I added some reconstructed ciphers in "Cardinal Mazarin and Ciphers". Thus far, the article mentioned "BnF fr.23202, which I have not seen" at one place, but recently I noticed it is available online.
One of the reconstructed ciphers is used in many letters in the period overlapping Mazrin's two exiles (1651, 1652).
11/02/2024
Reading Scrolls Carbonized in Ancient Vesuvius Eruption without Unwrapping
PC's newsfeed made me aware of deciphering of text on carbonized ancient scrolls without unwrapping (NBC News). (This "deciphering" has nothing to do with cryptography.)
A whole library of an ancient villa survived centuries under the earth, carbonized by the heat of the volcanic eruption of Mount Vesuvius in AD 79, which engulfed Pompeii and Herculaneum, where the villa was located. The library contained more than 1800 papyrus scrolls, of which some were presented to Britain and France in the nineteenth century (Wikipedia).
Early attemps to read the content were a destructive study, breaking the carbonized scrolls into pieces ("The Library of the Villa dei Papiri at Herculaneum", reviewing David Sider (2005), The library of the Villa dei Papiri at Herculaneum; "Twelve Books at Herculaneum That Could Change History" by Richard Carrier).
In 2011, Brent Seales et al. reported "virtual unrolling" of a Herculaneum scroll (Brent Seales et al., "Analysis of Herculaneum Papyri with X-ray Computed Tomography" (NDT, Semantic Scholar)). Virtual unrolling or virtual unwrapping is a non-destructive modality and begins by scanning a scroll with X-ray tomography (as in a CT scan) to produce a 3D volumetric image of the scroll, in which single layers are identified by intensive manual labour ("segmentation"). Once layers are segmented, they can be mapped onto planar images.
At this stage, Seales could not detect any text because carbon-based ink could not be contrasted from carbonized papyrus with his X-ray scan. In 2016, Seales succeeded in revealing text of a scroll found at En-Gedi, Israel, that had been charred in a fire, in which the ink contained lead readily identifiable with X-rays.
Detection of carbon-based ink was achieved by using X-ray phase-contrast tomography, whereby a slight difference in thickness caused by the presence of ink can be detected in phase difference of X-rays. With this method, successful decoding of some fragments of two Herculaneum scrolls (PHerc. 375 and PHerc. 495) was reported in 2016 (Bukreeva et al. (2016), "Virtual unrolling and deciphering of Herculaneum papyri by X-ray phase-contrast tomography" (Scientific Reports), Stabile et al. (2021), "A computational platform for the virtual unfolding of Herculaneum Papyri" (NIH)).
Seales' team adopted an approach that combines scanning with high-energy X-rays with identifying ink by machine learning trained with image data with visible text in ink (The Guardian).
Seales' work inspired the Vesuvius Challenge, launched in March 2023 by Nat Friedman, Daniel Gross, and Brent Seales. It offered a Grand Prize for recovering 4 passages of 140 characters as well as smaller prizes for contributions on the way. It was reccognized that the task was not easy, and the organizers hired a segmentation team to manually identify and label the papyrus surface in the volumetric data (PHerc.Paris. 4) and provide the flattened segments as an open source. It led to discovery of "the first directly visible evidence of ink and letters" by Casey Handmer (his blog) and close collaboration between the in-house segmenters and contestants drove the work (Vesuvius Challenge 2023 Grand Prize awarded).
In October 2023, Luke Farritor won First Letters Prize, a progress prize that required finding at least 10 letters in a 4 cm2 area. He used machine learning to find ink patterns. Detected patterns were fed back for training the machine learning model, which thereby learned to detect letters that cannot be recognized with his eyes. In his discovery, papyrologists on the organizing team could immediately recognize a word "porphyras" (First word discovered in unopened Herculaneum scroll by 21yo computer science student). Youssef Nader won a second-place prize for independently finding the same word shortly later.
In February 2024, the Grand Prize was awarded to a team of Youssef Nader, Luke Farritor, and Julian Schilliger. They achieved much more than required for the prize, revealing more than 2000 characters in total.
Still, about 95% of the scroll are yet to be read, and the Vesuvius Challenge offers further prizes for 2024!
A whole library of an ancient villa survived centuries under the earth, carbonized by the heat of the volcanic eruption of Mount Vesuvius in AD 79, which engulfed Pompeii and Herculaneum, where the villa was located. The library contained more than 1800 papyrus scrolls, of which some were presented to Britain and France in the nineteenth century (Wikipedia).
