Summary. János Pápai served as Ferenc Rákóczi II’s envoy at the Ottoman Porte, at first together with Ferenc Horváth. DECODE holds seven of his ciphered despatches, dated 1 May 1706 to 8 May 1710, an eighth filed as from an unknown sender (R784, 1707), and four more of January–March 1706 filed as from “Constantinople envoys” (R793–R796). The Hungarian National Archives also keep the key issued to him: DECODE R580, headed Nemzetes Vitézlő Pápai Jánosnak adott Clavis, with two further copies (R581, R452). Nothing on DECODE connected the key to the letters. Applied unchanged, it gives values for 16,346 of 16,861 groups in eleven letters: R795, R796, R793, R794, R805, R823, R765, R784, R757, R741, and the cipher passage of R740, which was transcribed here from the image; 306 further groups in R793–R796, above the table’s range, are nulls that the contemporary interlinear decipherer passed over. The text is continuous Hungarian, with Latin in the audience report of December 1706. The letters deal with Imperial troops crossing Ottoman ground, audiences with the Vezír in the presence of the Imperial internuncio, the French ambassador’s failure to bring the Porte to war, the French and Muscovite ambassadors, and money. A second pass splits run-together groups, re-joins split ones, treats every 3-digit group above 322 as a null and gives the unlisted 220 the value a. That brings the numeric letters to 16,641 of 16,698 groups (99.7%); the 57 left are digit fragments no key value fits. The twelfth letter, R731 (Belgrade, 6 March 1710), hides six short passages in a graphic alphabet. They were transcribed tentatively and attacked with a substitution solver that recovers a control exactly, and they gave no Hungarian. Benda’s 1963 edition of Pápai’s Turkish diaries is not online.
01 The documents
| Record | MNL OL G15 | Date | Groups | In key |
|---|---|---|---|---|
| R795 | C 36, pp. 18–21 | 1 Jan 1706, Constantinople | 1,617 | 1,356 + 251 nulls |
| R796 | C 36 | 20 Feb 1706, Constantinople | 1,025 | 1,004 + 16 nulls |
| R793 | C 36 | March 1706 | 1,226 | 1,194 + 24 nulls |
| R794 | C 36 | early 1706, undated | 1,175 | 1,152 + 15 nulls |
| R805 | C 33, pp. 34–37 | 1 May 1706, Constantinople | 336 | 336 (100%) |
| R823 | C 33, pp. 140–143 | 19 Dec 1706, Constantinople | 1,632 | 1,609 (98.6%) |
| R765 | E 80, pp. 2–11 | 5 Jan 1707 | 4,864 | 4,787 (98.4%) |
| R784 | D 80, pp. 110–111 | 1707, unsigned | 594 | 587 (98.8%) |
| R757 | E 109, pp. 85–90 | 16 May 1708 | 2,758 | 2,695 (97.7%) |
| R740 | H 226, pp. 85–88 | 2 Jan 1710, Constantinople | 168 | 168 (100%) |
| R741 | H 226, pp. 89–96 | 8 May 1710, Constantinople | 1,466 | 1,458 (99.5%) |
| R731 | H 226, pp. 31–33 | 6 Mar 1710, Landor Fejérvár (Belgrade) | graphic signs, not read | |
The group counts are what the decoder parses from DECODE’s manual transcriptions. Groups without a value are mostly one-digit slips. R823 also opens with a few sums in plain figures (400–1000), which are not cipher.
02 The key issued to Pápai
DECODE R580 (G15 Caps. C Fasc. 43/39) is a Hungarian syllabic nomenclator with values from 10 to 322:
- Letters, two homophones each: A 219|229, B 209|239, E 179|269, N 100|304, T 79|309, Z 60|312. The same pattern covers ö, ü and the digraphs ny, ly, gy, ty, cz, sz, nk and rt.
- Syllables in columns by final digit: 11 ab, 21 ad … 192 ge, 245 va … 236 üz.
- Names: 246 Király, 256 Fejedelem, 257 Fényes Porta, 267 Török Császár, 277 Vezér, 287 Bassa, 297 Bécs. Countries run from 17 to 207; towns, rivers and months from 18 to 298.
The letters have no word division and no nulls. DECODE had R805 as “partially decrypted” from a few interlinear syllables. The first test of the key on R805 gave 336 of 336 groups as running Hungarian, and the other letters followed without any change to the key.
03 What the letters say
The key gives continuous text. Word breaks and accents below are editorial, and every value is grade H (a direct table value). The full decoded text of each letter is in the repository.
- R805, 1 May 1706. The Porte should refuse the Germans passage across Ottoman ground: az németnek szabados általmenetelt ne engedgye … az Aradi Bassának ordere nélkül meg ne engedgye az németnek Szeged felé való általmenetelt. The envoys have also spoken to the Vezír about Muscovy.
