaller churches of a nave only. The East end is composed of three apses, frequently enclosed in a rectangular chevet. There is no ambulatory, with the exception of the Church of the Holy Sepulchre, a special instance in every way. The naves are covered with barrel vaulting, the aisles with quadrapartite groined vaults. There is no triforium, and the window spaces are small throughout, as was natural in the heat and bright light of Palestine. The type approximates to the Auvergnat and Burgundian churches of the late eleventh and early twelfth centuries. The Byzantine basilica, the ordinary type of ecclesiastical architecture that the crusaders found in these districts, was hardly ever copied by them." They brought their own familiar notions, and, whoever were the builders, the designing minds were Western.3 Only in smaller works, the domed baptistery at Giblet, the rotunda of the Ascension and the octagons of the Temple Baptistery and the Throne of Jesus, do they show a feeling for local forms, the small domed tombs of Islam, which in their turn go back to Hellenistic models such as the so-called Tomb of Absalom in the valley of Kidron.

One feature in their building has occasioned some discussion. The

1 Recent excavations there have been concerned as yet with sites of the Greeco- Roman period only.

2 The cathedral at Hebron seems to have approximated to the basilican form, though with Gothic vaulting, but its form was largely determined by the pre-existing en- ceinte of Herodian buildings into which it was inserted. Hebron has only been acces- sible to Christians since the war. This most interesting church was probably built be- tween I 115 (when the site was described by the pilgrim Daniel) and I 167 when it became a cathedral. The curious document, Trac- tatus de inventione sanctorum patriarcharum, an account of the discovery of the secret tombs of Abraham, Isaac and Jacob in 111 9 (Recueil : Hist. Occid. V., pp. 302-16) is not clear as to the stage the building had reached, though clearly there was some

church at that period. Cf. Vincent, Mackay and Abel, Hdbron, Paris, 1923-

3 The cathedral of Tarsus is one of the most curious in this respect. Captured by Baldwin of Boulogne in 1097, it was held by the rulers of Antioch till 1138, and, after a brief interval again till 1183. It was however an outpost, cut off by a strip of Armenian territory, and probably with only a small Latin garrison. The cathedral is small and crudely built: there was no vaulting, and the roof is supported on alternate columns and square pillars : there is no apse but it ends in a square sanctuary. Enlart (II, p. 378) finds it closely allied to a group of churches in Lorraine. The other crusaders' church, St. Peter's, is the common apsidal type, and must be later than the cathedral.

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 5

use of the pointed arch was almost certainly more general in Palestine in the first half of the twelfth century than it was in the West. The crusaders found in Arab work many examples of it, and the lack of wood and consequent need for stone vaulting may have hastened the process. EnlartI held that the earliest pointed arch he could date with certainty in France was that of the priory church of Le Wast, founded by Ida, mother of Godfrey of Bouillon and Eustace of Boulogne, about I I Io, that is after the return of many of their followers from Palestine. This remarkable building includes also a doorway ornamented with open saw tooth decoration, which is all but unique in France, and which recalls the famous gate of Bab-el-Foutouh at Cairo. This gate we know was seen by a crusading embassy as early as io99, and it appears to have exercised considerable influence in Palestine. It is not impossible that Le Wast also is a direct borrowing from it.

The most notable example of the mingling of styles is however the faqade of the Church of the Holy Sepulchre. In this building the crusaders had a particular problem, and in meeting it used a range of means found in no other of their churches. The holy sites had already seen much of building, destruction and rebuilding. As they stood in og99, they consisted of a domed rotunda covering the Sepulchre, while to the East there were the ruins of the basilica over Golgotha; between was an open atrium. They were largely recent work, built following the renewal of the Byzantine protectorate in the mid-eleventh century, after the overthrowing of the Christian shrines by the Caliph Hakim in ioo8. The Sepulchre was the central aim of the whole crusading movement and therefore became their earliest care.2 The Abbot Daniel when he visited it between 1 13 and I I 15, found the Rotunda of the Sepulchre and a chapel with fine mosaics over the actual rock of Golgotha,3 but the only work he describes specifically as made by the Franks was the canopy on pillars, surmounted by a figure of Christ in silver above the ordinary height;4 of which earliest work of crusading art nothing survives, and it is difficult to form a clear opinion of it or to find any parallel in the West which comes close to the Abbot's description. The next undertaking, after an earthquake in III14 had threatened the security of the existing buildings even further, was to build a domed choir and vaulted sanctuary connecting with the existing rotunda

1 C. Enlart, "L'6glise du Wast en Boulon- nais." Gazette des Beaux-Arts, 1927, pp. I-II. The whole question of Eastern influences is also dealt with in J. Ebersolt, Orient et Occident, Paris, I929.

2 There is a vast literature on the subject of the Church of the Holy Sepulchre. The most authoritative account is however to be found in Vincent and Abel, Jirusalem, Vol. II, 1926. Harvey, Structural Survey of the Church of the Holy Sepulchre. Oxford, 1935, gives details of the present dangers threatening the building, and has many valuable photo-

graphs. See also L. Marangoni, La Chiesa di Santo Sepolcro. Custodia di Terra Santa, 1937-

3 The great mosaic of Christ (cf. Harvey, op. cit., frontispiece) in the Chapel of Calvary is possibly a survival of these.

4 It does not seem to have survived the pillaging of Jerusalem by the Kharismian Turks : there is no reference to it in II87 and, despite its value in metal, it was probably left unharmed at the time of Saladin's conquest.

6 T. S. R. BOASE

and absorbing the basilican ruins. The sanctuary had an ambulatory round it, in plan and execution similar to that of St. Sernin of Toulouse and of the abbey church of Cluny which Urban II had dedicated at the time of the Council of Clermont. The choir was entered by a transept to the South, and the facade of this transept was the main entrance to the whole group of buildings. The work was almost certainly completed by II149 when a solemn service of dedication took place. At a later date, between i 149 and 1187 the bell tower was added.

