6

The sedentary sultan

IT WAS NOT until some three weeks after Sultan Süleyman’s death that Selim reached Istanbul from his governorship at Kütahya. Sokullu Mehmed Pasha had so deftly concealed the Sultan’s passing that many were surprised by his son’s appearance in the capital on 29 September. Immediately proclaimed sultan – as Selim II – he set off for the Hungarian front three days later, but was warned by Sokullu Mehmed that he must not continue beyond Belgrade because there was too little money in the campaign treasury to pay the troops the donative customary on the accession of a new sultan. On the pretext of completing repairs to the fortress, Sokullu Mehmed detained the army at Szigetvár until he was able to announce that it was too late in the season to continue the campaign, and the order was given on 20 October to begin the arduous march home.¹

News of Süleyman’s death and Selim’s accession had still not been officially proclaimed to the troops when the army set off the next day towards Istanbul. Following his death, Süleyman’s body had been washed and temporarily buried under his tent; now it was exhumed in order to be taken home. One of the late Sultan’s pages was chosen to sit in the Sultan’s carriage and gesture to the troops in his character. The chronicler Mustafa Efendi of Selanik (Thessalonica), who as a young man was present on the Szigetvár campaign, was one of six people chosen to recite passages from the Koran by the side of the carriage as it progressed; he describes the page chosen to be Süleyman’s double as white-faced, hawk-nosed, with a sparse beard and a bandaged neck, and an appearance of ill-health. Mustafa Efendi reported that although everyone was by then aware that Süleyman was dead, no formal proclamation was made until 48 days after the event, by which time the cortège was nearing Belgrade where the new sultan was waiting.² The funeral prayers performed there in Selim’s presence were later repeated in front of his father’s newly-built mosque in Istanbul, to give the people of Süleyman’s capital a last chance to remember the Sultan and his works. Sultan Süleyman was subsequently buried in the place he had chosen – not, as was customary, in front of the prayer-niche wall of the mosque bearing his name, but in a tomb built in its garden, next to that of his wife Hürrem Sultan.3

The ceremony accompanying the accession of an Ottoman sultan was customarily modest; the new ruler was set on the throne and his statesmen swore allegiance to him. Selim II, however, unwittingly set a precedent for the future. Like his father and grandfather before him, after his confirmation as sultan he had visited the shrine of Ayyub Ansari (the Companion of the Prophet Muhammad whose tomb had been miraculously rediscovered during the siege of Constantinople in 1453) to seek the saint’s blessing before departing for the front; as it fell out, Selim made this pilgrimage immediately after his enthronement, and thereafter every new sultan visited the shrine as an integral part of his accession ceremony.⁴ One benefit of the pilgrimage was that it gave the new sultan an opportunity to progress triumphally through the city in full view of his subjects.

Sokullu Mehmed Pasha had barely been able to contain the troops at Belgrade where, on their first encounter with the new sultan, they demanded their accession donative – he handed out a small amount sufficient to pacify them and, promising the balance later, raised their pay and that of the various bureaucratic and service personnel also on the campaign. The march home went smoothly enough but as the army reached Istanbul, the janissaries mutinied. The Sultan and his entourage entered the city through the Edirne Gate, in the shadow of the mosque sponsored by Selim’s sister Mihrimah Sultan, but when they reached their parade ground near the Şehzade mosque, the janissaries refused to continue towards the Topkapı Palace. For an hour they stood their ground; then they moved off again, then again came to a halt, this time before the baths of Sultan Bayezid II. Here one of Selim’s vezirs and the grand admiral Piyale Pasha remonstrated with them; both were knocked from their horses, and the stand-off was only broken by the distribution of handfuls of gold coins. Those janissaries appointed to duty at the palace proceeded there, but once inside, they barred the gates against the Sultan. Sokullu Mehmed Pasha resolved the crisis by advising Selim that the only way out of this potentially dangerous situation was to pay the balance of the accession donative forthwith.⁵

There had been janissary insurrections before, most notably during the first sultanate of Mehmed II in the 1440s. Selim’s humiliation demonstrated that to achieve a smooth passage to the throne it was not enough to be sole heir; it was still essential also to have the support of the janissaries and the other elite units of the army. In theory these corps were the servants of the sultan but in reality he was the prisoner of their whims, and without their support he could not exercise his sovereignty. Securing the allegiance of their troops was as essential for Ottoman sultans as it was for the crowned heads of Europe. As Ottoman history amply demonstrated, deposition or regicide was the fate of the monarch who forfeited this loyalty.

Like his brothers, Selim had received training befitting a warrior-prince, and had already been exposed to the rigours of campaigning. At the age of twenty, after a short spell in Konya, he was sent to Manisa in place of his dead brother Mehmed as prince-governor of Saruhan province – the post customarily held by the favourite to succeed to the sultanate – and remained there until the conflict with his brother Bayezid in 1558 when he was sent back to Konya. After the defeat of Bayezid in the succession struggle Selim was appointed to Kütahya, where he stayed until his father’s death. Süleyman had demonstrated the degree of his reliance on Selim in 1548 by leaving him as regent in Istanbul while he himself was absent on the Iranian front.6 On that occasion Selim seems to have acquitted himself well, yet he failed to embrace his new role with much energy when faced with the total responsibility of the sultanate. The fear inspired by the ever-present threat of mutiny among his troops was matched by his fear of a coup during his absence from the capital. After he became sultan Selim never again travelled further from Istanbul than the royal hunting grounds at Edirne.⁷

Sokullu Mehmed Pasha ran Selim II’s empire as the visible face of Ottoman rule, while the Sultan stayed aloof from the cut and thrust of decision-making. This remarkable man occupied the post of grand vezir continuously for fourteen years, under three sultans. He belonged to the minor Serbian aristocratic family of Sokolović (‘Son of the Falconer’), and was a product of the youth-levy. During Süleyman’s reign he moved smoothly up the palace hierarchy. His first significant post was as admiral of the fleet following the death of Barbarossa. He then held a number of important provincial governorships and military commands on the western and eastern frontiers of the empire until he was appointed grand vezir by Süleyman in 1565. The dangerous circumstances of Süleyman’s attempt to prevent conflict among his sons allowed Sokullu Mehmed Pasha to demonstrate his value to the dynasty. He was entrusted with suppressing the revolt of the pretender pseudo-Mustafa in 1555, and in 1559, as commander of the army sent by Süleyman to support Selim against his brother Bayezid, he proved his indispensability to the future sultan. Selim owed his victory to Sokullu Mehmed, and, while the Prince did not fail to reward him, Süleyman bound him still closer by marrying him, in 1562, to Selim’s daughter Esmahan Sultan (Sokullu Mehmed divorced his two other wives to free himself to accept this honour). The relationship of this able statesman and military commander with his master had echoes of that of Süleyman’s hapless favourite İbrahim Pasha; like İbrahim’s, Sokullu Mehmed’s special status was prominently advertised by the location of his palace, on the Hippodrome, close to that of his master.8

During the first years of Selim II’s reign the Ottomans were occupied with small but important operations on distant frontiers. When news of Süleyman’s death reached the province of Yemen in 1567 the powerful chief of the Zaydi clan, Imam Mutahhar ibn Sharaf al-Din, rallied his Shia followers in open rebellion. Ottoman authority in Yemen had always been tenuous. In this rugged and sparsely-populated terrain, it proved impossible to suppress independent local Arab chiefs, while a common profession of Islam was inadequate to secure their acquiescence to the imposition of an entirely alien regime. The construction and garrisoning of the fortresses essential to Ottoman attempts to subdue local dissent made the province expensive to control, and although the energetic governor Özdemir Pasha had made Ottoman rule more effective during his tenure between 1549 and 1554,⁹ his successors proved weaker. Yemen had been divided into two provinces in 1565, but the Ottoman governor of the southern province centred on San‘a was murdered and the many strongholds the Ottomans had previously secured were lost to Imam Mutahhar.¹⁰

Yemen was important because control of the spice route brought substantial customs revenues into the coffers of the Ottoman Empire. In 1568 a strong expedition was sent to pacify the province under the command of Sultan Selim’s former tutor and confidant Lala Mustafa Pasha, a choice which showed that Selim was not entirely the pawn of his grand vezir, for Sokullu Mehmed resented Lala Mustafa’s place in the Sultan’s affections. To put down the uprising in Yemen Lala Mustafa needed men and supplies from Egypt but the provincial governor, another rival, Koca (‘Great’) Sinan Pasha, refused his requests and made it impossible for him to pursue the campaign. In a spate of petitions to the Sultan the two defended their respective positions. Koca Sinan proved the stronger, and Lala Mustafa was dismissed from command of the Yemen campaign. To mark his continuing favour, however, Selim created for him the position of sixth vezir in the governing council of the empire. Koca Sinan took over leadership of the campaign, but the logistic demands of fighting in Yemen forced him to reach a settlement with the Zaydi. The two Yemeni provinces were reunited, and by 1571 Koca Sinan was able to return to Cairo.¹¹ The instability in the region caused the Ottomans to look again at the possibility of building a canal to connect the Mediterranean and the Red Sea. As an imperial order enjoined the governor of Egypt:

. . . because the accursed Portuguese are everywhere owing to their hostilities against India, and the routes by which Muslims come to the Holy Places are obstructed and, moreover, it is not considered lawful for people of Islam to live under the power of miserable infidels . . . you are to gather together all the expert architects and engineers of that place . . . and investigate the land between the Mediterranean and Red Seas and . . . report where it is possible to make a canal in that desert place and how long it would be and how many boats could pass side-by-side.12

But again the proposal went no further.

Muscovy’s annexation of the Muslim Tatar khanates of Kazan and Astrakhan in the 1550s adversely affected its cordial relations with the Ottomans and altered the strategic balance in the region. The gradual penetration of Muscovy into Transcaucasia added a third power as a potential focus of loyalty for local rulers whose vacillating preferences for Ottomans and Safavids exacerbated traditional rivalries in the region. The pressure from the Tatars of the steppe, prompting local people to seek Muscovy’s protection, was a further complication. In 1567, when Muscovite help was requested by one chieftain, Ivan IV obliged by building a fort on the River Terek, which rises in the central Caucasus and empties into the Caspian.¹³ The Özbek and Khivan Khans responded with an appeal to the Ottomans, complaining that Muscovite control of Astrakhan blocked the route south both for merchants and for pilgrims to Mecca.¹⁴

Sultan Süleyman and his vezirs had shown little interest in campaigns which would take them into territory even more inhospitable than the frontier zone with the Safavids, but Selim’s accession brought a change in policy. Encouraged by the Muslim rulers of the region, Sokullu Mehmed Pasha sought local advice on the feasibility of digging a canal between the Don and Volga rivers, and was encouraged to believe it would be possible. In Süleyman’s time reports had reached Moscow of talk in Istanbul of constructing a riverine route between the Sea of Azov and the Caspian, but no steps had been taken to realize the project. In the second year of Selim’s reign – a year after the Muscovites had built the fort on the Terek – an expedition to secure Astrakhan was in the making. Craft able to navigate the Don were built in the dockyards of Feodosiya and the necessary matériel and supplies were shipped from Istanbul to Azov. Troops were mobilized from Rumeli and northern Anatolia. Doubtful of the practicability of such a canal and afraid of the presence of Ottoman interlopers so close to his territory, the Tatar Khan of the Crimea was reluctant to participate – but unable to refuse. Muscovy offered the Safavids cannon and guns to create a diversion in the Caucasus, but the offer was not accepted.¹⁵

The Astrakhan campaign of 1569 was commanded by Kasım Pasha, the governor of Kefe province. The Don was so shallow in summer that even the special boats built at Feodosiya found progress up-river from the Sea of Azov difficult. The chosen site for the canal lay to the south of modern Volgograd, where the Don and Volga were still 65 kilometres apart. The land between them was hilly, and it became clear that a canal could not be cut through such terrain. The decision was therefore taken to transport the flotilla and supplies overland between the rivers, following the practice of the Cossacks of the Don. The effort needed simply to level the ground for this purpose was disproportionate, however, and Kasım Pasha decided to send his heavy equipment back down the Don to Azov, after which the force taking it there would march across the steppe to Astrakhan and meet up with him and his men who would have reached this city by following the Volga south. Ill-equipped and short of food as they were, the Ottoman forces were unable to make much impact on Astrakhan. They retired in September, sustaining further losses in men and matériel both as they returned to Azov and, owing to the seasonal storms, on their sea voyage back to Istanbul.16

The scheme to cut a canal to connect these two great rivers was consonant with Sokullu Mehmed’s penchant for ambitious engineering projects and his interest in military logistics, but although he found a willing and tenacious lieutenant in Kasım Pasha, Kasım’s intention to continue the campaign in the following year was rejected by Istanbul. Though the bold canal scheme had failed, it none the less had significant consequences. As little as the Ottomans did Tsar Ivan IV wish to get involved in a war in the steppe, and when the 1569 expedition was over he sent an envoy to Istanbul to congratulate Selim on his accession. The Russians abandoned the Terek fort, but Ivan refused to surrender Astrakhan.¹⁷

This amicable agreement between Tsar and Sultan reckoned without the Crimean Tatars, however. In 1571 the Tatars demanded the cession of Kazan and Astrakhan, and raided and burnt the capital city of Moscow. Selim took advantage of this new situation and sent a message to Tsar Ivan reiterating their demand, agreeing to support the Tatar Khan in a new expedition to retake the two cities. In the summer of 1572 the Tatar army again set out towards Moscow, but this time they suffered a great defeat close to the city, and the reconquest of the lower Volga was abandoned by the Crimeans and the Ottomans alike.^*18

*

Although Ottoman interest in distant adventures on land was not yet exhausted, Sultan Selim II’s reign was more noted for naval activity, as he continued the active forward policy pursued by Süleyman against the Spanish Habsburgs in the western Mediterranean. Nor was Spain alone in opposing the Ottomans: in the Magreb, the Sa‘di dynasty of Morocco and the Hafsids of Tunis offered Muslims alternatives to a dynasty based far away whose ability to protect its North African territories depended on the security of its maritime routes. Blocking Ottoman passage into the western Mediterranean were the Knights Hospitallers at Malta, Sicily – governed by a Spanish viceroy – and the Spanish outpost of La Goletta near Tunis.

