The Logic of Kleinkrieg: The “Book of Halil Beg” in Habsburg-Ottoman Diplomacy, 1550–76

Abstract

The Ottoman conquest of Szigetvar and Gyula (1566) exposed the weakness of the Habsburg monarchy. Unable to mount a military response, Maximilian II depended on his diplomats to ward off the imposition of a conqueror's peace along the Hungarian border. But an official register, the Book of Halil Beg, put forward the sultan's claims to a wide swath of towns and villages, including many still held by Habsburg loyalists. Moreover, the Ottomans said, His Majesty had best accept a settlement on these terms; victorious Turkish captains wanted more and were restrained only by Selim II's orders. This essay describes the efforts of Maximilian's ambassadors to forestall additional losses in peacetime. After years of tough negotiation, they managed to obtain minor revision of the relevant treaty text. Although the Book of Halil Beg would indeed be the basis for any border settlement, as the Porte insisted, the Ottomans acknowledged that Vienna disputed certain of its provisions. This mere protest may have been enough to assure loyalist magnates in Hungary that the dynasty had not abandoned them, or their hereditary lands. Hence the paradox of Habsburg-Ottoman diplomacy. The 1576 treaty of peace, formally ending hostilities, was at the same time a charter for Kleinkrieg.

Soon after his accession in 1564, Maximilian II had a problem on his Hungarian frontier. Troops loyal to the voivode of Transylvania conquered the fortress-town of Szatmár (Satu Mare),¹ the anchor of Habsburg territory east of the river Tisza.² By his own account, the emperor reacted cautiously, lest he seem to violate the Habsburg-Ottoman peace treaty of 1562. He sent no troops until Ottoman regulars supported a further attack from Transylvania.³ His forces retook Szatmár and some towns hitherto held by Transylvania, but their success helped provoke a reaction. On 18 November 1565, the Sublime Porte announced that Sultan Suleyman would march to Hungary in the spring.⁴ In 1566, the Habsburgs suffered their worst defeat in the long history of the Ottoman wars. In Hungary, Szigetvár was lost and so was Gyula. Maximilian's army, assembled at great cost, accomplished little and broke up in disarray.⁵

For the rest of his reign, Maximilian II's policy toward the Porte was guided by a recognition of Ottoman military superiority. Reversal of the recent losses was not possible; even before the catastrophic campaign of 1566, no one thought the Habsburg monarchy capable of a full-scale invasion of Ottoman territory.⁶ Some advisers recommended a coalition with other Christian princes, but there were always obstacles to an alliance.⁷ In 1571, Maximilian's adherence to the Hispano-Venetian Holy League would have required lengthy consultation with the Imperial Diet.⁸ When Poland-Lithuania was to elect a new king in 1572, Maximilian approved the organization of a candidacy on behalf of one of his sons, Archduke Ernst, but support in the Polish Sejm for the Ottoman candidate, Henri of Valois, was too strong. Two years later, when the French prince returned to a vacant throne in Paris, Maximilian advanced his own candidacy and was elected by the Polish Senate; but the lower house of the Sejm blocked the Senate's wishes, securing the election of another Ottoman candidate, Voivode István Báthory of Transylvania.⁹ The alternative to an anti-Ottoman military coalition was negotiation with the Porte. As this essay shows, Maximilian's Ottoman diplomacy addressed one key problem: How to prevent Ottoman conquerors from dictating a settlement to Habsburg subjects along the border, as prescribed in a Turkish document, the Book of Halil Beg?¹⁰

What to Do about an Ottoman “Book”?

In late 1566, following a suggestion from the paşa of Buda,¹¹ Maximilian sent a courier to the Porte to ascertain the conditions under which the new sultan, Selim II, would open discussions. After an encouraging response, he dispatched two ambassadors, Christoph von Teuffenbach, a member of his Hofkriegsrat, and Antun Vrančić, a humanist diplomat soon promoted to the archbishopric of Esztergom. On returning to Vienna, they reported Ottoman demands that were scarcely acceptable, notably that the border be regulated according to a certain Ottoman “book.” Nonetheless, the ambassadors were sent back with new instructions.¹²

The Treaty of Edirne was eventually signed on 17 February 1568. As on other occasions, there was no common text of the treaty; each side prepared its own version. Habsburg officials apparently failed to see—until sometime later¹³—that the Turkish text included a clause stating that taxes would be collected only from places listed in the Ottoman tax registers.¹⁴ Meanwhile, points that were not resolved in the treaty still had to be hammered out, for purposes of the required letters of confirmation.¹⁵ In March 1568, Teuffenbach and Vrančić reported that the Ottomans insisted that “common villages”—currently paying taxes to both sides—must be divided up according to a certain “book” in Turkish. Sokullu Mehmet Paşa, the grand vezier, tried to persuade them that this would be to Maximilian's advantage: if this book were taken as the standard, any claims by Ottoman officials for places not inscribed therein would be treated as illegal. But the ambassadors protested: Maximilian had “not only not seen” the book in question, he had “never known about it until now.” When the conversation grew heated, they adopted a fallback position given in their instructions: His Majesty would allow the issue of common villages to be settled by border commissioners, who could take “the book” into consideration. In response, Sokullu Mehmet Paşa “commanded that this article be changed in keeping with our wishes.”¹⁶ Vrančić and Teuffenbach sent back a copy of the Turkish-language book in question,¹⁷ and letters of confirmation were duly exchanged. Book or no book, it seems no one in Vienna thought border disputes could be settled by a commission. But what must at all costs be avoided was an official admission by His Majesty of the full implications of his recent defeat. Endless discussion was far preferable to a definitive settlement that allowed the Ottomans to claim lands and even castles they had not captured in the 1566 war.¹⁸