The fagade itself consists of a double doorway, with carved lintels, plain tympana which were originally painted and elaborate voussoirs : the second storey has two large windows, balancing the doorways and with similar voussoirs : both storeys have heavily carved cornices. Such are the main features : the details of them provide many interesting associations. The two lintels (now preserved in the new Museum recently opened at Jerusalem) are very dissimilar in treatment. The left hand one has a continuous, though curiously arranged, narrative; the raising of Lazarus, Martha and Mary imploring the aid of Christ, Christ giving instructions to Peter and John as to the Passover, the preparations for the Last Supper, the choice of the ass, the Palm Sunday entry,' the Last Supper (P1. Ic). The style of the carving has often been compared to that of the frieze of St. Gilles, and it is certainly allied to Provengal work : but the figures are more elongated than those of St. Gilles (though there considerations of perspective and view-point may account for deliberate foreshortening.) The lintel is much weathered, but must have been finely carved : some of the details in particular are very carefully worked out : the palm tree, from which palms for the procession are being plucked, is realistic and may have been studied from nature. It would be natural to expect this series of scenes to be continued by those of the closing days of Passion week : instead the right hand lintel is of completely different style. In long curves of foliage birds and small human figures are interlaced in an elaborate and highly stylised pattern (Pl. Ia). It has close parallels in the decoration of Toulouse, though nothing perhaps on this scale, certainly nothing of superior workmanship (P1. Ib). This panel is made up of three pieces of marble, with the centre one fitted as a keystone, and when they were removed from position it was found that the reverse of one slab was carved with Arab scroll work, probably of the tenth century, a curious instance of crusading quarrying methods and a proof that, whatever the Western influence, the lintel was carved in the East.2 A figure of Christ stood between the doorways, but whether as a trumeau or between the arches is not certain. The capitals of the columns of the porch are corinthian in type, finely carved, and could be

paralleled by many examples from Southern France (P1. 2d). The voussoirs on the other hand, of both the porch and the upper windows, are more unusual. They are composed of the form of decoration known as godrons, which consists in deep bevelling of the arch stones. The earliest known

1 Badly broken-the missing piece however was found by Clermont Ganneau built into the wall of a house and is now in the Louvre.

2 Quarterly of Dept. of Antiquities in Palestine, Vol. I, I931, p. 2.

1

a--Jerusalem, Church of the Holy Sepulchre. Right Lintel, Detail (p. 6)

b-Toulouse, St. Sernin, Capital (p. 6)

c-Jerusalem, Church of the Holy Sepulchre. Left Lintel, Entry into Jerusalem, Last Supper (p. 6)

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 7

use of this motive, which was much used in Palestine, is in the Bab-el-Foutouh gateway of Cairo, which can be dated to the period

lo87-1o91. This could

have been seen by the embassy sent to Cairo during the first Crusade, and, as noted above, seems to have impressed Western imagination. Certainly the developed use of godrons appears to have been Palestinian rather than Western, and though they are found both in France and Sicily it seems probable that they were popularised by their conspicuous use on the greatest pilgrimage church of Christendom. Above the arches of the porch and of the windows run elaborate cornices : these have sometimes been considered re-employment of antique material. In style they are Hellenistic and strongly reminiscent of some of the mouldings at Baalbek. That they are however work contemporary with the building seems clear from the exactness with which the design fits the course of the cornice and adjusts itself to the turns of the corners. It must be considered as an example of direct inspiration from those gorgeous remains in which Syria is so rich.

Within, the Church of the Holy Sepulchre offers many other fragments of decoration, nearly all of which present points of interest which are discussed at length either by Enlart or by Vincent and Abel and which cannot be dealt with here : but recently one particular group of decorative fragments has been written of somewhat provocatively in an article by Strzygowski which appeared in Speculum.) Enlart had already drawn attention" to the fact that several archways in the precincts of the Temple, including one in the cave of the rock itself, appeared to be of Frankish workmanship, and that certain of their details corresponded closely to the drawing of the tomb of Baldwin V made in the first half of the eighteenth century by the Franciscan Elzear Horn.3 The resemblances, which are very striking, consist in a particular type of tendril, convolving to form rosettes, of shell canopies in the niches, and of curious interlaced pillars, like the linking of metal work which Enlart terms 'un d6cor qui imite le tricot,' and Strzygowski less elegantly (the article is in English) "gut-like resolved shafts." This somewhat disagreeable fancy seems to be found in few other places : there is an example in a mosque at Hamah, which may be re-employment of earlier material (the mosque is probably fourteenth century). In the West similar interlaced patterns are found in relief, inspired by Carolingian models, but as actual column treatment there seem to be no parallels to the Jerusalem examples. The nearest approach to it that I know of is, curiously enough, in the Hohenstaufen palace at Wimpfen,4 where there is in the arcading a knotted column and also one composed of three twisted shafts,5 which might easily be memories of these crusading patterns. Baldwin V died in September 1186, less than a year before Saladin's victory at Hattin. His ornate tomb thus represents the final development of crusading art in Jerusalem. It is to be presumed that it remained in existence

1 Vol. XI, 1936. 2Enlart, op. cit., I, pp. 99-1oo, 167;

II, p. 210. 3 Golubich, Iconographiae Locorum Terrae

Sanctae descriptae a P. Elzearis Horn, I902. 4 The twisted pillar was also popular in

Jerusalem and there are examples of it in the Mosque El Aksa and in the Porte de la Chaine.

5 Photograph in Bruhns, Hohenstaufenschl's- ser, Blaue Biicher, 1937.

8 T. S. R. BOASE

till the destruction of the tombs of the Latin Kings by the Greeks in i8o6. It is possible that parts of it were then taken over and embodied in various mihrabs of the Temple area, and certainly the pieces in question correspond convincingly both in design and measurement, though their present positions suggest an earlier date for their incorporation. If we do not have in them actual parts of the tomb, we have certainly examples of the same school of carving, and a striking indication of its marked individuality. Strzygowski, who does not appear to be aware of Enlart's discussion of the subject, considers the motives of the columns to be Iranian in origin, and that out of the mingling of Eastern and Western culture 'something like the later Renaissance was produced in Jerusalem out of the close of the Romanesque style.' It is an obscure statement : Romanesque art had classical influence as well. But if interpreted as meaning that Jerusalem saw a synthesis of Hellenistic, Arab and Romanesque motives, which in some cases reached high artistic achievement, it is a statement which has many indications to bear it out.-

If the Holy Sepulchre still retains in its lintels an example of figure sculpture, it is peculiarly fortunate in doing so, and its survival is a curious example of the respect and tolerance with which the Mohammedans treated this centre of Christian pilgrimage. For the human form was by the teaching of Islam not to be portrayed2 and elsewhere all figure sculpture was purged from the mosques that had once been Christian churches, so thoroughly that but little survives to show its nature. Jerusalem has a few pieces, which Enlart and Deschamps have carefully examined, but many of them are too damaged to provide much grounds of comparison.3

It would in fact have been easy to think that the crusaders were in this respect influenced by Islam, and that the decorative sculpture had been mainly foliage or abstract design, the capitals of Tortosa or the rosettes and zig-zags of Giblet, were it not for a discovery made at Nazareth in 1908. Under Father Viaud archaological works were in progress on the Church

1 It is notable that the fragments discussed above are all in the Temple area. The Dome of the Rock became of course a crusading church, and the buildings of the Order of Templars adjoined it, but of crusading work in the Temple Area, or of the Temple itself under their care little is known. The iron grill, which still surrounds the rock itself, is generally accepted as crusading work, and is one of the finest pieces of mediaeval iron work in existence. The metal links would be at hand as a model for the pillars.