In 1568 the Ottomans were sowing discord within the Sa‘di clan in an attempt to undermine the dynasty’s hold on Morocco, when the corsair captain Kılıç (‘Sword’) Ali – also known as ‘Uluç’ (‘Barbarian’) Ali in reference to his non-Muslim, Italian origin – in Ottoman employ at the time, sent a small army overland from Algiers which defeated the Hafsids in battle to win their territory of Tunis, although not yet the key fortress of La Goletta. Kılıç Ali had timed his expedition against these Spanish vassals well, for the Spanish armies were occupied in the Netherlands or in suppressing the Morisco revolt on the Spanish mainland. The Moriscos petitioned the Sultan for help but their uprising was put down by King Philip II’s troops19 which led to further Morisco migration into Ottoman domains.²⁰

The major events of these years were the capture of Cyprus from Venice in 1571 and the defeat of the Ottoman fleet at Nafpaktos in the battle of Lepanto the same year. Venice had ruled Cyprus since 1489, having been invited there to protect the last, weak Lusignan kings against Ottoman attack. In those days, when Mamluk Egypt was a power to be reckoned with, Venice had paid Cairo an annual tribute for its most easterly possession, as it now paid the Ottomans. Friction between the Ottomans and Venice was never completely absent but outright war was usually avoided. According to contemporary Ottoman historians, it was Venetian protection for the corsairs who plagued Ottoman vessels sailing the route to Egypt which drove Selim to mount a campaign to conquer Cyprus.²¹ Naval preparations were under way during 1569, the year of the ill-fated Astrakhan campaign. Sokullu Mehmed Pasha cautioned against such an undertaking, so soon after the Malta debacle of 1565, but his rivals convinced the Sultan to seek a juridical opinion authorizing the expedition, which was in breach of the peace treaty with Venice that had been renewed following his accession. Sheikhulislam Ebüssuud conveniently declared himself of the opinion that an attack on Cyprus was legitimate if the intention behind the declaration of war was the recovery of lands which had once been under Muslim rule – as Cyprus had been, briefly, in the early Islamic period. The problem was framed thus:

A land was previously in the realm of Islam. After a while the abject infidels overran it, destroyed the colleges and mosques, and left them vacant. They filled the pulpits and galleries with the tokens of infidelity and error, intending to insult the religion of Islam with all kinds of vile deeds, and by spreading their ugly acts to all corners of the earth . . . When peace was previously concluded with the other lands in possession of the said infidels, the afore-named land was included. An explanation is sought as to whether, in accordance with the [sacred law], this is an impediment to the Sultan’s determining to break the treaty.

ANSWER:
There is no possibility that it could ever be an impediment. For the Sultan of the people of Islam (may God glorify his victories) to make peace with the infidels is legal only where there is benefit to all Muslims. When there is no benefit, peace is never legal. When a benefit has been seen, and it is then observed to be more beneficial to break it, then to break it becomes absolutely obligatory and binding.22

This was the only occasion during the sixteenth century on which a peace treaty was broken by the Ottomans.²³

The considerable amount of money needed by the Ottomans before they could embark upon the conquest of Cyprus was partly met by the sale of monasteries and churches belonging to the Orthodox Church in the European provinces of the empire. Orthodox Christianity had an honourable past as a bulwark against the Latins, earlier in the shape of Venice and the Papacy but now exemplified by the Catholic Habsburgs, and the functioning of the Orthodox Church within the Ottoman Empire was generally unproblematic: as long as the institution acted within the prescribed bounds of its relationship with the state, it had little cause for complaint. Sultan Selim’s confiscation of Church lands in 1568 was not aimed at the destruction of the Church, but was in line with the continuing efforts of Sheikhulislam Ebüssuud – who, until his death in 1574, served Selim as he had served Süleyman before him – to streamline the land-holding system in the Ottoman domains. Once forfeit, churches and monasteries could be bought back, to the treasury’s benefit. But the effects of the confiscations were uneven, the richer monasteries surviving while the poorer ones were sold to new owners who could afford to pay the price.²⁴

The Ottoman treasury thus enriched, Lala Mustafa Pasha was appointed commander-in-chief of the Cyprus land forces while the fleet was commanded by the grand admiral Müezzinzade (‘Son of the Prayer-caller’) Ali Pasha – who, in the words of one modern historian, ‘had never in his life directed even a caique’;25 he was fortunate to have at his side Piyale Pasha, who had previously served as grand admiral for fourteen years. Though the European powers had been aware for some time that a well-equipped and numerous fleet was being readied they were not certain whither it was bound. Rumour favoured Cyprus, and there was palpable unease in Venice in 1568–9 because the administration of the island was recognized to be corrupt, and unable to stand against any Ottoman attack; by the time the rumours were confirmed, some improvements to defences and supplies had been made. The Sultan’s envoy reached Venice in March 1570 with an ultimatum: Cyprus must be surrendered, or the Ottomans would mount an attack. By September they had occupied the inland city of Nicosia.

Venice was hard-pressed to find allies for the defence of Cyprus. The Austrian Habsburgs and the Ottomans had made peace in Hungary in 1568; the Spanish Habsburgs saw no strategic value in the island and owed Venice nothing, for Venice had failed to provide support against the Ottomans at Malta in 1565. In past times, indeed, Venice had always preferred to maintain good relations with the Ottomans rather than to participate in leagues against them. On this occasion much effort, especially on the part of the Pope, resulted in May 1571 in an agreement between Venice, the Papacy and Spain, a condition of which was that Venice would go to Spain’s aid in North Africa.

In September 1571 a fleet under the command of Don John of Austria, the illegitimate son of the former Holy Roman Emperor Charles V and half-brother of Philip II of Spain, sailed east from Messina. It reached the Ionian island of Cephalonia only to learn that Mağosa (Famagusta), the last Venetian stronghold on Cyprus, had fallen to the Ottomans on 1 August after an eleven-month siege. The reconquest of Cyprus rather than its defence must now be the aim of the Christian allies. In the Gulf of Patras, however, at the mouth of the Gulf of Corinth, Don John’s fleet found an Ottoman fleet that had spent the summer raiding and even capturing Venetian islands and possessions on the Adriatic coast. Don John seized the opportunity, and the two fleets engaged off Nafpaktos on 7 October.

Like the Ottoman failures before Vienna in 1529 and 1683, the battle of Lepanto is an event graven on the western consciousness as having only narrowly saved Christendom from being overrun by the ‘infidel Turk’. It was much described by eye-witnesses and later historians alike, but no Ottoman contemporary thought to preserve his recollections for posterity²⁶ – indeed, there were few survivors among the Ottoman seamen. Don John had over two hundred galleys – oared warships armed with cannon – and six galleasses – in effect, large galleys armed with bigger cannon – while the Ottomans had an even greater number of vessels but lacked galleasses. A shift in the wind meant that the battle was fought on calm seas and heavy cannon could achieve maximum effect; fired relentlessly at the Ottoman fleet at close quarters, they proved decisive for the Christians. Most of the Ottoman fleet was burnt and sunk; wounded men and corpses rendered ‘tutto il mare sanguinoso’. A violent storm blew up after the battle was over – the battle had lasted for four hours – and finished off any who might have hoped for rescue.27

Don John took to sea again in 1572, but Christian euphoria and plans for future attacks on Ottoman territories were to prove short-lived. The Ottomans had spent the winter rebuilding a fleet to replace that lost at Lepanto. Müezzinzade Ali Pasha having been killed in that battle the command was given to Kılıç Ali Pasha as grand admiral.²⁸ The two fleets skirmished inconclusively off the Peloponnese but the anticipated victory eluded the Christians. The league was beginning to fail, and did not sail again against the Ottomans in 1573 as planned. Instead, Venice sought peace through its representative in Istanbul, who had been held under house arrest since hostilities began in spring 1570.²⁹ As well as accepting the loss of Cyprus, Venice paid an indemnity of 300,000 ducats to the Ottomans; prisoners were exchanged and the Adriatic coast frontier set at pre-1570 limits.³⁰

One man at least on the winning side did not receive what he had hoped for from this treaty: Sultan Selim’s close confidant, the Sephardic Jewish banker and merchant Joseph Nasi. In recognition of his support in Selim’s struggle against his brother Bayezid Nasi had been rewarded with the title of Duke of Naxos, along with the considerable customs revenues from the island’s wine trade. It was said he wished now to be made King of Cyprus – contemporary European historians certainly credited him with having encouraged Selim to declare war against Venice in 1569, and it was rumoured that he had ready a standard embroidered with the arms of Venice and the legend ‘Joseph Nasi, King of Cyprus’ in gold letters.³¹ But the Sultan chose to keep the revenues of the island for the treasury, and Nasi was disappointed.

When it came to settling their own people in Cyprus the Ottomans experienced serious difficulty in promoting voluntary immigration from Anatolia. The island lacked the attractions of newly-conquered territory on the Rumelian front, for instance, or the consolations of Istanbul following the fall of the city to Mehmed II in 1453. Moreover, the climate was uncomfortably hot in summer and grazing land was scarce. Some volunteers there must have been, but forcible resettlement predominated: single women were sent as brides for the soldiers garrisoning the island’s fortresses; skilled peasants of good reputation were transported there with promises of land and tax relief. Many of those selected as settlers hid from the authorities before they could be apprehended, while many others succeeded in returning to the mainland, a matter of obvious concern to the government, which resorted to banishing undesirables there – an early precedent for the British policy of deporting petty criminals to Australia, perhaps. Those suspected of Kızılbaş sympathies, against whom there was a new wave of vigilance in the later part of the sixteenth century,32 were sent there, as were others considered a threat to the stability of society, including unruly religious students, brigands and minor officials who had fallen from favour.³³

Don John – having failed in the unrewarding season of 1572 to engineer the collapse of Ottoman naval power that had been anticipated after Lepanto – retook Tunis in 1573 with a Spanish armada, and built a new fortress at La Goletta. With a fleet larger than that lost at Lepanto the Ottomans again captured Tunis in 1574, in a combined operation with the land forces of the provinces of Algiers, Tripoli and Tunis. In diplomatic moves made before the fleet set out the Ottomans sought the assistance of the Moriscos of Spain, suggesting that they ally with the Protestants of the Netherlands; an Ottoman agent was also sent directly to the Netherlands to propose an alliance for a combined strike against Spain, but nothing came of it.³⁴ Agents and spies throughout Europe kept Ottoman statesmen well-informed about political alliances, while Joseph Nasi’s commercial connections provided the Sultan with another effective intelligence-gathering network across a wide area.