Despite the exchange of ratifications, the common villages question was not so easily settled. During the negotiations, Ottoman sancakbegs sent out troops to prohibit villagers from paying their accustomed taxes to Habsburg lords. According to the paşa of Buda, they acted lawfully because the villages in question were inscribed in the sultan's tax registers (plural).¹⁹

Vienna had again failed to notice that the Turkish text of Selim II's letter of confirmation differed from the accompanying Latin translation.²⁰ The Turkish text specified that “His Imperial Majesty has conceded [to the sultan] the subjects and common villages that pay taxes and dues to both sides, but are inscribed in the book.”²¹ Moreover, according to the Ottoman diplomatic style, a treaty expired with the death of a sovereign; hence new treaties had to be negotiated after the death of Selim II in 1574 and after Maximilian II's death in 1576.²²

The book was attributed to Halil Beg (or Halil Bey), who was defterdar (treasurer) of the province (eyalet) of Buda in 1550, and who died in 1568/69.²³ Ottoman officials seem not to have called it “the Book of Halil Beg” until after his death.²⁴ No “Book of Halil Beg” has been found, and no such book came up in discussions leading to the five-year Treaty of Edirne in 1547.²⁵ But a defter described in similar terms figured in negotiations prior to the eight-year treaty of 1562.²⁶ Under Maximilian II, Ottoman insistence on the “book” beclouded the meaning of the 1568 treaty and obstructed agreement on new treaties concluded in 1574 and 1576.²⁷

Hence this essay poses three questions. First, what may be learned about “the book of Halil Beg” from the correspondence of Ferdinand I's reign? Second, how did Habsburg and Ottoman negotiators hash out the issue during Maximilian's reign? Finally, how did this contentious dispute reflect larger interests of state on both sides?

The Book of Halil Beg during the Reign of Ferdinand I

For newly conquered land, the Ottomans conducted surveys to determine the income to be expected from each locality. Compiling a defter or register took time. The financial official in charge estimated revenue from agriculture and trade by taking the average for the last three years. Some defters were mere lists of sums due, others were detailed enough to indicate how many households there were of the various taxpaying categories (e.g., Muslims and infidels) in each village.²⁸ Once a district had been surveyed, it might not be surveyed again for decades.²⁹ Different kinds of surveys were conducted in the eyalet of Buda.³⁰ For this period, the compilation of records has been described in detail for one of the eyalet's component sancaks.³¹

The “book” or registrum ³² of Halil Beg seems to have been a defter of the summary kind, of which one copy would be kept in the capital of the eyalet (Buda), and another at the Porte.³³ Notwithstanding the later claim that Maximilian II had never heard of it, it can plausibly be identified with the “book” mentioned in the correspondence of Giovanni Maria Malvezzi, Ferdinand's ambassador at the Porte from 1548 to 1553. In 1548 Malvezzi was instructed to protest the paşa of Buda's use of a certain register to extort money from villages that had never paid Ottoman taxes.³⁴ Two years later, Malvezzi reported that the grand vezier documented the sultan's claim to two disputed towns by showing him “a book of incomes that was compiled before the truce”; that is, before the treaty of 1547.³⁵ Malvezzi was told not to worry: “They write places in their books because Ottoman soldiers have asked for them from Sultan Suleyman, but these places were in Christian hands before the treaty, and still are.”³⁶ Ferdinand and his advisers did not grasp the distinction—important for the Ottomans—between the official record kept by the defterdar of Buda, and the boastful claims of frontier commanders.

Places that the Ottomans had indeed conquered were certainly in the defterdar's book, and as Ottoman Hungary expanded, new towns and villages were added. When Szolnok was taken by the Ottomans in 1550, many rural magistrates in the vicinity were “compelled by fear of the Turkish sword to inscribe their villages in the said book.”³⁷ In 1552, after Ferdinand made arrangements for Transylvania to pass under Habsburg rule, the Ottomans invaded, conquering enough territory to establish a new eyalet in Temesvár.³⁸ According to a later account, the townsfolk of Debrecen agreed at the time of their submission to pay the sultan seven thousand thalers a year, a sum that was then “inscribed on the Book of Halil Beg.”³⁹ Ottoman officials always described inscriptions as voluntary. As one grand vezier told Ferdinand's ambassador, villages added to the book had “submitted to the sultan at their own initiative, and by their own will.”⁴⁰

By now, Ferdinand recognized the book as a problem. In June 1553, he sent new ambassadors to the Porte: Antun Vrančić, then bishop of Pécs, and Ferenc Zay, commander of the Danube gunboat fleet at Komárom. Their main assignment was to persuade Suleyman to grant Transylvania to his “beloved son” Ferdinand as a free gift—surely an impossible task, but one in which they dutifully persisted for three years.⁴¹ They were also instructed to insist that peasants continue paying their traditional census to Habsburg landholders, even if their villages were inscribed in “that book of the prince of the Turks.” Villagers must keep making payments to Habsburg as well as Turkish lords, as they had done prior to the Ottoman conquests of 1552, “so that each side receives an equal share of the ordinary tribute.”⁴²