Early attemps to read the content were a destructive study, breaking the carbonized scrolls into pieces ("The Library of the Villa dei Papiri at Herculaneum", reviewing David Sider (2005), The library of the Villa dei Papiri at Herculaneum; "Twelve Books at Herculaneum That Could Change History" by Richard Carrier).
In 2011, Brent Seales et al. reported "virtual unrolling" of a Herculaneum scroll (Brent Seales et al., "Analysis of Herculaneum Papyri with X-ray Computed Tomography" (NDT, Semantic Scholar)). Virtual unrolling or virtual unwrapping is a non-destructive modality and begins by scanning a scroll with X-ray tomography (as in a CT scan) to produce a 3D volumetric image of the scroll, in which single layers are identified by intensive manual labour ("segmentation"). Once layers are segmented, they can be mapped onto planar images.
At this stage, Seales could not detect any text because carbon-based ink could not be contrasted from carbonized papyrus with his X-ray scan. In 2016, Seales succeeded in revealing text of a scroll found at En-Gedi, Israel, that had been charred in a fire, in which the ink contained lead readily identifiable with X-rays.
Detection of carbon-based ink was achieved by using X-ray phase-contrast tomography, whereby a slight difference in thickness caused by the presence of ink can be detected in phase difference of X-rays. With this method, successful decoding of some fragments of two Herculaneum scrolls (PHerc. 375 and PHerc. 495) was reported in 2016 (Bukreeva et al. (2016), "Virtual unrolling and deciphering of Herculaneum papyri by X-ray phase-contrast tomography" (Scientific Reports), Stabile et al. (2021), "A computational platform for the virtual unfolding of Herculaneum Papyri" (NIH)).
Seales' team adopted an approach that combines scanning with high-energy X-rays with identifying ink by machine learning trained with image data with visible text in ink (The Guardian).
Seales' work inspired the Vesuvius Challenge, launched in March 2023 by Nat Friedman, Daniel Gross, and Brent Seales. It offered a Grand Prize for recovering 4 passages of 140 characters as well as smaller prizes for contributions on the way. It was reccognized that the task was not easy, and the organizers hired a segmentation team to manually identify and label the papyrus surface in the volumetric data (PHerc.Paris. 4) and provide the flattened segments as an open source. It led to discovery of "the first directly visible evidence of ink and letters" by Casey Handmer (his blog) and close collaboration between the in-house segmenters and contestants drove the work (Vesuvius Challenge 2023 Grand Prize awarded).
In October 2023, Luke Farritor won First Letters Prize, a progress prize that required finding at least 10 letters in a 4 cm2 area. He used machine learning to find ink patterns. Detected patterns were fed back for training the machine learning model, which thereby learned to detect letters that cannot be recognized with his eyes. In his discovery, papyrologists on the organizing team could immediately recognize a word "porphyras" (First word discovered in unopened Herculaneum scroll by 21yo computer science student). Youssef Nader won a second-place prize for independently finding the same word shortly later.
In February 2024, the Grand Prize was awarded to a team of Youssef Nader, Luke Farritor, and Julian Schilliger. They achieved much more than required for the prize, revealing more than 2000 characters in total.
Still, about 95% of the scroll are yet to be read, and the Vesuvius Challenge offers further prizes for 2024!
10/02/2024
Encryption for Security of Satellites
Thousands of satellites are orbiting the Earth. Naturally, they are wirelessly controlled with signals from a terrestrial station. Malicious control signals from an unauthorized entity might result in serious consequences ("Cybersecurity Principles for Space Systems" (2020)). Frequencies used in communication with satellites are not disclosed and messages are encrypted. But hackers might access such information, or might simply hack the ground station (車も衛星もハッキングされる時代！？注目が集まる衛星のサイバーセキュリティ (2021)). There are reported instances of actual hacking of satellites ("Hack a Satellite while it is in orbit" (2007), "AsiaSat accuses Falungong of hacking satellite signals" (2004) cited in Wikipedia).
CCSDS (Consultative Committee for Space Data System), an international standardization body since 1982, has a security work group, which has issued documents such as "Green Book on use of security in CCSDS" and "CCSDS Encryption Algorithms and authentication algorithms" among others (CCSDS Overview by NASA, p.14).