2 Mohammedan art has itself not always been strict in the observance of this, but it seems to have had no tolerance for Christian figure sculpture.

a P. Deschamps, "La Sculpture franqaise en Palestine et en Syrie

' 1'6poque des

Croisades." Monuments et Mimoires (Fonda- tion E. Piot), XXXI, 1930, pp. 9i-ii8. The most important are a capital, coming from Damascus, but certainly Frankish, showing a variant of the Constantine sub-

ject (it is now in the Louvre) : four capi- tals from Sebastyah, with the story of the

Baptist (they are poorly carved but in fair

preservation) : six foliage capitals with

curiously placed heads, coming from La- troun, which are of splendid workmanship, cut in marble, and unusually well pre- served: all these are in the Museum of

Constantinople, and photographs are repro- duced in Deschamps' article. There is a

striking figure of a Madonna, unfortunately headless, from the voussoir of St. Mary Latin, preserved with other fragments in the Museum of the Greek Patriarch at Jerusalem.

2

a, b-Nazareth, Church of the Annunciation, Capitals (p. 9)

c-Plaimpied (France), Church. Capital (p. II)

d-Jerusalem, Church of the Holy Sepulchre, Porch (p. 6)

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 9

of the Annunciation, works which were resulting in a clear plan of that building being made possible. In the course of these investigations a small chamber was opened up to the North of the church, in itself an object of interest for it was clearly an ancient dwelling regarded by the crusaders as of some sanctity and importance. In the course of clearing the entrance of this chamber, five capitals were discovered, of the most striking workmanship, and, apart from a few breakages, in perfect condition (P1. 2a, b). In style they are the most developed Romanesque, and their delicacy of line is not blunted by any appearance of weathering. One is a quadrilateral carved on three sides, the others octagonal carved on six sides: the blank sides show no appearance of having been placed in position.' The scenes depicted on the capitals are placed under an archi- tectural canopy of arcades resting on a line of corbelling; these arcades support a diaper patterned roof on which at the corners stand smaller arcades. The figures are slender with a pronounced facial type and with very linear treatment of the drapery, which forms clearly marked circles over the pro- jecting parts of the body. The subjects of the capitals present some diffi- culties, but on the whole have been fairly certainly identified. On the central face of the quadrilateral one a crowned woman bearing a cross leads forward a bearded man; on one side devils pursue the man, on the other they draw back, still in threatening form, from the outstretched cross. It is a generalised theme in which faith or the church leads an apostle through the dangers that the devil sets in his path, and this interpretation sets the note of the other capitals, all of which deal with the lives of the apostles. One of the octagonal capitals represents Christ with eight of the apostles, while St. Thomas reaches out to touch his wound (P1. 2a) : the next has two scenes : Christ by the Sea of Galilee, after his resurrection, with St. Peter advancing from the lake, while, behind, others in a boat draw in their nets, and on the remaining three sides the raising of Tabitha by Peter. The fourth capital is more difficult, but the accepted explanation convincing. St. James the Great is led to execution, and, by healing a paralytic, converts his accuser Josias, whom he baptises : this is represented on two faces, the next shows the execution of St. James : the remaining faces however have a new story, drawn from the legend of St. Bartholomew in India (P1. 2b). The last capital, with its crowned figures and mysterious devils, has been identified as scenes from the legend of St. Matthew's mission in Ethiopia. Writing at the end of the twelfth century, Gervase of Tilbury describes 'India,' that is, the unknown East beyond Syria, as divided into three parts, superior, inferior and meridiana, the first converted by St. Bartholomew, the second by St. Thomas and the third by St. Matthew.2 These were

1 Good photographs of the capitals are reproduced in an article by P. Egidi, "I capitelli romanici di Nazaret." Dedalo, II, 192O, pp. 761-776. For an account of their finding and a discussion of their iconography cf. Viaud, Nazareth, Paris, 19Io (including a long letter on the subject from R. de Lasteyrie). Deschamps discusses them in the

article already cited and in his French Sculp- ture of the Romanesque Period., Pegasus Press, I930, PP- 95-99- Cf. also Clapham, Roma- nesque Architecture in Western Europe, Oxford, 1936, p. I 13 and Enlart, Architecture religieuse, I, pp. 129-33 and II, pp. 292-310.

2 Gervase of Tilbury, Otia Imperialia. Scriptores rerum brunsvicensium, I, p. I9I.

IO T. S. R. BOASE

legends that earlier had enjoyed some popularity: Ordericus Vitalis has an account of them in his History of the Church'; but during the twelfth century the crusading settlement of Syria had brought new contacts which heightened their interest. There were constant rumours of the mysteries of the Christian East, of the kingdom of Prester John, with whom as early as I177 Alexander III had sought to correspond, and whose imagined letter to the Emperor Manuel, circulated in so many popular variants.2 The unusual iconography of the Nazareth capitals is a further indication of this twelfth century interest in the Far East : Peter, with his miracles at Galilee and Jaffa, James, martyred at Jerusalem and patron saint of wars against the infidel, supply the local interest : but the other apostles particularly commemorated, and in the case of Bartholomew and Matthew com- memorated with no known Western iconographical models, are the apostles of the East. The capitals seem to accord with the ground plan of the West doorway. It would have been expected that the Church at Nazareth should show scenes of the life of the Virgin, and probably they figured amongst some of the sculptures of the porch. The fact that, despite local association, the apostolic missions were given such prominence, is all the more striking testimony to the careful thought behind the choice of subject. The Nazareth capitals foreshadow the missionary journeys of the following century, and the new teaching of Francis and of Ramon Lull.

The capitals are not alone amongst the fragments of carving at Nazareth. In the Museum of the Greek Patriarchate at Jerusalem there are two heads of bearded men carved on a single block, found in Nazareth in 1867. Rather larger than life, their scale is very different from that of the capitals, but they show a very similar stylistic convention and facial type, and if not from the same hand, they are clearly from the same workshop. Like the capitals, they appear never to have been placed in position.3 The Museum of the Franciscan Convent at Nazareth, where the capitals are preserved, has also some other fragments of carved figures from the excavations on the site of the church, others from the site of the church on Mt. Tabor, and all, as far as can be judged from such small pieces, of the same type of workmanship. It seems evident that in connection with the building of the church at Nazareth, and probably also with the great contemporary work on Tabor,4 there was a sculptor of the first rank working either alone or with pupils; and that much of his work was not placed in position, but at some period when the town was threatened, was hidden away, and remained hidden or but little disturbed (at some period it received some damage, which has chipped one or two faces) till Father Viaud's discovery.