Lasting control of North Africa posed both Habsburgs and Ottomans a challenge upon which the prestige of each depended. The Ottomans had a mission to protect their co-religionists but there were clear risks in repeatedly running the gauntlet of the Spanish navy. On the other hand, though Philip II could not accept an Ottoman presence so close to the heart of his kingdom, he chose to give priority to the suppression of the Protestant revolt in the Netherlands, a complex and hugely expensive logistical undertaking. In 1575 Spain declared herself bankrupt.³⁵

In 1574, the year of the Ottoman reconquest of Tunis, Sultan Selim II died after a fall in the bath. He was 50 years old. He had emulated his forebears by building a prominent, monumental mosque complex, but had broken with tradition by siting it in the old Ottoman capital of Edirne in Thrace, where he loved to indulge his passion for hunting. His father’s Süleymaniye mosque symbolized the power of the Islamic religion and the Ottoman dynasty in the imperial capital; Selim carried this message beyond the confines of Istanbul. Edirne lay on the military road to Europe and the overland route travelled by the envoys of European powers bound for Istanbul on diplomatic missions; the Selimiye – situated on rising ground at the heart of the city, on the site of a palace built in the 1360s by Sultan Murad I36 – may be seen from all approaches. The mosque’s four minarets, soaring more than seventy metres into the sky, long impressed those who passed through. A century and a half later Lady Mary Wortley Montagu, wife of the English ambassador in Istanbul, remarked that the Selimiye mosque was ‘the noblest building I ever saw’.³⁷ The seventeenth-century traveller Evliya Çelebi adduced a characteristically Ottoman justification for Selim’s choice of Edirne, relating that the Prophet Muhammad came to Selim in a dream and directed him to build there.³⁸ The Prophet’s visitation seems to have taken place even before the death of Süleyman, since the chronogram for the Selimiye’s foundation yields the date 1564–5,^*39 but it was finished only after Selim’s death. It was built with the spoils of the Cyprus campaign, as the Süleymaniye had been with those of the Belgrade, Rhodes and Malta campaigns.⁴⁰ A departure from his usual scheme of a mosque with a central dome surrounded by semidomes, the Selimiye mosque with its single dome that is broader than that of Ayasofya is considered the masterpiece of the imperial architect Sinan, in which he aimed to demonstrate his skill and virtuosity by surpassing his Byzantine model.⁴¹

Selim’s mosque complex may have been built in Edirne, but he also left his mark on the architecture and skyline of Istanbul: in 1572 he embarked upon the first major repair of Ayasofya undertaken since the appropriation of the church by Mehmed II. In the century since the Conquest the structure had become ringed around with houses and other domestic buildings, and Selim ordered that they be demolished. Subsequent inspection revealed that the buttresses were crumbling and in urgent need of repair: the historian Mustafa Efendi of Selanik noted that the building was tilting. Selim inspected the mosque in the company of Sinan and issued orders for an extensive renovation. One of the two minarets added at the Conquest had been constructed of wood, and this was to be rebuilt in brick; two further minarets were added. The Sultan roundly condemned those who considered the works unnecessary because Hagia Sophia had been built by non-Muslims.⁴²

In the precincts of Ayasofya Selim ordered the construction of two theological colleges, as well as a mausoleum for himself. He died before his mausoleum was finished and he was buried under a tented awning on its site, the first sultan to die in Istanbul. The theological colleges were never built, and the completion of the minarets and mausoleum fell to his eldest son and successor, Murad III.43 Selim’s choice of Ayasofya for his burial was not surprising. He could hardly be buried elsewhere than in the imperial capital, nor in a mosque built by any of his predecessors, and Ayasofya was hallowed through its association with Mehmed II, the Conqueror. Selim’s decision to carry out repairs to Ayasofya was neither entirely connected with his plans for his burial nor entirely fortuitous. In directing his attention to this former Christian basilica so soon after the Ottoman conquest of Cyprus he was reinforcing the Muslim ascendance demonstrated by his victory over the Christian powers, and countering any suggestion that Ottoman might had been diminished by defeat in the battle of Lepanto. By the time of Selim’s death the Pyrrhic nature of this Christian victory was becoming apparent.

Sultan Selim also continued his parents’ involvement with Mecca, and his work gave the Great mosque the distinctively Ottoman appearance it retains today. The enclosure lacked the space for a monumental mosque like those in Istanbul, so the galleries surrounding the courtyard were remodelled in the Ottoman style and given domes in place of their original flat roof. These works were continued during Murad III’s reign, serving to impress pilgrims from all over the world with the power and munificence of the new protectors of the Muslim Holy Places.⁴⁴

Selim’s death was unexpected and the transition to power of the new sultan was again handled by Sokullu Mehmed Pasha, who sent secretly to Manisa to inform Prince Murad of his father’s death. The exigencies of ensuring dynastic continuity again imposed a delay before the new sultan could stake his claim to the throne; meanwhile, Selim’s body was preserved on ice in the palace.⁴⁵ In another departure from custom, the funeral prayer was performed in the confines of Topkapı Palace, rather than in public in a mosque. In death as in life: the intimate and private ceremony reflected the growing distance between the sultan and his subjects. Moreover, the ritual prayer at Selim’s funeral was said by the Sheikhulislam, indicating the more visible and formal role given to this official in Süleyman’s reign and setting a precedent for the future.⁴⁶

Selim’s eldest son – by some twenty years – happened also to be the eldest surviving male member of the dynasty. Murad III was clearly destined to follow his father, yet to prevent any challenge to his right to succeed he took the precaution of ordering the execution of his young brothers on his accession; they were buried alongside their father Selim.47 Murad’s Jewish ‘Third Physician’, Domenico Hierosolimitano, described his master’s qualms over the impending murders:

But Sultan Murat, who was so compassionate as to be unable to see blood shed, waited eighteen hours, in which he refused to sit on the Imperial throne or to make public his arrival in the City, seeking and discussing a way first to free his nine brothers of the blood who were in the Seraglio . . . In order that he should not break the law of the Ottoman state . . . weeping, he sent the mutes to strangle them, giving nine handkerchiefs with his own hands to the chief of the mutes.⁴⁸

Murad’s successor, Mehmed III, was his eldest son by some nine years and on his accession in 1595 he ordered the execution of his surviving brothers, of whom the eldest was more than twenty years younger than him. The many small sarcophagi of Murad’s and Mehmed’s brothers demonstrated that murder was the price of avoiding the civil strife which often attended the succession of a sultan; the public was profoundly shocked. One can be thankful that subsequent generations were less prolific and never again did so many young princes die to ensure that their brother would have a smooth passage to the throne.

The office of grand vezir had been held by Sokullu Mehmed Pasha throughout the eight years of Selim II’s reign, and he continued in post under Murad III, until in 1579 he was assassinated by a disgruntled petitioner in the imperial council chamber. After his death the prestige of the grand vezirate declined: seven men occupied the post during the 21-year reign of Murad III, and as they fell in and out of favour the office changed hands among them eleven times. The grand vezir became subject to the sultan’s whim, to be replaced whenever he proved unable to fulfil his master’s demands. Neither Murad III nor his son Mehmed III cared to be personally involved in the running of the empire, but that was not to say that they no longer made decisions. Quite the contrary: the grand vezir’s independent decision-making authority was curtailed, even in routine administrative matters. Direct contact between sultan and grand vezir became less usual, replaced by written correspondence in which the sultan indicated his decisions on a range of affairs of state – appointments, salary payments, bureaucratic administration – decisions based on summaries of the issue in question presented to him in the form of petitions.⁴⁹

Even more than Selim II, Murad III and Mehmed III preferred to pass their days in their private quarters, rather than in the council chamber where state business was discussed, and in their own apartments they were more susceptible to the influence of favourites over whom the bureaucratic procedures of government had little control. Subject though he was to the limitations imposed by palace functionaries and favourites, Sokullu Mehmed Pasha was able to rein in the worst excesses of the factions which flourished in this era of weak sultanic authority, reserving many influential offices for his own protégés and family members.50 After his death, competition between those in the sultan’s orbit became entrenched.

From the time of Süleyman, who publicly favoured his wife Hürrem Sultan to the exclusion of the concubines of his harem, the status of the senior women of the royal household was transformed. Continuing the trend started by Hürrem they became more visible, and more permanently so through their public building works. Some also gained a new and powerful role, that of valide, queen-mother, mother of the ruling sultan. Hürrem had died before her son Selim came to the throne, but Selim’s concubine Nurbanu Sultan dominated the life of her own son Murad III after his accession and until her own death almost ten years later. Nurbanu was queen-mother in the fullest meaning of the term – she was the first to use the title officially.⁵¹ Long thought to have been born to a noble Venetian family, captured when a child by the grand admiral Barbarossa and consigned to the imperial harem, it seems, rather, that she was a Greek from Corfu.⁵² Hürrem had played only a modest part in diplomacy through her correspondence in Süleyman’s name with the King of Poland and the sister of the Safavid Shah, but Nurbanu’s influence on the international relations of the Ottoman state was more open. Foreign envoys knew how important it was to gain her favour; a visitor to Istanbul in the suite of the Venetian ambassador, Jacopo Soranzo, invited to witness the circumcision festivities of Prince Mehmed in 1582 remarked, ‘the wife . . . with the queen-mother governed everything . . . one had to depend on them, or at least not have them against you’.⁵³

With Murad III the royal household became still more of a ‘royal family’. On his accession he moved his household to Istanbul from Manisa where as prince-governor he had lived with his consort Safiye Sultan and their children. In Istanbul, Nurbanu made her place again beside her son, moving into the harem at Topkapı Palace from the Old Palace – whither she had retired on Selim’s death. As queen-mother, she saw to the smooth functioning of the harem and stood at the peak of its hierarchy. Her daily stipend was the highest in the empire, three times that of the Sultan himself. Nurbanu Sultan’s transfer of residence from the Old to the New Palace was celebrated with a public procession through Istanbul.⁵⁴ Within ten years of Murad’s accession the number of women in the harem of Topkapı Palace – concubines and servants – had doubled to more than a hundred. The harem quarters were rebuilt to provide more splendid apartments for the Sultan’s mother, and extra space for the growing number of women housed there. For himself Murad built a two-storey, domed bedroom–pavilion, the inner walls clad with the finest İznik tiles, and to this he added baths and a domed throne room next to his bedchamber.55

Whereas Selim, like Süleyman, had lived in a separate residence in the third court of the palace, only paying visits to the harem, Murad was immersed in domesticity – ample indication of the momentous changes occurring in the conduct of the sultan and the character of his empire. Murad was not a martial sultan, eager to lead his army on campaign, but a ruler who preferred to spend his life in the company of women. During his ten or so years in Manisa Murad had lived with Safiye Sultan as his only sexual partner in a nuclear family with their three children (of whom the future Mehmed III was the eldest). His sister Esmahan and his mother considered this monogamous relationship inadequate to ensure that there would be a candidate for the throne when the time came, however, and, probably in the early 1580s, they encouraged Murad to take concubines. He died leaving 49 children.⁵⁶

The increase in the size and importance of the harem and in the authority and visibility of the queen-mother also served to enhance the position of the harem’s custodian, the most senior among the black African eunuchs who supervised the women living there.^* Sultan Murad created (or if not created, certainly enhanced) the office of chief black eunuch soon after he came to the throne, and charged the incumbent with oversight of the endowments which supported the Muslim Holy Places, hitherto managed by the chief white eunuch of the palace who supervised the pages, the male counterpart to the harem.⁵⁷ The extensive endowments of former sultans Mehmed II, Bayezid II, Selim I and Süleyman I also soon came under the chief black eunuch’s auspices, and he began to hold weekly audiences to deal with these matters. He controlled the flow of significant amounts of money – and enjoyed the power that went with it.⁵⁸ The grand vezir and other government ministers were also losers in the redistribution of power which resulted from the gains made by the harem and its senior officer. In his account of the reign of Murad III, the intellectual and bureaucrat Mustafa Ali of Gelibolu pointed out the deleterious effects of the practices which were developing, noting that because of their proximity to the Sultan the eunuchs and concubines of the harem were now in a position to influence appointments by exerting pressure on the political process; they even began selling appointments.⁵⁹ This would not have dismayed the Sultan who was doubtless well pleased with the first fruits of his scheme to enhance the power of the palace; Mustafa Efendi of Selanik tells us that, soon after the assassination in 1579 of Sokullu Mehmed Pasha – until Murad III’s reign, at least, the epitome of the all-powerful grand vezir – he even considered dispensing with the post altogether.60

Parallel to the world of the harem in which Murad III spent so much more time than his predecessors was that of his male favourites. Among these were men who had been with him since his years in Manisa: his tutor Hoca Sadeddin Efendi, his controller of finances Kara (‘Black’) Üveys Çelebi, and his spiritual adviser, the Halveti sheikh Şuca.⁶¹ Hoca Sadeddin’s father had been an intimate of Selim I while he himself had been an assistant to Süleyman and Selim’s sheikhulislam, Ebüssuud; he is known today for the Ottoman history which he dedicated to Murad.⁶² His son Esad Efendi became sheikhulislam in due course, maintaining the place of his dynasty at the heart of the state. Soon after Murad’s accession Sokullu Mehmed set himself up in opposition to this inner circle by accusing Kara Üveys of financial irregularity. Sokullu Mehmed’s scheme backfired, and his loss of prestige was plain to all: Kara Üveys was put in control of the financial administration of the empire with a seat in the governing council, while Sokullu Mehmed’s appointees were purged and their possessions confiscated. Most pointedly, his cousin Sokullu Mustafa Pasha, the able governor of Buda, was executed and Kara Üveys appointed in his place. Reluctant though Kara Üveys was to accept this post so far from the seat of power, he had little choice.⁶³ He was later appointed governor of Egypt, and during his tenure a revolt broke out in response to the harsh financial controls he imposed on the army. The mutineers entered his council chamber and pillaged his private quarters; he was personally attacked, and members of his retinue were killed.⁶⁴

Sheikh Şuca, an unlettered man, fostered Murad’s interest in mysticism. The Sultan relied on the Sheikh to interpret his dreams and forecast his destiny. This was by no means unusual, for hand-in-hand with the official promotion of Sunni orthodoxy went an eager seeking for esoteric knowledge, and the Halveti had become the most ‘orthodox’ of dervish sects in the sense that they found wide acceptance among the Ottoman establishment. Sokullu Mehmed Pasha, indeed, had provided a lodge for his own Halveti spiritual adviser, attached to the mosque complex he built for his wife Esmahan in the Kadırga quarter of Istanbul.⁶⁵

