> u φ ʒρ · Ɨь→πʒρ · |ʒρΔωμʒρбьη
B u d a · V e g l a · c a p i t a n e o
Δ = p · б = n · ь = e · ʒρ = a · 8 = f · “quid” = perchè

Pozsony → Ferrara · sign cipher, monoalphabetic with homophones · July–November 1482 · read in part

Sadoleto at Pozsony, 1482: the Venetian counter-offer

Archivio di Stato di Modena, Cancelleria, Ambasciatori, Ungheria b. 1/9 · four ciphered despatches of Nicolò Sadoleto, Ferrara’s envoy to Matthias Corvinus, catalogued in DECODE as partly decrypted (R1101, R1102, R1103, R1106)

The key is rebuilt from two sibling letters that carry contemporary clear copies · two of the four targets turn out to have been deciphered at the time · the seven unread lines of 16 July 1482 are read in gist · one faded block gives only fragments

Daniel Bourdeau · solved in part

Summary. In the spring of 1482 Venice attacked Ferrara, and Ercole I d’Este sent Nicolò Sadoleto to Matthias Corvinus to bring Hungary into the war on the side of the League (Naples, Milan, Florence, the Pope). His despatches from Buda and Pozsony were never printed: the nineteenth-century Hungarian copyists passed them over. DECODE lists four of the 1482 letters as partly decrypted. The images, on DECODE and on the Vestigia database, show a sign cipher with word divisions kept and clear words mixed in. Two sibling letters survive with the Ferrarese chancery’s clear copies beside them (Vestigia 1283 for 15 July and 1295 for 31 August 1482); aligning them sign by sign gives the alphabet, about twenty-five signs with homophones for the vowels and n, r. With it, two of the four DECODE targets turn out to be read already: 15 July has its clear copy (no. 7a) and 27 November has two sheets of decipherment (nos. 26a, 26b). The letter of 16 July 1482 has no clear copy, and its seven cipher lines are read here in gist. Sadoleto learnt in secret that the other side had written to Francesco Quirino at Matthias’s court offering the King Veglia, the command of the fleet for his natural son, and a hundred thousand a year. The faded four-line block of 17 August (the copy of his letter to King Ferrante) gives only fragments. The twelve 1483 letters DECODE also lists (R1107–R1118) all turn out to have contemporary decipherments, as copies or as slips pasted to the leaf.

01 The documents

The file holds forty-seven reports of Sadoleto, June 1482 to November 1483, most in his own hand, with a handful of ducal minutes in reply (Mátyus 2020). Vestigia catalogues all of them with images; DECODE has eighteen of those with cipher. The ones that matter here:

ItemDate, placeFormStatus found
DECODE R1101 · b. 1/9 no. 7 (Vestigia 1284)15 July 1482, Pozsonyfour pages, long cipher passagesno. 7a (Vestigia 1283) is the full contemporary clear copy: read at the time; used as a crib
DECODE R1102 · no. 8 (Vestigia 1286)16 July 1482, Pozsonyclear letter, seven lines of cipher at the foot of p. 2no clear copy: read here in gist
DECODE R1103 · no. 11 (Vestigia 1294)17 August 1482, Pozsonycopy for Ercole of Sadoleto’s letter to King Ferrante; four faint cipher lines on p. 2no clear copy: fragments only
DECODE R1104 · no. 13 (Vestigia 1298)31 August 1482, Pozsonyone page, mostly cipher“Decrypted”: clear copies 13a and 13b (Vestigia 1295, 1297); used as a crib
DECODE R1106 · no. 26 (Vestigia 1318)27 November 1482, Pozsonyfour pages, partly ciphertwo sheets of decipherment, 26a and 26b (Vestigia 1319, 4004): read at the time

DECODE’s “partially decrypted” on R1101 and R1106 reflects that the decipherments are filed as separate items and were not attached to the records; the contemporaries read both letters in full.

Seven lines of cipher signs at the foot of a letter page.
The seven cipher lines of 16 July 1482 at the foot of p. 2. Line 1 reads essendo in Buda per partire per venire a Posonia; the last line per mare, et gli voleva dare ogni anno cento milia. ASMo, Ambasciatori, Ungheria b. 1/9 no. 8; image DECODE (R1102).

02 The key, from two cribs

No Este key for this embassy is known. The archive’s Cifrario series holds a key “per corrispondere con Nicolò Sadoletti a Napoli tra 1479 e 1480”, rebuilt by nineteenth-century archivists (Cremonini 2017, n. 39), but it is not digitised. The key was therefore rebuilt from the letters.

The cipher keeps word divisions, which makes alignment straightforward. The clear copy 7a of 15 July runs “… volesse unirse e pigliare questa impresa, e non lassasse pretermittere questa opportunità laquale mai più fo, né forse sera, e che lo facesse per lo amore …”, and the first lines of the cipher on the original’s second page match it word for word. The one-page letter of 31 August and its clear 13a confirm it: “epsa me dice che è certa che la liga non … perchè Millano et Fiorenza non voriano …”. One habit shows up at once: where 13a has perchè the original writes quid in ordinary letters, which works as a code word hiding in plain sight.