1 Hist. Eccles, II, cc. 9 and I0. 2 The Prester John legend has been

examined in great detail by Zarncke in the K6niglich Sdchsische Gesellschaft der Wissen- schaften, Abhandlungen, Philologisch-historische Klasse, VII, VIII, XXIX, XXX : it is also summarised, along with much other important information in J. K. Wright, The Geographical Lore of the Time of the

Crusades, American Geographical Society, 1925-

3 Photographs reproduced in Deschamps' article, op. cit., and in his French Sculpture.

4 The Cluniac Church of Tabor appears to have resembled closely in plan the Abbey Church of Graville Sainte-Honorine near Havre : see detailed discussion in Enlart, op. cit., II, pp. 385-95-

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 11

The dating is no hard matter. Various texts, borne out by archmeology, prove that the crusaders at once rebuilt the ruined Byzantine Church at Nazareth, and that probably as early as 1I15 there was a new church standing there. Excavations however show many modifications and enlarge- ments of the plan : and certainly building continued through the century up till the taking of the town by Saladin in I187. The church was apparently not seriously damaged on that occasion, but survived till it was destroyed by Beibars in 1263. The capitals must belong to the period preceding the capture by Saladin, and may have been hidden in the period 1183-87 when the town was being constantly threatened. It was not till 1204 that the Christians reoccupied the town, and their hiding place may well have been lost sight of.

Where did the sculptor come from : was he a local product, or some new crusader from the West? The subject has been much discussed, and the discussions have led to a most interesting discovery. Deschamps has published' a capital from the Church of Plaimpied which most satisfactorily contains all the most distinctive features of the Nazareth capitals (P1. 2c). The subject is the Temptation; Christ, holding the book of the scriptures and seated under a canopy closely resembling those of Nazareth, repulses two devils. The pronounced long-nosed facial type, the large halo, the free use of drilled ornament, the linear drapery,2 the dramatic gesture all are strikingly similar: but even more remarkable are the devils; both are fantastically thin and contorted, and one is covered with curiously cut feathers, the like of which appear to be found nowhere save in the Bartholomew scenes on the Nazareth capital. Few instances of medieval carving can give such a completely satisfactory stylistic agreement. The Plaimpied capital is from the interior of the church, excellently preserved and oddly unlike other sculpture of the locality, save for a damaged tomb in the same building. This tomb is of a canon Sulpicius, and a later hand has carved on it the date I142, which must represent some local tradition.3 In the mid-twelfth century then a sculptor of a marked and personal genius was working at Plaimpied, of whose work there are no other surviving examples in the neighbourhood,4 but whose hand is all but unmistakable in work at Nazareth, carried out most probably in the seventies. There our knowledge ends. Archbishop Letard of Nazareth, who was elected in 1158 and held his office certainly till the eighties, had previously been prior. It is impossible to

1 "Un Chapiteau Roman du Berry." Monuments et Mimoires (Fondation Piot), XXXI, 1930, pp. II9-26. Enlart had already suggested the tomb at Plaimpied as a parallel to the capitals.

2 The drapery is more finely drawn in the Nazareth work, whereas at Plaimpied it is done in bold single lines. The use of circles and the general scheme is however exactly similar. The fine quality of the Nazareth stone may have encouraged more elaborate cutting.

3 De Roffignac (L'e'glise St. Martin de

Plaimpied, Bourges, 1928, p. 24) suggests that the tomb stylistically must be later in the century, but it is too individual both in lettering and treatment of drapery to make stylistic dating easy.

4 The portal of Plaimpied was destroyed in I740 and rebuilt: it had capitals with scenes from the life of Christ, fragments of one of which are in private hands at Plaim- pied. It is described as being mutilated, but having remains of an architectural canopy, with scenes below it. De Roffignac, op. cit., p. 13-

12 T. S. R. BOASE

trace through him any local connections in France, nor are there any other known ties between Nazareth and Berry. But some journey there must have been between the two, whether with the crusade of Louis VII, which mustered at Metz in II147, or, more independently, at some later date. The find of the buried capitals has given us another link in the tale of relation- ship between East and West and the knowledge of an artistic personality not least among those of the great Romanesque sculptors.

Second in importance to the Church of the Sepulchre only was the Church of the Nativity at Bethlehem.' Here the buildings, reverenced by Mohammedans as well as Christians, had remained intact. When the first crusade came in sight of Jerusalem, Tancred, the leader to whose energy the settlement owed so much, made a sudden raid and secured Bethlehem before siege was laid to the Holy City. The church he found there was that which Justinian had restored, and the crusaders made little alteration to the structure itself, though much was done to the monastic buildings, and the cloisters, with their Gothic arches springing from twin columns, still remain, disfigured but evident, as an example of their building. The interior decoration of the church however was one of the most striking artistic enterprises of the Latin Kingdom. An inscription, the Greek version of which is still largely preserved,2 gives us unusually exact details as to the manner and time of its being carried out :

"This work was finished by the hand of Ephrem, historiographer and artist in mosaic, in the reign of Manuel Comnenus the Great, born in the purple, and in the time of the great king of Jerusalem, the lord Amaury, and the most holy Bishop of sacred Bethlehem, the lord Raoul in the year 6677, second indiction [1I69]." The occasion is an interesting one. In I 167 Amaury had been married

to Manuel's niece, Mary, in continuation of the policy of close co-operation between Syria and Constantinople. At the same period Manuel was in close touch with Rome on the question of a possible re-union of the churches, and discussions on doctrinal differences and on their common Christian faith were once more prominent. The adroit Anglo-Norman,3 Raoul, Arch- bishop of Bethlehem, who was chancellor of the kingdom, may well have found occasion to combine the re-decoration of his church with a symbolic statement of these plans and hopes. The West wall was covered with

1 The most recent accounts are W. Harvey, Structural survey of the church of the Nativity at Bethlehem, Oxford, I935. H. Vincent, Bethlehem in the Revue Biblique, XLV, pp. 544-74 and XLVI, pp. 93-121. E. T. Richmond, "The Church of the Nativity." Quarterly of the Dept. of Antiquities in Palestine, VI, pp. 63-72. Vincent and Abel, Bethld- hem (Paris, 1914) still however remains the authoritative work. See also Byzantine Re- cord Fund, Church of the Nativity at Bethlehem,

edited by R. W. Schultz, London, I9I0. 2 The Latin version of the inscription is

known from travellers' accounts : Amaury is

given first place in it and there is a catalogue of his virtues: Manuel, however, is des- cribed as "dator largus." cf. Vincent and Abel, op. cit., p. 157-

3 'Anglicus natione.' Wm. of Tyre, XVI, c. 17. William described him as much favoured by the court and by his 'com-

patriot', Adrian IV, but 'nimis saecularis.'