- R823, 19 Dec 1706. An audience in the presence of the Imperial internuncio, reported partly in Latin: praesentia autem vestra non est ingrata Portae. The hereditary-kingship argument follows: legibus imperatoris hereditarii regis sunt Angliae, Daniae, Sveciae, et tamen legibus subsunt, with the envoys’ answer that Hungary will not give up its old laws.
- R765, 5 Jan 1707. Submission to the Prince’s will and the state of Poland. On the Prince’s order the envoys went to the French ambassador: a Francia Ország orátorához elmentünk.
- R795, R796, R793, R794, January–March 1706 (added 21 Sept 2026). Catalogued as two ciphertexts of “Constantinople envoys”, partly deciphered interlinear at the time; they are four despatches of Pápai, decrypted in full with his key. R795, signed Constancz die prima Januarii Ezer het szaz hat, Pápai János, written with Ferenc Horváth, reports the Prince’s discretion money handed to the Vezír’s Kiaya, the Tatar Khan, the Pasha of Temesvár and a new route for letters through Moldavia. R796 (20 February) answers the Prince’s letter from the Tisza: the Germans spread false news, the confederates are scorned, the Muscovite envoy is at the Porte and nothing is resolved on the propositions. R793 (March) reports secret talks with the Porte’s chancellor on the Muscovite quarrel, the King of Sweden and a Hungarian confederation with the Porte certis conditionibus, and the Prince’s manifesto translated and given to the Vezír. R794 (undated) concerns recruiting horse in Wallachia and the French King’s money, which the ambassador will not pay without the King’s order. In R795 every group above 322 is a null.
- R784, 1707 (added 21 Sept 2026). Catalogued as “unknown sender to Rákóczi”. Unsigned and undated, but in Pápai’s key, which decrypts it unchanged. A Francia Ország orátora jelentette … hogy semmi ratiókkal nem inducálhatta a Fényes Portát az hadakozásra: the French ambassador reported that no arguments could bring the Porte to war. The grandees of the last war have sworn to see that the Sultan resolves on it, and the whole Turkish nation leans that way. France means to turn Turkey wholly against the Germans (ezt a Török Országot egészen a német ellen fordítsa), from which Hungary would draw harm rather than profit. The ambassador has written the same to the Tatar Khan, and if Muscovy and Poland do not second it, the Germans will overrun Hungary. A note on p. 2 in another hand asks the Prince to keep the letter secret and decipher it himself (… titokban tartani s maga decifrálni).
- R757, 16 May 1708. A message from the Muscovite ambassador, and Pápai explains why he cannot send money from Constantinople.
- R740, 2 Jan 1710. The Kiaya’s answer, which is the only enciphered passage: legyen patientia, [ne] siessen, várakozzék … ha a Fényes Porta szívére nem veszi … ne kellyen illy haszontalan költeni s alkalmatlankodni Felségednek a Fényes Portán. That is, the Prince should be patient and stop spending money and effort at the Porte to no purpose.
- R741, 8 May 1710. A visit to the Defterdar. The Pasha of Belgrade’s letter had concerned only the merchants’ business.
04 What remains open
- R731 is signed Pápai János at Belgrade, 6 March 1710. It is in clear apart from six short passages in a graphic alphabet: three on p. 1 and three on pp. 2–3, about 330 signs. They were transcribed here at 1.8× zoom (
R731_signs.txt). The transcription is tentative: the low swash stroke may be a pen link rather than a sign, and the dots above and below the bars may be vowel marks. A swap-annealing solver with a 4-gram model of Pápai’s own Hungarian recovers a same-length control exactly, but on R731 it finds no Hungarian under any transcription variant tried. So either the transcription merges signs or the dots carry letters, and the camera images cannot settle which. The row of signs in the corner of the key copy R581 is ciphertext of the same kind, not an alphabet. - 57 groups in the eleven numeric letters still have no value (
decode2.py,retry_groups.tsv). Scored in context against every key value, none wins clearly. About 41 more tokens are marked “?” in DECODE’s transcriptions. - Prior print. Kálmán Benda’s edition of Pápai’s Turkish diaries (1963) was not checked for these despatches.
Image rights. No manuscript image is reproduced. DECODE states: “The image is not in the public domain. Publishing it is only possible with the permission of the archive.”
05 Sources
- DECODE records R731, R740, R741, R757, R765, R784, R793–R796, R805 and R823. The transcriptions were made by DECODE transcribers in 2020.
- The key: DECODE R580 and R581 (MNL OL G15 Caps. C Fasc. 43/39–40), and R452 (Ráday Archives C64-4d2-25/11).
- Kálmán Benda, ed., Pápai János törökországi naplói (Budapest: Szépirodalmi, 1963).
Decoder, key transcription and decoded text of every letter: papai1706/ in the repository. Manuscript: Magyar Nemzeti Levéltár. Images not reproduced.