Introduction STRUCTURE, STYLE AND CONTEXT IN THE MONUMENTS OF OTTOMAN JERUSALEM Robert Hillenbrand 1. The historical setting In 1453 Mehmed the Conqueror had fulfilled the longterm Ottoman ambition to capture Constantinople and to take over the territories of Byzantium. But his successors, not content with this, had sought to extend Ottoman power to the south, which resulted in their conquest of the Levant and Egypt, and-almost simultaneously-to the east. They had thus become embroiled in long-running, inconclusive and costly hostilities with Iran. Sulaiman the Magnificent had then embarked on a series of campaigns in eastern and south-eastern Europe. Thus by 1536, when-some twenty years after their conquest of Jerusalem-the Ottoman authorities began their first projects of serious architectural patronage in the city, their empire was firmly committed to a war on two fronts, both of them far from the Levant. The intensification of Ottoman interest in the European and Iranian theatres of war could not fail to give Sulaiman, as the Ottoman sultan, a perspective on Jerusalem which differed radically from that of his Mamluk predecessors although, like them, he saw himself as the leader of the Sunni Muslim world, which gave an extra edge to his conflict with Shi'ite Iran. For the Mamluks, Jerusalem was centrally placed in an empire which stretched the full length of the Levantine seaboard and also encompassed Egypt and Western Arabia. Moreover, it was they who had definitively liberated the Holy Land from Frankish hands and who had thus set the seal on a campaign of jihad inaugurated in the early 12th century and fostered by such legendary Ayyubid and Mamluk rulers as Zengi, Nur alDin, Saladin and Baibars. Just as the Crusaders had celebrated their capture of Jerusalem with a building boom there, so too did the Mamluks inaugurate a comparable boom when they took over control of the city. They had fought hard for Jerusalem and it had a special place in their affections. In the 16th century there was no reason for the Ottomans to share this view. Nevertheless, like the Mamluks they maintained the useful fiction of allegiance to a puppet caliph, and as self-styled rulers of the orthodox Muslim world they took over the guardianship of the Holy Places, which naturally included Jerusalem as the site of the First Qibla . But they did so from a very different geopolitical stance. Their pressing interests lay to the north-west and to the east, not in the Levant. For them, Syria, Palestine, Egypt and Arabia were provinces of a much larger empire whose centre of gravity lay in Anatolia and the Balkans. These provinces were a source of revenue, manpower and raw materials; otherwise they were of secondary interest. They were governed by appointees from Istanbul, where the metropolitan court and administration were Turkish in speech and largely Persian in culture . The relatively modest scale of Ottoman architectural patronage in Jerusalem makes sense only in the context of this seismic shift in geopolitical realities. Stripped of its special status under the Mamluks, reduced to the role of a minor provincial town, ruled by Muslims who - however sympathetic they were to Arab culture-nevertheless owed their first loyalty to the Turkish sultan in distant Istanbul, Jerusalem under the Ottomans had definitively relinquished its position at centre stage . The architecture built there by the Ottoman authorities inevitably betrays this dramatic change in its role. It is this context that best 1 Ottoman Jerusalem explains the unprecedented emphasis on repair work rather than spectacular brand-new buildings in the Ottoman contribution to the architecture of the city. 2. The differences between Ottoman and Mamluk Jerusalem The differences between Mamluk and Ottoman architecture in Jerusalem reflect this sea-change in the role of the city. From a religious point of view, very little had changed, at least outwardly; the Ottomans, as noted above, took over from the Mamluks the self-styled ·office of Guardian of the Holy Places, and like them embellished Mecca, Madina and Jerusalem. But the nature of their patronage in Jerusalem was very different, as will be discussed in Section 6 below . Its most spectacular expression was two -fold. One element was the refurbishment of the Dome of the Rock, which appears quite simply to have been unnecessary since there is no evidence that the Umayyad external mosaics which rendered the building unique had suffered serious damage over the centuries. Thus the Ottomans seem to have destroyed something exceptional in order to replace it by something commonplace-for by the mid-16th century the major imperial Ottoman religious buildings were frequently clad in lavish tilework. The work of Sulaiman the Magnificent could therefore be seen as an attempt to stake a claim to this especially holy site, and to ottomanise its appearance in an unmistakable way- for glazed tilework could not fail to have been recognised at the time as a foreign kind of decoration in the local context . Significantly, the interior, which naturally lacked the exceptional visibility, and therefore the propaganda potential, of the exterior, was left virtually untouched. The second element of Ottoman patronage was the walls, including the Citadel (probably 938 / 1531-2). Here, too, the political statement is unmistakable (see Section 3 below) . The third most holy city of Islam, which had remained without continuous walls, and therefore obviously vulnerable, since Ayyubid times, was now fully protected - this was the long arm of the Ottoman sultan in action. As in the case of the Dome of the Rock, the visual impact was what mattered. Hence, perhaps, the warlike battlements, more for show than use. Much of the new walling was too low and too frail to offer serious protection against a determined enemy or artillery bombardment. Happily the city was not called upon to endure such tests . These, then, were the two most public expressions of Ottoman patronage, and there can be little doubt that they had an incomparably greater impact on the local population, and on visitors and pilgrims-always an important category in Jerusalem-than any two Mamluk foundations in the city. But, with the single exception of the Khassaki Sultan complex, the rest of the Ottoman 2 buildings reveal a very modest financial commitment to the city's architecture on the part of the Ottoman regime and its functionaries. Their rule lasted almost twice as long as that of the Mamluks, and yet it produced only a fraction of the Mamluk output in terms of quality and scale of buildings. True, there are plenty of them; but as a group they are undistinguished. What is the explanation? The basic factor seems to be geopolitical. For the Mamluks, ensconced in Egypt with important provincial capitals in Damascus and Aleppo, Jerusalem was relatively close to the centres of power . It straddled the route to Syria. It therefore had strategic as well as religious importance and was an integral part of a relatively small and self-contained state. With the Ottomar,i conquest, all this changed. As already explained, the new masters of Jerusalem had much wider horizons than the Mamluks had possessed. They had interests in the Balkans, Anatolia, Iran, Iraq and along the North African coast. The significance of Jerusalem could only dwindle in this vast perspective. Above all, the Ottoman capital was in Istanbul, not Cairo, and in an age of slow, laborious communications this effectively rendered Jerusalem out of sight, out of mind. Moreover, the principal theatres of war in Ottoman times were ~he Balkans , Iran , the Mediterranean Palestine fomented a minor rebellion on occasion, but nothing to justify the intervention of the imperial army in force. After a brief spate of Ottoman interest in the city in the generation after the conquest in 1517-a period which saw most of the significant Ottoman contributions to the city's architectural heritage-Jerusalem sank to the status of a minor provincial town. It lost touch with the great world. An economy long boosted by the intense building activity of the Mamluk period gradually stagnated . The city turned inwards, resting on its past glories while its current affairs gradually became the preserve of a few leading local Arab families who dominated the religious establishment and the administration of the waqfs of its many pious foundations. Even in the Levant itself, Jerusalem under the Ottomans was significantly downgraded vis-a-vis its role a century earlier. Under the Mamluks, Damascus and Aleppo were indeed more important politically, yet Jerusalem rivalled them in its architectural heritage because of the special religious significance which it held to that dynasty. Under the Ottomans Damascus and Aleppo, in accordance with their much larger populations and their greater political and economic importance, were both graced with many major buildings. This was not true of Jerusalem. And thus a paradox emerges: the Ottomans disposed of much larger cash revenues than did the Mamluks, and-as already noted - ruled Jerusalem for considerably longer, and yet they built far less; and what they did build was on a much smaller scale . This can only mean that they gave Jerusalem a lower priority than did the Mamluks. Yet the city was by no means neglected; and it is Structure, Style and Context in the Monuments of Ottoman Jerusalem one of the ironies of its history that under the early Ottomans its population tripled, a process perhaps due in part to Sultan Sulaiman's repair in 1532 of the aqueduct originally built by the Amir Tankiz, who was viceroy of Syria between 1312 and 1340. The restoration of the aqueduct was completed in 1541-42; it conducted water from the Pools of Solomon to nine public fountains in Jerusalem, to others on the Haram and to certain hammams in the city. His reconstruction and enlargement of the bazaar area revitalised the urban economy. But it also had wider horizons-thus the new spice bazaar was part of the Ottoman response to the challenge which Portuguese activities in the Indian Ocean presented to the lucrative spice trade, traditionally a Mamluk monopoly. Above all, Sulaiman's rebuilding of the walls resulted in a complete circuit of fortifications and thereby set Jerusalem apart from most other Palestinian towns, which would have created a welcome sense of security and thus an added incentive to settle there . Unlike today, when much of the Mamluk heritage has vanished altogether and most of what remains is dilapidated or diminished by later constructions, in 151 7 and the following few decades it must have been largely intact, for the most part of recent origin and of a splendour that was very plainly hard to beat . These buildings, and sometimes the structures surrounding them too, were still well protected by functioning waqfs which prevented unauthorised re-use or adaptation. All this may well have made the Ottomans disinclined to enter into retrospective competition with the Mamluk achievement; but it may also have suggested to them that the city simply did not need an architectural transformation. The afterglow of Mamluk patronage must have lingered for generations. Its implicit challenge helps to explain why the best Ottoman architecture in Jerusalem is concentrated into the 16th century-though it is only fair to note that in the Ottoman capital, Istanbul, this was also the period which saw the apogee of Ottoman architecture. The fact that Jerusalem was still a relatively new Ottoman possession may also have played its part as an incentive for architectural activity. On the whole the Mamluk amirs had concentrated their building activity on the immediate surroundings of the Haram, and this area gradually filled up with public monuments in the course of the Mamluk centuries. Thus by 1517 there was virtually no room left for further significant construction in this part of the city-quite apart from the fact that it effectively bore the stamp of Mamluk ownership. There was only one place left for the new dynasty-the interior of the Haram-and, for all the difficulties that this site presented, the Ottomans were so to speak in honour bound to use it. But the ineluctable need to respect the Dome of the Rock and the Aqsa mosque and not to intrude on their physical space in any way seriously inhibited prospective Ottoman patrons. One cannot help being struck by how low the Ottoman buildings on the Haram are in comparison with the Mamluk fac;:adeswhich border it; they are dwarfed in comparison. Yet as the new masters of Jerusalem the new rulers had to assert their power in some tangible, visible, architectural way. That political imperative-so often a factor in glamorous public architecture-excluded foundations in the outskirts of the city, where they would have been effectively out of sight, and pointed, so far as religious architecture was concerned, to the hitherto unexploited Haram. In the field of secular architecture, on the other hand, there was one obvious task to be undertaken, and it was one which the Mamluks had sedulously ignored: the rebuilding of the walls. 3. The walls Beyond question, the most striking testimony of Ottoman architectural patronage in Jerusalem is the circuit of walls repaired and in part erected by Sulaiman the Magnificent between 1537 and 1541, as its eleven inscriptions record . The supervisor of the entire enterprise was Muhammad (:elebi al-Naqqash . Several introductory remarks may be in order. First, this was a rare accolade for the Holy City, since its walls had been in an increasingly ruinous condition ever since their deliberate breach by the Ayyubid princes, al-Mu'azzam 'Isa and al-Malik al-Kami!, in 1219 and 1227 to prevent their use by the Christian enemy. Second, fully walled cities - as distinct from fortified citadels dominating unwalled cities-were a rarity in the Near East. Third, these walls largely followed the line of earlier walls erected when the city was more densely populated, and thus the Ottoman walls defined the area of Jerusalem generously, leaving plenty of room for renewed population growth and thus expressing faith in the future of the city. And for the first time in centuries they gave Jerusalem a physical unity - the necessary prelude to a more intangible sense of community and morale. Fourth, there is a more than local context here, for at almost the same time the selfsame Sultan Sulaiman was constructing a major wall around Madina, the second holiest city in Islam; and of course, his capital, the former Constantinople, was itself surrounded by late Roman and Byzantine walls which were one of the wonders of the medieval world. That fact, and the perennial association of those walls with great empires of antiquity, might well have motivated Sulaiman to emulate that achievement elsewhere, and to present himself as a latter-day Caesar. Lastly, it would be mistaken to attribute this costly project entirely to the pious motives of the sultan. Other factors, both local and international, were at work. The local factors were not confined to the city. Jerusalem, in common with Hebron and other Palestinian towns, had long suffered from the raids which marauding 3 Ottoman Jerusalem Bedouin tribes regularly conducted against travellers on the open road and even against the inhabitants of the outskirts of the larger towns. Thus a chronic insecurity afflicted the surrounding countryside for much of the Ottoman period. Hence there is a darker side to the building of these walls. Put briefly, the building of the walls absolved the Ottoman authorities of the need to police the environs of Jerusalem effectively, and indeed could be interpreted as an acknowledgement that the writ of the government did not run in these areas. This was not a new situation; early in the 16th century, at the end of the Mamluk period, no one could perform the hajj from Jerusalem for an entire decade because the Bedouin had rendered the roads so unsafe. Bedouin lawlessness in fact persisted for centuries and the protection of travellers, for example on the main Ramla-Jerusalem road, necessitated the provision of a heavy guard or the payment of protection money. As late as the early 18th century the then governor of Jerusalem, Mustafa Agha, negotiated an agreement with the inhabitants of a trio of villages to the north of Jerusalem that they would not molest travellers bound for the city. The story does not end here: there is an international dimension too. After the defining and traumatic experience of the Crusades, it is no wonder that in Muslim eyes the Christian pilgrims who continued to visit Jerusalem should represent the shadow of a possible threat against the city, all the more so as the Europeans had never abjured the resumption of hostilities. In the 1530s, rumours were circulating of a new crusade under the leadership of the Holy Roman Emperor Charles V, stung by the recent Ottoman campaigns in Austria. In the aftermath of the fall of Rhodes, the last Christian outpost in the Levant, in 1520, such rumours could not be ignored, and indeed Charles did mount an expedition against Algiers in 1538. Its failure may explain why work on the walls of Jerusalem was abruptly terminated, leaving the interiors of about half the towers unfinished . Nevertheless, the scale and nature of Ottoman work on the towers can be gauged by the Stork, Gaza and Sulphur Towers, and by Goliath's Castle . Some of this work was executed on earlier foundations. The unfinished state of the work - for while the circuit of the walls is complete, neither the height nor the treatment of the ramparts themselves is consistent - is perhaps not all that serious a deficiency, since the role of these walls was at least as much symbolic as military. They represented a prise de possession and a deterrent to any Christian designs on the city. It is certain that they would not long have resisted a determined attack by an ene my force equipped with up-to -date artillery: at some points, admittedly, the walls are almost three metres thick at the base, with a glacis in some places, but they are on the whole less than 1. 