Following his accession in 1574 Murad III continued Selim II’s aggressive policies in North Africa and the western Mediterranean. No immediate and radical shift in orientation was to be expected, indeed, while Sokullu Mehmed remained in his post as grand vezir. Events moved fast in these years. Ottoman military support enabled the Sa‘di ruler to be driven from power and in his place a disaffected member of the family was installed in Morocco as an Ottoman client. This victory gave the Ottomans control of the whole North African littoral, and brought them into contention with the Portuguese on the western frontiers of their empire as well as on the eastern. Even as a Spanish envoy arrived in Istanbul to seek peace with the Sultan, King Sebastian of Portugal sought help against the Ottomans from his cousin Philip of Spain; Philip equivocated, but finally provided both men and ships. In 1578 Portugal invaded Morocco; King Sebastian was killed at the battle of Alcazar and although the Ottoman client ruler was also killed, Ottoman assistance enabled his brother to succeed him. The long period of Ottoman–Habsburg warfare in the western Mediterranean came to an end in 1580 with a treaty which freed Spain to concentrate its attention northwards.66

Tripoli, Tunis and Algiers were by now all provinces under the nominal authority of Ottoman governors, but local leaders continued to put their own parochial interests first, and undermined any notions the Ottoman central government might have had of bringing the administration of these provinces in line with the bureaucratic norms elsewhere in the empire. The relationship of the Ottomans with these Magrebian provinces was a ‘marriage of convenience’ in which both parties had low expectations of the other: Istanbul anticipated little revenue from the Magreb but hoped for assistance against common enemies in the Mediterranean, while nominally Ottoman subjects in the Magreb by no means sought ‘Ottomanization’ or integration into the empire, and expected little in the way of central government investment and infrastructure.⁶⁷

The Ottoman–Habsburg peace of 1580 indicated a certain equilibrium in the balance of sea power between the two empires. At the same time active Ottoman involvement against the Portuguese in the Indian Ocean was gradually waning; their attempts in 1585 and 1589 to break the Portuguese hold on the coasts of Mozambique represented the last gasp.⁶⁸

The Ottomans were glad of the respite in the Mediterranean, for in 1578 they had become embroiled in a full-scale international crisis, war with Iran in the Caucasus. This war, which dominated Murad’s reign, inaugurated a long period of belligerency between the two powers which continued until a lasting peace was finally agreed in 1639. There had been peace on this frontier since the Treaty of Amasya in 1555 but the death of Shah Tahmasp in 1576 was followed by factional infighting and a revival of Kızılbaş activity. Sokullu Mehmed Pasha had argued strongly against reopening hostilities with Iran. He was known to favour an Ottoman presence in the Caucasus to check Muscovite expansion there but he was also alive to the logistical problems involved, and wary of the expense of campaigning in the region. Sokullu Mehmed had had enemies even before the new clique around Murad came to prominence, before the queen-mother Nurbanu Sultan and the harem attained their unprecedented power. His old rival Lala Mustafa Pasha revealed himself as ready to capitalize on the situation, and provided a figurehead around whom the courtiers who hated Sokullu Mehmed could coalesce. They calculated that if Lala Mustafa, the hero of Cyprus, were to be victorious in war again this would suggest that Sokullu Mehmed should be dismissed and Lala Mustafa elevated in his place. Following the decision to go to war against the Safavids Lala Mustafa was appointed commander jointly with another relentlessly ambitious rival from the days of the Yemen campaign, Koca Sinan Pasha. But their inability to work together soon led to Koca Sinan’s dismissal, leaving Lala Mustafa in sole command and poised to reap the rewards for the success he anticipated.69

As the Caucasus was the theatre of war, the eastern Anatolian frontier city of Erzurum was the forward base for the Ottoman campaign against Iran. Having sailed to Trabzon and then marched south over the mountains, Lala Mustafa and his army mobilized at Erzurum during the summer of 1578. The Safavids and their client states in the Caucasus were in such disarray that the Ottomans were able to advance through Georgia, occupying Tiflis (capital of modern Georgia) as they went, to reach the principalities to the north. By the end of the summer, several of the princes of the region had submitted to the Ottomans, who now occupied parts of the Shirvan area on the western shore of the Caspian. Özdemiroğlu Osman Pasha – son of the former governor of the province of Habeş Özdemir Pasha – was given the unenviable task of governing this distant and vulnerable new province. With Tatar assistance he routed local and Safavid resistance but his supply lines to the Ottoman occupiers of Tiflis were cut, and he was obliged to withdraw for the winter from Shemakha, the main city of Shirvan province, to the Caspian coastal fortress-city of Derbent.⁷⁰

Leaving his old enemy Koca Sinan Pasha behind in Istanbul proved fatal for Lala Mustafa Pasha. When Sokullu Mehmed Pasha was assassinated in 1579 the second vezir, Semiz (‘Fleshy’) Ahmed Pasha, husband of Mihrimah Sultan and Rüstem Pasha’s daughter Hümaşah Sultan, took his place as grand vezir. Moved up the hierarchy to become third vezir, Koca Sinan was well-placed to arrange matters to his own advantage. Semiz Ahmed recalled Lala Mustafa from the front and appointed Koca Sinan commander in his stead. Lala Mustafa’s protégés were accused (in some cases with justification) of corruption and dismissed from government service. He managed to retain the post of second vezir, however, and when Semiz Ahmed died after only a few months in office, it seemed that the grand vezirate would at last be his.71

But Lala Mustafa did not after all achieve the promotion he so much desired: although he undertook the duties of the grand vezir, Koca Sinan was able to prevent his confirmation in office, and after a three-month hiatus it was Koca Sinan who was appointed grand vezir in August 1580. Lala Mustafa died soon thereafter. The deaths of Sokullu Mehmed Pasha, Lala Mustafa Pasha and Semiz Ahmed Pasha marked the closing of an era, for they were the last links with the reign of Süleyman. Koca Sinan was of a younger generation, the one that had risen to maturity and power under Sultan Selim II. He was more easily able to accommodate himself to government by faction, and indeed did so with acumen, becoming grand vezir a total of five times.⁷²

Soon after Koca Sinan, who was still also commander-in-chief of the army, reached Erzurum in November 1580, the Safavids were suing for peace.⁷³ Assuming hostilities were at an end, Koca Sinan returned to Istanbul, when in fact continuing war in Georgia had prevented the conclusion of an Ottoman–Iranian peace; he was dismissed some months later and the second vezir, Siyavuş Pasha, appointed in his stead. The following years were spent securing Ottoman control in the Caucasus. In contrast to earlier campaigns against the Safavids on this frontier, this time the Ottomans attempted a permanent occupation of the region. Shirvan was only one of four new provinces created at this time.⁷⁴ The problems faced in the Caucasus were similar to those in Yemen, namely the volatility of local rulers and the inhospitability of climate and terrain. The expansion of the empire into these peripheral regions depended on the Ottomans establishing and holding fortresses, outside which they enjoyed little control. The fortress of Kars, today in north-eastern Turkey, was rebuilt as an advance base from which to supply the garrisons of the newly-conquered territory; Yerevan was also rebuilt and other smaller strongholds were secured.

Özdemiroğlu Osman Pasha remained in Derbent until reinforcements arrived from Rumeli by way of the Crimea in 1582, then marched overland to drive the Safavids from the eastern Caucasus. The Tatar cavalry of the Crimea were essential to his ability to maintain what was a tenuous grip on the region, but in defiance of his obligations as the sultan’s vassal Khan Mehmed Giray II now refused to provide sufficient troops to aid the Ottomans. Özdemiroğlu Osman therefore marched into the Crimea and, with the assistance of a fleet from Istanbul commanded by Kılıç Ali Pasha, installed a new khan. During the five years of his service on the Caucasian front Özdemiroğlu Osman was isolated from the intrigues at court; when he returned to Istanbul he was given a hero’s welcome and, to the anger of the clique around Murad who this time failed to influence the Sultan, was appointed grand vezir in the summer of 1584. He died a year later, having captured Tabriz in the course of a further campaign in the east – this time with significant Tatar help – and, for the first time ever, having held it for the Ottomans.75

Özdemiroğlu Osman Pasha’s conquest of the former Safavid capital of Tabriz in 1585 inaugurated a new phase in the war with Iran. So great was Ottoman confidence that statesmen for a while looked favourably on a proposal from the Khan of the Özbeks, whose territory of Transoxiana lay to the north-east of Iran across the river Oxus, that his armies should push north to recapture Astrakhan from Muscovy.⁷⁶ Another front against Iran was opened in the south as the new governor of Baghdad, Cigalazade Sinan Pasha – a scion of the Genoese family of Cicala who had been captured at sea as a boy and converted to Islam – went on the offensive. He took part of south-west Iran for the Ottomans, and carved out two new provinces.⁷⁷

The internal upheaval in Iran which followed Shah Tahmasp’s death eventually ended in 1587 through the determined efforts of his grandson who came to the throne in that year. During his long reign Shah ‘Abbas would go on to establish a reputation as the epitome of Safavid grandeur and greatness, but in 1588 and 1589 he was vulnerable: the Özbeks attacked Iran across the Oxus and captured the cities of Herat, Mashad and Nishapur. Shah ‘Abbas sued the Ottomans for peace, the terms of which forced him to settle for the status quo. It was an expensive peace for the Safavids, as the Ottomans now held much of the Caucasus and Kurdistan – albeit at great cost. For the Ottomans, it was a more decisive result against the Safavids than they had achieved at any time since Selim I’s victory at Çaldıran in 1514. The treaty established a border farther to the east and north than had been possible before, and it seemed that the century-old contest between the Safavid and Ottoman states had come to an end.

Meanwhile, Muscovy, having once established a foothold in the Caucasus, continued its expansion apace: this ever more confident state set out to colonize the region by insisting that the local chiefs – whether Christian, as the Georgians were, or Muslim – swear allegiance to the tsar; those who did not were threatened with the imminent arrival of an army. The urgency of the situation was expressed in dramatic language by the ruler of Dagestan to the Ottoman Sultan in 1589:

. . . the cities you took from Persia . . . will not be able to defend themselves; and the Russians will unite with the Persian shah and the Georgian king, and then they will march on Istanbul from here and the French and Spanish kings [will march in] from the other side and you, yourself, will not survive in Istanbul, and you will be captured and the Muslims will become Christians, and our faith will come to an end, if you do not intercede.78

This plea seems to have gone unheard but the spectre of encroachment into the Caucasus by the Muscovy of Ivan IV meant that the Ottomans could not ignore this region: like south-east Anatolia in the opening years of the century, it was one that attracted the strategic interest of a trio of outside powers.

With the confidence born of success on their eastern border, few Ottoman voices were raised against a renewal of conflict with the Habsburgs on the empire’s European land frontier. Although peace had officially been made in 1568 – and the treaty renewed in 1574 and 1583 – localized hostilities and skirmishes along the extensive Croatian–Bosnian border had persisted. The Habsburg authorities had repeatedly protested about raids by local Ottoman forces, and reorganized their border defences to protect their people, but were anxious to maintain the peace and assiduously paid the annual tribute or gift – depending on the point of view – agreed in 1568, so as to give the Ottomans no pretext for launching a formal campaign. In the years since their last encounters both Habsburgs and Ottomans had come to realize that neither could hope for a decisive victory.⁷⁹

In 1591 Hasan Pasha, governor of Bosnia, took a number of forts on the western, Croatian, section of the Ottoman–Habsburg frontier – on the face of it an independent action, but probably supported from Istanbul – and a new Ottoman fort was built on the Kulpa river at Petrinja. The Habsburgs, well aware of the neglected state of their border defences, saw this as a hostile act, but attempted by diplomatic means to avoid an escalation of the fighting. In 1593 Hasan Pasha crossed the Kulpa and besieged the fort of Sisak. Whoever held Sisak held the route along the Sava to Zagreb and into Austria; a hastily assembled relief force routed the Ottoman attackers, many of whom, including Hasan Pasha, were killed.⁸⁰

Here was the excuse for a full-scale campaign Koca Sinan Pasha, once more grand vezir, had been seeking: in July 1593 he set off westwards at the head of the army. From the rapidity with which the Ottomans were able to react to the failure of the siege of Sisak it is clear that, just as the Ottoman navy had recovered quickly after the losses incurred at Lepanto, so the Ottoman military machinery had swiftly rebuilt itself after the recent Iranian wars. With a major offensive in central Europe in prospect, all other issues were set aside, including a naval campaign against Spain for which the Protestant powers had been lobbying, and which in 1590–1 was already in preparation. To abandon this last in particular probably caused the Ottoman government few regrets, for their strategic goals in the western Mediterranean had been accomplished and their disengagement from that arena, symbolized by the 1580 truce with Spain, subsequently renewed. It was, in any case, clear both to the Venetian ambassador and to Koca Sinan himself that the Ottoman navy would have been unequal to another series of distant encounters at sea owing to the neglect of the fleet in recent years.81 Unlike naval campaigns, Koca Sinan Pasha observed, land-based warfare could be far more immediate: ‘One can launch a campaign on land by a mere command: everybody mounts his horse and sets off. A naval expedition is not like that . . . however great the material investment and human effort, it can only be realized in seven to eight months’.⁸² The war in central Europe thus begun in 1593 in such carefree haste lasted until 1606; neither side gained much from it, while the costs to both in financial terms and in damage to the fabric of the state were enormous.