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 13

a mosaic of the Tree of Jesse' : the South wall above the architraves had a row of half length figures, the ancestors of Christ after St. Matthew, and above them a series of architectural designs figuring the seven general councils, with statements in Greek setting forth the principal doctrinal point with which each was concerned. Only the second council of Nicea, which condemned two emperors and three patriarchs, had its inscription in Latin not in Greek. The ancestors are labelled alternately in Latin and in Greek. On the opposite wall a similar scheme is carried out, only the ancestors are after St. Luke, the Councils Provincial not General. Between the windows are mosaics of angels, which on the North side are fairly well pre- served and bear the signature Basilius Pictor. On the South side a fragment of mosaic has the letters B and C, possibly a Greek contraction for the same name. The apses had scenes from the life of Christ, apparently following a normal scheme, and, from what remains of them, strongly Byzantine in feeling.2

Besides the lavish mosaic decoration, the round columns of the basilica were painted with the single figures of saints, now faded into indistinctness, but still decipherable. Their titles are inscribed on each painting, generally in both Greek and Latin. As with most of the surviving painting in Palestine, the general style is Byzantine, though in some figures the treatment of the drapery is reminiscent of Romanesque carving.3 The selection of saints however shows a real catholicity of interest. They include a group of hermit saints, Macarius, Anthony, Onofrius, Euthymius, Theodosius, Sabas, who may be taken as a representative gathering of Eastern coenobites, the last three of whom had strong local associations. The greater figures of Christendom are represented by the Virgin, St. Anne, Elias, St. John the Baptist, St. John the Evangelist, St. James the Greater (with a kneeling pilgrim in attendance, a clear reminder of Compostella and the sister war against the Moors), St. Bartholomew and St. Stephen. None of these are in any way unexpected-St. Bartholomew enjoyed a wide cult in Southern Italy and also in Armenia.4 Anselm had recently brought one of his arm bones to England and there too he was growing in popularity.5 Interspersed with these greater figures are various other saints: George, Leonard, Cosmas

1 M. Emile Male considers the Tree of Jesse as Western in origin, and that in the accounts extant of this Bethlehem example it corresponds to Western types. L'Art reli- gieux du XIIe sidcle en France, p. 175. Cf. also A. Watson, The Early iconography of the Tree of Jesse. (Oxford, 1934), PP- 18-19, 25, 169.

2 There are photographs and reconstruc- tions of them in Schultz, op. cit. Stylistically the decorative panels of the Councils have points in common with the tomb of Bald- win V.

3 The most important extant series of frescos is that in the church of Qaryat el Enab, near Jerusalem. Here the painting is Byzantine in style, but has both Latin

and Greek inscriptions. The pillars, are frescoed as at Bethlehem. According to pilgrims' accounts there were striking pain- tings at both Nazareth and Tabor. Some frescos have recently been found in a bap- tistery chapel below Crac des Chevaliers: they deal with the life of St. Pantaleon and seem Western in origin. Deschamps, Comptes rendus de l'Acadimie des Inscriptions et Belles-Lettres, 1935, p. 365-

4 Armenian interest in Bethlehem is attes- ted by the doors presented in 1227 by Hayton of Armenia.

5 St. Bartholomew's Smithfield was begun in I 123. He had 165 dedications in England, cf. F. Bond, Dedications of English Churches, p. 135-

14 T. S. R. BOASE

and Damian, Cataldus, Vincent, Canute Rex Danorum and Olaf. Most of the Saints have their names in Greek and Latin characters, but these last two in Latin only. St. Olaf was a popular figure of many legends, among which was one of a journey to Jerusalem : but coupled with Canute his presence must be taken as clearly indicative of some Scandinavian influence, a tribute to the frequent assistance that the crusading states had received from Northern Europe': St. George represents the local cult of Lydda, and was the warrior saint of the crusades, who had miraculously led their forces at Antioch : Leonard on the other hand is an English and Aquitanian saint, much invoked for the release of prisoners and therefore a most suitable patron for crusading interests: Vincent, with a predominantly Spanish cult, was a fairly popular saint in Southern France and Italy : Cataldus, with his cult centring in Tarento, was well known throughout South Italy and Sicily : Cosmas and Damian had a wide popularity throughout the Eastern Mediterranean. The choice of subjects therefore represents a considerable mingling of local interests.

The work at Bethlehem leads on to another central problem of crusading art history. The angels in the mosaics are signed Basilius Pictor. The same name Basil occurs as signature on one of the illuminations in the cele- brated Milisande Psalter in the British Museum (P1. 3). The history of this MS. can be briefly summarised : it was purchased by the Museum in 1845 coming from the collection of Dr. Commarmont in Lyons, who was said to have obtained it from the Chartreuse of Grenoble.2 The evidence for it being of Palestinian workmanship are three entries in the calendar, the death of Baldwin II under August 21st, that of his wife Emorfia under Ist October and the statement under the I5th July that on that day Jerusalem was captured. Baldwin II, who died in I 131, was succeeded by his daughter Milisande, who married Fulk of Anjou, and as the entry of Baldwin's obiit points to their reign, the Psalter has been given the Queen's name. Further, on one of the ivory covers, under the carving of a bird, occurs the word Herodius (P1. 3a), which is an alternative name for the fulica (heron) and may possibly be a rebus for Fulk.3 As Fulk's death, which took place in

1144, is not entered in the calendar, the Psalter is usually assigned to the period 1131-44. If however the possible connection between Herodius and Fulk is not held conclusive, the Psalter, the prayers in which are designed for a woman, might as well have belonged to Milisande's sister, Yvette, abbess of the monastery of St. Lazarus at Bethany. Were that so,, Fulk's death need not have been added to that of Baldwin and his wife, and, as Yvette was alive as late as i I80,4 a later date for the Psalter would be possible,

1 Olaf however had two dedications in Constantinople.

2A. du Sommerand, Les Arts au Moyen Age, Paris, 1841, Vol. V, p. 162.

3 C. Cahier, Mdlanges d'Archeologie, d'Histoire et de Littirature, Paris, 1874, pp. 1-18 : a full account of the ivory covers. Apart from the account in Sommerand, op. cit., there

are discussions of the Psalter in J. A. Herbert, Illuminated Manuscripts (London, 1911), pp. 57-60, and 0. M. Dalton, Byzantine Art and Archaeology (Oxford I91I), pp. 231-3 and pp. 471-73.

4 Chartes de 1'Abbaye de Notre-Dame de Josaphat. Rev. de l'Orient Latin, VII, 1899, p. I50.

a-Ivory Cover b-Initial Page

3

c-Entry into Jerusalem

~p;, wm? _77:

+r+ I.A

~'+T ] C .......... .: +++ ' " ... " . .... ........ ....

.

IYIIb CIA. + 1', pito+ ,

in

,er,

:i:i

d-Virgin and Child

M61isande Psalter, British Museum (p. I4f)

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 15

and one which might allow for the mosaicist Basil of Bethlehem and the illuminator being one and the same person.

The Psalter contains twenty-four full page illuminations, dealing with the life of Christ in a series of scenes' (P1. 3c, d) which, where they can be compared with the fragments of the Bethlehem mosaics, seem iconographically not unsimilar to them. Stylistically the nearest approach to them in illumina- tion work is in a Syriac lectionary, the work of Joseph of Malatea, on the Euphrates, made under a Bishop John, who is probably to be dated to II93-I22O.2 These are inferior to the Jerusalem Psalter, but repeat the same facial and architectural types. M6lisande and her sisters, whose mother was a Syrian by birth, were known as patrons of the Jacobite church as well as of the Latin, and this school of illumination may represent some such local interest.