5 metres thick at the level of loopholes and less than half that thickness at the height of the ramparts . This reduction is sometimes managed by a succession of 4 retreating jogs. Thus the upper sections of the wall, festooned as they are with bosses, parapets and merlons, are largely for show-hence, for example, the sudden outcrop of decorative forms in the battlements of the Damascus Gate-though they would assuredly have foiled any attacks by raiding Bedouin. Numerous staircases give access to the upper part of the walls . The vari ations in height-from 5 to 15 metres approximately, that is 17 to 50 feet- tell the same story of protection against casual raids rather than a professional army with state-of-the-art cannons. There was no attempt to rebuild the walls from scratch; earlier sections of finished wall, from Herodian times onwards, were incorporated without change wherever possible. This explains the extremely varied stonework, which ranges from the cyclopean blocks of the ancient foundations to the small, neat and relatively uniform squares which characterise Ottoman work. The earlier wall-lines were used wherever this was practicable, as were the existing ditches. Some basic facts will clarify the size and ambition of the whole enterprise. If all the angles, re-entrants, towers and gates are included in the calculation, they extend a total of 4.325km (2.7 miles ). For some reason Mount Zion was excluded, despite its strategic importance, and the Citadel, itself of Mamluk origin, which Sulaiman had fortified and garrisoned as early as 1531, had its own system of defences, though in places these blend with the outer city walls . Like man y a medieval mon arc h before him, then, Sulaiman's first concern was to establish a strong military presence in the city. The military aspect of the walls, at least so far as their overall visual impact is concerned, is driven home by no less than 34 towers, which vary in their height, design and angle of fire. These are supplemented by 344 loopholes or embrasures intended for gunfire, and by seventeen machicolations , some of them clearly intended in a decorative spirit to judge by the muqarnas decoration at their base. There are seven open gates, all but one of them inscribed with the date of their construction. In some cases earlier Mamluk elements are incorporated in secondary use, perhaps for ta lismanic purposes , as in the lions - the heraldic symbol of Sultan Baibars, the scourge of the Crusaders -w hich flank St Stephen's Gate, otherwise known as the Lion Gate. Four older gates-the Golden, Single, Double and Triple Gates-lend still further monumentality to the ensemble, though the latter three gates, being located under the south wall of the Aqsa mosque, are not strictly part of the city walls. Access to the two main thoroughfare s whic h bisect Jerusa lem is through the four major gates: Jaffa, St Stephen's, Damascus and Zion (or Sion), but Herod's Gate and the Dung Gate, to the north and south respectively, handled the overflow of traffic and offered supplementary entrances to other parts of the city. Building on this scale taxed local resources in both Structure, Style and Context in the Monuments of Ottoman Jerusalem cash and craftsmen well beyond their natural limits. Thus over the 5-year campaign (1536-41) enforced contributions from the whole of Palestine, and even some of Syria, supplemented the expenditure incurred from central government funds. The sijills of the Shari'a court for the year 945/1538 give details of this. Similarly, craftsmen were imported from as far afield as Cairo and Aleppo. For reasons of security, work began on the north side, which was the most vulnerable, and ended on the west side. The urgency with which the project was driven forward helps to explain not only the short cuts outlined above but also the frequent use of spolia. That same desire to save time and expense explains why some of the roundels and circular medallions which ornament the walls at irregular intervals (for instance in the area of the Damascus Gate) turn out on close inspection to be the re-carved ends of cylindrical columns inserted into the walls as strengthening devices. Most of the carved roundels, however, were cut from square blocks specifically for use on the walls. Numerous inscribed plaques document the progress of construction; the bare facts and dates are interspersed with verses of thanksgiving. The walls of course have their later history, notably the alterations to the New and Dung Gates, to the glacis and to the moat of the citadel in the course of the 18th and 19th centuries; but their form today is sufficiently close to the way that Sulaiman's engineers left them. 4. The role of Jerusalem in the Ottoman Levantine context While there is no question that in matters of religious significance Jerusalem was unchallenged by any other city in the Levant, and was outranked only by Mecca and Madina in the Islamic world at larg e, its political and economic role was quite another matter. Egypt, in Ottoman as in medieval times, was the major force in the south-east Mediterranean, and was thus in many ways a law to itself. Comparisons with the major cities of the Levant in Ottoman times, namely Aleppo and Damascus, are therefore more to the point. Jerusalem shared with these cities the formula traditionally followed in Arab cities, whereby public buildings were concentrated in a city centre criss-crossed by major arteries, with the private residential zone further out and linked by smaller, irregular streets, alleys and culs-de-sac. But the main fact to bear in mind is that, by the standards of the major Levantine cities, Jerusalem for most of the Ottoman period was tiny, both in its physical scale intra muros and in terms of its population. In about 1525 Damascus, even after the destruction inflicted on it by Timur a little over a century earlier, had, according to the Ottoman registers, a population of approximately 57,000. In that same year, those registers show that Jerusalem had a population of some 4,000 inhabitants; this tripled during the reign of Sulaiman the Magnificent, but incompetent government and a resultant decline in revenues-for the city was essentially a tax farm administered by Ottoman officials in Egypt, Sidon or Damascus for their own benefit-steadily impoverished most of the inhabitants. This very small population suggests that, while there might well have been pockets of dense habitation in the area around the Haram, Ottoman Jerusalem as a whole cannot have been heavily built up. Supporting evidence for this assertion comes from the waqf of the Qadiriyya zawiya, which mentions a big square haud for plantation and 'a fair hakura (orchard) situated to the east of the zawiya, approached from the zawiya and planted with figs, almonds and grapevines.' As late as 1806 the number of inhabitants was estimated at 8,000, at a time when, according to Sauvaget, Aleppo had a population of about 250,000 . It is well to ponder these figures, and the economy which they reflect, when one tries to find an adequate context for Ottoman Jerusalem and its architecture. To put it brutally, the city was poor, and such money as it generated came mainly from non-Muslims. In the 16th century the two major sources of government revenue in the city were the toll levied on visitors to the Holy Sepulchre and the poll-tax imposed on Christians and Jews. The two-stage conversion of a local Christian church into the mosque of Nabi Da'ud, a process completed in 930 / 1524, illustrates another aspect of this anti-Christian policy. The commercial activity of the local Arab population could not begin to compete with that of Damascus or Aleppo, and thus could not support the kind of building programmes undertaken in those cities. It is therefore no wonder that the patrons for the buildings of Ottoman Jerusalem belonged largely to the Turkish governing class . Moreover, the relative obscurity of Jerusalem in this period and the short tenure of office which these officials could expect in turn meant that they did not on the whole wield significant power or wealth; they were not rich grandees by the standards of Damascus or Aleppo, let alone Istanbul. All this helps to explain why the buildings of Ottoman Jerusalem are so modest in scale and decoration. A wider context for these buildings can be found by examining Aleppo and Damascus in Ottoman times. Aleppo can serve as an exemplar of a great Levantine city under Ottoman rule . As the capital of a vilayet, it enjoyed a strategic site athwart a major trade route to the West, and this was reflected by the presence of permanent European mercantile enterprises, or factories, in the city. While European interest focussed on the international textile trade, which embraced both cheap cottons and luxury fabrics woven with gold and silver thread, Aleppo was also the major centre for regional trade, 5 Ottoman Jerusalem and remained so throughout the Ottoman period. To this day its gigantic covered bazaar, a labyrinth some 7km long, largely an Ottoman creation and perhaps the most impressive in the Arab world, has not degenerated into a tourist trap but is still a living, functioning organism, an entrepot for hundreds of commodities. Aleppo has remained a mercantile town with its face to the desert. Hence, in part, its prodigious economic vitality, especially in Ottoman times. It was helped in this commercial role not only by its favourable geographical situation, with easy access to the sea, to Anatolia and to Mesopotamia, but also by its large multi-ethnic population. Indeed, in Arab lands it was second in importance only to Cairo, as its large central zone, 10. 6 hectares in extent, testified. The Ottoman period saw a 50 percent increase in this area. The Ottoman architectural contribution to Aleppo is seen at its most characteristic in the residential quarters (mahalla; there were 72 in all) built of high-quality stone masonry. They include far more residences and palaces for the elite than does Jerusalem. The best houses were clustered nearer to the city centre, and were typically between 400 and 900 square metres in ground area, as compared with the 80-190 square metres of the houses built for the middle class and the haush system of multiple low houses clustering around a shared courtyard, or on either side of an alley, which-as at Damascus and Jerusalem - was the preferred housing for the urban poor. The standard pattern of elite housing in Aleppo is of an inward-looking structure dominated by a spacious reception room and an iwan overlooking a courtyard (qa'a) often embellished with a pool, vines and citrus trees. The decoration draws on local craft traditions which were already well established in the medieval period: inlaid marble in various colours and patterns, and woodwork both carved and painted. This domestic architecture is typified by Bait Jamblat and, in the Judaida quarter, a group of 17th-18th century residences-the Ghazali, Sadir, Ashikbash, Wakil, Balit, Dalla!, Sayigh and Basil houses. In commercial architecture, Aleppo is distinguished by a series of ambitious caravanserais. These include Khan al-Wazir, with its offices ranged around its courtyard, and its monumental portal executed in ablaq masonry, Khan al-Nahasin and Khan al-Kattin. Some contain mosques, and many are graced with windows whose frames display delicate carving with vegetal and geometric themes. Taken together, these caravanserais underline the scale of the city's trade in Ottoman times. But some have other dimensions too. A waqf of Dukaginzade Mehmed Pasha dated 1555 mentions a great mosque, three khans, three qaisariyyas and four suqs, the whole covering some three hectares. Scarcely less impressive is the Khan al-Gumruk (after 1574), or 'Customs caravanserai', which had two rows of suqs adjoining its fac;:ade, and with its 344 shops covered an area of 8,000 square metres. But 6 it also lodged the banking houses of the French, English and Dutch, who, following 'capitulation' treaties with the Sublime Porte, had settled in the city from 1562, 1583 and 1613 respectively; here too resided the consuls of these three powers. The Venetians, by contrast, had their own fondaco in another quarter. This was the age of the Marseilles merchants and the Levant Company; by 1662 the English factory in Aleppo numbered some fifty traders, and as late as 1775, even in a period of economic d~dine, there were 80 European firms represented in the city, mainly clustered in the Frankish quarter. Many baths, fountains and workshops for textile manufacture also survive from this period. The domestic and commercial architecture of Ottoman Aleppo is to this day the dominant accent of its urban environment, and is only lightly leavened by buildings of a religious purpose; in that respect, the contrast with Damascus, and even more so Jerusalem, is dramatic. While these vernacular, palatial and mercantile structures are firmly within the local Syrian tradition, the higher-profile public religious architecture in Aleppo, typified by various gubernatorial mosques-the 'Adiliyya (1555), Bahramiyya (1583) and Khusraufiyya (from 1537)-sometimes looked further afield; thus the latter complex, attributed to the court architect Sinan, exhibits close links with the metropolitan architecture of Istanbul. What of Ottoman Damascus? Like Aleppo, it was a provincial capital; participated in the European trade (for example by exporting the many kinds of damask made by the local women in their homes); had separate quarters for Jews and Christians; and expanded dramatically, especially to the south, as the Maidan suburb shows-in fact, the population doubled in the first three centuries of Ottoman rule. It was in the quarters that the life of the city was concentrated; indeed, Hanafi jurists distinguished between the public and the private zone of a city. These quarters were isolated, a fact symbolised by their being closed at night. The Muslim quarters typically had an oratory, mosque, bath and non-specialised markets, and were thus furnished with all that was needful for community life. Not surprisingly, this sense of closeness had its drawbacks; for example, conflicts with neighbouring quarters were common. Many quarters had a dynamic religious life centred on Sufi brotherhoods, probably because they were close to rural zones. In all these ways Damascus conformed to a pattern that was widespread in the contemporary Near East. Like Aleppo, it boasted splendid palaces for local notables-for example, the Dahda palace of the 17th century, the house of Nur al-Din, now partially re-erected in the Metropolitan Museum, New York, and dated 1707, or the 'Azm palace of 1749-52, as well as smaller houses of quality like Bait Nizam and Bait Siba'i. Its favourable situation between the intensively cultivated Bik'a valley of Structure, Style and Context in the Monuments of Ottoman Jerusalem the Lebanon and the fertile Hauran plateau made it the natural market for central Syria, and it was also the nexus of several arterial routes and a natural entrep6t for the Mediterranean ports. Like Jerusalem, moreover, its commerce was invigorated by pilgrim traffic-for Damascus, the last major town before the dangerous desert crossing, was a major stage in the hajj route. Like Jerusalem, again, it benefited from lavish imperial patronage in the 16th century-notably two takiyas, complexes intended inter alia to accommodate pilgrims on the hajj route. These foundations show how seriously the early Ottoman sultans took the responsibilities attendant upon their honorific title 'Protector of the Two Holy Cities', and thus fit neatly into the pietistic context of early Ottoman architectural patronage in Jerusalem. The takiyas are named Salimiyya and Sulaimaniyya after the successive sultans who paid for them; the Salimiyya ( 1518) was constructed over the tomb of Ibn al-'Arabi, a noted Sufi, which brings to mind the carefully fostered Sufi connections of certain Ottoman sultans. The Sulaimaniyya (1554-5) is the work of the court architect Sinan, who also designed the 'Imara of Sultan Sulaiman, which was intended for the distribution of food to indigent pilgrims visiting the tomb of Ibn al-'Arabi. Completed in 1552, it is used to this day as a bakery; the parallels with the Khassaki Sultan complex in Jerusalem are instructive. Other expressions of government patronage include several Jami's whose local striped masonry in limestone and basalt is only a veneer for the underlying metropolitan character revealed in such details as hemispherical domes, pencil-shaped minarets and courtyards surrounded by multiple domed bays. Many of the great religious foundations of the Ottoman period were situated outside the city walls, a marked contrast to the situation in Jerusalem and clear proof that Damascus had expanded well beyond its medieval limits. The mosque and mausoleum of Darwish Pasha (1571-5 and 1579 respectively), the Khan al-Harir (1572), also due to him, the mosque of Sinan Pasha ( 1586-91), the Qaimariyya mosque (1743) built by Fathi Efendi, an official of the Ottoman treasury, the Khan al-Gumruk built by Murad Pasha in 1608-9 and the palace, madrasa and khan of Asad Pasha al-'Azm all testify to the patronage of enlightened nobles. Well over a dozen Ottoman khans survive; these, unlike so many of the important religious buildings, are intra muros. Their size and number clearly reflect the intense commercial activity of Damascus in this period. Only Aleppo, as already explained, among the other cities of the Levant can match this mercantile investment; the almost total absence of such buildings in Ottoman Jerusalem speaks for itself and goes far to explain the much more modest scale of the Ottoman architecture there . The emphasis on tilework in many of the Ottoman buildings of Damascus again brings Jerusalem to mind; Damascus in the 16th and 17th centuries was a thriving centre for the manufacture of glazed ceramics and tiles which are a subset of Iznik wares. But Damascus signally lacked stable government. The bare facts are sufficiently telling: between 1516 and 1697 the city was governed by no less than 133 pashas. Few of them did the city much good, for all that the mosque and mausoleum of Darwish Pasha (16th century), and the palace and khan of Asad Pasha al-'Azm helped to beautify the city. It is no accident that the greater prosperity of the city in the 18th century was co-terminous with the tenure of the governorship by members of the 'Azm family for most of that time. These brief sketches of Ottoman Aleppo and Damascus are enough to show that Jerusalem belongs less with them than with, say, Gaza, Hebron, Nablus and Tripoli in this period. For example, in the later 17th century Gaza, then enjoying a period of prosperity , acted as the capital of Palestine, and possessed an attractive set of buildings. As for Hebron, the religious significance conferred on the city by the tombs of the Patriarchs (e.g., Adam, Joseph, and especially Abraham), and the quantity of Muslim pilgrims visiting them - for just as in popular Islam Jerusalem developed as a substitute for Mecca as a pilgrimage destination, so did Hebron develop as a substitute for Madina - ensured that the Ottoman sultans oversaw its upkeep . 