Using Script against undesirable Readers III: Vrančić’s Coding Aids

Među rukopisima Antuna Vrančića u Nacionalnoj knjižnici Széchényi u Budimpešti nalazi se pet priručnika za kodiranje, svaki na jednom listu papira. Riječ je o četirima službenim dokumentima što ih je pedesetih godina 16. st. Vrančiću dodijelio bečki dvor i o jednom Antunovu pomagalu za osobnu upotrebu. Svih pet načinjeno je po istom obrascu: sastoje se od šifriranog alfabeta i dodatnih šifriranih riječi; jedino Antunov privatni priručnik nema dodataka. Najzanimljivijima su se pokazala četiri popisa kodiranih riječi, broj kojih postupno raste, od pedesetak u najstarijem sustavu do više od dvije stotine u posljednjem. Oni sadrže najvažnija imena i pojmove austrijske vanjske politike onoga vremena. Najstariji i najmanji predan je Vrančiću i Zayu prilikom njihova odlaska u Tursku 1553. godine, drugi im je dodijeljen tri godine kasnije u samom Carigradu, treći je povjeren Antunu kratko nakon njegova povratka iz Turske 1557. s ciljem izvješćivanja budućega češkoga kralja Maksimilijana o prilikama u Transilvaniji, četvrti je sastavljen negdje u isto vrijeme, oko 1558, a bio je vjerojatno namijenjen Vrančićevoj misiji na Zapadu (do čega međutim nije došlo). U radu smo pokušali identificirati sve natuknice s popisâ. U njima se zrcale glavni akteri i problemi s kojima se suočavala austrijska grana Habsburgovaca, od osmanlijske opasnosti preko erdeljskog pitanja, protestantizma i papinstva do španjolsko-francuskog rata u Italiji, od Livonije na sjeveru do Sicilije na jugu te od Engleske do Perzije. U prilogu studiji objavljuje se prvo izdanje svih pet Vrančićevih priručnika za kodiranje.Coded letters in the 16th century were common in European diplomatic communication. Much involved in this practice was the Croatian-Hungarian humanist Antun Vrančić (1504-1572), as diplomat first in the service of the Transylvanian and then from the middle of the century of the Austrian court. Vrančić’s secret correspondence with the court of Vienna and with his brother Mihovil has come down to us (for the latter, cf. papers of M. Lončar and D. Sorić in CM XXII and CM XXIII). But in the Vrančić Papers in the Hungarian Széchényi National Library five of his coding aids are kept, four of them official and one private. The official manuals were given him by the Viennese court on various occasions, and he himself compiled the other for correspondence with his brother Mihovil. We have named them from the marks on them (or from their want of any mark). 1. Older codes of Antun Vrančić and Ferenc Zay 2. Newer codes of Antun Vrančić and Ferenc Zay 3. Codes of King Maximilian 4. Codes with no mark of belonging 5. Codes of Mihovil Vrančić All four official code books are composed in a similar way. They consist of obligatory sections: alphabets, doublet characters, empty characters (used for duping uninvited readers), declinations of the expression Maiestas Vestra and a list of various words – all with pertaining codes; bigger manuals also have the occasional added section. Only the Codes of Mihovil Vrančić contains no list of various words. Characters of the Latin and Greek alphabets are usually used as a secret script, either »pure« or with additions (lines, circles, crosses), in addition to Arabic numbers or particular characters.The list of coded words contains a selection of the most important names, titles and terms of Austrian foreign policy of the 1550s. The number of words on the list increased over the course of time, from fifty or so in the Older Codes to more than two hundred in Codes with no mark of belonging. The purpose was to make unauthorised decoding more difficult, for even if the actual alphabet were decoded, the most important concepts would remain concealed behind their own codes. 1. The Older Codes of Antun Vrančić and Ferenc Zay is originally entitled Ziffra, quae prior data fuit. Verantii, et Zaii. The title enables it to be dated quite simply, since Vrančić and Zay ran Ferdinand’s mission in Istanbul from 1553 to 1557. The manual must have been handed to them before they left for their journey. It is interesting that this manual was withdrawn from service in March 1557. It was also possessed by Ferdinand’s confidant Dominik de Gaztelu, but after his death it got into the hands of some secretary of his and was considered thus unreliable and replaced with another. It is uncommon that mistakes should have made their way into the alphabet. First of all the character l was missed out, and then added at the end of the table. At the bottom of the list is one more alphabet with a reduced number of characters for coding. But in it, the letter s is missing. The list of words in this