By the end of the sixteenth century the character of warfare was changing in both east and west. In former times, particularly in Iran, the ability of the enemy to avoid a pitched battle by vanishing into the countryside had often thwarted Ottoman efforts to attain a decisive victory, but the recent campaigns on that front had demonstrated that a more static style of warfare was now becoming the norm, involving fortresses to be reduced by lengthy siege if territory was to be won. On the Habsburg–Ottoman border, after the truce of 1568, the Habsburgs and their partisans took refuge behind a line of strongholds intended to protect the hinterland from enemy incursions, and these had their counterparts on the Ottoman side of the frontier. In the central section of this front, on the Habsburg side stood Nagykanizsa, Győr (Raab), Komárom, Nové Zámky (Neuhäusel) and Eger; ranged against these, on the Ottoman side were Szigetvár, Székesfehérvár (Stuhlweissenberg), Buda and Esztergom, with a second line of fortresses in an arc from Belgrade to Timişoara.⁸³

The gains and losses enjoyed and suffered by Ottomans, Habsburgs and Hungarians alike over the thirteen years of this war well illustrated its wearing and inconclusive nature. The first two years achieved nothing. Early in 1595 Sultan Murad III died and was smoothly succeeded by his twenty-nine-year-old son. Mehmed III inherited a state in disarray and it was clear that a new strategy was urgently needed to raise the prestige of both sultan and empire. At a meeting convened by Grand Vezir Koca Sinan Pasha it was agreed that, in a radical departure from recent practice, the new sultan, inexperienced as he was, must lead his army in battle⁸⁴ – something no sultan had done since Süleyman’s final campaign in 1566. Koca Sinan died in April 1596; in June the Ottoman imperial army set out to join the forces which had been holding the line on the frontier. Their aim was to capture the fortress of Eger which lay on the route between Austria and Transylvania which, together with Moldavia and Wallachia, had sought Habsburg protection. Eger fell and the Ottoman army with its Tatar reinforcements then encountered the Transylvanians and the main Habsburg army in the nearby plain of Mezőkeresztes on 25 October. From the closely-contested field battle which ensued, the only one of the war, the Ottomans emerged victorious – at first it seemed that they had lost, and only the fierce attack on the Habsburg forces as they plundered the Ottoman camp saved the day. The Sultan, not relishing his role as commander-in-chief, had suggested to his new grand vezir, Damad İbrahim Pasha, that he should return to Istanbul; he had to be chivvied to stand fast.85 The English ambassador to the Ottoman court, Sir Edward Barton, was taken on the campaign by Mehmed to accompany the Habsburg ambassador to the Sultan and ensure that he and his suite were safely conveyed back to their own territory. Barton’s secretary Thomas Glover described the scene following a skirmish in which the Ottoman forces came off worst:

At which time I leave to the world, to consider what fright the Grand Signior was in, seeing all his Armie flie; yet incouraged by some about him of his chiefe Officers, caused his Banners Imperiall to march forwards upon the Christians; and he with his Bow and Arrowes shot thrice, and as some say, slue three Christians.⁸⁶

The following years saw peace talks mooted as border fortresses continued to change hands. Wallachia again opted for Ottoman vassaldom. In 1600 the Ottomans broke through the Habsburg defensive line to the south with the conquest of the strategically important fortress of Nagykanizsa; Vienna feared another siege,⁸⁷ and the following year the Habsburgs tried and failed to retake Nagykanizsa. Pest was lost to the Ottomans and then retaken during the last desultory years of the war, and Transylvania again aligned with the Ottomans. The final year of fighting, 1605, saw Esztergom (captured by the Habsburgs in 1595) once more in Ottoman hands. By now all the protagonists were exhausted and eager for peace. The next year this was agreed – at the border village of Zsitvatorok rather than in Istanbul as was usual. This in itself denoted a concession by the Sultan, whose predecessors had customarily issued peace terms to their vanquished foes. Among a variety of provisions, each side was to retain the territory it then held – giving the Ottomans the meagre reward of only two new strongholds, Eger and Nagykanizsa – and Emperor Rudolf won the further advantage that he and his successors would henceforth be treated as equals by the sultan. The ‘tribute’ paid by emperor to sultan would cease, after the payment of a one-off sum of 200,000 florins.88

The military response on the Habsburg side had been bedevilled by the Counter-Reformation, which had cost them the support of potential Protestant allies; while the Ottomans, after 1603, were campaigning on two fronts, as Shah ‘Abbas of Iran attempted to regain the territory he had signed away in 1590. Within the empire, moreover, the multifarious domestic troubles which had been building up since the 1580s had come to a head: 1599 saw the first of a number of military campaigns directed against an alarming wave of rebellion in Anatolia and there was unrest in many other parts of the empire. That Sultan Mehmed III died in 1603 and was succeeded by his thirteen-year-old son Ahmed seemed of little consequence.

The last quarter of the sixteenth century was as difficult for European states as for the Ottomans. The warfare in which all were frequently engaged imposed budgetary strains from which each state sought its own relief, and the economies of Spain, France, England and Austria were all subject to severe disruption and over-extension; annual expenditures exceeded revenues, and innovative measures had to be sought to handle the crises. Inevitably these were accompanied by social and political upheaval – from which the Ottomans could not hope to be immune, although in their case it took a peculiarly Ottoman course.

The precise causes of the disruptions within the Ottoman Empire in the later sixteenth century are still poorly understood, and a clear hierarchy of causes and effects has yet to be distinguished. Until the sixteenth century, Ottoman economic and demographic expansion was underpinned by the revenues of newly-conquered provinces, but as conquest slowed, less coin was available to oil the wheels of this quite highly-monetarized economy. In the winter of 1585–6, in an attempt to raise more cash during the Iranian war, the Ottoman government debased the silver asper, the akçe, almost halving its silver content; in an economy where the amount of silver – or gold – mixed with alloy in a coin determined its value, this measure caused great financial instability.⁸⁹

The Ottoman economy had always been open to external influences. Since early in the sixteenth century, specie – coin – from the mines of the Americas had been making its way eastwards in the course of commercial transactions, driving indigenous coins of low silver content from circulation. Following the debasement of 1585–6 those on fixed salaries paid in aspers – such as members of the bureaucracy or the military – suffered greatly, for their money now bought only half as many goods as before. Such a drastic inflation of prices exacerbated social discontent to the point where the authorities found it impossible to force people to accept payment of their salaries in the debased coinage. Because taxes were paid largely in aspers, treasury revenues were also halved in real terms. The government tried to close the gap between state revenue and state expenditure by levying new taxes on the peasant population, and the system of tax-farming – whereby an individual or partnership paid the state in advance a sum equivalent to the tax revenue from a given source, then made the actual collection themselves (with an extra element of profit built in) – was extended. The treasury also borrowed from the personal fortunes of wealthy members of the establishment. This internal borrowing indicated a very different approach to financial management from that of the European powers – the Ottoman Empire did not raise foreign loans until the nineteenth century.* By contrast, had the Habsburgs not been supported by the financial resources of their Catholic allies, the war of 1593–1606 might have had a different outcome: the German princes within the Holy Roman Empire contributed more in 1594 to the war against the Ottomans than they had paid towards all Charles V’s anti-Ottoman campaigns put together.⁹⁰

By 1589 the adverse effects of the debasement of the asper produced a janissary revolt like that which had erupted in response to Mehmed II’s manipulation of the currency during his first sultanate in the 1440s. The janissaries, who had the support of the Sheikhulislam, blamed the governor of Rumeli and the director of the treasury for paying their salaries in debased coin. Sultan Murad III sacrificed these officials to the mutineers – the first of many such episodes in which a terrified sultan found himself utterly at the mercy of competing factions close to him. (Neither Selim II nor Murad III seems to have shared Süleyman’s sense of the portentous nature of the Islamic millennium; contemporary intellectuals, on the other hand, viewed this violent episode as emblematic of the breakdown of authority which would precede it.) Continuing financial instability in the last years of the century was blamed by the palace cavalry on Sultan Mehmed’s mother Safiye Sultan’s lady-in-waiting Esperanza Malchi, a woman who conducted the queen-mother’s transactions with the outside world, and in 1600 she was murdered by some of their number.⁹¹ Many leading statesmen of the seventeenth century became scapegoats for the continuing crisis at the heart of the state, and no sultan could afford to be at odds with the military. In a further complication, the two main corps of his elite troops, the janissary infantry and the palace cavalry, more often than not supported rival cliques. In 1582 a fight between these two groups had left several dead and brought a dramatic end to the lavish public celebrations for the circumcision of Mehmed III, then still Prince Mehmed, being held in the Hippodrome.92

The financial and social distress that accompanied the debasement of the coinage came at a time when the costs of warfare had reached unprecedented levels as methods of fighting changed. The provincial cavalry forces, supported by agricultural taxes paid in lieu of an obligation to appear on campaign, were less effective in an age of defensive, siege-based warfare, and as the borders of the empire expanded provincial cavalrymen lost their enthusiasm for campaigning: scarcely had they recovered from the exhausting Iranian war of 1578–90 when their services were required first on the Habsburg frontier, and then again on the Iranian. As Mustafa Efendi of Selanik wrote in 1597, they had not seen peace for twenty years.⁹³ Infantry – which in the Ottoman context meant primarily the musket-bearing janissaries – were more useful than cavalry in modern warfare, and their numbers grew accordingly, from almost 8,000 in 1527 to 13,500 by the time of Sultan Selim II’s death in 1574 and to almost 40,000 by 1609.⁹⁴ Like other salaried state employees (whose numbers also grew inexorably) they had to be paid in cash – and on time if trouble was to be averted.

The government was in a dilemma. The rapid increase in the numbers of salaried troops could not continue indefinitely, and other sources of manpower were sought. A solution which appealed because it was cheap was to enlist men from the peasantry – the prime requirement now being Muslims who could wield a musket – to serve for the term of a campaign, after which they would be demobilized. This innovation openly flouted the fiction that peasants were disbarred from serving in the combat army, being employed only to perform various auxiliary tasks alongside the sultan’s elite fighting troops and the provincial cavalry. It soon became clear, however, that even if they had not been troublemakers prior to enlistment, men already disaffected by over-taxation and their inability to make ends meet became a major disruptive element after demobilization. They retained their guns, and did not return to their previous occupation. The allegiance of these men could be bought by anyone who could pay them, be he bandit leader or fractious servant of the state. In an era of poor communications, local ties were usually stronger than loyalty to the government and its officers in far-away Istanbul, and the population’s experience of the demands of central government for taxes and manpower hardly more favourable than that at the hands of the so-called rebels among whom they lived. It was primarily in Anatolia that troops were raised from the peasantry, and it was Anatolia that suffered the most violent effects of the ensuing brigandage and outright rebellion. Having discovered provincial cavalry to be unsuited to the new style of warfare, and seeking as ever to raise money, the government ordered many of the cavalry not to present themselves on campaign but to pay a tax instead; they too became available to join in the unrest.

During its first centuries the Ottoman state had seen many challenges to its power – the resistance of the Anatolian principalities to Ottoman territorial control, the succession struggles of Ottoman princes, the rebellions of the Kızılbaş against Sunni orthodoxy, the outspoken stances taken by individual clerics or preachers, the king-making revolts of the elite army in Istanbul. The opening years of the second Islamic millennium were ones of deep crisis as uprisings against Ottoman rule were witnessed across the empire.