5. The role of the Haram in Ottoman times To build on the Haram al-Sharif - as distinct from building along the outer perimeter of its two terraces-was not a straightforward proposition. It seems probable that there was something of a taboo in operation so far as building within the precinct itself was concerned; hence the very tentative use made of this prime site in post-Umayyad times and even under the Mamluks. On the other hand, the Ottomans were new to Jerusalem and had not been inured to such a taboo. Nevertheless, the striking fact is that nothing at all was built in the city in the reign of Selim the Grim (1512 -20) or, more to the point, in the first fifteen years of the reign of his son and successor, Sulaiman; and when Sulaiman began to build, the Haram had only a minor role in his plans. Two sabils there document this stage of his patronage: Sabil Qasim Pasha (whose inscription exalts him as the Second Solomon) and Sabil Bab al-'Atm. Sabil Qasim Pasha serves for ablutions and is referred to in the sijills as Hanafiyya; in other words it provided the running water which this particular madhhab requires for ablution. This proves, incidentally, that not all sabils were built only to furnish drinking water. The major new foundations of the new dynasty were the complexes of Bairam Jawish and Khassaki Sultan, both of them in the 7 Ottoman Jerusalem immediate surroundings of the Haram and thus firmly in the Mamluk tradition, but both of them afflicted with a layout that at every turn betrays the acute shortage of space that cramped the architect's design. Indeed, the Khassaki Sultan complex incorporated into its design a hall which was part of the Mamluk foundation of Dar al-SittTunshuq; this was transformed into stables. Clearly there was very little prospect of significant new building activity in this area, where-as noted above-space was at a premium and the detailed provisions of scores of waqfs further hampered the free exercise of an architect's imagination. The gap sites were getting smaller all the time . This made it all the more important to extract maximum advantage from them. Thus even a modest little building like the Khalwat al-Dajani is visible from several different angles and is at an important intersection of traffic. Yet the decision to build within the precincts of the Haram had its own problems. The prestige of the Umayyad monuments was unmatched, for they had the whole weight of Islamic history-including salvation history - behind them. Above all, ample empty space was an essential aspect of the awe-inspiring presence exerted by the Dome of the Rock and the Aqsa Mosque . It was therefore not a serious option to build major new monuments on the Haram and thus to risk upstaging these numinous landmarks. Political considerations also had to be taken into account, for the Ottomans, foreign conquerors whose base was in distant Anatolia right outside the Arabic-speaking world, could not afford to alienate their Arab subjects by intrusive and insensitive construction projects on this immemorially hallowed site. The upshot of all this was that the Ottomans proceeded with extreme circumspection in their development of the Haram area. Sabils apart, not a single Ottoman sultan, it seems, erected a brand-new structure on the Haram . Given the massive imperial building programme between c. 1460 and c. 1620, which utterly transformed the face of Istanbul and of numerous cities in Anatolia and the Balkans, this restraint in the third holiest city of the Islamic world is truly remarkable. Yet by slow degrees the Ottomans did contrive to establish their physical presence on the Haram. The key point to remember is that virtually none of the buildings which did this were imperial. The trickle of minor, modest buildings never became a flood, but by 1650 or so they permeated the entire precinct. Moreover, quite apart from their impact as individual buildings, they also worked in concert by virtue of the way they clustered together. Nor should one forget the dimension of sound-prayers could be heard through their open windows and sanctified the entire environment. Thus the Ottoman claim to the Haram was staked very modestly and over many decades. This extreme caution brings out the extreme boldness of Sultan 8 Sulaiman's masterstroke, which was nothing short of a coup de theatre: the glamorous refurbishment of the exterior of the Dome of the Rock. It was the perfect way of establishing in visual terms the Ottoman dominion of the city, and the fact that it had come to stay. In rather different language, it reiterated the political message of the Umayyad caliph 'Abd al-Malik nearly nine hundred years earlier; and it was aimed at the same audience: the people of Jerusalem. But precisely because Jerusalem has for millennia had such wide horizons in religio-political terms, the message of the second Solomon, broadcast from the site of the first Solomon's Temple, reverberated throughout the Islamic world-and beyond it . The constant flow of pilgrims, Muslims and non-Muslims alike, to Jerusalem would have seen to that. Sulaiman's immediate audience, of course, was an Islamic one. He therefore had no need to claim for Islam as had 'Abd al-Malik-a place of worship long held sacred by other faiths. But he certainly renewed that claim and thus invited comparison with his great Umayyad predecessor. He sought, moreover, to improve on that aspect of the Umayyad achievement which was most obvious to all, namely the colourful exterior mosaic of the building . Especially from a distance, glazed tilework in blue and white (the predominant tones of this composition) has much more visual impact than the more muted palette of mosaic . No evidence has been found to suggest that this mosaic had suffered serious decay . But the history of Islamic architecture is full of examples of later patrons remodelling the decoration of much earlier monuments so as to put their own personal stamp on them . This is what Sulaiman did . It seems unlikely that the good health of the building-or that of the neighbouring Dome of the Chain, which was included in this refurbishment, though a little later, in 1561-2-demanded it. Indeed, he forbore to attempt any major structural intervention. Instead, by sheathing the exterior of the Dome of the Rock in glistening tilework he upstaged his Umayyad predecessor, brought up to date one of the most seminal structures in the Islamic world and established the Ottoman presence right in the middle of the Haram. This was something that not even the Mamluks had done, despite their sustained embellishment of the Holy City over a period of almost three centuries . They certainly carried out work on the Haram, particularly on its inner fa<;:ades,and even a certain amount of repair and refurbishing of the Aqsa Mosque and the Dome of the Rock, where the largest single project was Qala'un's repainting of the inner dome . But they attempted nothing so grand and of such public impact as tiling the lower exterior walls of the Dome of the Rock . It is perhaps a measure of Ottoman imperial pretensions that the project of loading this, the first great monument of Islamic architecture, with redundant ornament should ever have been conceived . It was only the Dome of the Rock that was Structure, Style and Context in the Monuments of Ottoman Jerusalem so singled out; no significant structural work was undertaken at the Aqsa mosque for almost six centuries after 1350, nor did it offer a comparable opportunity for redecoration. Scattered references in the sijills record such minor running repairs as the replacement of lead sheeting for the mosque' s roof in 996/1588 or the repair of its doors. It has been suggested that the motive for the redecoration of the Dome of the Rock was a misunderstanding of the themes found in the Umayyad mosaic programme, namely that the depictions of 'cherubim' (probably Sasanian winged motifs) described by Western travellers of the late 15th and early 16th century stimulated an iconoclastic reaction. If so, it was at second hand, for Sulaiman never saw the building, and it seems odd that it took over eight centuries to materialise. It seems more likely that the intention was simply to assert the Ottoman presence in Jerusalem by redecorating the city's most celebrated monument. Sulaiman gave the building a new meaning through the choice of epigraphy: the drum bears Sura 18: 120 and the date (in figures) 952/1545-6, while lower down the main octagon asserts that the work was carried out by 'the most able masters of their age' and is signed by one 'Abdallah al-Tabrizi, a Persian to judge by his nisba, and dated 959/1551-2. Nor did Sulaiman's work end there, for shortly before his death he caused the doors at the eastern and western entrances to the building to be covered with bronze plaques. The main Ottoman contribution to Jerusalem apart from the walls, the aqueduct and the retiling of the Dome of the Rock is the transformation of the upper terrace of the Haram al-Sharif, on which the Dome of the Rock stands. It is important to note from the outset that this transformation was not part of a master plan for the Haram; it happened piecemeal over more than a century. Before 1517, the edge of the upper terrace bore only the minbar of Burhan al-Din, the colonnades and the Ayyubid Turbat al-Nahawiyya. Even before the arrival of the Ottomans, every post-Umayyad structure on the upper terrace was small, discreet and had lots of space around it. This illustrates a sensitive response to the problems of the site and the Ottoman buildings respected this approach. It is noticeable, for instance, that the fa~ades of the hujras and similar structures on the upper terrace are uniformly single-storeyed whereas their fa~ades on the lower terrace or esplanade are two- or even three-storeyed. This is a very significant contrast. It meant that, in a most unobtrusive way, these khalwas were able to serve as a bridge between the lower and the upper terrace . Several other ways of sanctifying the Haram are worth brief mention here . The mastaba- a very slightly raised platform used for open-air prayer-helped assert a presence on the Haram without breaking up the familiar view.The finial of most of the cells and aedicules on the Haram is Mecca-oriented, another (if understated) way of emphasising their religious function, while the !wan alSultan Mahmud II is aligned to the Dome of the Rock. More distantly, high-rise buildings located relatively far from the Haram might nonetheless be designed to offer sight-lines to its monuments. Thus a view of the Dome of the Rock can be enjoyed from the Zawiya alMuhammadiyya, though from its side chambers rather than the main ones . It is worth reflecting briefly on the nature of these Ottoman structures. It is hard to interpret a single one of them as justified by an imperative religious, social or public need. In any event, their tiny size would have militated against their performing any such function. By and large, these are not buildings erected for the public good. Once again, one sees Ottoman patrons paying the price of arriving too late on a site already hallowed by Islam for almost a millennium. One suspects that some of these cells reflect a low-key competition among minor Ottoman functionaries for a place-any place-on the Haram. After all, some 60 percent of the Ottoman buildings in the city are concentrated there, and this trend accelerated markedly in the last two and a half centuries of Ottoman rule. All this is not to deny that these cells do serve a wider purpose in the architecture of the Haram. Their location shows a consistent desire to line the perimeter of the upper terrace-the prime focus of Ottoman building activity-with minor monuments. They form a kind of cordon sanitaire. The Kursi Sulaiman, strategically placed at the north-east corner of the Haram, seems to express the same idea. It remains an open question whether the parcelling out of the available lots on the site was governed by any hierarchy or notion of privilege . Perhaps this might explain why the west and north sides were built up rather more heavily than the east side, which is largely empty to this day. Fifteen cells were built around the west, north and east sides of the upper terrace of the Dome of the Rock. Five of these are of a higher standard than the rest (cat . nos. 21, 22, 23, 24, 34 and also the madrasa of Ahmad Pasha, cat. no. 25). They are the work of one patron and one decade and probably one local architect : 'Abd al-Muhsin ibn Mahmud ibn Nammar. The family produced at least three generations of architects (see end of cat . no. 22). Certainly the lower terrace contains far fewer minor buildings than does the upper terrace, and a difference in status between the two areas seems a possible explanation. What, then, were the options open to the Ottomans as successors to the Mamluks and guardians of Islam's third holiest city? It must be conceded that they were unpalatably limited. After all, the immediate surroundings of Haram were already crammed with impressive Mamluk monuments and there was no obvious area in the most prestigious part of the city in which an Ottoman enclave could be built from scratch. True, much of the Haram was empty; but, as already noted, that 9 Ottoman Jerusalem emptiness was a necessary element of its design and function. Thus the Haram was indeed available, but only for small-scale structures; and such buildings necessarily sent out a somewhat modest message. It is scarcely surprising that the patronage which produced them was not royal. It might be argued that, for all their modest size, these buildings could have had more impact had their decoration been more lavish. After all, there was the possibility of spectacular colour in the Ottoman as in the Mamluk architecture of Jerusalem because of the wide choice of coloured stone, for example yellow, red and pink. Such polychromy was indeed used, but only very sparingly, which again points to the modest ambitions if not also the modest means of the patrons . Moreover, the Ottoman architectural style did not favour applied ornament, in contrast to the Mamluk style which virtually depended on it. Thus there was no way that the Ottomans could have outdone Mamluk patronage by fancy decoration; they would simply have had to build on a much larger scale. Moreover, even if the money had been available to do this, it is a moot point whether the tiny population of the city could have justified such expense. For Sulaiman, Jerusalem belonged in a much wider pietistic context . Most of the major architectural projects of his entire reign were of a religious nature, and they serve to put his works in Jerusalem into perspective . At his orders, in Baghdad the tomb of Abu Hanifa - the founder of the favoured Ottoman madhhab--was restored and enlarged; so was the mosque over the tomb of Jalal alDin Rumi in Konya, an act which publicly affirmed that royal favour for Sufism which was foundational for the Ottoman state. Above all, in Mecca he restored the Ka'ba and the aqueducts and in Madina he restored the Masjid al-Nabi as well as building walls 12 metres high around the city, with a ditch in front of them. All this building work was clearly consistent with his activities in Jerusalem, and was of a piece with the Ottoman decision to arrogate to the sultan the appointment of the chief qadi of Mecca, formerly a privilege of the Grand Sharif of Mecca. Other sacred sites in Palestine also benefited from his patronage. Thus when, in 1552, the domes over the tombs of the patriarchs in Hebron collapsed-they had not been repaired since the time of the Burji Mamluks-orders for their repair were issued from Istanbul. Indeed, a stream of such orders for the repair of the Hebron shrine, together with provisions for creating extra waqf property to help with its upkeep and for looking after the stream of pilgrim traffic, and orders for skilled craftsmen to be sent there from Damascus, shows _ how seriously the Ottoman sultans took their obligations as guardians of the major Islamic shrines. Another example of a shrine that was extensively repaired and enlarged in the Ottoman period is the maqam ofNabi Musa. It is possible, incidentally, that -in undertaking this impressive sequence 10 of pious foundations Sultan Sulaiman may have had it in mind to match the achievements of the Mamluk sultan Qa'itbai, whose memory would still have been green in the early 16th century and who, besides erecting a sabil in the Haram al-Sharif, built aqueducts for Jerusalem; he also erected a madrasa adjoining the Masjid al-Haram in Mecca and reconstructed the Masjid al-Nabi in Madina and its domed tomb. That said, Sulaiman's patronage in Jerusalem pales in comparison with his architectural enterprises in Istanbul-though in both cities there is a notably strong welfare element in his foundations . Finally, the role of the Haram as a magnet for pilgrims needs to be taken into account. Nasir al-Din Rumi, writing in the mid-16th century, makes it clear that there is a set itinerary for the Muslim pilgrim to the holy places in Jerusalem. Perhaps the most famous guide to the Holy Places-Mecca, Madina and Jerusalem-is the Dala'il al-Khairat of the Berber mystic al-Jazuli (d . 