The encryption scheme specified for controlling satellites is, as expected, AES (Advanced Encryption Standard), a symmetric block cipher adopted for the US government in 2001 as a replacement for DES (Data Encryption Standard). The Green Book "CCSDS Cryptographic Algorithms" (2023) prescribes "AES is the sole symmetric encryption algorithm that is recommended for use by all CCSDS missions and ground systems." (The same expression is found at least as early as the 2014 version, but not in a 2012 Blue Book).
What was used before AES was recommended by CCSDS? One might think it was DES, but actually, "at first security was thought of as not required for civilian space missions by CCSDS", according to the 2008 Green Book "Encryption Algorithm Trade Survey". Indeed, one patent document JP 2000-341190 A1 says communication with satellites needs to be encrypted "for artificial satellites for specific purposes." Even today, researchers point out security measures for satellites are no more than "security by obscurity" ("Satellites Are Rife With Basic Security Flaws" (2023)).
CCSDS (Consultative Committee for Space Data System), an international standardization body since 1982, has a security work group, which has issued documents such as "Green Book on use of security in CCSDS" and "CCSDS Encryption Algorithms and authentication algorithms" among others (CCSDS Overview by NASA, p.14).
The encryption scheme specified for controlling satellites is, as expected, AES (Advanced Encryption Standard), a symmetric block cipher adopted for the US government in 2001 as a replacement for DES (Data Encryption Standard). The Green Book "CCSDS Cryptographic Algorithms" (2023) prescribes "AES is the sole symmetric encryption algorithm that is recommended for use by all CCSDS missions and ground systems." (The same expression is found at least as early as the 2014 version, but not in a 2012 Blue Book).
What was used before AES was recommended by CCSDS? One might think it was DES, but actually, "at first security was thought of as not required for civilian space missions by CCSDS", according to the 2008 Green Book "Encryption Algorithm Trade Survey". Indeed, one patent document JP 2000-341190 A1 says communication with satellites needs to be encrypted "for artificial satellites for specific purposes." Even today, researchers point out security measures for satellites are no more than "security by obscurity" ("Satellites Are Rife With Basic Security Flaws" (2023)).
07/02/2024
Japanese Cipher Machine Green (Not Japanese Enigma)
Japanese cipher machines include RED, ORANGE, PURPLE, JADE, etc. Relatively unknown of the series is GREEN. Some sources say GREEN is a US codename for 3-Shiki Kaejiki (三式換字機), known as the Japanese Enigma, but the picture on the book that seems to be the source of this association shows it is not the GREEN machine as recorded by US cryptanalysts.
GREEN is a codename for a cipher machine, 1-Shiki 1-Go Injiki (一式一号印字機), of the Imperial Japanese Army captured at Baguio in the Philippines in May 1945.
I learned all these from Chris Christensen, whose new paper, "The Japanese Green Machine" in Cryptologia is now published online. I believe this is the first detailed account of the real Japanese Green machine.
I learned all these from Chris Christensen, whose new paper, "The Japanese Green Machine" in Cryptologia is now published online. I believe this is the first detailed account of the real Japanese Green machine.
03/02/2024
Dutch Windmill Code?
Windmill Code Used by Dutch Resistance?
Windmills are part of the Dutch landscape. According to Wikipedia, the positions of sails of Dutch windmills were "used to signal the local region during Nazi operations in World War II, such as searches for Jews." If this text (present from the first version of 2 March 2023) was taken from Goleta Valley Historical Society) (present as early as in August 2022), the paragraph break in the latter suggests that the signalling was made by one specific position of the sails. If so, the information conveyed would have been no more than an on/off alarm.According to Traces of War, the windmill Vrijheid in Beesd "was used in World War II by the resistance to sent messages by the position of the wings. To commemorate this, the hitherto unnamed mill was named "The Freedom (De Vrijheid)" after the renovation in 1968." According to machine translation of the explanation board, the position of the sails was indeed used to pass on information, and the name "de Vrijheid" was given in commemoration of the efforts to regain freedom during the war, but the connection between the passing of information and war-time efforts is not clear. But Wikipedia confirms "During World War II, the mill was used to send signals to the Dutch resistance. This was done by the position that the sails were set at."