At the very time when the Ottoman army was engaged in fighting in both west and east, the Kurdish Canbulad clan, hereditary rulers in northern Syria, seized the opportunity to assert their independence. To declare his sovereignty Canbuladoğlu Ali Pasha, appointed Ottoman governor of Aleppo in 1606, had the Friday prayer read in his own name – and probably also minted his own coinage. Canbuladoğlu Ali had the support of anti-government rebels in Anatolia and was also encouraged by the Duke of Tuscany, Ferdinand I, who appreciated the commercial importance of Aleppo as an outlet for Iranian silk and other goods bound for European markets, and hoped to share in Canbuladoğlu Ali’s good fortune. Grand Vezir Kuyucu (‘Well-digger’) Murad Pasha was sent against Canbuladoğlu Ali with a large army in 1607; ignoring his protestations that he was a loyal servant of the Sultan, Kuyucu Murad continued southwards to meet the rebellious governor in pitched battle. Canbuladoğlu Ali managed to escape with his life and was subsequently pardoned and appointed governor of the distant province of Temesvar in Hungary. Kuyucu Murad Pasha was finally able to take his revenge in 1610, when he had Canbuladoğlu Ali executed in Belgrade.95

A one-time ally of Canbuladoğlu Ali, the Druze chief Fakhr al-Din Ma‘n, controlled most of the territory covered by modern Lebanon and northern Israel and was similarly ambitious. In 1608 Fakhr al-Din concluded a treaty with the Duke of Tuscany, and in 1611 Kuyucu Murad’s successor as grand vezir, Nasuh Pasha, mobilized an army to curb his growing strength. Fakhr al-Din fled to Tuscany in 1613 but returned five years later, by which time his brother Yunus had reached an accommodation with his Ottoman masters. Fakhr al-Din’s continuing career of defiance of the government in Istanbul and expansion of the territories under his control ended only with his execution in 1635.⁹⁶

The Ottoman central government also had to deal with many lesser revolts against its authority. Egypt was frequently the scene of mutiny: that of 1589 against the governor Kara Üveys Pasha was followed in 1598 by an attack on the governor at that time, Serif Mehmed Pasha; in 1601 troops broke into Serif Mehmed’s council chamber and murdered several officials. In 1604 Hacı İbrahim Pasha was murdered – the first governor of Egypt to die in this way. It took the despatch from Istanbul of the third vezir, Hadım (‘Eunuch’) Mehmed Pasha, to bring an end to the military uprisings of these years, for which he earned himself the soubriquet ‘Hammer of the Troops’.97 In the European provinces, famine was the pretext for the murder of the governor of Buda in 1590,⁹⁸ while in North Africa, Tripoli and Tunis were in revolt: Tripoli was almost lost to the Ottomans before Egyptian troops were sent to restore order in 1592,⁹⁹ and the governor of Tunis was murdered by a mob around a man who proclaimed himself the Mahdi.¹⁰⁰

But these events, although sufficiently alarming for the central authorities at a time when resources in money and manpower were stretched to the limit, were overshadowed by the rebellions which convulsed Anatolia during these years. They were known as the Celali rebellions, after Sheikh Celal who, as we have seen, had led a revolt in the early sixteenth century. Official documents contain vivid descriptions of a Celali raid:

. . . several hundred horsemen and musketmen including [twenty-four names] like brigands came to the province, pillaged the goods of the poor, burnt their houses, killed more than 200 men, ran off with the young boys and virgin girls, and stole more than 50,000 sheep, goats, horses and good camels, and took stores of barley, wheat, oil, honey and other commodities: then they captured more than 300 men, torturing them day and night.¹⁰¹

Over the course of the sixteenth century the Safavid state had lost the millennarian fervour of Shah Isma‘il’s reign, to reach, under Shah ‘Abbas, a modus vivendi with the Ottomans. As religious rivalry with the Safavids waned, those who were dissatisfied with life in the crisis-ridden Ottoman state rebelled against its authority in a different way, unencumbered by religious rhetoric. Although the term ‘Celali’ had originated in the context of a religiously-inspired protest movement, in the course of the sixteenth century it came to be used by Ottoman officials to describe a wide range of rebels against the state¹⁰² – even those with no obvious religious motivation for their discontent. In the opinion of one modern historian, ‘The characteristic tone of troublemakers before and during the Celali risings seems less an appeal to the higher values of society than a brash and cynical contempt for them’,¹⁰³ and scholars today tend to see the Celali rebellions of the turn of the sixteenth/seventeenth centuries as secular revolts of the disaffected from many walks of life – brigands, students of religion, provincial governors, demobilized soldiers, deserters, landless peasants. However, while it is true that, unlike many earlier Anatolian opposition movements, the Celali rebellions did not present their cause in religious terms, the fact that there were no similar uprisings in the Balkan provinces of the empire suggests, at the least, that they were more than simple armed rebellions with no link to the earlier history of religious and political opposition in Anatolia.

Problems arising from religion did not of course disappear: individuals were still charged with heresy – with having Kızılbaş sympathies, implying that the accused was Shia and owed allegiance to the shah rather than the sultan. A Kızılbaş purge was ordered before the outbreak of war with Iran in 1578 when the governor of the province of Baghdad – where there were many Safavid sympathizers – reported that there was ‘no end to the heretics and misbelievers’.104 The government in Istanbul feared fifth-column activities in this heavily Shia area. The only serious Kızılbaş uprising during the second half of the sixteenth century also occurred in 1578, when a man claiming to be Shah Isma‘il appeared among the Turcomans of southeastern Anatolia.¹⁰⁵

Among the first of the ‘new-style’ rebels to attract the attention of the Ottoman authorities was one Karayazıcı (‘Black Scribe’) Abdülhalim, employed alternately in the state’s salaried forces and in the retinue of a provincial sub-governor. Losing his job when his master was dismissed, he joined a militia band of which he soon became the leader and his name became linked with a series of upheavals which spread insecurity across Anatolia while most military men were campaigning in Hungary, and determined many people – both Christian and Muslim – to migrate to Istanbul. Karayazıcı issued orders as though he were the sultan and made appointments, including designating one of his followers his grand vezir, and organized his forces as though they were those of the central state. Like other leaders seeking to enhance their standing, he claimed that he was descended from the shahs and that the Prophet Muhammad, appearing to him in a dream, had bestowed upon him the right to rule. The governor of Karaman province, ordered in 1599 to put down Karayazıcı’s rebellion, joined him instead. An army was then sent from Istanbul under the command of Sinanpaşazade Mehmed Pasha, son of the late grand vezir, the formidable Koca Sinan Pasha. The rebels took refuge in the city of Urfa (Edessa) in south-east Anatolia. After besieging Urfa for some two months, Sinanpaşazade Mehmed lifted the siege having cut a deal with Karayazıcı to deliver up his unfortunate ally the governor of Karaman, who was taken back to a painful end in Istanbul. In spring 1600 Sinanpaşazade Mehmed pursued Karayazıcı across Anatolia to the city of Amasya; the rebel had, so it seems, been bought off with the governorship of the sub-province of the same name.106

Karayazıcı was soon moved to a new post at Çorum, and in 1601 an army under the command of Sokullu Mehmed Pasha’s son Sokulluzade Hasan Pasha beat Karayazıcı’s army in battle at Elbistan, south-east of Kayseri, inflicting a great blow on Celali morale. Karayazıcı died shortly afterwards and leadership of the rebels fell to his brother, Deli (‘Crazy’) Hasan. In spring 1602 a Celali army attacked towns in north-central Anatolia and besieged Sokulluzade Hasan in Tokat. From his base at Aleppo the Venetian consul in Syria, Vincenzo Dandolo, reported that Sokulluzade Hasan lost five million pieces of gold, his baggage train and his harem to the Celalis; he also lost his life. Deli Hasan and his men moved on to besiege Ankara and other towns: it was the eastern provinces that were worst affected by Celali depredations. Ottoman subjects of all classes supported them, however, as did the brother of the Khan of the Crimea, hoping for Celali backing to win the khanate for himself. In another desperate attempt to bribe their way to a solution, the government awarded Deli Hasan the rank of pasha and the post of governor of far-away Bosnia.¹⁰⁷

The policy of co-opting rebels into the ruling class of the empire as provincial governors and military commanders was an imperfect way of neutralizing their energies, as a glance back at the history of the fourteenth century might have reminded the government. The Ottoman establishment seemed to have forgotten how Cüneyd, the dispossessed prince of Aydın, awarded the governorship of Nikopol on the Danube, had then taken an active and disruptive part in the civil war among the sons of Bayezid I. Deli Hasan proved no more loyal a servant despite government hopes: after a couple of seasons spent campaigning against the Habsburgs he was suspected of treasonous dealings with them and executed.¹⁰⁸ The historian İbrahim of Peç (Pécs), who was on the Hungarian front at the time, had nothing but criticism for the conduct of Deli Hasan’s troops on campaign, in particular their insubordination: ordered to help with the building of earthworks in the course of an engagement during the last years of the Hungarian wars, they retorted, ‘We have fought for years in Anatolia and nowhere did we dig ditches or construct palisades,^* and we do not intend to do so now’.¹⁰⁹

The removal of Deli Hasan from Anatolia changed nothing. The Ottoman government was unable to deal with the Celali rebels because sufficient forces could not be spared to put down the unrest decisively: to wage war against powerful foreign adversaries on two distant fronts took all the strategic inventiveness the state could command. The prevailing insecurity disrupted trade routes, while the passage of armies destroyed agricultural land so that tax-payers could no longer pay the dues on which the provincial cavalrymen depended to equip themselves and their retinue for participation in military campaigns. The Celali depredations in Anatolia caused so many thousands to leave the land that this exodus came to be known as ‘the Great Flight’. The rich went to Istanbul, the less well-off to the relative security of the walled towns of Anatolia. Villages and agricultural lands were deserted, as drought affected much of Anatolia from 1603 and winters were unusually hard. Prices soared.110

Hostile military activity against the Ottoman state came from yet other quarters. The Cossacks of the northern Black Sea steppe could tie up the regular army for months at a time and, as with the Celalis, their predations – at first localized – took on ever more alarming dimensions. The Cossacks first appear in the historical record in the fourteenth century, as roaming bandits or adventurers inhabiting the ‘vacuum’ of the steppe, where central government authority was weak, and they could pursue a way of life at odds with that of the settled population – fishing and hunting, raiding the merchants plying their trade between the Black Sea and the cities to the north, and alternately fighting or banding together for joint exploits on the steppe with the Tatars of the Crimea against Poland-Lithuania and Muscovy.

With the appearance of the Ottomans on the northern Black Sea coast in the later fifteenth century, and the ensuing symbiosis between this leviathan and the Crimean Tatars, the politics of the steppe were transformed and the Polish and Ukrainian nobles of the borderlands (‘ukraina’ means ‘borderland’) – who themselves barely acknowledged central government – recruited Cossack warriors to defend their estates against Tatar raids. Soon after 1538, when Sultan Süleyman led a successful campaign against his insubordinate vassal Moldavia, he designated the northern Black Sea littoral from the Dniester to the Boh a new province. Subsequent Cossack attacks on strongholds, and on herdsmen and travellers, caused the Ottomans mounting concern, as thousands of captives as well as livestock, arms and all sorts of chattels were seized, in a region where the sultan’s writ was intended to bring security.¹¹¹

In the 1550s and early 1560s the Ukrainian prince Dmytro Vyshnevetsky led the Cossacks. He organized them against the Tatars and oversaw the construction of a fortress on one of the islands below the Dnieper rapids – some 375 kilometres upstream from the river mouth – which became their headquarters in the region. The creation of this administrative centre was the first step in the formation of a Cossack corporate identity. Vyshnevetsky offered his services to Muscovy and attacked both Moldavia and the new Ottoman province; in 1556 he attacked Cankerman – the fortress capital of the province – but left after causing much destruction to the town and its environs, and in 1563 he was captured and executed in Istanbul. These raids were considered by the Ottomans to breach the official state of peace between Poland-Lithuania and the Sultan: even before the creation of the Polish-Lithuanian Commonwealth in 1569, when the relations of the Polish king with the Cossacks as a defence force began to be regularized, complaints about Cossack incursions into zones the Ottomans considered within their sphere of influence were addressed by the Sultan to the Cossacks’ nominal overlord, the Polish king.

The Cossacks gradually widened the scope of their ravages, no longer confining their attacks to the steppe. The near absolute security of the Ottoman mare nostrum that was the Black Sea, and of its shores, was rudely disrupted from the last quarter of the sixteenth century as the Cossacks of the Dnieper region set sail in their highly manoeuvrable longboats to attack settlements on the Rumelian coast, and reached the mouth of the Bosporus. In 1614 they appeared on the north Anatolian coast, raiding the port of Sinop and causing, by all accounts, a great deal of destruction. In the words of the contemporary intellectual Katib Çelebi:

With the guidance of renegades who fled from the land of Islam, [the Cossacks] came to the fortress of Sinop on the Anatolian shore and entered that old castle by surprise and caused much damage . . . they took [with them] the goods and families that they had plundered and set out to sea.112

There was little the Ottomans could do against such a stealthy and swift foe.

Sultan Ahmed I came to the throne on the death of his father Mehmed III in 1603. Two years later he was persuaded by Nasuh Pasha, in charge of government efforts to suppress the Celali revolts, that only the Sultan’s presence at the head of the imperial army would convince the rebels to give up their struggle – his father had, after all, been on the battlefield at Mezőkeresztes in Hungary in 1596, when the Ottomans had triumphed. But Ahmed’s campaigning days ended almost as soon as they began. When he reached Bursa in November 1605 he became ill from drinking the water coming down off Uludağ mountain; as the English ambassador of the time, Henry Lello, expressed it,

. . . the Emperor him self coming into that ayre wch he ws not accostomed unto and dringing of that water wch cam from the snowy hills hurte his stomache that he becam sicke wthall and wished h. self at home againe for there was noe meanes to go forwards.113

A Venetian report also remarked on the young Sultan’s sensitivity:

Ahmed was happier in his garden than in Anatolia where wolves prowled, where people ate grass and stinking corpses of fallen horses and camels in their hunger, and asked for alms from those riding by.¹¹⁴

The Sultan’s uncomfortable experiences inclining him to favour conciliation rather than confrontation, he allowed some of the most prominent Celali leaders and their followers, who were camped near Bursa at this time, to enter the ranks of the Ottoman army.¹¹⁵ However, once the war in Hungary ended the following year the Ottoman government could at last turn its attention away from the western frontiers of the empire and concentrate all its vigour on the problems in the east – the Anatolian rebellions, and the continuing war in Iran.