The problems of the Milisande Psalter, however, do not end with the scenes from the life of Christ. The Psalter itself has elaborately gilded initial pages, with extremely complex decorative designs (P1. 3b). Some of these show possible Western influence and the interlacings which make up their patterns repeat themes of Romanesque art3; others however suggest strongly some Arab source. The Dixit insipiens has a winged griffon framed in a pattern of squares, placed at angles over one another: the Exultate Deo has a similar frame work, whose strictly geometric design is Arab rather than Romanesque. Finally there are marginal zodiac medallions to the calendar and nine half-page miniatures of Saints towards the end of the book. These seem all to be by the same hand, but not by that of Basil. The colour scheme is quite different, and lacks the harsh and vivid contrasts of the full page illuminations. They are, Herbert thinks,4 "plainly the work of a Western artist," and though there are some Byzantine details in them, the whole appearance suggests contemporary work in France. The manuscript, which is in Latin throughout, is written in small minuscules of great beauty, most certainly by a practised Latin scribe.

While recent work has done so much to recall the unexpected richness of the crusading kingdom in ecclesiastical building, its great achievement still remains that of fortification. In this, thanks to the facilities under the mandates, there has been a complete re-examination of much of the existing material. In Syria in particular the work of Dussaud and the various government surveys have made possible for the first time a detailed examin- ation of the strategic placing of the crusaders' defence system.5 Deschamps' researches at Crac des Chevaliers have provided the first exhaustive study

1 It is the last scene, Christ enthroned, which is signed on the throne "Basilius me fecit."

2 H. Omont, Monuments et Mimoires (Fon- dation Piot), XIX, 1911, pp. 20I-IO.

3 Similarly the interlacings on the ivory covers are Western in feeling.

4 op. cit., p. 58. 5 R. Dussau, Topographie historique de la

Syrie, Paris, I924. P. Jacquot, L'Etat des Alouites, Beyrouth, I929; and Antioche, An- tioch, I931. The recent publication (1938) by the Survey of Palestine of a map, Palestine of the Crusades, with a gazetteer by C. N. Johns is a valuable contribution on the English side. The volume Syrie et Palestine of the Guides Bleus (1932) contains several plans and summarizes much recent research.

16 T. S. R. BOASE

of a crusading castle, and, when, through his agency, in November 1933, the Etat de Lattaqui6 ceded the castle to France in return for a sum indemnifying the villagers who inhabited it, the most important piece of conservation work in the history of crusading archaeology was undertaken and successfully carried through. In Palestine the excavations undertaken by C. N. Johns at Athlit have also produced rich new material, and if it is ever possible to carry them to a conclusion will provide an English parallel to the French achievement at Crac.'

The central problem of crusading castle building can be briefly stated thus: did the crusaders, influenced by Eastern models, produce a new style in military architecture which was then brought to Western Europe, or did they bring with them the plans with which they were familiar in the West and work out a logical development of them contemporaneously with similar developments in Europe, borrowing minor details only from Greeks and Arabs?2 The answer depends on an accuracy of dating and attribution which it is hard to secure. In many cases there is no documentary evidence and the distinctions between Byzantine, Arab and Frankish work are hard to draw. Deschamps has provided a valuable analysis of types of masonry and of mason's marks, but these in a country of local workmen and, frequently, of captive labour, may well be misleading. Clermont- Ganneau writing in 18993 had claimed that one form of stone tooling was specifically crusading and a test for the date and origin of masonry. The detailed analysis of Crac has somewhat shaken this position. Greeks, Franks and Arabs all used the same tools, and varied them according to the nature of the stone employed. In the work of the Franks, the facing employed in the twelfth century was generally a raised boss, in the thirteenth a flatter surface, but the method of facing varies considerably : and, if the apparel of the stone has a finish rare in contemporary Arab work, even there there are exceptions and uncertainties.

The Castle of the Kurds, the original Arab fortification on the site of

1 Quarterly of the Department of Antiquities in Palestine. I, pp. 111-29; II, pp. 41-104; III, pp. 145-65 ; IV, pp. 122-37 ; V, PP. 31-74; VI, pp. 121-52.

2 This problem was stated very clearly in a remarkable essay written as a thesis for the B. A. examination by T. E. Lawrence, while still an undergraduate at Oxford. Lawrence had spent a vacation in Syria, where he had seen many of the most impor- tant sites, but he was travelling by himself, was very short of time and carried out his investigations under great difficulties. As he himself admits he spent a day only at Sahyun and was suffering from malaria at the time. It is hardly to be wondered at therefore that many of his conclusions are hasty, and that his dating of the masonry and attribution of it as between crusader and Byzantine are unsatisfactory. His main theses were: I. that

the Crusaders were strongly influenced by Western developments, and that the con- centric castle was an invention of the West transferred to the East : 2. that the innova- tions were made by the Hospitallers and that the Templars were less original and adopted the Byzantine plan of rectangular saliants in the main walls. (The excava- tions at Athlit are doing much to restore the Templars' credit as builders). The thesis was not published till after his death, but it raised considerable discussion at the time it was presented, and its arguments were well known to students of the subject in England. It was published in an idition de luxe by the Golden Cockerel Press, 1936.

3 Introductory remarks on the distinctive and specific character of crusading masonry. Archaeological Researches in Palestine, Vol. I, pp. 1-47-

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM I7

Crac, was taken during the first crusade, and was permanently occupied by Tancred in II Io. From then till its cession to the Hospitallers in I142 nothing is known of its history, but it is unlikely that, in its strategically important position guarding the gap between Hornms and Tripoli, it should have been left without further buildings. The earliest scheme of fortification, using some of the Arab foundations, probably dates from I1421 : it was that of two parallel lines of fortifications, with square salient towers, and two gateways, one protected by two towers, the other by an advance work in the side of which the gate opened. There seems to have been no keep, such as would almost certainly have dominated a contemporary building in France : the double curtain wall and the square towers are more in keeping with Byzantine examples, as is also the entrance gateway in the flank of the advance work. This advance work has a further peculiarity : on the outer wall there is a machicolation formed by three arches on which a stretch of wall is raised, leaving a space behind for missiles to be dropped on a shallow talus at the foot. The scheme shows signs of experiment : there are two rows of blind arches above the existing ones, and in the end the arch machicolation was abandoned and the wall crowned by corbelled machicolations of a later type. There is a similar arched machicolation at Sahyun : the earliest example of it in Europe is probably at Niort, near Poitiers, built by Henry II about I16o. Corbelled stone machicolations as opposed to wooden hoardings on stone brattices were certainly not known in Europe before the thirteenth century. Timber was scarce in Syria, and possibly the arch machicolations are an attempt to find a substitute for hoardings : the connections between Syria and Poitou were close and it would not be surprising to find a crusading experiment copied there. But in this branch of defence work the Arabs were more advanced than the occidentals : stone corbell machicolations have been found by Creswell on buildings as early as 5162 and the complete gallery of machicolation is probably an Arab invention. At Crac in the thirteenth century stone corbells were used, but not linked up in a continuous circuit till after the taking of the castle by Beibars in 1271.