Vrančić and Zay found Ottoman officials willing for “common villages” to keep paying tribute to both sides, provided that Ferdinand abandoned his claim to Transylvania.⁴³ To be sure of Ferdinand's will on this point, the Porte demanded that he send a new ambassador with new instructions. For this purpose, Ferdinand chose a young Flemish nobleman recommended by his agents at the court in Brussels, Oghier Ghislain van Boesbeek, better known as Busbecq. Stopping in Buda en route to Istanbul, Busbecq encountered a local version of a larger problem. The provincial governor, Semiz ‘Ali Paşa, had previously agreed to “perambulate” the disputed villages with a Habsburg envoy to settle outstanding differences. Now, however, the paşa reneged on his promise, apparently in deference to his military commanders. Once in Istanbul, Busbecq remained as ambassador until 1562,⁴⁴ while Vrančić and Zay returned in 1557. Ottoman border garrisons continued to harass villages that had not previously paid Turkish taxes.⁴⁵ Nonetheless, when times were propitious for discussing a new treaty, officials at the Porte did not include adherence to the Ottoman tax registers among their demands.

In January 1559, Busbecq and Grand Vezier Rüstem Paşa agreed on a draft treaty. It provided that the sultan's soldiers would not molest Habsburg villages, and that “common villages not be subject to unaccustomed burdens.”⁴⁶ For various reasons, this treaty did not become final. Negotiations then resumed in 1561, when Semiz ‘Ali Paşa became grand vezier; this time a treaty was concluded, in 1562, on more or less the terms agreed in 1559.⁴⁷ Busbecq carried the text back to Vienna, accompanied by Ibrahim Beg, the chief dragoman.

It is sometimes said that the Ottoman state, dedicated to expanding the sway of Islam, did not accept fixed boundaries until the Treaty of Karlowitz (1699). In fact, wars of this era between the Ottomans and Venice were concluded by marking out a new boundary between their respective territories in Dalmatia.⁴⁸ For years, Ferdinand had vainly sought fixed borders for his lands in Hungary and Croatia.⁴⁹ Now, Semiz ‘Ali Paşa offered to divide the common villages and fix a boundary between them. He seems to have had in mind a kind of no man's land. There could be no firm peace, he told Busbecq, “unless a certain space were left between the territories of both rulers, so that men divided by religion might be distinguished from one another.” But Ferdinand was no longer interested.⁵⁰ Ali Paşa accused Busbecq of harshness “because we would not suffer the unfairness of the common villages to be resolved by fixing a border.”⁵¹

Frontier soldiers had their own ideas. Even as the treaty was being concluded, Ferdinand complained that Ottoman soldiers were demanding tribute from villagers “on the pretext that these villages are inscribed in a certain book of their emperor.”⁵² The grand vezier might have deplored such unprovoked aggression by either side. Yet by his own telling, Semiz ‘Ali Paşa was more inclined to peace than his fellow veziers.⁵³ If the defterdar's register was not an issue for the 1562 treaty, there were officials at the Porte who had not forgotten it.

The Book of Halil Beg under Maximilian II, 1568–76

By the summer of 1568, the Ottoman understanding of the proposed new treaty was clear to Habsburg officials. A clause in the Turkish text provided that villages not inscribed in the registers did not have to pay Ottoman taxes;⁵⁴ by implication, those inscribed had to pay. According to the Turkish original of Selim II's letter of confirmation, Maximilian had “conceded” all places listed in the Ottoman registers.⁵⁵ Moreover, Vienna knew that the sultan had commanded the paşa of Buda to ensure that enrolled villages made no payments to landlords who were subjects of Maximilian II.⁵⁶ For the Porte, only one issue remained: there must be a fixed border, with boundary stones, based on the tax books.⁵⁷

If a border were drawn as Semiz ‘Ali Paşa envisioned it in 1562, there would have been a no man's land clearly separating the two realms,⁵⁸ implying that governments and leading men on both sides forfeited all claims to lands across the border. But what the Ottomans now had in mind was a different matter. On their side, castle-towns claimed by the Porte,⁵⁹ with their subordinate villages, would “render services only to the Turks.” On Maximilian's side, while Habsburg landlords received their accustomed census, Ottoman lords would also receive the payments “which villagers of their own accord are accustomed to make.”⁶⁰

After Vrančić and Zay returned to Vienna, Albert de Wijs, who had come to Istanbul as Busbecq's secretary, remained at the Porte as Maximilian's resident. In May 1568, Mehmed Paşa Sokullu told De Wijs that the common villages must be divided “according to his emperor's book.” He expected that Ibrahim Beg, then in Vienna, would bring back a definitive answer from Maximilian II.⁶¹ Because no such response was forthcoming, the grand vezier hinted at consequences. The sultan would have his beglerbegs keep their men within bounds, he said, “but if would be altogether better if the common villages were divided according to the sultan's book, so as to fix a limit that would not be lawful for armed men to cross.”⁶²

Ottoman officials had no local maps and seem to have thought of the border as a network of fortified positions. On the Habsburg side, military cartography of border districts, just beginning, made things clear to Maximilian and his advisers.⁶³ A settlement along lines proposed by the Porte would strip away villages and castle towns that Ottoman armies had not even approached in 1566, much less conquered. His Majesty would thus lose by the peace territory he had not lost in the war.⁶⁴ The task of Habsburg diplomats at the Porte was to ward off this further catastrophe.