manual is divided, as it is in the others, into several recognisable units. The first twelve places are occupied by persons from Ottoman political life (Imperator Turcorum, Rex Franciae, Hruztanus Passa...) and the most important Turkish cities (Constantinopolis, Adrianopolis, Belgradum...). It is interesting that the king of France should have been included into the Ottoman group. At this moment Henri II was on the throne (1547-1559), with whose father, Francois I, Ferdinand’s brother Charles V (1519-1558) had been at war for almost the whole of the first half of the 16th century for the sake of European hegemony. Henri II, following in his father’s footsteps, was allied with Suleiman the Magnificent (1520-1566). That the Older Codes was created expressly for the Istanbul mission can be concluded from the first part of the list being taken up with a Turkish circle of persons and cities. Then comes an equally large Austrian circle (Imperator noster, Rex Romanorum, Rex Poloniae..., Cracouia, Silesia, Hungaria...)in which Transylvania makes up a smaller circle (Regina Isabella, Reginae Isabellae filius, Petrus Petrowyth...); in it Austrian interests are opposed to local and Turkish. Alternating to the end of the list are two Austrian sets of concepts, one Turkish and a military political and Venetian. All these things are of recognisably great interest to the Austrians and so it was necessary to code them separately. 2. Newer Codes of Antun Vrančić and Ferenc Zay must be the codes that replaced the Older. It contains a non-standard section, that is a code for the name of Ferdinand’s confidant Augier Ghislain de Busbecq (1522-1592), who travelled to Istanbul in 1554 and once again in 1556. He was supplied with a coding manual superioribus annis (Ferdinand’s expression from a letter to Vrančić and Zay, 1557),which would mean – before he set out on his journey. This dating is confirmed by Bayazet and Selim (Sultan, 1566-1574) being on the list, and Mustapha, who was murdered on 6 October 1553, not (this date would be the terminus post quem). In addition, Haydar-Pasha, found on the previous list, was pensioned off in 1552 and was replaced in the Newer Codes by Sinan-Pasha, which would point to these codes being considerably later than the Older. But a confirmation of the hypothesis that this is a manual that was created instead of the Older can be found in the fact that letters from the Ottoman capital that are written precisely in these codes are extant. The list of concepts in the Newer Codes, as well as being more than half as long again, has in the first twenty places the names of persons and geographical areas from the Austrian and Western circle. This is a considerable difference from the older list, where the first part is occupied by the Turkish circle. (The reason for this might be that the Newer Codes was given to Vrančić and colleague at a later date, and improvised; it was really a list of Busbecq’s codes, which were almost certainly made for other purposes.) The beginning belongs to the Habsburgs, German princes, the Vatican, the king of Poland and Philip II of Spain. Judging from this, the main purpose might have been to track Protestantism and actions against it, for this is what links the first twelve entries. Many of the German princes had embraced the new religious movement, which enjoyed toleration even in Poland. Charles V and, still more, his son Philip were fierce opponents of it and undertook punitive expeditions against their heretical subjects, the most ardent urgings coming from Rome. (Listing the Roman Curia was an important novelty.)The Austrian branch of the dynasty was a little more permissive, particularly Maximilian (1564-1576), which in 1555 resulted in the Augsburg Settlement at which the celebrated principle cuius regio, illius religio was established. If the codes were not actually devised because of Lutheranism, the main orientation was nevertheless certainly Westward-looking. In the Turkish circle, among the twelve entries the most important appearance was made by the wife of the Turkish leader, in the third place. This is in line with the stories about the huge influence exerted on Suleiman by Roxelana. In several small thematic series in the sequel the Austrian circle of lands and cities, military-political ideas are listed, then the Turkish circle of lower-ranked persons and some other terms. 