In I170 Crac was severely damaged by an earthquake. The walls however seem to have stood, and the rebuilding, as traced by Deschamps, seems to have centred in the chapel and some of the living quarters. It is however between i 170 and the early years of the next century that the creation of Crac was completed. In 1188 it was strong enough to resist Saladin, but it seems most probable that these were its earlier walls, and that the main re-building took place in the 90's. On the square fortifications of the existing enceinte three round towers were raised on the South face, where the spur joins the main hill side. The greater part of the earlier wall was then covered by an enormous talus, leaving space within for a connecting passage way. Before the talus a moat was dug, half defence, half water supply. A new curtain wall was then set up, with round projecting

1 There are mason's marks on it which correspond to the marks on the first stage of building at the cathedral of Tortosa.

2 K. A. C. Creswell, Early Muslim Archi- tecture, I, pp. 345-47.

18 T. S. R. BOASE

towers and free use of machicolations. Each line of defence was controlled by the one above it, and the defences were skilfully proportioned to the natural strength and weakness of the site. In the use of round projecting towers, steeply graded talus, machicolations, and in the relationship of the various lines of defence to one another, the concentric castle as known in Europe in the thirteenth century was now fully in existence. Its Western parallel is however almost contemporaneous. Chiteau-Gaillard, which in many features recalls the design of Crac,1 was built in 1196-7. Ceeur

de Lion had never seen Crac des Chevaliers, but he may well have brought back builders with him from the East : or the journey may have been made the other way, and Crac represent a system which had been maturing in the minds of Western designers. All Deschamps' careful research cannot, tantalisingly, provide us with a date accurate enough to solve this vexed question of priority.

At Crac the main problems lie between crusading work of various periods and between that of crusaders and Arabs. The castle of Sahyun raises the question as between that of crusaders and Byzantines. Little is known of the history of this castle, and there is no record of the building of it by the Latins. It was taken by John Zimisces in 975, and presumably fortified by him. The date of its occupation by the Franks is uncertain. Robert of Sahyun2 was killed in 1119, after the battle of the Ager Sanguinis. He was a man of some importance, lord also of Balatonos and of the important frontier castle of Zerdana (Sardone). He had been in alliance with Tughte- kin of Damascus, but was, when captured, executed by his orders. His son William was killed in a skirmish during the dispute between Tripoli and Baldwin II in I132. His widow Beatrice married Joscelin II of Edessa. The lords of Sahyun were therefore considerable people, who might be expected to carry out castle building on a large scale. After the death of William, however, there is no further record of the family, though Beatrice is referred to as a woman of character who aided in the defence of her husband Joscelin's territory. The castle was stormed by Saladin in i188, and never reoccupied by the Franks. An inscription from the late thirteenth century refers to Arab repairs of the curtain wall, but these do not seem to have altered the scheme of the fortifications, though within the walls some buildings were added, including a mosque whose minaret still stands. There is there- fore much to be said for placing the main buildings in the first half of the twelfth century. The castle is known to have been at that time in the hands of a powerful family who would be unlikely to leave it without additions, and the Frankish buildings seem to represent only one undertaking, not several stages as at Crac. After the fall of Edessa (1144) Beatrice and Joscelin would be unlikely to have carried through such a large project, and, though the history of the castle from I144 to 1188 is a blank, all probabilities point to a date before that period. It would thus represent

1 The machicolations at Chateau-Gaillard are a development of the arch not the corbel type: the latter as used at Crac seems certainly borrowed from the Arabs.

2 Cahen, "Notes sur les seigneuries de

Saone et de Zerdana," Syria, XII, I931, pp. 154-59; and Deschamps, "Le Chateau de Saone et ses premiers seigneurs," Syria, XVI, 1935, PP. 73-88.

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 19

the most important early example of the castle building of the crusaders. The castle walls include an area of rather over 1o acres1 : in form it

is a triangular spur between two deep valleys. Where the spur meets the main hill side a ditch has been cut in the rock, 420 feet long, 60o feet broad and 84 feet deep : at one point in the cutting a pinnacle of rock has been left standing to support a bridge across the ditch. The line of the ditch was fortified by a curtain wall, in the centre of which was a three storied keep, 90 feet square, 76 feet high, with the thickness of the walls about 22 feet. The curtain wall, which continues round the spur, is fortified with round or square towers : on the ditch face the bases of the towers are buttresses carved in the rock, at the time of the ditch's construction, and the cutting out of the ditch and the building of the walls must therefore be part of one scheme. The keep and the main part of the curtain wall at this point are built of square blocks with bossed facings and are an admirable specimen of masonry. There is no doorway or entrance between the wall and the keep, which thereby, as always in Byzantine plans, becomes the last point of resistance. This however was not the original keep : behind it on the highest point of the spur, which slopes from the ditch to a slight hillock, was a square Byzantine tower, much of which still remains, and whose original scheme of defences (two rectangles of walls fortified with polygonal towers) can still be traced. This scheme is complete in itself but ends short of the ditch and the assumed Frankish keep, though there appears to have been some Byzantine work, an advance building of some kind, at the point where the ditch was later spanned by a bridge. This much is clear : the earlier Byzantine defence system was at some period completely revised: its main wall defence became internal lines, probably serving as part of the living quarters : the new scheme was accompanied by buildings of remark- able excellence in masonry, influenced by some Byzantine strategic ideas, combining a use of square and round towers, far in advance of anything known in the West in the first half of the twelfth century, and including a piece of ditch cutting which surpasses any of the known Byzantine examples, but which is entirely Byzantine in conception. If this engineering feat was carried out by the crusaders, it seems strange that no chronicler has mentioned it, yet the buildings must be those of the lords of Saone and Sardone, and if the Byzantines had an earlier ditch there, the crusaders at least completed and enlarged it.