Another prominent Celali leader was a man named Kalenderoğlu Mehmed, whose first significant rebellion was in western Anatolia in 1605. In 1607, when Grand Vezir Kuyucu Murad Pasha marched across Anatolia to suppress Canbuladoğlu Ali Pasha’s rebellion in Syria, he offered Kalenderoğlu the governorship of the sub-province of Ankara. Kalenderoğlu accepted, but the gates of the city of Ankara were barred against him. Frustrated by an unsuccessful two-week siege, he headed westwards again to attack Bursa: all but the inner fortress fell to the rebels. News of Kuyucu Murad’s victory over Canbuladoğlu Ali failed to calm the panic in Istanbul as the people of the city anticipated the imminent arrival of Kalenderoğlu.¹¹⁶ Grigor of Kemah, an Armenian priest who followed events closely, wrote of the fear that Istanbul might be set alight by Celalis who managed to enter the city undetected. The Sultan ordered that all suspicious-looking individuals be apprehended, and arrested if no one was willing to vouch for them.¹¹⁷ Manpower to resist the Celalis was sought wherever it could be found – including, as the French ambassador of the time, M. Salignac, reported, among the citizens of Istanbul: but they, he opined, would ‘die of fright before they fight’.¹¹⁸

After spending some time in the area to the west of Bursa, Kalenderoğlu and his men moved south to west-central Anatolia and, in summer 1608, turned east to resume their brigandage. Kuyucu Murad Pasha, on his journey home from Aleppo after his defeat of Canbuladoğlu Ali, hoped to catch the rebels in central Anatolia, between his own army returning from the south-east and an army marching from Istanbul to meet him, as renewed efforts to bribe the Celali leaders with promises of state appointments fell on deaf ears. Despite the inherent logistical problems, Kuyucu Murad’s skill and experience as a military commander and his ability to hold the loyalty of his army finally enabled him to defeat Kalenderoğlu’s forces in battle at a pass deep in the Taurus mountains, north-east of Adana, on 5 August 1608.119

Kalenderoğlu’s troops fled north-east with government forces in pursuit. Near Şebinkarahisar, north-east of Sivas, it seemed that the rebels might be annihilated but they escaped to confront their pursuers again east of Bayburt, where another battle ensued. The remnants of Kalenderoğlu’s Celali army, numbering some 10,000 ‘musketeers and fully-armed cavalry’, plus their servants and grooms, reached safety in Iranian territory in late autumn 1608. Any anxiety initially felt by their hosts at accepting these unruly fighters proved to be unfounded. Eskandar Monshi, one of the chief secretaries of the Safavid court, was a witness of their reception by the Shah’s envoy and of the arrival of five hundred of their number at the Safavid capital of Isfahan. By his account they were met with feasts and celebrations wherever they went,¹²⁰ the implication being that this was a humiliation for the Ottoman Sultan.

Iran and the Ottomans were still at war, but in spring 1609 peace was mooted – although the presence of Kalenderoğlu and the other Celalis in Iran bedevilled the relationship between Sultan and Shah. Determined to eradicate any remaining rebels from Anatolia, in the campaigning season of 1609 Kuyucu Murad Pasha sent his commanders against them rather than launching a full-scale attack on Iran, and the last Celali leaders still in Anatolia were killed, with their followers. Kalenderoğlu died in May 1610 and those of his men who had followed him to Iran returned to Anatolia under the protection of Nasuh Pasha, now governor of Diyarbakır, who forged those experienced fighters into an elite brigade of musketmen.¹²¹

Kuyucu Murad Pasha thus achieved a victory that had eluded the Ottomans for years, and received a hero’s welcome on his return to Istanbul. The many refugees who had fled the countryside for the safety of the capital, and those who had continued on into Thrace, were given three months to return to their homes. To reassure his Christian subjects, the Sultan gave orders for the repair of the churches and monasteries destroyed by the Celalis, and taxes were to be waived for three years. Grigor of Kemah, one of those affected by this order, described the perilous return journey of the displaced to their homes in north-central Anatolia, where they faced an uncertain future:

. . . we set out on the road like a flock of sheep without a shepherd . . . we were more than 7,000 people – Armenians and Turks – and there was not enough food or animal fodder to be found along the way . . . we reached Tosya without mishap but were here obliged to camp out of fear at the signs of Celali activity. [A Celali chief] with many soldiers was camped in the Hacı-Hamza plain and all the city and castle gates were closed. Later we moved off all at once, women, children and the less able men walking by an upper road and the rest of us, equipped with bows and arrows, following the main road. The Celalis were afraid when they saw how many we were and hid in their tents; some stood at their tent-door and greeted us. Thus we continued on our way without mishap to Merzifon and on to Niksar.122

The renewal of war with Iran in 1603 was devastating for a state whose resources were stretched to the limit; in particular, the Ottomans had no manpower available to defend territory recently won in the east. Shah ‘Abbas was able to take advantage of the Ottomans’ over-extension because since 1590 he had transformed his army by recruiting an elite corps of men answerable only to himself, in effect a military slave caste analogous to Ottoman practice. His aim was to reduce his dependence on the fickle loyalties of the tribal levies who had underpinned the Safavid rise a century before, and had also been responsible for much of the disorder that had attended his own succession.¹²³ Like Shah Isma‘il before him, during the years of peace ‘Abbas had also engaged in energetic but abortive efforts to gain diplomatic and financial support from the West.¹²⁴ A number of border incidents provoked the resumption of hostilities, and in September 1603, in twelve days, Shah ‘Abbas marched from Isfahan to Tabriz at the head of his army. He found the city empty of its Ottoman garrison. Tabriz retaken, the Safavid armies then won back Nakhichivan, and took Yerevan after a six-month siege.

Sorely-pressed though they were on their western frontier at this time, it was clear to the government in Istanbul that they could not afford to leave the task of confronting Shah ‘Abbas solely to the military forces of their eastern provinces, and in 1604 the vezir Cigalazade Sinan Pasha was appointed to head an army sent from Istanbul. He found the frontier zone depopulated and devoid of provisions, the result of the same scorched-earth tactics which had hindered Ottoman progress in the past. Reverting to the ways of his forebears, the Shah kept just ahead of his pursuers. Cigalazade Sinan and his forces, wintering in Diyarbakır and Van, came under Safavid attack and were forced to retreat to Erzurum. In May 1605 the two armies met near Tabriz. Uncharacteristically, the Ottoman army deserted the field, leaving behind most of its equipment and provisions.¹²⁵

As the Ottomans concentrated on their domestic problems in Anatolia during the following years, the Safavids chased the Ottoman garrisons from their remaining strongholds in the Caucasus and Azerbaijan. The fighting was over without the Ottomans having been able to put up any defence. The grand vezir Kuyucu Murad Pasha set out against Shah ‘Abbas in 1610 but failed to engage him in battle and died in Diyarbakır in August 1611. A truce was agreed in the following year by the terms of which the Ottoman–Safavid frontier would revert to that agreed at Amasya in 1555 – the Ottomans were to lose all their gains from the wars of 1578–90.126

Before the treaty could be ratified, however, two Georgian princes sought Ottoman protection, prompting a strike by Shah ‘Abbas which the Ottomans interpreted as a breach of the truce. The Shah, furthermore, had detained the Ottoman envoy at his court.¹²⁷ Nasuh Pasha, who had favoured peace with Iran, was executed in 1614 and replaced as grand vezir by a more aggressive man, Öküz (‘Ox’) Mehmed Pasha, who was betrothed to Sultan Ahmed’s daughter. His appointment presaged a change of policy, and in August 1616 he arrived before the fortress of Yerevan at the head of a large army. His siege failed, however, and he in turn was replaced. After continuing skirmishes the Ottoman forces fell into an ambush near Tabriz on 10 September 1618 and lost, according to an informant of Eskandar Monshi, Shah ‘Abbas’s secretary, fifteen thousand fighters.¹²⁸ The war was over – at least for the time being – and the peace which Nasuh Pasha had worked years earlier to achieve at last became a reality.

These years – and the ones that succeeded them – were truly a ‘time of troubles’ for the Ottoman Empire. The work of contemporary men of letters reflected their anxiety about the manifestations of crisis they observed, which seemed to defy resolution, and predicted the imminent collapse of all that had been built during the previous three centuries. Mustafa Ali of Gelibolu was one of those who expounded his analysis of the changes which had taken place in the Ottoman polity by the turn of the millennium. Naming the four main Islamic states then in existence – the regional empires of the Ottomans, the Safavids, the Mughals and the Özbeks, all of which had been forged since the Mongol invasions of the thirteenth century and had Turkish and steppe-nomadic roots – he noted that unlike the Mughals who claimed descent from the Timurids and the Özbeks who claimed descent from Genghis Khan, the Ottomans lacked a claim to legitimacy based on descent; nor did they have a religious ideology that could be traced back to the Prophet, as the Safavids did. Indeed, he observed, their earlier claims to legitimacy based on ideologies of their descent from the Oğuz clan of Central Asia, or as inheritors of the Seljuks, or as the only true warriors for Islam, were irrelevant. What had given the Ottomans their unquestioned legitimacy, he suggested, was first and foremost the more tangible attribute of a dynastic order committed to universal justice dispensed by a strong central authority; this, however, had been vitiated since the time of Selim II, or even earlier, when Süleyman allowed his favourites a hand in the affairs of state.129

As well as producing pessimistic analysis the Islamic millennium engendered a yearning for the well-ordered world which Ottoman intellectuals imagined had existed in the past: Sultan Süleyman’s attempt to create the illusion of a just polity became a victim of its own success. Territorial victories were now less easy to win and the sultan had retired from active involvement in the business of government: the vicious struggle for a share in the rewards accruing to those in power must have been profoundly shocking to men raised under the old order. The Ottoman Empire was not alone, however – government by favourites became matter for criticism in Spain and France after the death of Philip II in 1598 and of Henry IV in 1610. A true king, as the author of a contemporary treatise wrote reprovingly of Philip II’s son, Philip III, ‘should not be content with simply having supreme power . . . and then merely sleeping and relaxing; but should be the first in government, in council, and in all the offices of state’.¹³⁰

The image of the warrior-sultan at the helm of an ever-expanding empire was becoming harder to sustain. Mehmed III was the last sultan who as a prince had governed a province as preparation for the sultanate. The theatres of war in which the Ottomans now found themselves involved were so far from Istanbul that a sultan leading his army in person might be gone for months, if not years. As long sieges came to predominate over swift field battles as the typical form of imperial military engagement, moreover, victory became even less predictable: the blow to the prestige of the dynasty in the case of a failure was more easily deflected if blame could be laid at the door of an expendable officer of state, rather than the sultan. At the same time, some grand vezirs were equally keen to avoid the invidious position of being held to account for military defeat, and equally fearful of losing office while on campaign. In consequence, the task of commanding the imperial army in battle gradually devolved on lesser vezirs, or on military commanders appointed for the term of a particular campaign.¹³¹ Initially, contemporaries who saw the sultan first and foremost as a warrior viewed this break with earlier practice as a dangerous innovation, but by the early seventeenth century perceptions of the sultan’s role had shifted to accommodate the new reality, and it was regarded as prudent for him to remain in the capital.¹³²

Not all commentators moved with the times, however, and some who saw only the undesirable consequences of the sultan’s changed role expressed the opinion that he had become a marginal figure.133 Advice manuals written by Ottoman men of letters – called ‘mirrors for princes’ – had a long tradition stretching back to the late fourteenth century. Like the woes they catalogued, the blessings they looked for hardly varied: a strong, just ruler; a balance between the sultan’s salaried standing army of cavalry and infantry and the provincial cavalry; peace and security for the producing class so that they could afford to pay the taxes on which the smooth functioning of the state depended; and adherence to the structure of society laid out in the sultanic law as elaborated during the sixteenth century.¹³⁴ Recent events had revealed all too clearly the extent to which these ideal norms had been violated – if they had ever been even approximated. The sultan was no longer seen to be in control of state affairs; the manpower demands of the turn-of-the-century wars had necessitated the recruitment of men of the producing classes – indeed, of any who ‘had a horse and could equip themselves’¹³⁵ – and these had infiltrated the sultan’s elite forces, once staffed almost entirely by means of the youth-levy; and the need to pay the militia who functioned as the musketmen that were the backbone of the modern army imposed budgetary demands that severely strained the state’s coffers. The various ‘mirrors for princes’ dwelt at length on the need to prevent the entry of ‘unauthorized’ outsiders into the tax-exempt class, and spelt out who should be regarded as ineligible for the sultan’s service: Muslim Turks, nomads, Armenians, Jews, Kurds, Gypsies, and the ethnic groups of the Black Sea.¹³⁶ The compilers of these tracts could not accept that the gap between their ideal but imaginary state and the new reality was unbridgeable, that the traditional system of clearly stratified ‘estates’ they lamented was over – if it had indeed ever existed.