3. Codes of King Maximilian was – judging from the list of various concepts – meant for a more closely defined purpose, that is, the monitoring of the situation in Hungary and Transylvania. The Austrian Habsburgs wanted to put into effect their rights to Transylvania, which, out of fear of them, had accepted the embrace of the Ottomans. Suleiman’s intention was to advance on Hungarian ground, which in the middle of the forties and fifties had split with his military wedge into east and west. The composition of this codebook can almost certainly be dated to the last months of 1559, since in the list of words there are names of Transylvanian magnates Ferenc Bebek and the Kendy brothers, who were killed on September 1, 1558. Also given is the name of Queen Bona of Poland, although she had died in autumn, 1557, but seemed nevertheless to have stayed on the list from the two previous lists thanks to the inertia of the court administration. The year 1558 fits in well with Vrančić’s biography. At the end of the previous year he had returned from his four-year mission in Istanbul. He was immediately after that appointed bishop of Eger. From there he would have been able to give first-hand information about the situation in Transylvania, with which Maximilian was clearly particularly charged. A particular feature of the Maximilian Codes is that it is primarily concerned with Hungarian and Transylvanian geography, with somewhat fewer than twenty entries. After them comes an equally large Turkish block of peoples and supremos, and then a block of Transylvanian magnates. To the end there are some other military and political concepts and a set of rulers from the Austrian sphere. 4. Codes with no mark of belonging is the largest; this is primarily the result of a list of some two hundred ranks, names, titles and concepts. It can be seen how much the catalogue has been enlarged since the previous codebooks: it is three times as big as the Codes of King Maximilian and the Newer Codes, and almost fivefold the Older. It seems then that the system worked well and was being constantly updated, the reason for which was for certain the greater security of what was written. Codes with no mark of belonging is also the youngest version. At the end of the list we can find codes for 1560-1567. We think that this indicates the period meant for the use of the book. This in turn would indicate that it was composed in 1560 or just before, but certainly after 1558, since the first entry is for Cesarea Maiestas, which must refer to Ferdinand I, and he was crowned Holy Roman Emperor in 1558 (reigning until 1564). Since almost the first forty places in the list are given to the West and only the next score to the East (with the Turks heading it), it seems that this system of codes was not meant for Vrančić’s reports from the two diplomatic missions in Istanbul, which is confirmed by the timing being wrong. Vrančić’s first mission to the Bosporus lasted from 1553 to 1557, which is too early, and the second from 1567 to 1568, which is again out of the range in the other direction. But in 1559 Ferdinand had invited Antun to Augsburg and prepared him to undertake missions to England and Rome. This journey did not in the end come off, but it might have been the occasion for which Vrančić obtained this manual. Finally, on the list immediately after the immediate Habsburg circle and the most influential German estates came Regina Angliae – in other words, Elizabeth I (1558-1603).The high place that she occupied on the list is certainly to do with negotiations about her marriage to the third son of Ferdinand, Archduke Charles. These talks were being held precisely in 1559 (and again in 1564 and 1568). The Vatican was also at the top of the list. Politically perhaps the most important new departure was that those in power in the cities of North Italy and Naples and Sicily were included. This is clearly the consequence of the war that was waged in Italy between Philip II on the one hand and Pope Paul IV (1555-1559) and Henri II on the other, ending with the French signing their defeat at Cateau-Cambrésis in 1559. Also included in the list is the victorious general Emmanuel Philibert of Savoy (entitled Dux Sabaudiae). The rest of the list contains the usual subject units. As in the previous lists we first tried to identify all the entries, and then to understand why they were chosen, and why they were listed in the given order and what changes had taken place since the previous lists. 5. Codes of Mihovil Vrančić. Antun Vrančić used his professional skills for private purposes as well. He composed two copies of an aid to be used in his correspondence with his closest brother, Mihovil. The specimen preserved must all in all be that which he used for reading his brother’s letters. An edition for all the five coding manuals has been published by way of appendix to this paper. They have thus become accessible to researchers into European cryptography of the early Modern Period