The viceroy of Sicily to King Ferdinand, Messina, 27 April 1503: the cipher rebuilt and read in large part

Document

Paris, BnF, ms. Espagnol 318, no. 94, ff. 120r-121v (Gallica btv1b52503046q, views 452-455). Juan de Lanuza, viceroy of Sicily, to King Ferdinand, Messina, 27 April 1503.

New

A first reading. The key is rebuilt from the letter itself. It is a reading in large part, not a full one: 64 code words have no value. We found no earlier reading and no printed text.

Not seen

The original key ("Cifra del visorrey", Real Academia de la Historia, 9/15 ff. 1-6) is not online and was not seen. The published keys of the Catholic Monarchs do not fit.

The key

A symbol alphabet with several signs for each letter (44 signs for 20 letters), and code words of two to four letters for common words and names. Two code words stand for letters: otto is c, and malz is ll. The names of the signs are ours. The image of a sign is cut from the three lines below.

pa221 timeso-a121 times3a47 timesaotb24 timesbFc28 timescPd108 timesNd35 timesdxe181 timestbe121 timestboe18 timeseGg29 times5g9 timesEg1 timesg8h28 timesh&i89 times:.i58 timesZi9 timesiHl47 timeskl29 timeslC.m36 timesaom12 timesmssn171 times5on55 timesAn7 timesnRo163 timesalo67 timesXo29 timesoBp41 timesdlp31 timespVq21 timesqIIIr118 times=r61 timesMr15 timesr7s126 timesffs42 times9s18 timess6t89 times+t11 times9tt2 timestau112 timesßu47 timesuQQy27 timesQy2 timesy3oz15 timesz

In each box: the sign, our name for it, its letter, and how many times the letter has it. A box with no image: the sign is not in the three lines below. The shapes in words: transcription/SIGNS.md.

The code words of the three lines:

All 36 code words with a value, and the 64 with none: KEY_TABLE.md. The key as a table for the decoder: key.py.

Worked example

Each box holds one token of the first transcription pass: its image, its name, and its value by the key. The line under a group of boxes gives the word; the word division is ours.

f. 120r, line 1, tokens 13 to 33: after the rout

klauxe5galoluegoNdxedlpauxe7sdepuesmogdeldelNdtbe7sotb3aIIIrpa6ttbedesbarate

luego depues del desbarate

The viceroy wrote right after the "desbarate" (the French defeat near Gioia, 21 April 1503).

The line begins with code words that have no sure value; the example starts at the first sign of "luego". The two transcription passes have the same token in 20 of these 21 places.

f. 120r, line 7: the captains in the fortress

maselelpaffstbeGgau=rpaPdalokltbe7saseguradolesrazquequeBpauxe7s6talo7spuestosmysenenmublalapaaloIIIr6tpakltbe3ozo-aaortalezapaaPdalo5ondon&ialopa5onioanmuydede/stop 

el, asegurandoles que, puestos en la fortaleza a don Joan de

Don Juan de Cardona, Antonio de Leyva and Carvajal were put in the fortress so that an inventory could be made and a good account given.

Two letters differ. The key gives aseguradoles; the reading file writes "asegurandoles". The key gives aortaleza; the reading file writes "fortaleza": the transcription has one name, p, for the sign of a and the sign of f, and the reading restores f by context. The name "Cardona" follows in line 8. The two transcription passes have the same token in 41 of these 42 places.

f. 121r, line 21: the threat to don Hugo

namno value qayyaomxemenymhanhanPd:.iottoc8hRodicho7s6taußu:.ixeIIIrRossnstuuieronpetparaparamokdardarHlxelemuydedeBpaussnQQyo-aHlpaPdo-a7spunyaladas

y me han dicho stuvieron para darle de punyaladas

Some knights told him they came near to stabbing don Hugo.

The first token is a code word with no value. The two transcription passes have the same token in 34 of these 34 places.

How sure

Outside check. The names agree with the chronicles of the Great Captain; they were not used as a crib. Zurita (Historia del rey don Hernando, book 5, ch. 25 and 29) used the same matter: he has "Hernando de Valencia", "Alonso Guerrero", the anger of the Cardonas, and Gioia "puesto a saco, y quemado". He does not have the sack of the fortress by night, the inquiry, or the threat to stab don Hugo. A search of the printed sources found no clear text of the letter.

What is not resolved

• 64 code words have no value. They stand mostly in the stretches that the reading marks with "...".
• The transcription does not separate the sign for f from one sign for a. The letter f was restored by context.
• Three signs that look alike were confused by the passes. A third pass can change single words.
• A line that we do not show for this reason: f. 120r, line 12, where the reading has "la camara de mosse de Aubeni". The two passes give clmara and mosae or mousae there. Both passes are in DECODE_RAW.md.

Edited reading and English summary: READING.md. Folder: espagnol-318.