Traditional Meaning of Windmill Positions at Rest
At least, it seems certain that the position of the sails of a windmill at rest has been traditionally used to convey some meaning, though there is discrepancy among sources about the specific meaning. According to one source, a position slightly tilted to the left means joy; a position slightly tilted to the right means mourning; a straight position like "+" means a short rest (no operation for a few days); a diagonal position like "x" means a long rest (no operation for weeks), etc. (Olie Molen de Passiebloem). The same image (reproduced below) is posted on many websites (Google).This form of expression was actually used in some public occasions. Windmills were placed in "mourning position" when many Dutch people were killed in a Malaysian aircraft shot down in 2014 (CNN quoted in Wikipedia) as well as when Prince Friso died in 2013 because of a skiing accident (PresReader, wind mill sign language in Holland).
Long Distance Communication like Optical Telegraphy
Regarding communication more than a simple expression of some meaning, apart from Wikipedia, the magazine Popular Mechanics (June 1908), p.365 (Google) reports the Dutch government's experiment to use windmill signaling for military purposes between mills some miles distant. It says secret code had been used "from generation to generation of millers." (The idea is similar to the optical telegraphy known from the eighteenth century.)Use by German Agents?
According to Cathleen Small, Code Breakers and Spies of World War II (Google), p.13, during the First World War, German agents used starts/stops of the sails of Dutch windmills to transmit messages in Morse code.
02/02/2024
English Ciphers during the Restoration Period
Many English ciphers from the Restoration Period are preserved in TNA SP106/6 and BL Add MS 40677. I uploaded a new article "English Ciphers during the Restoration Period" about these.
Some of these ciphers use printed templates, which are already discussed in another article, which is now a bit supplemented and renamed as "Diplomatic Codes after the Glorious Revolution and Use of Printed Templates".
Some of these ciphers use printed templates, which are already discussed in another article, which is now a bit supplemented and renamed as "Diplomatic Codes after the Glorious Revolution and Use of Printed Templates".
29/01/2024
Preconcerted Code Words Used by Secretary of Mary, Queen of Scots
Mary, Queen of Scots, is known to have used various ciphers, including very simple ones as well as ones with quite a few code symbols. I recently learned Gilbert Curll, her Scottish secretary, also used preconcerted language (code words) such as "the merchant of London" (Queen of England). I added it to "Ciphers of Mary, Queen of Scots".
28/01/2024
What Code Was Used when Nisho Maru Oil Tanker Broke Blockade of the Royal Navy?
The Nissho Maru Incident (日章丸事件) (Wikipedia) was a 1953 incident in which the Japanese oil tanker Nissho Maru broke an embargo on Iranian oil posed by Britain. The mission required utmost secrecy, and coded messages played a role.
When in 1951 Iran nationalized the Anglo-Iranian Oil Company (AIOC), which had been drawing off most of the profits from oil business in Iran, Britain imposed economic sanctions on Iran and, in July 1952, the Royal Navy seized the Italian tanker Rose Mary, claiming that its oil was stolen property. However, seeing that the nationalization was being accepted internationally (『ペルシャ湾上の日章丸』p.53-54, 61, 148-149, etc.), the Japanese petroleum firm Idemitsu Kosan acted quickly to purchase the Iranian oil before other companies in the world still dared not challenge the British blockade.
The Nissho Maru set sail on 23 March 1953, purportedly for Saudi Arabia. When cruising in the Indian Ocean, a radio message from the home office revealed the real purpose of the voyage was to receive petroleum at Abadan, Iran. The sealed packet entrusted with the captain contained a message to the crew from the president of the firm, declaring his resolution to have a direct access to the oil resources without interference from the monopolozing major oil companies (ibid. p.152-155). After this, the Nissho Maru kept radio silence until 30 April, when it came back as far as the East China Sea. The Nissho Maru received a hero's welcome when she safely arrived at Kawasaki, Japan, on 9 May 1953 (ibid. p.172). No less welcome awaited her on her second, no longer secret, visit to Abadan in June (ibid. p.211-217).
Where does code come in?
The crucial message revealing the real destiation as Abadan is said to have been a coded message (NIKKEIリスキリング). The conclusion of a contract with Iran had been reported in code (『ペルシャ湾上の日章丸』p.138). When the Nissho Maru was returning to Japan, the home office sent a coded message to tell her to lower speed so as to arrive at noon on Saturday, 9 May 1953, because then even if AIOC requested seizure of cargo, a provisional disposition of the court would not be issued at least until Monday.
A businessman, Takeshi Hotoku (宝徳健), notes on his blog that it was his mother who actually operated sending of the coded messages. It appears the coded messages are preserved in Idemitsu Kosan.
I have not found detailed description of the code.