Vienna's strategy was to seek revision of the Ottoman version of the February 1568 treaty, avoiding reference to the sultan's “book,” or minimizing its impact. De Wijs was to bring up an idea Mehmed Paşa had seemed to accept in 1567: disputed points could be settled by a bilateral commission, with consultation of the sultan's “book.”⁶⁵ According to De Wijs, the grand vezier “nodded calmly” at this suggestion; this procedure would not displease him, provided that one could “establish firm peace and good neighborliness among subjects on both sides.”⁶⁶

As paşa of Buda, Sokullu Mustafa Paşa (r. 1566–68), a kinsman of the grand vezier,⁶⁷ occupied a key position in the Ottoman hierarchy. With fifteen sancakbegs reporting to him, he was a de facto viceroy of Ottoman Hungary and the natural intermediary for talks between Vienna and the Porte.⁶⁸ Mustafa Paşa had given “the first impulses” for peace negotiations in 1568,⁶⁹ and early in 1569 Maximilian sent an envoy to sound him out; now too the paşa seemed amenable to resolving problems. The envoy highlighted two names on a list of villages wrongly harassed by Ottoman tax collectors. The paşa “had the book of his emperor brought to him”; not finding the two villages, he promised to check the rest of the envoy's list. But on the idea of a bilateral commission, Mustafa Paşa was evasive: “We will see,” he said.⁷⁰

De Wijs died in October 1569; initially promised a one-year term, he had been at the Porte since 1562.⁷¹ Vienna now had to send an envoy having the rank of ambassador, and men of the required standing were not eager to live for untold years among Ottoman “barbarians.”⁷² Maximilian also had to respect national sensibilities: Would Germans trust a Hungarian as ambassador? Or Hungarians a German? He turned for advice to a Fleming, Busbecq, now the tutor of his grandchildren. Busbecq's contact in Brussels recommended a young Flemish jurist named Karel Rijm, and he accepted.⁷³ Rijm's instructions (February 1570) summarized previous disputes on common villages, the Ottoman tax register, and the text of Selim II's letter of confirmation.⁷⁴

When Rijm arrived, the Porte was preoccupied by preparations for the invasion and conquest of Venetian Cyprus. In response, Spain, Venice, and Pope Pius V formed the Holy League after six months of tense negotiations in Rome.⁷⁵ Maximilian was not a member of the League, but the Ottomans still had suspicions.⁷⁶ As Spanish and Venetian fleets mobilized in the summer of 1571, Rijm had troubling news from Flanders, where William of Orange had raised the banner of revolt. Rijm's brother asked that he be allowed to return home, but the Habsburg official who coordinated diplomatic affairs, Obersthofmeister Johann von Trautson, would not consent; these were dangerous times, he said, “and we have no one with whom to replace him [Rijm].”⁷⁷ Meanwhile, Rijm had made little headway. The sultan and his veziers would not under any circumstances set aside a tax register that recorded what they saw as their gains in the 1566 war. The best Rijm could do was to fall back on the usual idea of border commission that could take the sultan's book into account.⁷⁸

The Ottomans might have been more concerned about their western frontier after the disaster at Lepanto (7 October 1571),⁷⁹ but if so, it was hard to tell from discussions recounted by Rijm. His one success was to have the grand vezier send a written command for Mustafa Paşa of Buda to investigate Habsburg grievances about border violations. Given a translation of the command, Rijm saw that it contained “the old song” about checking the “book” to determine which villages were Ottoman. But Rijm also found something he thought was new:⁸⁰

What they mean by the book is restricted to the register of Halil Beg, which I understand was drawn up a number of years ago. Newer books have been compiled and are still being compiled, in which many villages hitherto immune and unregistered have been inscribed; these are to be abolished, and … the book of Halil Beg is to be the norm.

The language that caught Rijm's attention was not in fact new. To be sure, there had been some ambiguity in official Ottoman texts. While the Turkish version of Selim II's letter of confirmation stated that taxes were not to be collected from villages not listed in “that book” (singular), the much-disputed clause in the Turkish version of the February 1568 treaty referred to “tax registers” in the plural.⁸¹ At some point, the Ottoman government chose to resolve any ambiguity by making the “Book of Halil Beg” definitive for the Hungarian frontier. This did not mean accepting the territorial status quo of the 1540s, when the defter was first compiled, because other places had been registered as Ottoman conquests proceeded.⁸² What it meant was that decisions about the frontier were to be made at the Porte, not by local commanders.⁸³

Transylvania offered an example of how “the book” could be used. Voivode István Báthory (r. 1571–76) had good relations with its Ottoman overlord and also with Vienna.⁸⁴ In June 1572, Sultan Selim II sent to Buda a list of ninety villages claimed by Transylvania, along the eastern border of the Buda eyalet; Mustafa Paşa was to check the list against the Book of Halil Beg. Villages inscribed therein were to be retained for the sultan; concerning the others, the paşa should make inquiries because these villages might have been seized by Ottoman sipahis without authorization.⁸⁵ The Habsburg monarchy never stood in the same relation to the Porte as Transylvania did. But Selim II's command to Buda showed that the Book of Halil Beg could be an instrument for reining in the ambitions of Ottoman military men.