869 / 1465), which, with much other material such as prayers for the Prophet, outlines these itineraries . Illustrated copies of this work with detailed views of the three Holy Cities became very popular in Ottoman times . Thus a list of key religious sites was easily available, and was often used by those seeking to acquire merit by such pious visitations, and by those paying for pilgrimage by proxy. The parallel with the Stations of the Cross is striking . 6. The nature of Ottoman in Jerusalem patronage There is an obvious distinction to be made between royal and local patronage. To the first category belong the walls, the aqueducts, the re-tiling of the Dome of the Rock, the sabils and the Khassaki Sultan complex . With the exception of the sabils-which though beautiful enough in all conscience, and undeniably Islamic, are cheap - these all represent major financial investments in the city. All of them, moreover, the sabils included, are high-profile enterprises; they affected directly the daily life of every citizen of Jerusalem. The aqueducts and the associated sabils brought clean water within reach of everyone; the walls and the rebuilt Citadel assured the security of the whole population; the re-tiling of the Dome of the Rock could have been seen as a lavish gesture to honour the city's most famous and charismatic monument; and the Khassaki Sultan complex (like its eponymous counterpart in Istanbul, datable to 1536, which also fitted into a gap site and is also still functioning) was a massive welfare project targeting the poorest members of society. The sheer scale of that project is reflected in the size of the cauldrons from the kitchens, which are preserved in the Haram museum. All this is patronage on a grand scale. How is one to interpret this continuous enterprise? Structure, Style and Context in the Monuments of Ottoman Jerusalem Since all of these works were carried out within the space of half a generation, their cumulative effect is hard to exaggerate.' They far outshone any single Mamluk contribution to the embellishment of Jerusalem, and seem to lack that element of direct personal self-interest that marks so many medieval Islamic charitable foundations. They illustrate a type of patronage distinctively different from that of the Mamluks, for they are aimed at society at large rather than at a small segment of it-those people who would, for example, benefit from a ribat, madrasa or khanqah. Most Mamluk patronage was of this more specialised kind, although it did consistently lean towards private charitable institutions. Naturally there are exceptions: the work of Sultan Hasan in rebuilding the north-east part of the Aqsa and of Sultan al-Malik al-Nasir Muhammad on Suq al-Qattanin do have wider perspectives. But Ottoman patronage was different even from this. It smacks of a plan drawn up in Istanbul rather than one worked out on the spot. This patronage at arm's length, so to speak, makes it unlikely that Sulaiman was actuated by a desire to match the patronage of the Mamluks in Jerusalem, or at least to continue their work. Enough has been said to indicate that royal Ottoman patronage fostered to a degree unprecedented in Islamic Jerusalem both the practical daily welfare of its people and their lively sense of the special destiny of their city. He was its last and greatest patron and he never set foot in it. Moreover, his interest in the city spanned almost his entire reign, from the Sabi! Bab al-Mahkama (Qasim Pasha) of 1527 to the refurbishment of the Dome of the Chain in 1561-2. No later Ottoman sultan achieved very much in the way of architectural patronage in Jerusalem, and this too highlights the scale of his work there. Yet, as already noted above, this undeniably impressive achievement is dwarfed by the scale of his foundations in Istanbul itself, notably the gigantic kulliye named after him, of which the mosque is the core, half-a-dozen other large mosques, several madrasas and-as at Jerusalem-the reconstruction of the water system. The fact that no major new mosque was built in Jerusalem by any Ottoman sultan, while in Istanbul a whole series of them was erected in the 16th century, speaks for itself. Nor was it a matter of new mosques being confined to Istanbul; Cairo, for example, saw several major foundations in this period, such as the mosques of Sinan Pasha (1571) and Malika Safiyya (1610), to say nothing of literally hundreds of smaller buildings, including scores of sabil-kuttabs . In fact, this type of monument-virtually unknown in Ottoman Jerusalem was almost a trademark of Ottoman Cairo; fifty of the sixty-three Ottoman sabils in Cairo have a kuttab or maktab in the upper floor, which comprises a single chamber . Local patronage in Jerusalem was of course a much more modest affair. Indeed, anonymous patronage was virtuous according to some religious teaching . But whether the patrons were named or not, the critical factor was, as ever, not a sudden drop in the capacities of the craftsmen, but in the amount of cash available for building activities. After c. 1600 political and economic decline, which had set in after the death of Sulaiman the Magnificent, accelerated, and restricted still further the scope of architectural patronage. But even so, in the previous century the absence of foundations by Sinan, the Chief Architect (mi'marbashz) to the Ottoman court from 1538 to 1588, is marked, especially as his work is so widely found in the Ottoman provinces. In the case of non-royal foundations there is no evidence of strict control imposed from Istanbul. Most of the buildings of Ottoman Jerusalem were built by local men for local patrons . The city was given by the Ottoman authorities to its successive governors as a tax farm, and since their term of office was often no longer than two or three years, they had a strong incentive to make as much money out of the post as possible before relinquishing it . This was emphatically not a situation which encouraged such men to spend their money on public works. Similarly, in Ottoman Cairo the brief tenure of office allotted to the Turkish governors discouraged them from erecting major buildings. In Jerusalem, the major exception to this trend favouring small structures are the foundations of Bairam Jawish . His maktab follows a familiar Mamluk pattern in that its prime function was not its ostensible one-i.e. a school for orphans - but to act as a burial place for Bairam Jawish himself, and the eastern chamber of the ground floor is given over to this purpose. Thus anyone entering the building could not have failed to notice the founder's tomb. But this tomb is not a self-contained structure; it is merely a cross-vaulted lower room in a much larger structure. Thus the Mamluk obsession with massive mausolea has diminished. But it remains a problem as to how children were taught in the context of this building, how big the classes were, and what age groups were catered for here. The simplicity of the facilities - essentially a succession of large halls or chambers - would lend itself to multiple uses. Perhaps this was indeed the original intention . Bairam Jawish was the administrator in charge of the neighbouring Khassaki Sultan, and thus he lived 'above the shop'. He would therefore have been well placed to oversee how his own foundation was being used. The standard method of securing the future for a charitable foundation was, as it had been for many centuries, the pious endowment or waqf. Although the central government supported the Aqsa mosque and several other buildings (such as Kursi Sulaiman) by a sum (al-surra) sent annually to Jerusalem from Istanbul, most waqfs, including the best-documented ones, are those of the traditional Jerusalem families . Many others are recorded in the siJills, and a few specific examples will reveal something of how the system worked. Thus the waqf 11 Ottoman Jerusalem of Bairam Jawish for his ribat (which functioned as a hospice for the poor) and maktab (here, a school for orphans) was originally registered in Gaza, even though it referred to a property in Jerusalem. Much of the waqf property in this case was located in Bethlehem. If Bairam Jawish's family died out, the endowment stipulated that the inspector of his waqf was to be the Nazir al-Haramain alSharifain (i.e. superintendent of Aqsa and Hebron). In this particular case, the rate of interest charged on the capital sum was 15 percent. This rate continued for another generation; the waqf of the hujra of Islam Beg, dated 1002 / 1593-4, was a capital sum of 500 gold sultanis, to be administered so as to yield a legal profit of 75 (i.e. 15 percent); the mutawalli was to avoid usury (riba). The money was to be spent on administration and on seven reciters of the Qur'an (the benefit of the readings and associated prayers was to go to the founder, his brother, the founder's family and all deceased Muslims). About a century later, Yusuf Pasha made a waqf of 95 ghirsh as'adi (gold coins) to pay for two reciters in the Dome of the Rock, oil to illuminate the cave of the Dome of the Rock and for people in charge of lighting the lamp, and for the administrator of this waqf; the annual amount was to be 14 and one third ghirsh, i.e. about 15 percent. Soon after, however, in 1665-6, waqf money was attracting interest at the standard rate for all transactions: an extortionate 20 percent, as the case of the Khanqah al-Maulawiyya shows. Here part of the original capital had been lost in the course of successive administrations. The running costs of this building were met in that year by the income from two houses, a single room, a stable, three chambers, a storage place, and a plot of land-all in the Maulawiyya complex or nearby. The waqf of the Sabi! al-Khalidi is typical in the scrupulous care with which it defines the exact site of the building to be endowed; indeed, the setting for this vanished sabil has scarcely changed to this day from what it was in the early 18th century . Another waqf for the Khanqah al-Maulawiyya specifies 3kg of bread and 6kg of meat (presumably provided daily ?) and a reduction of the allowance to 1,470 piastres a month to feed all members of the order. The waqf of the North-Western Khalwa of Ahmad Pasha specifies that some of the money should be set aside for oil for the lamp (this is a common provision in the waqjs of Jerusalem) and for restoring the doors of the building. The waqf of the Qubbat al-Arwah, too-an openplan aedicule in the middle of the largely empty upper terrace of the Haram-provides for a lamp to be lit throughout the doors of darkness, though the purpose of this provision is not specified. The waqf of the Khalwat Bairam Pasha mentions that qusurmil (building material) was provided for the monument, while that of the Zawiya al-Muhammadiyya covered such unusual annua l expenses as mats for the floor of the zawiya and posts to support the 12 trellis of the grapevine over the main entrance and south wall of the zawiya. In other cases cash in the form of direct salaries, grants and stipends predominated. Thus the waqf of the madrasa of Ahmad Pasha states that the teacher (mudarris) will have a salary of 500 silver misriyya per year, and each of the four students a stipend of 50 silver misriyya per year. Teacher and students were to pray for the patron and his ancestors. The sabil of Mustafa Agha had a waqfof 40 gold coins, and the caretaker was allotted four gold coins annually, to be paid so long as this capital sum remained intact. In some cases cash and kind were combined. For instance, a waqf of 1038 / 1628-9 for the hujra of Muhammad Agha provided money for a Qur'an reader who every morning was to recite Sura 36 (Ya Sin) in the upper part of the hujra; the merit of the reading was for the soul of the Prophet Muhammad. For this he got 12 gold pieces a year, plus permission to live in the lower part of the building and to profit from cultivating the small garden attached to the hujra. Similarly, the waqf of the Sabil al-Shurbaji, which was based on the income from three and a half shops, paid for two people to run the sabil: a caretaker to keep it clean and illumine it during Ramadan and Sha 'ban, and a water-carrier; while the waqf of the Odat Arsan Pasha stipulates that the reader had to recite the Qur'an and the normal petitions plus prayers to the Prophet and invocations to Allah. The waqf of al-Zawiya al-Naqshbandiyya is one of the more interesting and informative examples and shows how a waqf could be rooted in the local neighbourhood and affect the local economy. It provides for only four members of the order to be housed in the building, and by the second half of the 19th century it was serving the Jerusalem poor with accommodation and food. Half of the waqf was for the descendants of the donor. Ampng the properties which were made waqf was a vegetable garden in the Ghawanima quarter. The income from the waqf was I62 piastres a year. The donor reserved for himself the position of caretaker (not a pension, because only 6 piastres a year were allocated for this post). Each Sufi was to receive I piastre per month for food and half a piastre for accommodation (also per month). By contrast, the 1633 waqfiyya by Muham mad Pasha, the governor of Jerusalem, in favour of the Zawiya al-Qadiriyya, is an entirely cash operation: 1,000 silver ghirsh as'adi. The administrator of the waqfwas enjoined to deal with this sum legally, producing 12 ghirsh for each ten ghirsh, and he was to avoid riba. Yet the rate of interest here is extortionate. The annual income was to be 200 ghirsh as 'adi. The principal calls on this sum were the 11 Sufis resident in the zawiya (6 ghirsh annually each, i.e. 66 in all) and the food cooked for them every Friday .night (72 ghirsh). How they ate the rest of the time remains unclear. The provisions of the waqf ensured that the 11 Sufis stayed Structure, Style and Context in the Monuments of Ottoman Jerusalem put, not being allowed to travel or to transfer their rights to another;and it was also mandatory for them to gather after everyprayer and recite the Qur'an, for the benefit inter alia of the donor and his family. Thus the stipulations of the waqftranslated into strictly enforced rules for the daily life of the beneficiaries. As to the kind of buildings that were put up, they are for the most part small and insignificant, as if the intention were that they should blend into their surroundings and make no splash. In any case, the shortage of space meant that architects had to build upwards and not sideways,and not only had to take account of existing buildings but also to incorporate parts of them. This is especiallytrue of the area around the ribat and dar of Bairam Jawish, which was a prime site for Ottoman building projects. Sometimes only a small detail, like a carveddecorative band on that building, which betrays the remains of the original cornice, gives clues as to how the original structure looked. But small as most of these Onoman buildings are, they are still a prise de possession:like the much more splendid Mamluk buildings, they assert the presenceof an external authority. The difference is that the nature of these buildings suggests that the Ottomans saw themselvesas guardians and even caretakers rather than as conquerors. For much the same reason; they did not lay claim to the local Mamluk buildings. If one excludes the Khassaki Sultan complex, the walls and the foundations providingJerusalem with water, the great majority of these Onoman structures do not serve a public or welfare purpose; again, competition with the Mamluks was just not feasible,so a different message had to be transmitted. The effects of the developments