An event concerning the construction of an observatory in Istanbul illustrates the flimsiness of an individual’s grip on power at this time. Sokullu Mehmed Pasha was the enterprising and persuasive grand vezir during whose tenure such projects as the Don–Volga and Red Sea canals were proposed; in 1574 he convinced Murad III, the newly-enthroned sultan, to order the construction of an astronomical observatory in Galata, across the Golden Horn from Istanbul. In furthering the scientific study of the cosmos – astronomy – it would also improve the accuracy of the astrology used to predict auspicious moments for various undertakings by the Sultan. But Sokullu Mehmed was assassinated in October 1579, and in January 1580 the observatory was razed after the sheikhulislam of the time, Ebüssuud Efendi’s son Ahmed Şemseddin Efendi, advised Murad that observation of the stars brought bad luck, as evidenced by the ills that had befallen the Ottoman state during the past few years.¹³⁷

As the fate of the Istanbul observatory also helps to show, the sheikhulislam was one of the state office-holders whose prestige was enhanced as that of the grand vezir diminished. As the religious hierarchy lost its former distance from secular affairs, he and his advisers became the representatives of the interests of the clergy in the political arena. They did not shrink from fighting out their struggles for patronage in public, to the distress of such traditionalists as Mustafa Ali of Gelibolu, who bemoaned the days when, as he saw it, the religious hierarchy was above politics and possessed the moral authority of the disinterested.138 During the reign of Ahmed I the juridical opinions of the sheikhulislam became a source of law, particularly where land was concerned, and he came to be relied on for decisions which had earlier been the responsibility of the chancellor.¹³⁹ Contemporary sources show that between 1550 and 1650 almost half of those holding the three highest posts in the religious hierarchy came from eleven families: indeed the power of the family of Mehmed III’s former tutor, later sheikhulislam, Sadeddin Efendi – who had been by the timorous Sultan’s side at Mezőkeresztes in 1596 – was second only to that of the Ottoman dynasty itself.¹⁴⁰

As the new millennium and the new century unfolded, there were hints that the state religion of the Ottomans was taking a puritanical, even dogmatic, turn. The episode of the observatory was one sign of this tendency towards inflexibility. There were others: laws restricting the apparel of Christians and Jews were issued, and the drinking of alcohol banned – albeit briefly.¹⁴¹ Yet although it became increasingly intolerant of latitudinarianism in the practice of Islam, and was prepared to pursue those who went beyond the bounds of acceptability in their expression of the Muslim faith, the Ottoman Empire remained remarkable for its toleration of non-Muslim minorities whose separate – and unequal – status was guaranteed by law in exchange for payment of a poll-tax. Jews were prominent in commerce, and many were successful tax-farmers. During much of the sixteenth century they were close to the Ottoman dynasty, employed as physicians to the sultans and as diplomats; the Jewish banker Joseph Nasi advised both Süleyman and Selim II. Following the financial crisis of the 1580s, however, and the social and economic upheaval which ensued, the position of such prominent Jews altered. Envy of the wealth they had accumulated, for example, resulted in the levying of extraordinary taxes upon them, in contravention of the exemptions they had hitherto enjoyed. Mehmed III’s mother Safiye Sultan’s lady-in-waiting Esperanza Malchi, murdered by members of the palace cavalry in 1600, was Jewish, which might have exacerbated their fury at what they saw as her part in the currency manipulation of the mid-1580s; they also accused her of meddling in the tax-collection process. Mehmed’s immediate reaction was to appease the palace cavalry by imposing further restrictions on the customary rights of the Jews, but he lifted these two years later.142

Murad III continued the practice of relying on a Jewish confidant, in his case the Ragusan merchant David Passi, who had been Joseph Nasi’s interpreter. Passi was at court from the mid-1580s; his advice was sought on financial matters, and also on domestic and foreign policy. It was this last which brought about his downfall, for he made an enemy of Koca Sinan Pasha, five times grand vezir, who had different ideas about who were the enemies of the Ottoman state. Koca Sinan launched a violent campaign of defamation against Passi; he not only wrote to the Sultan about Passi’s supposed crimes but fulminated against all Jews and declared them unsuitable to hold positions of influence in an Islamic state; he blamed the economic problems of the time on Passi, and sought to have him executed. In 1591 Murad ordered that Passi be deported to Rhodes, which became a recognized place of exile for those who narrowly escaped execution.¹⁴³

At the turn of the century, one of the most populous Jewish communities in the city of Istanbul was dispersed to make way for the mosque complex of the queen-mother Safiye Sultan. Situated between the port facilities on the Golden Horn and the bazaar on the hill above, this was the commercial hub of the city, where Rüstem Pasha had built his mosque and shops. A synagogue and many Jewish houses were compulsorily purchased to provide a site for Safiye’s complex; it was a controversial proceeding, and the project was criticized for its expense.¹⁴⁴ The first stone was laid on 20 August 1598,¹⁴⁵ and had the work continued, it would have been the first mosque built by a queen-mother within the walls of Istanbul. The building works lapsed, however, because Mehmed III died in 1603, followed by Safiye in 1605.^*

The Greek Orthodox subjects of the empire also came under pressure during these years. In 1587 they lost the Church of the Pammakaristos in Istanbul, which had housed the Patriarchate since soon after the Ottoman conquest of the city. In celebration of Ottoman successes against Iran in the Caucasus the church became the ‘Mosque of Victory’ (Fethiye Camii) and the seat of the Patriarchate moved to the Church of St George in the Fener district on the Golden Horn (where it still is today).¹⁴⁶ Provincial governors who sought to follow this example by converting churches into mosques were forbidden to do so,¹⁴⁷ but Koca Sinan Pasha, who founded many institutions for the benefit of the Muslim community, such as mosques, bridges, fountains and bath-houses, in 1590–91, year 999 of the Islamic calendar, provided the goods and money necessary to establish a charitable foundation and make the alterations needed to transform the Church of St George Rotunda in Thessalonica into a mosque. The mosaics in the former church had not yet been covered up in 1591, when they were seen by the Venetian envoy Lorenzo Bernardo as he passed through the city.148

Up until the reign of Murad III, the sultan’s library still contained over one hundred Greek manuscripts, and Greek manuscripts could easily be bought in Istanbul by visitors such as Ferdinand I’s ambassador Baron Ogier Ghislain de Busbecq, who was in the city in the middle of the sixteenth century. John Malaxos, a Greek scholar who moved to Istanbul following the conquest of his native Nafplio by the Ottomans in 1540, catalogued eight private libraries during the reign of Selim II and found some 555 Greek manuscripts; their subsequent fate is unknown, but it is clear that the last years of the sixteenth century saw the end of Ottoman interest in their Byzantine heritage.¹⁴⁹

Ahmed I undertook extensive repairs to Ayasofya, where mosaics of several Biblical figures had remained visible from the central prayer space since the Conquest. Many of these he now had covered over, choosing which to conceal according to Koranic principles. The Pantocrator in the dome, the image of Christ as God, was quite unacceptable from an Islamic perspective, but Mary is a revered figure in Islamic teaching, so the Virgin and Child in the apse were left untouched.¹⁵⁰ The message conveyed by this iconoclastic impulse was of a sultan seeking to find a new role appropriate to the times.

Neither Selim II, Murad III nor Mehmed III built an imperial mosque in Istanbul. Selim built in Edirne, Murad in Manisa where he had been prince-governor; Mehmed did not build a mosque at all. All three were buried, each in his own mausoleum, in the garden around Ayasofya. Sultan Ahmed I may have failed to emulate the military command of his father, however hesitant, but he followed the precedent of his illustrious warrior ancestors, constructing a monumental mosque complex in Istanbul; he was the last sultan to do so until the mid-eighteenth century. Work began in 1609, and there were those who opposed it, pointing out that it was inappropriate because imperial mosques should be built only with the proceeds of conquest.¹⁵¹ This was an objection that had been raised in the past, in relation to other mosque complexes; Mustafa Ali of Gelibolu regarded such extravagance as contrary to divine law.¹⁵² In a time of severe financial straits when foreign and domestic wars were draining the imperial exchequer, it was a project that could hardly fail to excite opposition.

The date when work began on Sultan Ahmed’s mosque was not without significance. The Ottomans were smarting from the loss of face implicit in the terms of the peace treaty made with the Habsburgs in 1606, and the loss of territory to the Safavids in the ongoing war, but the grand vezir, Kuyucu Murad Pasha, had just accomplished the suppression of the Celali rebels. It was not a conquest of the sort traditionalists recognized, but it was the only Ottoman military success of recent times, and Sultan Ahmed’s mosque was its celebration. The extensive complex, commonly known today as the Blue Mosque from the predominating colour of the tiles cladding its interior, was finished in 1617, the year of Sultan Ahmed’s death. It stood on a prominent site on the south side of the Hippodrome; to make way for it the palace of Sokullu Mehmed Pasha – built over the remains of the Byzantine imperial palace, opposite that of Süleyman’s one-time favourite İbrahim Pasha – was demolished.

Ahmed’s choice of a prominent site on which to build his monumental mosque was not the only way in which he sought to emulate Sultan Süleyman: Ottoman and foreign contemporaries alike remarked on his obsession with his great-great-grandfather. Like Süleyman, he promulgated a law-code of his own; he constructed a garden at Dolmabahçe where Süleyman had once had a garden; he ordered new editions of literary works commissioned by Süleyman; and he rode through Istanbul in procession on a richly bejewelled and caparisoned horse as his ancestor had done before him.153 The young Sultan’s mimicry of his illustrious forebear could not change the reality of the times, however, and new themes were also addressed, and new styles adopted, in the representation of the royal personage in both literary and artistic productions of the period. The eulogistic forms of the mid-sixteenth century which glorified a sultan’s military successes gave way to narratives portraying the life of a sedentary ruler, through a wide range of events at court and across the empire. Writers now based their narratives on ‘sources’, both the eye-witness experience of themselves or others and the actual documents put out by a burgeoning bureaucracy. Their histories focused on the state as a whole rather than on the dynasty in and of itself, chronicling the day-to-day governance of the empire and emphasizing appointments and political affairs. By the time Ahmed I came to the throne, the post of official historiographer, charged with lauding the person of the sultan, had been dispensed with as an embarrassing irrelevance.¹⁵⁴

Selim II showed no interest in historical manuscripts; his son Murad III, by contrast, was an avid patron of the arts of the book, and during his reign some of the finest Ottoman manuscripts were produced. These were illustrated with miniatures in the traditional style as well as portraits of Murad and his forebears – he was the first to commission a series of likenesses of the sultans to illustrate a historical text. This series, completed before the assassination of Sokullu Mehmed Pasha in 1579, was based on portraits of the sultans commissioned by the Grand Vezir in Venice, possibly from the workshop of Paolo Veronese. At the same time, representation of the sultan on his throne became an established convention, a visual shift – from the sultan on horseback at the head of his army to an enthroned sultan – that echoed contemporary reality.155

Ambition for conquest was not dead, however, as an album of sultans’ portraits from the reign of Ahmed I indicates. This introduced a new iconographic element in the shape of the ‘Red Apple’, the symbol of world conquest which had once referred to Constantinople, then to Rome, Buda and Vienna in turn.¹⁵⁶ When Ahmed’s son came to the throne as Sultan Osman II some months after his father’s death, he is said to have set off on campaign against the Commonwealth of Poland-Lithuania dressed in the armour of Sultan Süleyman,¹⁵⁷ as if this talisman would restore his troubled domains to their former greatness.

* The project for a canal to join the Don and Volga was finally realized in 1952, along the route the Ottomans had planned to follow four centuries earlier.

* A chronogram is a piece of writing, often an inscription or poem, in which certain letters – in the case of Ottoman chronograms these are usually in the final line – are each assigned a known numerical value; when these values are added together they produce a date.

* Like the Mamluk sultans the Ottoman sultans employed eunuchs to guard their private household; apart from the sultan himself they (and mutes and dwarves) were the only adult males permitted to enter the harem. Little is known of their recruitment and training.

* Despite the Islamic prohibition on usury Ottoman lenders – at least in the Balkans and Anatolia – collected interest on loans; they employed various subterfuges to disguise it. In the later nineteenth century, when the empire was heavily indebted to the European Powers, European rules were applied, and the Ottoman government perforce abandoned all pretence of adhering to the canonical injunction.

* A screen of close-set pointed sticks lashed together, set up as a defensive shield or enclosure.

* As we will see, the mosque was eventually completed as the Yeni (‘New’) or Valide (‘Queen-mother’) mosque in 1665, and still dominates the square at Eminönü.