Yet Rijm found his interlocutors unwilling to consider any changes to the language to which the sultan had agreed in 1568. In July 1572, Sokullu Mehmed Paşa told him that the paşa of Buda would soon receive a command concerning villages along his western border, but the ambassador should not even think about raising the question of a treaty revision in a public divan, or in an audience with the sultan.⁸⁶ A month later, Rijm had a copy of the mandate to Buda. He had hoped that reference to “the book” might be omitted altogether, or at least remanded to the consideration of a bilateral commission, but on both counts he was disappointed, and so lodged a complaint. Mehmed Paşa told him to be content with the fact that a mandate for regulating the border had at last been sent, and then “started harping once more on that book.”⁸⁷ Maximilian and his advisers had apparently decided that the time for discussing revision of the treaty was not opportune. Mehmed Paşa was certainly of this opinion.⁸⁸ According to Rijm the grand vezier “seemed to nod” at his suggestion that Maximilian could best defend his lawful rights not by “memoranda and disputations” but “in fact and deed” (reipsa factoque), as he was now doing.⁸⁹ Up to a point, both capitals understood that some disputes between ambitious frontier lords would only be settled by a trial of arms.

Maximilian was by now searching for a new ambassador. Rijm “most eagerly awaited” the someone he was told was coming to succeed him.⁹⁰ Names of possible candidates, mostly Hungarian and some German, were proposed by the Hungarian Diet and by magnates lay and clerical.⁹¹ This time the emperor chose a German, David Ungnad, a member of the Herrenstand in Styria who later served as president of the Hofkriegsrat. One of the eight languages Ungnad reportedly spoke was Hungarian, and his close ties with Lutheran theologians of the Philippist party will also have recommended him to the mainly Protestant Magyar aristocracy.⁹² Ungnad accepted in March 1573 on the understanding that he would serve for one year only⁹³ (he ended up staying until 1578). Rijm was now given the name of his successor but was also asked to stay on for half a year more to “educate” Ungnad in the ways of the Porte.⁹⁴

From the start of his embassy, Ungnad showed an aristocratic punctiliousness about upholding his dignity as an ambassador, a posture he believed the emperor's dignity required.⁹⁵ He arrived in the Ottoman capital in August, and for the next fifteen months most of the correspondence was signed by both ambassadors and written in Latin.⁹⁶ At the time, there were three Venetian diplomats at the Porte seeking clarification on the terms of a treaty ending the Cyprus war. One of them⁹⁷ told Maximilian's envoys that Venice knew how villages around Szolnok had been forced against their will to submit to Ottoman rule. Rijm and Ungnad then wrote to ask for Venice's help in obtaining a revision of the language about common villages that was deceptively inserted into the Turkish text of the 1568 treaty.⁹⁸ But Venice, hard-pressed to protect its own territories, offered no more than a sympathetic ear.

On the main point of their mission, two ambassadors fared no better than one. The Porte's assumption that Maximilian had accepted the Book of Halil Beg was made clear when Rijm and Ungnad were given a list of seventy villages that refused to pay Ottoman taxes, even though they were said to be listed in the book.⁹⁹ Seconding a letter sent by both men, Ungnad wrote separately to Maximilian, begging his pardon for having accomplished “less than we hoped … because the circumstances of the Ottomans are now favorable” since the Holy League had broken up.¹⁰⁰ Still, the ambassadors were nothing if not persistent. They suggested once more that the status of the book might be referred for decision by a bilateral commission, but Mehmed Paşa insisted that a commission was pointless because the villages in question had already been “ceded” to Selim II, as was said in the sultan's letter of confirmation.¹⁰¹ When they asked him to convey their request to Selim II, they were told that the padishah would countenance neither a revision of the treaty nor new language about a commission. It now seemed that pressing things might endanger the peace, such as it was.¹⁰² The ambassadors tried yet again when reports came that the Ottoman naval expedition to reconquer Tunis was going badly, only to get the same answer. Their dispatch of 8 January 1574 concluded this latest report with a confession of hopelessness: “I, Carolus Rijm have not been able to change [the grand vezier's] mind, and I, David Ungnad, doubt that I can do better if I remain here alone.”¹⁰³

Reports of Ottoman progress in the Mediterranean—Tunis was in fact recaptured in August 1574—led the ambassadors to despair; the government with which they had to deal had now vanquished all its foes.¹⁰⁴ But Maximilian had a new strategy. In January 1574, he sent an envoy to Buda to sound out Mustafa Paşa's willingness to help the Habsburgs.¹⁰⁵ Apparently on the basis of this discussion, Maximilian then wrote to Sultan Selim. Instead of seeking revision of the 1568 text, he proposed a “prolongation” (prorogatio) of the existing treaty for eight years, starting in January 1576.¹⁰⁶ Commenting on this letter, Ungnad and Rijm explained to Mehmed Paşa that Maximilian preferred for the Book of Halil Beg not to be mentioned in documents proclaiming a prolongation, but at this the grand vezier grew so angry that Mahmud Beg, the court dragoman, warned them to drop the topic. This time, their fallback proposal was that a bilateral commission would take the Book of Halil Beg as the basis for discussion, not merely as something to be considered. Mustafa Paşa of Buda would have “responsibility for the interpretation of that book,” together with a commissioner named by Maximilian. This, at last, was a formula to which the grand vezier was not averse.¹⁰⁷

Mehmed Paşa had the chancery prepare a command along these lines for the paşa of Buda,¹⁰⁸ of which the ambassadors were given a translation. Ungnad and Rijm reported that the wording was not what they wanted, but they nonetheless found reason for satisfaction because the command going to Buda included the terms “controversy” and “disagreement.” By definition, they said, “these words indicate that in accepting a prolongation, Your Majesty has not accepted the book [of Halil Beg], because ‘controversy’ and ‘disagreement’ contain the idea of contestation.” This language was thus an improvement over the Turkish text of Selim II's letter of confirmation, which specified that Maximilian had “ceded” the disputed villages.¹⁰⁹