sketched in broad outline in Section 2 above can be read in the record of Jerusalem's Ottoman architecture. It is this that explains the multitude of minor foundations, often anonymous. These were not built to perpetuate the memory of some amirand perhaps provide an income for his descendants. Instead they reflect the personal piety, the desire to procure religiousbenefit for the Muslim community at large-and, incidentally, the straitened financial circumstances-of their donors. The numerous cells and open-plan aedicules commemorating some holy event or traditional Islamic saint were too small to cost much. They were erected piecemeal,as the series of khalwas on the Haram esplanade shows, at the private instigation of modest patrons, for example members of the imperial militia and often local notables such as qadis or scholars. Occasionally a bolder or richer patron might, so to speak, think more laterally. Thus the North-Eastern Khalwa of Ahmad Pasha is the perfect match for his North-Western Khalwa; the patron had secured for himself two prize locations abutting either side of the north-eastern mawazin colonnade. In other words, the two foundations were conceived together, to balance each other. This is, however, an exception to the general trend of piecemeal, small-scale patronage. Larger building campaigns of the kind financed from the imperial purse would have resulted in a less haphazard distribution of these buildings. The constant re-use of earlier material points in the same direction. And this in turn follows from the disjointed nature of Ottoman patronage. True, an individual structure can display workmanship of an impeccable quality, but such excellence reflects the personal skill of that particular craftsman. It is not the natural product of sustained investment in new building. The results of that kind of investment can be seen in Mamluk Jerusalem: it raises the overall standard, from techniques of stereotomy to varieties of vaulting, from the design of an entire fa<;:adeto the execution of a capital. It can be seen in the capacity to exploit a gap site or to incorporate standing elements into a new design without strain. As the building boom which had given Mamluk Jerusalem its distinctive character petered out, so did the level of expertise decline. People must have left the construction industry for lack of work . Not surprisingly, therefore, the best of Ottoman architecture in Jerusalem is to be seen in the two generations following the Ottoman conquest, a period roughly co-terminous with the reign of Sulaiman the Magnificent, when the accumulated expertise assembled in the course of the Mamluk period could still be drawn upon and had yet to be dissipated. Typically, too, the best Ottoman work can be found in projects which involve a single master rather than an entire team. The sabils are a case in point. Certainly they offer challenges of design and execution. But those challenges are well within the capacity of a single master. And the skills they require are those of the mason and the sculptor, not those of the architect. With the single exception, as always, of the Khassaki Sultan complex (and perhaps that of Bairam Jawish as well), the buildings of Ottoman Jerusalem are deficient in the very lifeblood of architecture: a sense of space. This is a cruel irony when it is remembered that imperial Ottoman architects in the 16th century experimented more audaciously with spatial values than their counterparts in perhaps any earlier school of Islamic architecture. That was at least in part the result of massive investment in building campaigns, though of course it had much to do with the kind of building favoured by the sultans. Not even a distant echo of the hum of activity and eager experimentation in Istanbul can be heard in Ottoman Jerusalem. Its minarets attain barely a third of the height of those in the capital. Its monuments do not reflect the direct involvement of the top architects of Istanbul. The spatial experiment, the glamorous decoration, the precise stereotomy, the cascading curvilinear volumes of the great imperial Ottoman mosques evoked no comparable response in Jerusalem. And the provincial nature of this architecture became steadily more marked with the passing of time, and was compounded by the gradual decline in 13 Ottoman Jerusalem building activity. In the course of the 18th century, for example, not a single monument of significance was erected in Jerusalem. It is perhaps no coincidence that it is from this fallow period that some of the musallas on the Haram - mere platforms of dressed stone, sometimes furnished with a mihrab----date. Who were the major patrons, and what manner of people were they? They include Bairam Pasha, who, while governor of Egypt, donated 1,000 gold coins (ghirsh) to the buildings on the Haram; it was spent on porticoes for the Aqsa and near Bab al-Nazir and on Sabi! Sha'lan. He gave a further 1,000 silver coins (qit'a misriyya) to the tomb of al-Nabi Da'ud, plus textiles for it. His other major donations were for thirty-two Qur'an readers on the Dome of the Rock platform, and for the mu 'adhdhin of the Haram. A later waqf of his indicates that he sent ready money yearly for his khalwa. The other major patrons were Bairam Jawish, Ahmad Pasha, Khudawirdi Abu Saifin and Muhammad Pasha . Several of them, as well as other lesser patrons, held appointments as Ottoman governors, and it may be that patronage of local building campaigns was a case of noblesse oblige for such men. Sulaiman Pasha, who built the Iwan al-Sultan Mahmud II, an open-plan pavilion of Turkish type, in 1233 / 1817-8 and also restored the Maqam al-Nabi Da'ud for Sultan Mahmud II in the same year, was governor of Sidon and Tripoli . Ahmad Pasha, who built the North-Western Khalwa, was governor of Gaza ; Yusuf Pasha, the patron of the sabil, mihrab, and mastaba which all bear the name Sha 'lan , was governor of Jerusalem; al-Hajj Arslan Pasha, who restored the small building now known as the Shurta al-Gharbiyya, was governor ofJerusalem, Nablus and Gaza; Muhammad Beg, who ordered the mihrab niche in the Qubbat wa Mihrab alNabi, was governor of Gaza and Jerusalem ; Mustafa Agha Baraunazadeh, who ordered the domed tomb of one of the Naqshbandiyya shaikhs, Muhammad al-Salih al-Uzbeki, in the zawiya of that order, was governor of Jerusalem, as was Mustafa Agha, whose sabil in the Haram is the undisputed masterpiece of local 18th-century architecture. Muhammad Pasha, another governor of Jerusalem, endowed land for the Khanqah al-Maulawiyya and a plot of land on the Mount of Olives for the Sufis who came to Shaikh Muhammad al-'Alami, and the Zawiya alAs'adiyya, also on the Mount of Olives, in 1623. Indeed, all his patronage was for Sufi orders and in this he is unique among the great patrons of Ottoman Jerusalem. The reasons for this remain to be discovered, but may have something to do with the way that Sufis had risen to prominence in Jerusalem in the later 17th century, a time when the city had become a specially favoured place of pilgr image . Among the other great patrons of Ottoman Jerusalem, Ahmad Pasha favoured the 'ulama and Bairam Jawish the needy. 14 7. Architectural style in Ottoman Jerusalem Yet this is not the whole story . As Eric Schroeder has wittily said, 'To eschew the sacred wafer of Genius is no hardship to a man who chews the beefsteak of honest performance'. Good architecture is often a matter of good manners. And Jerusalem has long had a solid tradition in this respect. The ready availability of good building stone can be recognised at a glance in many an old street of the city. A city in which dressed stone is the standard building material will never lack stone-cutters and masons and the centuries of practical experience which they bring to their trade. Hence, at minimum , the sense of comfortable rightness, and more often the sober elegance, that characterises most Ottoman public buildings in Jerusalem . This architecture is admittedly not spectacular, but it is consistently good to look at. The architects took care that however widely rubble masonry and concealing layers of plaster were used inside a building, the best-quality stonework would be saved for the exterior. Who were these architects? One of the major discoveries made by Dr Mahmud 'Atallah, and confirmed by Dr Natsheh, in the sijill records was the existence of nothing less than a dynasty of local architects whose activity spanned five generations and almost two centuries: the Ibn Nammar family . Many of them bore the prestigious title mi'mar bashi. A few of them played a major role in local financial and administrative affairs . Altogether, Dr Natsheh has unearthed the names of sixty master builders, mainly local men but also hailing from Hama, Aleppo, Cairo and Istanbul as well as such local towns as Hebron and Ramla. They included Christians in their number. It was standard practice for them to work in close concert with the qadiin the inspection of buildings and in carrying out the repairs that he decreed. An efficient system of checks and balances was in operation, and - to judge by the numerous cases heard before the court - it seems to have safeguarded the interests of the ordinary citizens, whatever their confessional loyalty, with conspicuous success . The picture that emerges is one of fruitful and harmonious teamwork between the law and the building trade. (a) Building types: public architecture The survival of almost sixty public structures of Ottoman date in Jerusalem, many of them accurat ely dated or datable by endowments preserved in the sijills, makes it possible to assess the local style in the round . Happily, too, these buildings are of the most varied type: open-plan aedicules commemorating some venerated prophet; simple local mosques; zawiyas serving the Sufis living in or visiting the city, sometimes so placed that buildings for separate tariqas adjoin each other; tiny residential open-air Structure, Style and Context in the Monument s of Ottoman Jerusalem madrasas,which also served to house the shaikh in charge; sabils; mastabas; minar ets; simple domed cells; and two extensive charitable foundations fulfilling a wide range of functions. Only a few of these building types are numerous enough to justify a general discussion. The sabils are an obvious case. Doc um entary evidence shows that they were intended as a package: Sabil Bab al-Nazir, which was made waqf in the name of Sultan Sulaiman five years after it was constructed, was one of nine sabils-which indicates that some of them have vanished in the course of the centuries. What was their purpose? Some refer in their inscriptions to the hope of the water of Paradise and compare their water to the water of Paradise . But other factors must also be considered. Several sabils are dated to the month of Muharram, and the semantic connection of that word with anctity, and hence purity, is entirely appropriate to their function of aiding ritual purity. Not to be overlooked, either, are the Shi'ite pilgrims who thronged to the city; for them, there would be the added associations of the awful thirst suffered by the martyrs of Karbala . Indeed, the now vanished Sabi! al-Husaini of 1724-5 refers in its surviving inscriptionto Husain the son of 'Ali ibn Abi Talib, and thus to the battle of Karbala. The implication is that all who drank from the sabil were expected to pray for al-Husain. The family name of the founder, as it happens, was alHusaini. It may also be relevant to some of these sabils that- as the local people would have known - the 16 Muharram was the day that Jerusalem was selected as the qibla.Both Sabi! Sha'lan and Sabil Bab al-'Atm have prayer facilitiesattached (in the former case both a mihrab and a mastaba);perhaps it was usual to pray after drinking, but in any case the sabil in these two cases seems to have been intended for ablution as well as consumption . The latter was surely the primary purpose of such structures; an inscription on the Sabi! al-Shurbaji of 1685-6 proclaims "Abd al-Karim al-Shurbaji built the sabil so that thirsty people might drink.' Hence, too, their location at traffic nodes so that th ey could serve the largest number of people.The fact that the se sabils are of such different sizes and designs suggests that they were the work of various craftsmen, and that there was little overall supervision. Certainlythis is implied by the contrast between the most elaborate (Sabi! Bab al-Silsila) and the simplest of all (Sabil BabSitti Maryam) . Not all of the surviving sabils were the work of Sultan Sulaiman; Sabi! Bab al-Maghariba, for example, may be one of three built, according to Evliya Celebi, by Danya! Pasha. Like other sabils, it was endowed- in this case, with a brass cauldron and with money for buckets, ropes, and the daily hire of a watercarrier. The sabil has a strong claim to be the most typical building type of Ottoman Jerusalem, for it spans most of the Ottoman period, from Sulaiman's foundations to the sabilof Mustafa Agha some two centuries later. It would be hard to choose a more fitting envoi for the type . It is a poised and elegant building, delicately proportioned, whose high plinth is an integral part of the design . A little gem and easily the finest building of the 18th century in Jerusalem, it fits perfectly into its context - the long arcade bordering the Haram precinct . Still commoner are the domed squares which border the Haram or are scattered around it. The Qubbat Yusuf Agha of 1681 is typical of many of these little buildings in its extremely plain domed structure and in its isola tion . Although certain formal sub-divisions for such structures do suggest themselves, these buildings follow a formula deeply rooted in Ottoman architecture from the early 14th century onwards . It was a notably flexible formula: effective on both a small and a large scale, it could accommodate adjoining structures without losing its character, and it readily lent itself to repetition . These qualities are well illustrated in the Hujrat Muhammad Amir Liwa' al-Quds, a pair of adjoining domed squares which formerly had a three-domed portico, a canonical form in Ottoman Anatolia . The Khalwat Junbalatiyya, on the other hand, formerly had a two-bayed porch (the stone here has weathered only a little, which suggests that the porch was removed only recently), while in the case of the madrasa of Ahmad Pasha only a platform remains. Indeed, the disappearance of the porticoes of these cells seems to be a general and as yet unexplained pattern . In a few cases, the porches have been rebuilt and the bases of their columns now serve as capitals. This cavalier attitud e to reused material, which will shortly be discussed in greater detail, is typical of the architecture of Ottoman Jerusalem . As for their purpose, some of them probably housed the staff of the Haram, as is still the case for many of them, such as Khalwat Qitas, or Khalwat Barwiz, which is now the office of the Guards of the Noble Sanctuary, or the Qubbat Yusuf Agha, now the ticket office to the Haram . Other khalwas served for teaching and for Qur'an recitation. On the other hand, Evliya C:::elebiwrites of pilgrims being lent a cell for the duration of their visitpresumably pilgrims of a certain social standing. A rather different fate befell the Qubbat al-Khadr: in time it lost its original meaning, and was then used to store material belonging to the Aqsa mosque; other khalwas also served as storage space . This shifting of meaning, associations and functions from one site to another is typical of the Haram, as is the use of multiple names for the same site, which points in the same direction . The inscription of the Khalwat al-Dajani (1138/1725-6) calls it a maqam - 'place' - as if the inten tion were to leave its actual function as broad as possible. Perhaps al-Dajani's primary aim was to erect a building-any building-on this coveted site. Its fa<;:ades are grossly irregular, suggesting that its position, tucked into a corner of the staircase to the upper terrace , was the key factor in its design . Further information is given in the 15 Ottoman Jerusalem waqf for the North-Western Khalwa of Ahmad Pasha, which stipulates that he has 'constructed and endowed cells in the mosque of Jerusalem . He has appointed to each cell a group of scholars from Jerusalem and allocated expenditure ... ' He specifies the scholars, who are members of the Qadiriyya order, by name. Some of the money is for oil for the lamp and for restoring the doors. Another common form was the open-plan octagonal (or occasionally hexagonal) aedicule, a type encountered all over the Muslim world at least from the 10th century onwards and functioning not only as a commemorative building-the category to which the many examples at Jerusalem belong - but also as a fountain (Ma'arrat al-Nu'man, Syria) or a mausoleum (Van, Turkey). The examples on the Haram are commemorative and the case of the two structures erected by Yusuf Agha in 1681-the QubbatYusuf (an open-plan aedicule) and the Qubbat Yusuf Agha (a closed domed square)-clearly suggest that, despite the identical terminology, different forms connoted different functions in Jerusalem at that time. Nor should one forget structures of a more vernacular or industrial type, though these are not catalogued in this book: over a score of bakeries (here close dating is particularly difficult); hammams, though the few that survive are much denatured; two khans ( one mentioned in the endowment deed of Khassaki Sultan as part of that complex, the other, Khan