SignValueSignValueSignValue
ʒρ, –1a→, )g3, η, round 6o
>b2hΔ (stem through apex)p
| (plain stroke)cωiΔ with looped apexq
φdπlſ, Γr
ь (6 with a stem), small Γe4m7s
8fб (flat-topped 6)nμ, ρt
u, иuƗ (v with a loop)vʉρet

The table is in sadoleto1482/key_working.md with the aligned words for each sign. It is a plain alphabet with a few alternative signs for common letters; no syllable signs were needed. The nomenclator is small or rarely used: one group in the letter of 16 July does not spell out and is almost certainly a code word (below).

03 Pozsony, 16 July 1482: what Venice offered the King

The letter’s clear part is routine news of the court. The last seven lines are in cipher.

Essendo in Buda per partire per venire a Posonia, sapi per gran secreto da […] Francesco como [X] haveva scripto ad Francesco Quirino, che è qua, fo[sse] a … gli fa[cesse] … como [X] voleva havere seco bona a[micit]ia, et voleva dargli Vegla, et voleva lo figlio (e bastardo) per loco capitaneo de la [arma]ta per mare, et gli voleva dare ogni anno cento milia.

“Being in Buda about to leave to come to Pozsony, I learnt in great secret from […] Francesco that [X] had written to Francesco Quirino, who is here, […] that [X] wanted to have good friendship with him, and wanted to give him Veglia, and wanted his son (the bastard) as captain-general of the fleet at sea, and would give him a hundred thousand every year.”

The anchors are words that could be nothing else: Buda, Posonia, Francesco twice, Quirino, Vegla, capitaneo, per mare, ogni anno, cento milia. The sense fits the moment. Veglia (Krk) had been taken by Venice from Count Giovanni Frankopan in 1480, and the count was at Matthias’s court that summer (13a: “le qui lo Conte Zohane da Vegla”). The “son, the bastard” can only be János Corvin, Matthias’s natural son and intended heir. The sum is the telling detail. Six weeks later, in the letter of 31 August, Sadoleto told Ercole that Matthias “per cento milia ducati ogni anno pigliaria la impresa”, that is, would enter the war for the League for a hundred thousand ducats a year. The cipher of 16 July shows where that price came from: it was the figure the other side had already put on the table.

[X] is a five-sign group that occurs twice, after como both times, as the subject of haveva scripto and of voleva havere seco bona amicitia. It does not spell a word in the letter key, so it is a nomenclator entry, and the context makes it the Venetian side, most likely “la Signoria”. The informant’s title before Francesco is another three-sign group not yet read.

04 Pozsony, 17 August 1482: a faded block

R1103 is the copy, sent to Ercole, of Sadoleto’s letter to King Ferrante of Naples. Four lines of cipher in the middle of p. 2 have faded almost to the colour of the paper. Contrast stretching on the blue channel brings out perhaps half the signs, and they give fragments only: “… se tener che … et … lo credo …”, “… suoi …”, “… vostro … meco … che …”. The clear text around the block concerns the King’s price for joining the League and the hundred thousand ducats, so the block probably says something about that, but no reading is offered. It needs better imaging of the original.

05 The 1483 letters: all read at the time

DECODE lists twelve more Sadoleto letters of 1483 from Buda and Zagreb (b. 1/10, R1107–R1118), several as partly decrypted. None is open. Four have clear copies filed beside them (nos. 4a/4b, 20a, 21a, 24a on Vestigia); two of the DECODE records are those copies themselves (R1108 = 9a, R1113 = 20a, R1115 = 24a); and the three letters with no filed copy carry the chancery’s decipherment on slips pasted to the leaf: R1109 (4 April, to King Ferrante) over both cipher pages, R1116 (Zagreb, 3 October) and R1118 (30 November) at the foot, keyed to the cipher by a caret. The slips show on DECODE’s photographs; on Vestigia’s later ones they have been lifted, which is why the pages look unread. The 1482 key reads the cipher of R1118 in agreement with its slip (“ma come scripsi a V.S. non gli [s]panza questo accordo col turcho”).

06 Method

All images are 4320 × 3240 two-page spreads, fetched from DECODE with a logged-in session and from Vestigia, which needs no login. The cipher lines were cut into strips, glyphs segmented by connected components and numbered on the strip (the segmentation script is the one written for the Milanese letters of 1489, see Buda 1489), and read by eye against the clear copies. The pairs of signs most easily confused were settled on high-magnification crops: the two Δ forms (p and q, which differ only in the loop at the apex), the stemmed and flat-topped sixes (e and n), and the long ſ against Γ.

07 What remains uncertain

08 Sources

The key with the aligned words, the transcription of 13a, the glyph-by-glyph reading of 16 July, the DECODE and Vestigia record data, and the segmentation scripts are in sadoleto1482/ in the repository. The archive images are not redistributed; the one crop on this page is shown for identification.