Other issues now intervened. There would be no “prolongation,” Ottomans insisted, until Maximilian razed his new fortress at Kálló,¹¹⁰ near Szolnok. They also demanded that he deliver to the Porte two rebels who had found refuge in Vienna: Bogdan IV, a former voivode of Moldova,¹¹¹ and Gáspár Bekes, who challenged István Báthory in Transylvania.¹¹² Maximilian's military advisors thought Kálló too important to be torn down.¹¹³ Mehmed Paşa was somewhat flexible on this point, and in the end Kálló continued to stand, even though Selim II, in the letter of prolongation that was eventually sent to Maximilian, presumed that it was to be torn down.¹¹⁴ As to Bogdan and Bekes, Ungnad suggested it would be better if they were no longer seen at the court in Vienna. Maximilian made excuses; he did not know where Bogdan was, and he did not think he had violated the terms of the treaty by extending hospitality to Bekes.¹¹⁵

By August 1574, Rijm and Ungnad were hoping to influence the language of the two letters they would carry back to Maximilian, one from Selim II, the other from Mehmed Paşa, and were frustrated because the grand vezier would not discuss the matter.¹¹⁶ When they finally got to see translations, in October, they were dismayed once again. It seems Selim II's letter still referred to the “cession” of disputed villages by Maximilian.¹¹⁷ It also called for the destruction of Kálló, and the ambassadors were told the sultan would not hear of any change to the text he had approved.¹¹⁸ There was, however, an interesting bit of court gossip. The paşa of Buda's resident secretary at the Porte was heard telling a friend that the Book of Halil Beg ought to be set aside because it did not include many villages that Ottoman sipahis currently possessed as part of their timar allotments.¹¹⁹ In other words, not all the territories that the Ottomans de facto occupied had been entered into this register. Thus, if Maximilian agreed to take the book of Halil Beg as normative, his noble subjects might in theory be able to recover some of their lost lands.

Ungnad and Rijm persisted in seeking modification of the two letters. By their account, they obtained three changes and decided not to seek a fourth.¹²⁰ By November, the sultan's letter no longer spoke of a “cession” of villages, and language about the destruction of Kálló had been deleted from Mehmed Paşa's letter if not from the sultan's.¹²¹ On 11 November, the last joint letter from both men called attention to the implications of the new language. The Ottomans had for the first time agreed that a bilateral commission was to address “controversies” about the villages, a formulation that was “diametrically opposed” to the claim that Maximilian had simply approved the book of Halil Beg and ceded the villages; thus “Mehmed Paşa will no longer be able to allege” that offensive clause from Selim's letter of confirmation because it was in effect annulled by his new letter prolonging the peace. Because this important change would not have been possible without the grand vezier's cooperation, the ambassadors recommended sending Mehmed Paşa the twelve thousand thalers he had been promised if a good result were achieved.¹²²

Before the agreement could be ratified by both parties, it was invalidated by the death of Selim II in December 1574. Murad III (r. 1574–95), the first sultan to have been raised in the harem,¹²³ apparently needed some coaching on how one conducted negotiations with infidels.¹²⁴

As time went on, Ungnad, now Maximilian's sole representative at the Porte, suspected that the resumption of formal negotiations was being delayed so that Ottoman commanders could seize more fortresses and land in the interim.¹²⁵ In the end, Maximilian would not live to see a new treaty (he died on 12 October 1576). But Ungnad had an agreement by November 1576, in terms essentially the same as what had been agreed two years earlier. Rudolf II (r. 1576–1612) promptly approved both the treaty and Ungnad's work in negotiating it.¹²⁶

Some years earlier, Albert de Wijs thought that what happened along the border did not matter: “Turkish border commanders and their men think of nothing except plunder and slaughter, but they cannot hurt us so long as Mehmet Paşa [Sokullu] shows favor to us, for everything is done by his nod, and at his will; his goodwill must be maintained by all means, and especially by large gifts.” Even if Maximilian in his response seemed to agree,¹²⁷ the more considered opinion of the Habsburg government was that the Ottomans could not be allowed to regulate the border as they wished, according to their own records. But the net result of eight years of tough bargaining (1568–76) was a minor concession from the Ottomans: the Book of Halil Beg would indeed form the basis of discussion for a border commission, but on the understanding that the Habsburgs disputed some of its provisions.

Neither Maximilian nor his advisers expected border issues to be settled by a commission,¹²⁸ if it ever convened. Yet Rijm and Ungnad took satisfaction in their work.¹²⁹ This slight change in the Ottoman position arguably diminished the dread possibility of another full-scale invasion. Because the Porte now recognized that its land records were not accepted without qualification, border skirmishes were less likely to be elevated to the status of a casus belli. This was perhaps what Ungnad meant in writing to Maximilian that the new agreement had “achieved most of Your Majesty's wishes.”¹³⁰ In fact, Maximilian was in no position to make demands of the Porte, he could only make requests. The end result, modest as it was, satisfied the conquered Habsburgs as well as the conquering Ottomans. Why this was so can be made clear by a brief look at the strategic aims of Selim II's Ottoman Empire and Maximilian II's Habsburg monarchy.