al-Sha'ara, in the present Jewish quarter); oil presses (though only two now survive, whereas more than a score were in operation a generation ago); and markets, notably Suq Khan al-Zait , Suq al-'Attarin, Suq al-Lahhamin, Suq al-Khuwajat, and Suq al-Husur. (b) Building types: private architecture At least thirty examples of the dar or private house for the elite survive . Some are extremely rare, like the Dar Bairam Jawish, of 953 / 1546, which is the only house datable to the 16th century in Jerusalem. Its north and south fronts both have an arch with a double tier of voussoirs, as does the fa~ade of the ribat of Bairam Jawish. The lack of comparable structures makes it difficult to assess the originality of this feature. More generally, there are no known extant Mamluk precedents for this dar; the most appropriate parallel is the dar of Sitt Tunshuq, which is a palace rather than the private house of a notable. Another exception is the Dar al-'Izz of 1790-1 , perhaps built as a private commercial guesthouse; this is a rare example of a dated secular private building in Ottoman Jerusalem, notable for the presence of a courtyard and garden on the first floor. Since elite Ottoman housing does not figure largely in the book, it may be useful to give here a generic description of the type. A courtyard at ground floor level 16 surrounded by utility rooms is reached by a single external entrance. The principal rooms are at first floor level and are reached by a staircase which gives onto a balcony overlooking the courtyard . These are the rooms which contain ornament: external carved window frames, carved and painted plaster, and domes on octagons, so low that they are often only just visible from the street. Some of these houses are up to three storeys high, and narrow single or paired rectangular windows and suspended buttresses (that is, buttresses built into the wall but standing proud of it) enliven their outer walls. Although they are scattered all over the Muslim quarter, clusters of two or more of them are to be found in 'Aqabat al-Shaikh Rihan, Tariq Bab al· Silsila, Tariq al-Hakkari, 'Aqabat al-Khalidiyya, 'Aqabat al· Maulawiyya and especially 'Aqabat al-Saraya. Nor should one forget more humble types of dwelling, notably the haush. Many of these units still survive in Jerusalem; they house as many as six families apiece . They represent a remarkably economical exploitation of the very limited space for building available in the Old City, for they are fitted into the irregular spaces left over between larger structures, and are capable of expansion not only laterally but also vertically for several low storeys . Haush al-Hilu,for example, is entered by a low vaulted passage from 'Aqabat al-Saraya; this leads into the first courtyard. Off this courtyard are passages which lead to further dwellings which seem eventually to back on to Khassaki Sultan and Haush Shawish. Thus every crevice of space is utilised. (c) The nature of the cityscape Nonetheless, for all this variety of public, industrial, commercial, vernacular and private buildings, there are some telling absences which are enough in themselves to define Jerusalem as an economic and architectural backwater. The lack of major Ottoman mosques, palaces, and caravanserais clinches this unflattering description. The picture is very different from that of Aleppo and Damascus in Ottoman times, as already mentioned. The buildings in those cities are not only much more numerous - as is only to be expected given their much greater size-but of much higher technical quality, larger , more ambitious, and much more lavishly decorated. Thus their impact in the urban environment is much more assertive than those of Jerusalem. Yet this very remoteness from the world of the great provincial cities, let alone from the imperial Ottoman metropolis itself, brings concomitant advantages in its train. Jerusalem offers an excellent opportunity to investigate the physical environment of a small provincial Ottoman town of only minor commercial importance. Despite the ravages of war and the relentless pressures of commercial and industrial development, despite its ballooning population and its flawed record of conserving historic buildings, the Old City, at the beginning of the new Structure, Style and Context in the Monuments of Ottoman Jerusalem millennium, can claim-against all odds-to preserve the c ence of a pre-modern Palestinian town. It is wonderfully all of a piece. This is due in large measure to its wealth of domestic and vernacular architecture built in the local stone, amidst which the Ottoman public buildings are randomlyscattered. In other words, it is precisely because they have not lost their original setting-the workaday context out of which they grew-that these public buildingsseem so natural a part of the cityscape. They are constructed of the selfsame local stone as the streets and houses all around them. Those streets maintain a comfortable human scale; apart from the main arterial thoroughfares,they are neither so wide as to be grand nor o narrow as to be unpleasantly constricting. Designed for pedestriansand animals only-cars are tolerated only close to the city gates- they follow the natural topography in their rise and fall, with short flights of wide, easily negotiablesteps to mark transitions or to modulate a long vista. At second-storey level or above, arches like flying buttressesbridge the street (hence their name: qantara) and upport passageways and rooms, thus making it possible for a single house to spread comfortably across a public thoroughfare.It is also common for the bridge to belong exclusivelyto the house on one side of the street, so that the wallof the house opposite serves only as a support for one end of the bridge. Over two dozen of these bridges survive. (d) Vaulting The small scale of most Ottoman architecture in Jerusalem discouraged significant innovations in vaulting technique and even militated against a lively interest in methods of roofing. Indeed, the domes of many of the smaller tructures are grossly over-built. The frequency of minor infelicitiesand adjustments suggests that many masons workedby eye, confident in their ability to muddle through somehow. Cramped and irregular sites generated asymmetricalspaces to be vaulted, and later modifications, additions and rebuildings created still further problems of thiskind.Yetthe solid competence of these masons ensured that such vaults worked efficiently, even if they were not of drawing-board exactness. The most popular types were pointed barrel vaults, groin vaults, domes on squinches, domes on pendentives and pendentive domes, ribbed saucer domes, cross vaults over rectangular spaces (sometimes with tiercerons), domical vaults, umbrella domes, sail vaults, folded vaults with a hexagonal or octagonalcrown or a saucer dome, and star vaults with 12, 16, 20 or 24 points. The 'Mamluk' cell (North-Western Khalwaof Ahmad Pasha) in the north colonnade even has a dome featuring 32 whorled ribs with 4 pendentives and 8 folded triangles; this ribbed dome rests on a star-shaped vault whose points touch the springing of the supporting arches,which in turn are of uneven height and thus have their apices follow a rising and falling rhythm. Qubbat Yusuf has a dome on pendentives whose 38 ribs have a double zigzag frame. The Zawiya al-Muhammadiyya has a domical star vault with a scalloped outer rim, and also a dome on pendentives, whose collar is marked by a continuous open arcade. The interior acquires monumentality from the four arches which carry the dome and whose springing begins just above floor level. It would be a useful study to collect the terms used by local masons for these very varied forms of roofing; similar work done in Iraq by Reuther, Langegger and Herzfeld or in Morocco by Paccard suggests that every form had its own name: 'father of four', 'almond', 'spider's web', or 'four oil lamps'. Naturally the more complex types tended to be those in the most ambitious foundations, such as Khassaki Sultan and the ribat and dar of Bairam Jawish. On the other hand, the visible surface of the vault, with its lines picked out sharply in plaster, may not always correspond to what lies beneath, especially in cases where elaborations of the design are executed only shallowly in the plaster. In a few cases the dome bears external ribs in high relief. (e) Capitals and plinths The readiness of Ottoman architects in Jerusalem to use spolia explains the frequency of non-matching columns and capitals in their buildings. The variations in scale are especially intrusive. This suggests that the savings in time and money represented by the re-use in a new setting of existing Islamic capitals and columns counted for more than visual harmony. That said, the simple re-use of earlier capitals, such as basket- and bowl-shaped capitals, is rare. So too is the partial re-use of antique capitals, for example those with foliate scrolls in the upper part of the capital and Ottoman lanceolate niches below. All this is remarkable in view of the many scores of re-used antique capitals in the Dome of the Rock and the Aqsa mosque, whose presence shows that in the early Islamic period at least there was no inhibition about re-using earlier building material. A typical example of the variety of capitals encountered in a single building is provided by the Qubbat al-Arwah. Here some capitals are admittedly similar to each other, but only two are identical; they illustrate basket, muqarnas (some fully carved, some with large blank spaces between the muqarnas cells, some with plain cells, some whose cells have added internal ornament of vegetal or facetted type, some double-tiered and some triple-tiered) and debased and simplified Corinthian. In one case an extra block of white stone has been inserted below the capital to raise it to the required level. The abacus varies from one capital to the next, but the shafts are identical-which suggests that only the capitals were re-used. Very little research on the forms of capitals in Islamic architecture has been carried out; there is, for example, no study devoted entirely to Ottoman capitals. This is a pity, for their variety is remarkable. Perhaps the 17 Ottoman Jerusalem easy availability of antique models in the great Islamic shrines of the city-which, despite the fact of their re-use, were nevertheless sanctified by their Islamic context and thus became suitable models to be copied-inspired the general layout of the standard Ottoman capital. That typical Ottoman capital can be explained quite naturally as a schematised and abstracted version of the successive tiers of foliation in a Corinthian column. The space created by two half-leaves and the meeting-points of their tips resolves itself into a pointed niche which in turn becomes an element integrated into a muqarnas system. Similar forms are known in the capitals of late Ottoman Baghdad. Even the volutes of a Corinthian capital may be recognised, much reduced and denatured, in those capitals whose uppermost tier has a middle section which is empty, leaving elaborately carved corners, or conversely is crammed with ornament and leaves the corners quite plain. In both types the ghost of the classical volute can be sensed. It is an instructive example of how Islamic craftsmen understood and re-interpreted the classical heritage, and it is no less revealing even though the model was at that time more than a thousand years old. The sense of natural growth which permeates the classical original has been replaced by a love for regular repetitive pattern; there is variety enough, but each pattern is confined by its niche and cannot develop freely. In similar fashion, the essential threedimensionality and spatial freedom of the Corinthian capital is toned down to such a degree that the volumes of these Ottoman capitals are compressed, even imprisoned, and thereby lose that suggestion of burgeoning life so appropriate to vegetal ornament. This Ottoman vegetal ornament is rigidly geometricised - for example, the treatment of the cypress motif-and it is not suffered to stray beyond the boundaries set for it. Even the sense of interpenetrating levels which makes classical foliate capitals so lifelike is ironed out so as to separate each tier from the one above or below it. Accordingly it is not surprising that some capitals display a clear hierarchy in their visual vocabulary, with a steady growth in complexity of ornament from the lowest to the topmost tier. Often a heavily ornamented level is set against a plain surface, such as a dosseret. Basically the interest of Ottoman capitals resides in the experiments made with the muqarnas form. When these variations are closely analysed, they prove to be very close to each other, but their sculptors managed to make them look different by various means. One was to alternate plain fields for the individual cells with vegetal (e.g. almond or cypress shapes) or geometric infill for them. Another was to contrast a richly carved upper structure for the capital with a plain lower one, or vice versa. Yet another was to leave the central upper portion of the capital plain . Occasionally the muqarnas theme is confined to chamfered corner cells. Other and somewhat aberrant types include open papyrus 18 capitals, globular capitals with scalloped bases, jeux d'espru that combine the Ionic with the palm leaf capital, and capitals with plunging angular folded planes , a form possibly inspired by the folded vaults so popular in Ottoman Jerusalem. The abacus is often strongly emphasised, often with an intermediate cyma recta or torus moulding. The latter two mouldings are standard in plinths too, but muqarnas and basket designs are also encountered there. (j) Arches Such a wide range of arch types was in use during the Ottoman period in Jerusalem that no one type could be termed standard. The varieties include round or hemispherical, slightly pointed, four -centred, lancet, ogee, trefoil, equilateral, transverse and depressed arches. Segmental arches, often stilted, shouldered and flattopped, are repeatedly found crowning the niches of sabils , and the horseshoe form turns up occasionally , though the return is much less pronounced than in Maghribi and Spanish architecture. These multifarious arch forms acquire extra visual interest by the use of two-colour voussoirs and by various mouldings, sometimes as manyas six per arch. On occasion the keystone stands out by virtue of its rhomboidal shape, or because it breaks the contourof the arch, or thanks to the carved medallion which it bears. All these arches are of stone masonry. Given the prevailing sobriety of Ottoman architecture in Jerusalem, the special attention paid to doors and windows is particularly striking. A wide rangeof lintels is employed for both forms. Flat lintels may consist of a simple horizontal stone beam, often singled out from the surrounding masonry by its smoothness and by its different colour and material - for example, marble. But they often stand out by means of joggled bi-coloured voussoirs, tapering stepped blocks at their centres, or by the flaring or irregular cut of their constituent blocks. Sometimes a shallow relieving arch, itself perhaps in twotone masonry or bearing a carved medallion, surmounts the lintel. The arches of the door or window may exhibit some of these characteristics. (g) Stonework This stonework repays close attention. It is noticeable that subsequent repairs have frequently disfigured these walls, especially in that later repointing has tended to enlarge the apparent width of the joints and thus to blur the previously sharp edges of the stonework. This may seem a small detail, but it is enough to transform a wall and to blunt its sense of mass. Over-lavish use of mortar also serves to mask how often stones were cut precisely to fit some irregularity in the wall. A brief look at any one of the fa~ades of the North-Western Khalwa of Ahmad Pasha is enough to reveal that the masons used stones of many different Structure, Style and Context in the Monum ents of Ottoman J erusalem dimensions and deliberately exploited this variety to add life to the wall. There are sufficient examples of the transgression of the main horizontal masonry courses to indicate that these breaks in rhythm, too, were deliberate . The eye is not allowed to dismiss the wall as a piece of mechanically accurate coursing made up of blocks of equal size with regular vertical and horizontal accents. Thus stones of exceptiona lly large size are framed by others which are exceptionally small. The rising joints zigzag unpredictably along the vertical axis. Nor are the differences in stonework confined to size and colour; texture is an equa lly significant factor. Thus the extreme smoothness of marble-particularly the grey marble used for lintels and sills, whose natural horizontal graining emphasises this tactile quality - contrasts with the pitted urface of the bulk of the stonework. Rustication was not part of the repertoi re of the 16th -century Palestinian tonemason, but the outer surface of some stones is worked with such deliberate roughness that some sculptural effect seems to have been intended. Some of these stones are pock-marked or pitted, others veined, yet others relatively mooth. Rustication is employed in the hall of the Zawiya al-Qadiriyya,but since it is found only there it seems likely that it dates from some time other than the rest of the building (see below). An assured stonework technique is the key to this architecture. Stonewo rk can also have wider implications. Thus the different size of the blocks and the method of tonecutting employed on them sometimes has dating implications; for instance, the presence of different stonework at the lower level in many khalwas suggests that many of these monuments are mere superstructures to earlier buildings which respected the emptiness of the upper platform. In that case it would have been the location of these earlier buildings that dictated that of the Ottoman khalwas. No studie s detailed enough to establish dating criteria on the basis of tools used, stone sizes and treatment of surface have yet been carried out. Nevertheless, it is plain that for much of the Ottoman period local fashion preferred a pitted surface to smooth ashlar, though sometimes a compromise was effected whereby the two types of surface were juxtaposed for greater contrast. Thus marginal drafting is employed for the outer edges as a framefor the roughly finished square or rectangular central area. The resultant cloisonne appearance has parallels in earlier Ottoma n architecture, for example at Bursa, where-following Byzantine precedent-brick was used to framestone blocks. In late Ottoman times the technique of rustication was introduced, presumably from Europe; the fa~adeof the Dar al-'Adl is a good example. It would be wrong to characterise Ottoman masonry technique as even. The wall of the bakery at the KhassakiSultan complex shows stones with a wide range of sizes and of smoothness; some are heavily, some lightly pitted, and there is no pattern to where these variations occur, nor to where blocks are slightly recessed into the wall. Occasionally a course of narrow horizontal blocks breaks up the wall, or long and short blocks are juxtaposed. By such means the wall itself is animated. The stones of th e Khalwat al-Dajani have several different types of dre ssingone variety is roughly facetted and betrays the us e of a pointed chisel - and this suggests that they were not quarried together but were in secondary use. Special care was taken with the external cladding of domes, for which stones of a uniform size were selected , as the roofscape of Dar Bairam Jawish shows. The stones used in thi s fashion to 'pave' the exterior of a dome were laid so that their largest surface area was visible. Nevertheless, the tiny size of the stones used for so many Ottoman buildings implies that speed and cheapness of construction were paramount factors . 