The reporting of the contract may have been sent by telegram. Telegrams were commonly sent by using a codebook. It was also common for businesses to use their proprietary codebooks. Radio messages to/from ships at sea were commonly sent by Morse code until about 1978 (Idemitsu Tanker). The radio messages to/from the Nissho Maru could have been encrypted with some special codebook or cipher.
I first learned of this episode in relation to a best selling novel, A Man Called Pirate (海賊と呼ばれた男) by Naoki Hyakuta (百田尚樹) (Wikipedia). Although his writing tends to have many historical inaccuracies, his description about coded messages in the novel may be of some interest. (The names in the novel are pseudonyms.)
『海賊とよばれた男』下 より
p.130
「翌日は帰国予定だったが，数日ずらすことにし，本社にその旨を電報で送った．今回，正明らと本社のやりとりは，通信の秘匿のためにすべて暗号を使用していた．暗号文を知らされているのは限られた者だけだった．この暗号を作ったのは武知と元ラジオ部の部長である藤本壮平である．旧中野学校の教官であった武知にとって暗号はお手のものだった．かつてラジオ修理を持ち込んで国岡商店に入った藤本も元海軍大佐である．二人は専門家でなくても使える便利な暗号を作った．」
イラン行きの文脈．
p.172
四月五日正午，日章丸がセイロンの南，コロンボ沖にさしかかったとき，国岡本社から無電が入った．日本を出て十三日目のことだった．
「SAKUR NXRQT LPRDX BFNOW TXKPJ」
通信長は意味のわからない無電に首を傾げながら，船長の新田に電文を持ってきた．
新田は「とうとう来たな」と思った．これは事前に打ち合わせておいた暗号電だった．彼はすぐに手帳の暗号解読表を見て電文を読み解いた．
そこにはこう書かれていた．
「アバダンへ行け」
p.205
日章丸は川崎に到着すると聞かされていた徳山の従業員たちは驚いた．慌てて，九州からも助っ人を要請してタンクの清掃に取り掛かった．
にわかに徳山が日本中の注目を浴びた．報道記者たちも続々と徳山に集まった．
しかしこれは鐡造の陽動作戦だった．彼は前もって，日章丸に向けて，暗号電文で次の指令を送っていたのだ．
「川崎において差し押さえの懸念あるにつき，九日午後より揚荷開始の予定．九日正午，検疫錨地に入港するよう適宜考慮して航海せよ」
初めから鐡造は九日に日章丸を川崎港に入港させるつもりだった．九日にこだわったのは，その日が土曜日だったからだ．
When in 1951 Iran nationalized the Anglo-Iranian Oil Company (AIOC), which had been drawing off most of the profits from oil business in Iran, Britain imposed economic sanctions on Iran and, in July 1952, the Royal Navy seized the Italian tanker Rose Mary, claiming that its oil was stolen property. However, seeing that the nationalization was being accepted internationally (『ペルシャ湾上の日章丸』p.53-54, 61, 148-149, etc.), the Japanese petroleum firm Idemitsu Kosan acted quickly to purchase the Iranian oil before other companies in the world still dared not challenge the British blockade.
The Nissho Maru set sail on 23 March 1953, purportedly for Saudi Arabia. When cruising in the Indian Ocean, a radio message from the home office revealed the real purpose of the voyage was to receive petroleum at Abadan, Iran. The sealed packet entrusted with the captain contained a message to the crew from the president of the firm, declaring his resolution to have a direct access to the oil resources without interference from the monopolozing major oil companies (ibid. p.152-155). After this, the Nissho Maru kept radio silence until 30 April, when it came back as far as the East China Sea. The Nissho Maru received a hero's welcome when she safely arrived at Kawasaki, Japan, on 9 May 1953 (ibid. p.172). No less welcome awaited her on her second, no longer secret, visit to Abadan in June (ibid. p.211-217).
Where does code come in?
The crucial message revealing the real destiation as Abadan is said to have been a coded message (NIKKEIリスキリング). The conclusion of a contract with Iran had been reported in code (『ペルシャ湾上の日章丸』p.138). When the Nissho Maru was returning to Japan, the home office sent a coded message to tell her to lower speed so as to arrive at noon on Saturday, 9 May 1953, because then even if AIOC requested seizure of cargo, a provisional disposition of the court would not be issued at least until Monday.
A businessman, Takeshi Hotoku (宝徳健), notes on his blog that it was his mother who actually operated sending of the coded messages. It appears the coded messages are preserved in Idemitsu Kosan.