Reasons of State

De facto, Sultan Suleyman became the ruler of south-central Hungary after his victory at Mohács in 1526. De jure, by Ottoman law, he became lord of the entire kingdom because he had defeated Hungary's previous king in open battle.¹³¹ In Ottoman Hungary, officials followed Hungarian law and custom insofar as it helped them consolidate their position. For example, paşas and sancakbegs took advantage of the rule that lordship of a castle entailed lordship of its dependent villages; in some cases, this principle put Habsburg negotiators on the wrong side of an argument about local custom.¹³² Although the Porte described submissions to the sultan's lordship as voluntary,¹³³ Ottoman law, and the law of conquest in particular, was always in the background,¹³⁴ and Ferdinand I's ambassadors had to adapt to it. As a prelude to their petition that the sultan demonstrate his liberality by granting Transylvania to Ferdinand, they first acknowledged that all of Hungary belonged by right to Suleyman.¹³⁵ In Transylvania, as in other lands not directly ruled by the sultan, his grant of authority to the voivode was revocable at will. But the title could not be taken away without his permission. This was a form of law, a conqueror's law, and it provides a context for the Porte's regulation of the border.

In the wake of Ottoman victories in 1566, the Porte gave free rein to its commanders along the southern sector of border, against Croatia,¹³⁶ but not in Hungary. Almost from the start of the post-Szigetvár discussions, Ottoman language about the Book of Halil Beg had a dual aspect: officials insisted that this tax register be taken as normative, but they also stipulated that villages not listed would not be molested.¹³⁷ The first part of the message came through loud and clear, but it took Vienna some years to listen carefully to the second part. Buda's Mustafa Paşa was willing to grant that some villages claimed by Transylvania were not taxable according to the Book of Halil Beg.¹³⁸ Might he do the same for the Habsburgs? Perhaps not. The paşa's agent at the Porte seemed opposed to a policy that would have taken land from Ottoman sancakbegs or sipahis.¹³⁹

Why did Selim II's government place limits on the ambitions of its fighting men in Hungary but not in Croatia? The most obvious reason was that, as Ottoman military planners surely knew, Habsburg defenses were stronger in Hungary than in Croatia.¹⁴⁰ But other considerations were also in play, including the law-bound character of Ottoman government, a point that recent scholarship has stressed.¹⁴¹ While the law recognized “the absolute independence” of the sultan's “supreme authority,” it presumed that even the sultan ought to be guided by the “objective rules” of established law.¹⁴² From this perspective, making the Book of Halil Beg normative was a way of subjecting the violence of border life to a legal regimen.

A policy of restraint could also help tighten the Ottoman chain of command. From a strictly military standpoint, it made sense to keep soldiers busy at their trade, while not allowing the men or their commanders to take undue liberties. Without seeking permission from the Porte, paşas and sancakbegs could authorize raids of some size, if not major attacks.¹⁴³ But they sometimes took it on themselves to reward sipahis with an extra village or two, and to have such transactions recorded in a defter kept by a local financial official. In so doing, they might offend a Habsburg lord who had the means to fight back, and thus provoke a quarrel that was neither desired nor intended by the Porte. Officials in Istanbul alluded to this practice—and the need to suppress it—as they attempted to convince their interlocutors that Maximilian would serve his own interests by accepting the Book of Halil Beg.¹⁴⁴

Finally, further expansion into Hungary was not a current priority. During his tenure as grand vezier (1566–79), Mehmed Paşa addressed major issues one after another: putting down a rebellion in Yemen (1567–68); attempting to dig a canal between the Don and the Volga so as to strike at Iran across the Caspian Sea (1569); and fighting a naval war in the Mediterranean (1570–74). He also kept an eye on developments that might have adverse implications, like elections for a new king in Poland-Lithuania or the emergence of an independent Muslim power in distant Morocco.¹⁴⁵ From this quasiglobal horizon, the Book of Halil Beg was a useful vehicle for keeping a rein on border commanders, and thus diminishing the likelihood of disturbances in Hungary that might require diverting troops from elsewhere.

Peasants who tilled the soil or pastured their flocks suffered the most from border violence. To judge from written threats issued by rival commanders, it seems villagers might face a choice between being burned out if they failed to pay Habsburg taxes, or carried off into slavery if they failed to pay Ottoman taxes.¹⁴⁶ In the long run, it would prove that the only way to keep the frontier settled was to create a class of free peasants holding land in return for military service.¹⁴⁷ Meanwhile, more and more villages were depopulated, and Habsburg fortresses were left short of provisions and without the peasant labor (robot) that repair and upkeep required.

In political terms, what counted most was that border violence threatened the great landed families on whose continued support the monarchy depended. Magnates and nobles, often in separate chambers, spoke for the various provincial estates that voted taxes for the frontier. Magnates and provincial estates provided no small amount of the credit that kept the government afloat.¹⁴⁸ They also held most positions of military command; one finds a virtual roster of Habsburg Hungary's great families in Miklós Istvánffy's contemporary account of the Ottoman wars.¹⁴⁹ Finally, by intermarrying among themselves, the great clans formed a supraterritorial elite that helped to counterbalance national and regional antipathies among the monarchy's subject populations.¹⁵⁰ Despite memorable conflicts over magnate privileges, the interests of the dynasty are hard to distinguish from the interests of its great families.