'Big ' in this context is 40cm x 1.92m (e.g. cat. no.25). Moreover, ashlar masonry, even in the buildings of higher quality, served only as a facing for much rougher masonry. Indeed, the madrasa of Ahmad Pasha uses plastered rubble for the area behind the fa~ade - a 'cheap and cheerful' solution for surfaces not intended for normal public view. (h) Windows It is not surprising that an architectural style that developed in a crowded urban context should rely so heavily on ways of accentuating the street fa~ade . Many devices to this end had been developed by the Mamluks, but under the Ottomans other original features also evolved. Among them the oriel window deserves special mention. Essentially a combination of balcony and window, and situated high up on the outer fa~ade, it was not only a source of light but also gave the inhabitants of a house an excellent vantage point from which to view the street without being observed themselves. The balcony itself may be of stone or wood, and above it may be as many as five windows on the same level (Dar Muhtadi). Many windows were fitted with iron grilles which not only served to modulate the light that entered a room but also served to articulate a fa~ade. Sometimes the window is framed by quoins, or a broad horizontal lintel may crown it. This suggestion of a boundary may be heightened by the sparing use of ablaq masonry or of joggled voussoirs. In the case of windows, it is not the treatment of arch or lintel that is visually decisive, but its surroundings, and especially its superstructure. The ensemble of which the window is the centrepiece comprises a narrow recessed vertical panel. Apart from a richly moulded frame, the window itself is apt to be flanked by quoins alternating with stone blocks of a different colour or texture. Above the window proper an elaborate multi -tiered design unfolds, in which contrasting colours of stone, joggled voussoirs, glazed blue faience insets, carved roundels, medallions and 19 Ottoman Jerusalem stars, flat muqarnas panels and a stalactite cornice may all play a part. (i) Pottery screens Often the bridges or private pathways over the streets - of which more than two dozen survive - are lightened by being constructed over tiers of hollow baked clay pipes whose openings are so disposed as to form triangular patterns. These pottery screens also serve to modulate light filtering into the areas behind them. Such pottery walls are also employed as screens marking off the top of courtyards and demarcating boundaries at parapet level. They are not peculiar to Jerusalem but are a feature of pre-modern Palestinian architecture, as at Ramla. Similar forms were also common in late Ottoman Baghdad. (j) Applied ornament It was standard practice to reserve the best decoration for key locations; thus the madrasa of Ahmad Pasha displays more decoration on the qibla side than elsewhere. But of course applied ornament has obvious financial dimensions, and in view of the obviously limited budget for most Ottoman buildings in Jerusalem it is not surprising that most of them rely principally on the aesthetic effect of good-quality stonework for their appearance. Blank recessed panels sometimes articulate a fac;:ade in an understated way. But the entire dimension of colour (see below) is much reduced, for no painted plaster survives and the only colour is that of the stone masonry itself, which is red, ochre and white; the latter has sometimes weathered to grey but often has a creamy tint. The black hue of much of this stonework is also not its natural colour; it has various causes, among them the effect of water, wind, shadow or fungal growth . The resultant contrast of surfaces can be quite startling: thus the hujra of Islam Beg has a main fa<;:ade of predominantly white stone and a rear fac;:ade of predominantly black stone, all apparently due to weathering. A very few buildings, notably the sabils, display varied and complex ornament . Qubbat Yusuf (1681) is another good example . This is a smaller building than the Qubbat Yusuf Agha by the same patron, but much more richly decorated. Why? The reason may be that it is on the upper terrace, where space is much more limited, and where a bigger building that might upstage the Dome of the Chain , let alone the Dome of the Rock, was excluded. But the very choice of the upper rather than the lower terrace (where the Qubbat Yusuf Agha is located) created an extra prestige and aura of sanctity for the building and made lavish decoration all the more appropriate. Indeed, this was the best available way of expressing extra importance where size was not an option. The inclusion of an inscribed Ayyubid plaque gives the qubba an ancient pedigree and, as van Berchem has argued, extra prestige. 20 Perhaps the favourite decorative theme of Ottoman architects in Jerusalem was the muqarnas. It turns up in all kinds of guises. Thus Sabi! Birkat al-Sultan hasa tympanum with three tier s of muqarnas, each one differentiated from the others; this is a highly developed use of this theme. Sometimes, as at Sabi! Bab al-'Atm, the muqarnas is even used upside down . At the Zawiya alQadiriyya, as at Nabi Da'ud, a shallow muqarnas niche just above eye level marks the junction where two external fa<;:ades meet at right angles: a modest but elegant boundary indicator. The maktab of Bairam Jawish has on its first-floor elevation a band of ten and a half triangular lancet panels of muqarnas type, all different. The muqarnas also serves a useful role as a support for the gallery of local minarets. The range of mouldings is somewhat limited. The denticulated frieze used at the Qubbat al-Khadr to outline (and over-emphasise) the main arches on the external fac;:ade recurs in several other buildings of the period of Sulaiman and later in Jerusalem: Maqam Nabi Da'ud, the minaret of the Citadel, Sabi! Bab al-Silsila and Sabi! Bab al-Nazir. Other mouldings are more or less accurate versions of such classical types as the cavetto, cyma rectaor cyma reversa. Occasionally they are serrated or bear applied ornament such as diaper-work. Roll and billet mouldings were also popular . Cornices, too, follow set patterns. That on the Qubbat al-Khadr consists of a miniature arcade with a leaf form in the field of each arch . Another common form for cornices was a muqarnas system, whether single- or multitiered, or a band of facetted triangles or denticulations. The formula customarily involves a long or short perpendicular plane at the top of the cornice, with convex , concave or indented planes immediately below. Much use is made of stone roundels with carved geometric (often stellar) designs. These are used as grace notes rather than as a regular element in the mason's repertoire . In this they conform to the taste for the sparing use of decorative accents. The fullest collection of these roundels is to be found on the city walls, but they also occur within each of the lower lobes of a trilobed arch, at the centre of the tympanum of a sabil, and more generally to highlight spandrels, to mark or crown the apices of arches, and to emphasise the key central and outer points of the entablature crowning doors and windows . Similar forms, but pierced, also serve as windows. A few examples of stone sculpture also deserve mention . The sabil of Mustafa Agha has hexagonal columns, of which four sides are decorated (all differently) and two are plain, and the same system is found at the mihrab of the Iwan of Sultan Mahmud II. Thus the capacity for applied ornament in stone was there, as in contemporary Damascus and Aleppo, ready to be called forth by the right kind of patronage. Structure, Style and Context in the Monuments of Ottoman Jerusal em What of the dimension of colour? This has many aspects. In general, it is much reduced, for no painted plaster survives and the only colour is that of the stone masonry itself. The nature of the local stone is therefore crucial. Its colour, as already noted above, is far from consistent, varying as it does from a blackish grey to offwhite. Often quite marked variations occur on a single block of stone. Sometimes it takes on a mottled brownish tinge;in many cases it has weathered to black. The masons proved adept at varying the colour harmonies of their fa~adesand creating accents at specific or random intervals by the placing of a single lighter or darker block, or by varying the colour as well as the tonality. The patchwork result bears a startling resemblance to certain types of modern abstract art . Grey, white, red and yellow stone or marble used in contrasting colours-a continuation of Mamluk ablaq- comes into its own for the decoration of window frames, mihrabs, the fields of blind arches, two tone or joggled voussoirs, or even simple single or double string courses, arch profiles, keystones, jambs and quoins, as at the Nor th-Western Khalwa of Ahmad Pasha. The mihrabof 'Ali Pasha uses four tones : white, grey, red and black, while two-tone marble floors are quite common. Coloured stone or marble, sometimes heightened by the use of tiny blue faience inserts in the shapes of stars or triangles,was ideal for the six-pointed star - often known as Solomon's Seal and perhaps intended too as a flattering reference to the great Ottoman sultan himself-which is such a leitmotif on these fas;ades. The masterpiece of such marble compositions is to be found in the door-frame of the ancillary chamber of the North-Western Khalwa of Ahmad Pasha, whose beauty lies in its very understatement . The muted pastel tones set each other off to perfection, and because they are applied in such broad bands the cartouches and stars placed at such intervals along them function as grace notes to an abstract colour composit ion rather than as the focus of interest. This is also the role of the square glazed tesserae used so sparingly in these buildings; it is perhaps a reflection of the poverty of Ottoman Jerusalem in comparison with, say, Ottoman Damascus that- apart from the very special case of the Dome of the Rock, which represents an intrusion of toplevel imperia l patronage-the Iznik-style tilework of Damascus is so conspicuously absent. This is truly remarkable,especially as the enormous task of re-tiling the exterior of the Dome of the Rock meant that local workshopswere established. Clearly their production was set aside for that building alone. Such tilework as survives elsewhere in the city is not only so rare as to be insignificant, as shown for example by the tiles over the doorway of Maktab Bairam Jawish, but of indifferent quality. Colour is also used somewhat unexpectedly to create mihrabsin floors. Sometimes this theme is echoed in the elevation; thus the Qubbat wa Mihrab al-Nabi has a diminutive niche with a central red band running through a line of concave cylindrical indentations, a blind arcade , thereby emphasising - and in the same colours - the qibla already announced by the Mamluk mihrab laid in reserve on the floor. It is not easy to explain the preference for reserved mihrabs (mihrab makhtut Ji 'l-ard), for some of which two- or three-tone marble or stone is used . This is not an Ottoman idea; it was already in use at least in Mamluk times if not earlier. One may suggest that this type of mihrab was used so as not to encroach too obviously on the more holy monuments all around. (k) Spolia Spolia are part of the stock in trade of the local architects . Indeed, the sijills are full of references to the sale and purchase of ready-made building materials. A sijill reference from 1530 indicates that a block of dressed stone fetched a silver piece at that time . Clearly, then, there was no lack of incentive for a thriving trade in these materials to develop. Re-used material came from several sources: whether spolia from destroyed buildings, gleanings from ruins or merely fragments from the stock kept by local masons. The prevalence of spolia is not surprising because Jerusalem was such a rich quarry for buildings of preOttoman times, many of them non-Muslim and inadequately protected from despoliation. In a few cases, such as Sabi! Bab al-Silsila, a foreign origin is manifest: here a truncated Crusader rose window is crammed into the tympanum. Similarly, at the Sabil Bab Sitti Maryam a Crusader arch and Crusader columns with a plaited central braid sit rather uneasily alongside Ottoman muqarnas work, while Sabi! Bab al-'Atm uses a non-Ottoman chevron motif, and the Zawiya al-Naqshbandiyya has a chevron arch. It remains an open question why these spolia were such a marked feature of the sabils founded by Sultan Sulaiman. In any event, this use of spolia follows precedents set in Nurid, Ayyubid and Mamluk times, as illustrated by the antique patinas (sigma-shaped tables) reused in 12th-century Syrian mosques and madrasas, the Turbat al-Nahawiyya on the Haram, Christian columns in Maghribi architecture, Byzantine capitals in the Aqsa mosque, and-most eye-catching of all- the Gothic portal of the church of St John at Acre, which was incorporated into the funerary madrasa of Sultan al-Nasir Muhammad in Cairo. Nor were such spolia confined to the sabils. The Qubbat al-Arwah boasts not only a chevron and gadrooned arch but also two Crusader columns and capitals. Indeed, it is a hotchpotch of borrowed elements, and shows how variously, therefore, Crusader spolia were incorporated into the Islamic repertoire. A triumphalist intention here, though politically quite outdated, is possible in this case because the concentration of foreign material is so strong and because this aedicule is close to the Nahawiyya, which 21 Ottoman Jerusalem suggests an Ottoman sensitivity to making the same point as that building and in the same visual language. Often enough, however, the spolia carry no extra charge of triumphalism. In the Hujrat Islam Beg, the shafts are in secondary use, which implies the desire to save money and an indifference to appearances even on the Haram. These are jerrybuilt columns, made up of slightly uneven blocks, and they carry disproportionately huge capitals. Columns are of different height, diameter and colour in a single building; sometimes a structure with only eight capitals will have no two that are identical. This readiness to use second -hand material betrays a certain indifference to the visual aspect of a building, which in turn may help to explain the gross asymmetries and disparities which disfigure fa~ades that in other respects display exemplary craftsmanship, such as the four fa~ades, each one quite different from the next, of the North-Western Khalwa of Ahmad Pasha. Such details bespeak not only a lack of coherence, a lack of vision about the project as a whole, but also an absence of oversight. The structural dimension of using spolia should not be forgotten. Very often metal bands are applied around the columns. This is not only a sure sign that the elements thus joined had not been carved to be used together, but also suggests that the structure was perceived to be weak, perhaps even damaged, in which case such bands could serve to conceal the fact. They may also be a precaution against earthquake. Further uses for