In Northern Syria in fact the Byzantine and Latin motives in military architecture are inextricably interwoven : sometimes as at Bourzay the main Byzantine form is completely retained: sometimes as at Shogr there is a double system, a Frankish superimposed on a Byzantine : or again at Baghras there is Armenian building with variants of its own, while in the Southern boundaries of Antioch the Assassins maintained their more primitive, but not unimpressive, forms of military architecture.2

* * *

1 Deschamps, "Le Chateau de Saone." Gazette des Beaux-Arts, 1930, II, pp. 329-64.

2 Several detailed articles on particular

buildings have appeared in various period- icals: the following are amongst the most important : Deschamps, "Deux positions

20 T. S. R. BOASE

In sculpture, painting and architecture, the West and the East meet and effect exchanges in Palestine : there was also a similar interaction in literature. It has long been recognised that in the West the Crusades were an important impetus to the writing of history and the composition of Chansons de Geste. They had news value on a quite unprecedented scale, and the demand was met not simply by the ordinary clerical chronicle but by lay memoirs such as the Gesta Francorum, and eventually by vernacular prose such as that of Villehardouin or Ernoul; while in verse forms there were a bewildering selection of chansons, whose popularity led to their constant re-editing, so that their original twelfth century form is difficult to determine. They, however, have as yet been mainly studied as part of the literature of the West, and little attempt has been made to differentiate these writings which were produced in Syria itself from those composed in Europe. William of Tyre's great work is recognised as a classic example of mediaeval historiography, but its background, William's own training, his library and his general culture, remain obscure and unstudied. The thirteenth century vernacular writers who carried on his work are roughly classified into Eastern and Western, but there has been little investigation as to any generalised differences between them. Yet it is becoming abund- antly clear that the literary culture of Syria was a considerable matter, and underwent influences and developments very distinct from those felt in the typical Parisian culture of the West. In such studies pioneer work has been done by Anouar Hatem in his book Les Poemes e'piques des Croisades.' He examines in particular the three epics, the Chansons of Antioch and of Jerusalem and the poem known as Les Cheitifs. These poems were edited and linked together, probably in the late twelfth or early thirteenth century, by Graindor of Douai, who inserted also exhortatory verses which show that he was writing at a time of crusading recruitment. The most important basis of his work, as Hatem clearly establishes, was the poem on Antioch by Richard the Pilgrim, which must now be accepted as a contemporary account including details found nowhere else. Of the other two poems, the Chanson de Jdrusalem is a mere imitation of that of Antioch, inventing parallel events for the taking of the Holy City. The interest of it however and of Les Che'tifs lies not in their historical worth, but in the fact that they are of Syrian origin. The Chanson de Jerusalem Hatem places between I 130 and I 135 and stresses its connection with the pilgrim church of St. George at Lydda. Les Chetifs was composed in Antioch at the Court of Raymond of Poitou sometime before his death in I 149. If then their facts are roman-

stratigiques des Croisis ' I'Est du Jourdan."

Rev. Historique, CLXXI, 1933, PP- 42-66. Deschamps, "Les entrees des Chateaux des

Croisis." Syria, XIII, pp. 369 ff. J. Bar- thoux, "Une forteresse de Saladin decouverte au Sinai." Syria, III, pp. 44 if. Du Mesnil du Buisson, "Les anciennes defenses de Beyrouth." Syria, II, pp. 235 ff. and pp. 317 ff. G. Wiet, "Les inscriptions de Saladin." Syria, III, p. 307. M. Barres, "Le voyage aux Chateaux des Assassins."

Rev. des Deux Mondes, I933, PP. 481 ff. C. N. Johns, "The castle at 'Ajlun'." Quarterly of the Department of Antiquities in Palestine, I, p. 21 ff. C. N. Johns, "The Citadel at Jerusalem." Q. D. A. P., V, p. I21 ff. G. Beyer, "Das Gebiet der Kreuzfahrerherrschaft Cesarea." Zeitschrift des Deutschen Palaestinavereins, LIX, 1936. P. Savignac, "L'isle de Graye." Rev. Biblique 1913, p. 588 ff and 1936, p. 256

ff. 1 Paris, 1932.

THE ARTS IN THE LATIN KINGDOM OF JERUSALEM 21

ticised, their local colour is irreproachable, and they contain, Hatem claims, valuable information on the habits and customs of Frankish life in Syria, particularly, in the case of Les Che'tifs, with regard to the close relations that were rapidly formed between the settlers and their Arab and Turkish oppo- nents. The story of Les Che'tifs has all the colour and marvels of romantic tradition. It is the story of some prisoners captured by Corbaran (Kerbogha) at the destruction of Peter the Hermit's camp-in which needless to say Kerbogha had in fact no part. It deals with individual combats, in which the heroes disperse armies, kill dragons, and convert their captors; and it breathes the mixture of brutal warfare and enlightened tolerance and curio- sity which characterised the mixed population of Antioch, Greek, Armenian, Syrian, Norman, Poitevin, in their growing wealth, living luxuriously within their huge circuit of walls, whatever hardships faced them without. Such, with a wealth of illustration are Hatem's arguments. They have however already been carried further in an important review by R. Goosens.' In an extremely skilful analysis he relates the romantic adventures of the poem to the actual happenings at Edessa under Baldwin I, and establishes con- clusively the historical basis underlying this Chanson. Edessa, where the Franks were never more than a garrison, where Baldwin married an Armenian princess, was above all the principality of toleration; its history provided themes which, embroidered in the romantic manner, would be popular with all classes and races of the two Northern principalities. Further, on this basis of local history, the narrative is completed by borrowing from Byzantine epics, which themselves were influenced by Arab tales. In the long neglected fable lies an arresting example of the Franco-Byzantine-Arab synthesis, around which so much dispute has inconclusively been waged, and about which so much still remains to be discovered.

1 Byzantion, VIII, pp. 706-26.

Article Contentsp. 1p. 2p. 3p. 4p. 5p. 6[unnumbered]p. 7p. 8[unnumbered]p. 9p. 10p. 11p. 12p. 13p. 14[unnumbered]p. 15p. 16p. 17p. 18p. 19p. 20p. 21

Issue Table of ContentsJournal of the Warburg Institute, Vol. 2, No. 1 (Jul., 1938), pp. i-iv+1-84Front Matter [pp. i-iv]The Arts in the Latin Kingdom of Jerusalem [pp. 1-21]The Poetical Sermon of a Medival Jurist: Placentinus and His 'Sermo de Legibus' [pp. 22-41]Trinitas Creator Mundi [pp. 42-52]Blake's 'Ancient of Days': The Symbolism of the Compasses [pp. 53-63]Miscellaneous NotesA Medival Formula in Kant [p. 64]Blake's 'Glad Day' [pp. 65-68]A Simile in Christine de Pisan for Christ's Conception [pp. 68-69]Mary in the Burning Bush [pp. 69-70]The Seal of St. Nectan [pp. 70-71]A Portable Altar in the British Museum [pp. 71-72]The Literary Sources of the 'Finiguerra Planets' [pp. 72-74]'Apollo and the Swans' on the Tomb of St. Sebaldus [pp. 75]The Four Elements in Raphael's 'Stanza della Segnatura' [pp. 75-79]A Note on Michelangelo's Pieta in St. Peter's [p. 80]'Adamas Mourned by the Nymphs' in Schedel's 'Liber Antiquitatum' [pp. 80-81]'Good Counsel': An Adaptation from Ripa [pp. 81-82]'Grammatica': From Martianus Capella to Hogarth [pp. 82-84]