Some magnate families sought to consolidate their lands, but others found it useful to maintain holdings across a number of Hungary's counties.¹⁵¹ To reclaim family patrimony, or to vindicate the property rights of offices they held, the great men contested the expansion of Ottoman dominion, before and after the Szigetvár war. On one of his missions to the Porte, Antun Vrančić dropped his diplomat's mask and spoke out in the divan as an aggrieved archbishop of Esztergom: even though Gyula had been lost to the sultan in 1566, if necessary he would assert his proprietary rights over villages in that region and even beyond, as far as Belgrade.¹⁵² The commander of Kanizsa, which had replaced Szigetvár as the anchor fortress for southwestern Hungary, insisted on collecting taxes from villages in the sancak of Pécs, conquered by the Ottomans in 1543.¹⁵³ The next commander at Kanizsa was Juraj Zrinski/György Zrínyi, son of Count Nikola IV Zrinski, commander of Szigetvár's heroic defenders. In keeping with his lawful rights, as he understood them, the younger Zrinski collected taxes from villages that had once been subject to Szigetvár.¹⁵⁴ For their part, Ottoman commanders were not timid about defending lands assigned to them by the Book of Halil Beg, and probably a bit more besides. The sancakbeg of Zolnok sent out one hundred cavalrymen to warn villages that had hitherto paid taxes to both sides that they faced “grave threats” if they did not stop making payments to Habsburg lords. In Slavonia, the northern part of Croatia, Ottoman tax collectors traveled as far as twenty miles beyond the ostensible frontier to inform villagers of their new obligations.¹⁵⁵

For the Porte, the restlessness of border lords was a problem that could be controlled, as noted in the preceding text. These men held their lands by virtue of participation in the sultan's conquering armies; if need be, those who gave the army its orders could rein in their ambitions. For Vienna, the restlessness of border magnates was a problem of a different order: it threatened not only a precarious peace with a more powerful enemy but also the continued survival of the Habsburg Kingdom of Hungary. The magnates held their lands by rights and customs long preceding the advent of Habsburg rule. There had been a time when a Habsburg sovereign could discipline a border princeling and still retain his loyalty.¹⁵⁶ Now, after the Szigetvár war, Maximilian II may not have had the luxury of doing so. Had he commanded his magnates to abandon all hope of recovering what they had lost, these local potentates might have chosen to bargain with the Ottomans for a better deal, as men like them had done in the past.

To avoid another war, Maximilian II had to be seen by the Porte as paying Selim II the deference that a conquering sultan demanded. To avoid further discouragement of loyal magnates, His Majesty had to be seen by his subjects as protecting their rights. The treaty language finally agreed to was in one sense little more than a face-saving fiction; the Ottomans showed not the slightest inclination to relinquish any towns or villages listed in their defter. But it was also a way of satisfying the conflicting imperatives that Vienna faced. Together with the annual tribute, continued from previous treaties,¹⁵⁷ Maximilian II's formal recognition of the Book of Halil Beg, implying a recognition of Ottoman law, satisfied the Porte's demand for deference. The conditions attached to this recognition, implying that rival claims along the border were not definitively settled, may have been enough to show the great families that His Majesty had not abandoned them. After all, both governments accepted the inevitability of continued skirmishing along the border.¹⁵⁸ With this treaty in place, a magnate who asserted his claims might be less likely to provoke a full-scale invasion by the Ottomans, and less likely to incur the wrath of His Majesty. Hence the paradoxical interdependence of peace and war. Maximilian II's Ottoman diplomacy, meant to forestall another major war, provided at the same time a charter for Kleinkrieg.

Toward a New Strategy

In December 1576, Ambassador David Ungnad sent a long circular letter to the Austrian estates. With Maximilian's consent, he had sent a previous report to the estates in January.¹⁵⁹ The letter painted a grim picture of Habsburg-Ottoman relations. Ungnad acknowledged the gravity of recent Ottoman conquests in Croatia, a sensitive matter for nearby Austrian duchies.¹⁶⁰ He referred repeatedly to the failed Habsburg candidacy in Poland-Lithuania to show that Maximilian II had not failed to make a serious effort; rather, the fault lay with the insolence of what he called “the rabble” (pöbl, here meaning the lesser nobility, the lower house of the Sejm).¹⁶¹ He also warned his peers in the estates not to expect any relief from the recently concluded Capitulatio with the Porte: the Ottomans never give back conquered territory, and raids now came almost on a daily basis, even with a treaty in place.¹⁶² In the end, Latin Christians could only trust in God's mercy to spare them from the Untergang that had befallen Greek Christians. After some thirty pages, the ambassador came to the point. Following the failure of his plans for Poland, the dear departed emperor had the wisdom to summon the Reichstag, so as to ask the imperial estates for a recurring contribution to help defend the frontier. The Diet had now approved a generous Türkenhilfe, yet the sums involved would not be sufficient. The Austrian estates must therefore do their part to build a Gegenwehr against the Turks by agreeing to the Defensionshilfe requested by the new sovereign, Emperor Rudolf II.¹⁶³

In 1574, Ungnad wrote to Maximilian that changes to the proposed treaty text achieved most of the sovereign's goals.¹⁶⁴ Now, as part of an argument for higher taxes for the border, he acknowledged that the treaty did nothing to prevent Ottoman commanders from expanding the sultan's dominion, bit by bit. Rudolf II was in the process of moving his capital from Vienna to Prague, leaving his brothers, Archduke Ernst and Archduke Karl, to represent the dynasty's interests in its Austrian lands along the Hungarian border. Ungnad's circular letter was an early sign of a sustained effort by the Habsburg government to mobilize support for the costly, long-term project of hardening the frontier.¹⁶⁵ Upon his return from the Porte in 1578, Ungnad became president of the central organ for coordinating military policy, the Hofkriegsrat in Vienna.¹⁶⁶ After five years of negotiating with the representatives of a superior power, he knew better than most that diplomacy was not enough.