metal clamps are to be seen in the Qubbat al-Arwah, where rivets are employed to join some of the blocks forming the raised border of the floor and a metal tie-beam has been inserted at springing level. (l) Inscriptions It is too early to state definitively that a standard Ottoman local form for inscription panels evolved. The necessary detailed research remains to be done. Nevertheless, some continuity from earlier periods is quickly recognisable. In the Masjid al-Saif, the inscription in the name of Sulaiman continues Mamluk modes, especially the Qa'itbai style in its use of naskh not thulth and its uneven tripartite design. Further noteworthy features here are the thick forms of the letters and the unusual positioning of this inscription at the keystone of the mihrab arch . Equally unusual is the thick and chunky naskhi inscription on the Kursi Sulaiman. While a rectangular format for inscription plaquesanother Mamluk heritage-was standard, the octagonal form was also known. Such panels were lightly ornamented with rounded corners or a plain fillet border; lobed cartouches, some of them richly ornamented with flowers, branches and leaves of various kinds, are also known. Sometimes these panels are recessed. A few inscriptions, 22 including those with a date, are in Ottoman Turkish, and sometimes both languages are used together. Many inscriptions are in Arabic poetry, and it is standard practice to give the date by chronogram as well as in figures. Two buildings only-the Qubbat Yusuf and the Qubbat Yusuf Agha-have two dated inscription panels of similar content. The content of these inscriptions has not received separate study as a body of connected material. Not surprisingly, however, there are references or allusions to the special sanctity of Jerusalem, to the hope of paradise, to the performance of ziyara or to prophets who have a particular link with the city. Thus an inscription from Kursi Sulaiman quotes from the Qur'anic story of the encounter between Solomon and Bilqis, the Queen of Sheba, and thereby confirms that its popular name today accurately reflects the original purpose of the building, namely to commemorate Solomon. (m) Summary The local style changed remarkably little in the course of the Ottoman period. This in itself is an index of the remoteness of Jerusalem from the fashions which came and went in larger centres. The baroque and rococo elements which transformed later Ottoman architecture elsewhere occur only intermittently, like the rustication on the fa~ade of the Dar al-'Adl. This architecture possessed the virtue of straightforward good manners, expressed in simple forms, clean, consistent, sober stereotomy, a generally uniform colour and material, restrained articulation and parsimonious ornament. Its preferred forms were simple, strong, cubic . These characteristics, though they may sound rather unexciting, do bear further scrutiny, and they work together very harmoniously. Formal inventiveness is not the strong suit of local architects. It is well, however, to remember the modularity of Ottoman architecture, with its unflagging focus on the domed square as the base unit of the design. Thus the architecture of Jerusalem is thoroughly in tune with the spirit of Ottoman architecture generally. That domed square unit recurs repeatedly both in open -plan designs like the so-called iwan of Mahmud II and in the constricted space of the typical khalwa on the Haram terrace. Moreover, its visual impact could vary quite dramatically, not only according to its size but also to whether it was isolated, doubled, or attached to other forms such as the great arcades of the Haram. A popular fashion was to place a multi-domed arched portico in front of one or more domed squares, a formula very popular for small mosques from early Ottoman times onwards. No matter whether the buildings are large or small, the aim of their architects is consistent: to achieve maximum monumentality with minimum means. Structure, Style and Context in the Monuments of Ottoman J erusalem 8. Conclusion The present book represents the first serious attempt to record the buildings of Ottoman Jerusalem in a comprehensive manner, but even so it cannot hope to say the last word on any of them, and should rather be regarded as a foun dation on which future research can be based. It does show, however, that the architecture of Ottoman Jerusalem, which grows so naturally out of the rich Mamluk heritage in the city, repays close study. The absence of major buildings of international importance is less significant than the preservation, in largely excellent condition, of an entire pre-modern town. That town is overwhelminglyan Arab one, and to this day it is cherished by its Muslim inhabitants. Written records corroborate oral traditions that many of its distinguished families have been there for centuries . This book owes a great deal to the love which the local Muslims have for Jerusalem, and it is a matter for celebration that the city's heritage of Ottoman buildings has now been recorded in detail here. Ottoma n architects in Jerusalem had inherited from their Mam luk and still earlier predecessors an enviableconfidence in the working of the local stone and a keen awareness of its possibilities. In general, their taste was austere, even minimalist, in comparison with Mamluk work. Buildings are often generically akin with their Mamluk counterparts, but have as it were lost their decoration.The beauty of the stonework and the simplicity of the basic forms-a rcades, aedicules, domed squares, fa~adeswith recessed vertical panels - is allowed to speak for itself. The art iculation is pared down to mouldings, capitals, plinths, embrasures and the like, and is consistently set off by expanses of plain stonework. The Ottomans were the last major Islamic dynasty of international stature to rule Jerusalem. Theirs was by far the most sophisticated and powerful of the late medieval to earlymodern Islamic states; indeed, in some sense this was the supreme Islamic dynasty . It brought to fruition many trends and ideas which were latent in earlier Islamic polities. It imposed its stamp on much of the Islamic world and for centuries it symbolised that world to the West. The Ottomans are the final link in a chain which, apart from a brief rupture under Crusader rule, stretches back well over a millennium to the beginning of Muslim rule in 638. The Islamic presence, then, is no transient episode; it is fundamental to an understanding of the city, past, present and future . Indeed, despite the evidence of Christian and Jewish presence in the Old City over the centuries, this is a Muslim town through and through . The intrinsically Islamic nature of the physical fabric of the Old City is frequently undervalued and even ignored in the West. It is sometimes demoted in the media to mere local colour. Yet even the most casual and uninformed visitor to the city cannot fail to notice this pervasive Islamic dimension . Most of the fabric which is an integral, tangible part of that dimension is of Ottoman date . Every stone of that fabric is part of a preciou s heritage and it deserves loving protection. Note I would like to express my deep indebtedness to Dr Sylvia Auld , who proved a fount of wisdom, information, ideas and inspiration at every turn in the preparation of this chapter. Her extensive knowledge of Jerusalem, built up over more than three decade s, which she shared so generously with me and which is nourished by her deep personal commitment to the city and its people, was absolutely indispensable. My warm thanks go also to Martin Dow for much useful information on vernacular architecture in the city, and to Dr Yusuf Natsheh for his kindness in showing me around the Ottoman buildings to which he has devoted so much of his life. This chapter should be read in conjunction with those of Dr Natsheh and Mr David Myres; while all three inevitably cover a good deal of the same ground, the differences of approach and emphasis should ensure that the picture of Ottoman architecture in Jerusalem that finally emerges from these three contributions is a fully nuanced one. 23 Ottoman J erusalem • I Haram al-Sh arif, Jerusa lem, with the Dome of the Rock and al-Aqsa Mosque. Hajj Vehaletnamesi, dated 951 / 1544 -45, in th e Topkapi Sarayi Library. II Haram al-Sharif, Jeru salem , with the Dom e of the Rock and al-Aqs a Mosque. Nur -i vahhaj li tahsil alilaj, decorated by the copyist Mustafa Kashif (muzehhib) in 1253/ 1857, Ms.Vat.Turco 125, f.26r. III Sandal of the Prophet with a kufic inscription which reads 'jimjim e al-nabi '. Nur -i vahhaj li tahsil al-ilaj , decorated by the copyist Mustafa Kashif (m uzehhib) in 1253 / 1857, Ms .Vat .Turco 125 , f.6 v. ·-----i-~----- -~ ('.;._, IV Sandals of the Prop h et and the Hara m al-Sharif , Jerusalem . Jerus alem , N ation al an d Un ivers ity Library, Yah.m s.Ar.117, f.4lr. Ottoman Jerusalem V Sandals of the Prophet in an architectural frame . Ceramic panel from the Darwishiyya Mosq ue , Damascus, dat ed 982/1574 -75 . ""·· VII Dervish scene around the Dome of the Rock. Falnama attr ibut ed to Aqa Mirak. Tabriz 1550 , Pozzi Collection, Musee d 'Art et d 'Histoire , Geneva , no.1971-107 /34. VI Footprints from the sandals of th e Prophet . Falnama attributed to Aqa M irak. Tabriz 1550 , Pozzi Collection, Musee d'Art et d'Histoire, Geneva, no.1971 -1 07 /3 4. VIII Marble slab with the representation of the footprint of Jesus Christ in the Quo Vadis Chapel , Rome. Ottoman Jerusal em IX Solomon's Library in the form of a labyr inth . Nur -i vahhaj li tahsil al-ilaj, decorated by the copyist Mustafa Kashif (muzehhib) in 1253 / 1857, Ms.Vat.Turco 125, f.20r. XMundus imaginalis. Cosmic representation of the heavenly world. Nuri vahhaj li tahsil alilaj, decorated by the copyist Mustafa Kashif (muzehhib) in 1253 / 1857, Ms.Vat.Turco 125, f.45r. XI Double frontispiece from the Qur'an endowed in 964/1556 in the name of Bayazid, son of Sultan Sulaim an al-Qanuni. Although never in fact sultan, Bayazid is so named here. Ottoman Jerusal em XII Man's qumbaz from Jerusalem, c. 1930 . White silk with red and yellow stripes, bound with ochre-coloured silk braid at the neck opening. A Turkish entari, alth ough similar in construction, would not have h ad the tie at the front , and would most likely have had more elaborate trimming . International Folk Art Foundatio n Collec tion in the Museum of Ne w Mexico, Accession no. FA. 72-25- 7 . XIII Woman 's wedding dre ss, thub abu qutba, from Jerusalem, c. 1850. The body of the dr ess is made up of piece s of handwoven red and green silk, arranged to form stripes. Gold silk panels are used on the sleeves. The shape of the dress , with its long triangular sleeves, round neck opening and slit at the neck is typical of dresses of the central region of Palestine . Apart from the decorative stitching u sed to piece the dress, the garment is decorated with a central panel or qabba, sparingly embroidered with scalloped silk in red , green and yellow, and silk tassels on the neckline cord. Museum of New Mexico Collection, Accessio n no. 3370. XIV Dre ss from the Jerusalem region , c. 1920s or earlier. The main body of the dress is a striped silk produced in Syria, qasabiya. There are also three red and green taffeta skirt inserts, as well as a yoke of patterned velvet, perhaps of European origin. The dress is further decorated with an elaborately embroidered qabba and additional embroidery on the sleeves and skirt inserts. Both this dress and the one worn in the photograph in col. pl. XIII would ha ve been part of a trou sseau. London, Museum of Mankind , 1969 ASS 9 . XVI Detail of a knotted wool carpet with a red ground; the pattern of blue -green six-petalled flowers with yellow centres is reminiscent of the same device found on stone roundels on the 'Imara al-'Amira, Dar Bairam Jawish and city walls. Islamic Mu seum of al-Aqsa Mosque , Jerusalem (Photograph © S Auld) . Ottoman Jerusalem XVII Dome of the Rock. The present wooden ceiling of the inner ambulatory (Courtesy of the Aq \ a M osqu e Restoration Committee) . Ottoman Jerusalem XVIII Fragments of th e origi n al Umayyad mosaics from the parapet niches, uncovered in 1966 (Ph otograph © John Carswell). XX Fragments of the origina l Umayyad mo saics from the p arapet niches, uncovered in 1966 (Photograph © John Carswe ll). XIX Fragments of the orig inal Umayyad mosaics from the parapet niches, uncovered in 1966 (Photograph © John Carswell) . XXI One of the orig in al waterspouts from the Dome of the Rock , decorated in black under a transparent glaze, c. 1560 (Photograph © John Carswell). Ottoman J erusalem XXII Examples of the original 16th-centur y glazed bricks used to screen the windows of the Dome of the Rock, removed in 1966 (Photograph © John Carswell) . XXV Three underglaze tile panels, used in sub sequent repairs to the Dome of the Rock ; two are dated, AH 1233 and AH 1234 . Probably Syrian manufactur e (P h otograp h © John Carswell). XXIII Hexagonal, triangular and square tiles decorated in black under amber or green transparent glazes, from Sulaiman 's 16thcentury restoration, removed in 1966 (Photograph © John Carswell). XXVI (Top ) Four cuerda seca 16thcentury tiles, removed in 1966. (Bottom) A tile showing the transition from cuerda seca to true undergl aze decorati on (Phot ograph © John Carswell) . ,;. ,· ..- ~: ,- XXIV Cuerda seca tiles used to decorate the drum of the Dome of the Rock in the mid-16th century, removed in 1966 (Photograph © John Carswell). XXVII Two underglaze decorated tiles, and plain glazed white, turquoise and manganese purple bricks, from Sulaiman 's mid16th century restoration , removed in 1966. The bricks clearly show the original copper pin s used to lock them together, in turn fastened to the masonry fa<;:ade(Photograph © John Carswell). Ottoman Jerusalem XXVIII Thi s stud y of the (essentially ) Ottoman Holy City was taken before the aesthetic and integrated balance of the skyline was marred by hi gh-ri se buildin gs in the second half of the 20th century (Photograph © Alistair Duncan ) . XXXI Detail of a cuerda seca tile from the Dome of the Rock, c. 1560 (Photograph © John Carswell). XXIX Tw o ra re imported Iznik tiles used in the redecoration of the Do m e of the R ock, c. 1560 . Removed in 1966 (Photograph © John Ca rswell). ｾ＠ XXX De tail of a cuerda seca tile from the Dom e of the Rock , c. 1560 (Ph otog raph © John Carswell ) XXXII Tile from the Dome of the Rock, c. 1560,painted in underglaze cobalt blue , turquoise and black (Photograph © John Carswell). Ottoman Jerusal em XXXIII Tile from the Dome of the Rock, c. 1560 , painted in underglaze cobalt blue , turquoise and black (Photograph © John Carswell) . ---,·· XXXIV Tile from the Dome of the Rock , c. 1560 (detail) (Photograph © John Carswell). XXXVI The Dome of the Rock during the latter part of the restoration work of the late 1960s . In the foreground, Qubbat alArwah (Photograph © Alistair Dun can ) . XXXV Tile from the Dome of the Rock, c. 1560 (detail) (Photograph © John Carswell ) . Ottoman Jerusalem ｾ＠ ... ·'.la'"-~ XXXVII Ce iling of the Pasha Room , Rabah Efendi Hussaini House (n ow T h e American Colony Hotel ), in the Shaikh Jarrah Quarter, from 186 5- 76 (restored ) (Photograph © Sharif alSharif). XXXIX Detail with grotesque ornament from vaulted ceiling in the entrance of Imperial Hotel , near the Jaffa Gate, Jerusalem, built by the Greek Orthodox Patriarchate in 1884-89 (Phot ograph © Sharif al-Sharif) . XXXVIII D eta il of ceiling , dec oration with baroque quatrelobe outline and ba mb oo octagon , from C optic Patriarchate, Old City (P h otograph © Sharif al-Sharif) . XL Ceiling corner detail , St Saba Monastery, with St Matthew in a baroque cartouche and a vase of flowers (Photograph © Sharif al-Sharif) . XLI Detail of a frieze in the Azar family house in Nazareth, showing a panorama of a seashore with two Ottoman castles and a steamboat behind a Palestinian village (Photograph © Sharif al-Sh arif) . Ottoman Jerusalem XLII Sabi! Bab al-'Atm against the backdrop of the Aminiyya fa;ade (730 / 1329-30). XLIII Tile panel in the Islamic Museum of al-Aqsa Mosque. XLIV Hujrat Islam Beg, west elevation . Ottoman J erusalem XLV Hujrat Islam Beg, south elevation . XLVI North -West Kha lwa of Ahm ad Pasha , view of south and east fac;:ades with further khal was in the background . XLVII Dar al-B airam Jawish , look ing north, built as a qantara ab ove Tar iq al-Wad . Ottoman Jerusalem XLVIII North-West Khalwa of Ahmad Pasha, east fa<;:ade. LI Pavilion of Sultan Mahmud II (Photograph © Michael Burgoyne ) . XLIX Mastaba of Bab al-Maghariba. L Mihrab 'Ali Pasha , inscription. Lill Mastaba and Sabi! Mustafa Agha. The mastaba is also known as Mastabat al-B asiri and the sabil as Sabi! al-Shaikh Budair (Photograph© Michael Burgoyne). Ottoman J erusalem LIV No rth -West Kha lwa of Ahmad Pasha , int ernal d oorway. Ottoman Jerusal em -- ~.i~'~ LVII QubbatYusuf (Photograph © Michael Burgoyne ) . LV Mihrab 'Ali Pasha (Photograph © Michael Burgoyne). 4~~-·-:,..._..,ｾ＠ ,... < ~-~·4· ---= :-_'-' -,I ｾ＠ - ---~ LVIBurj al-Laqlaq (Photograph © Michael Burgoyne) . r ---~ ｾ＠ - .._._ :_, ｾ＠ -r L · / ~ LVIII Sabi! Mustafa Agh a (al-Sh aikh Budair) (Photograph © Michael Burgoyne). Ottoman J erusalem LIX Sabi! Mustafa Agha (al-Shaikh Budair ): detail of SW column . LXII Qubbat Yusuf LX North-West Khalwa of Ahm ad Pasha . LXI Sabi! Tariq Bab al-Nazir (Photograph © Michael Burgoyne). LXIII North-West Khalwa of Ahmad Pasha. LXIV Sabi! Bab al-Silsila (Photograph © Michael Burgoyne).