I have not found detailed description of the code.
The reporting of the contract may have been sent by telegram. Telegrams were commonly sent by using a codebook. It was also common for businesses to use their proprietary codebooks. Radio messages to/from ships at sea were commonly sent by Morse code until about 1978 (Idemitsu Tanker). The radio messages to/from the Nissho Maru could have been encrypted with some special codebook or cipher.
I first learned of this episode in relation to a best selling novel, A Man Called Pirate (海賊と呼ばれた男) by Naoki Hyakuta (百田尚樹) (Wikipedia). Although his writing tends to have many historical inaccuracies, his description about coded messages in the novel may be of some interest. (The names in the novel are pseudonyms.)
『海賊とよばれた男』下 より
p.130
「翌日は帰国予定だったが，数日ずらすことにし，本社にその旨を電報で送った．今回，正明らと本社のやりとりは，通信の秘匿のためにすべて暗号を使用していた．暗号文を知らされているのは限られた者だけだった．この暗号を作ったのは武知と元ラジオ部の部長である藤本壮平である．旧中野学校の教官であった武知にとって暗号はお手のものだった．かつてラジオ修理を持ち込んで国岡商店に入った藤本も元海軍大佐である．二人は専門家でなくても使える便利な暗号を作った．」
イラン行きの文脈．
p.172
四月五日正午，日章丸がセイロンの南，コロンボ沖にさしかかったとき，国岡本社から無電が入った．日本を出て十三日目のことだった．
「SAKUR NXRQT LPRDX BFNOW TXKPJ」
通信長は意味のわからない無電に首を傾げながら，船長の新田に電文を持ってきた．
新田は「とうとう来たな」と思った．これは事前に打ち合わせておいた暗号電だった．彼はすぐに手帳の暗号解読表を見て電文を読み解いた．
そこにはこう書かれていた．
「アバダンへ行け」
p.205
日章丸は川崎に到着すると聞かされていた徳山の従業員たちは驚いた．慌てて，九州からも助っ人を要請してタンクの清掃に取り掛かった．
にわかに徳山が日本中の注目を浴びた．報道記者たちも続々と徳山に集まった．
しかしこれは鐡造の陽動作戦だった．彼は前もって，日章丸に向けて，暗号電文で次の指令を送っていたのだ．
「川崎において差し押さえの懸念あるにつき，九日午後より揚荷開始の予定．九日正午，検疫錨地に入港するよう適宜考慮して航海せよ」
初めから鐡造は九日に日章丸を川崎港に入港させるつもりだった．九日にこだわったのは，その日が土曜日だったからだ．
23/01/2024
Mary Stuart's Ciphers in Papers of Earl of Moray
When I browsed the stack of papers in my study, I came across a photocopy of an old book about ciphers of Mary, Queen of Scots. I looked for this when I uploaded "Ciphers of Mary, Queen of Scots", but I couldn't find it at the time because it was put in a wrong folder.
I had to go to a university library in Tokyo to make the photocopy, but now the book is available on line.
Now, I added a section "Ciphers in Papers of Earl of Morray" in the above article.
I had to go to a university library in Tokyo to make the photocopy, but now the book is available on line.
Now, I added a section "Ciphers in Papers of Earl of Morray" in the above article.
17/01/2024
Venetian Ciphers with Superscripts in ASVe
I'm collecting specimens of ciphers with superscript figures or letters in "Venetian Ciphers with Superscripts". I made additions about materials preserved in the State Archives of Venice (ASVe) by consulting the DECODE database and Paolo Bonavoglia's papers.
I launched this page hoping to find a clue to solve unsolved ciphers from the 1520s (see the section "Older Use"), but the added materials are from the second half of the century, and do not seem to help.
I launched this page hoping to find a clue to solve unsolved ciphers from the 1520s (see the section "Older Use"), but the added materials are from the second half of the century, and do not seem to help.
15/01/2024
Specimens of Caselle Cipher
The caselle cipher (cifra delle caselle) is a Venetian cipher with superencryption, adopted in 1578 and used for almost two decades by the ambassadors to Germany, France, Spain, and Constantinople (Paolo Bonavoglia (2021), "The ciphers of the Republic of Venice an overview", Cryptologia).
Related materials are in the DECODE database.
The four grids used for superencryption are in R1788 (State Archives of Venice [ASVe], "IT ASVe 0045 010 (Serie) Busta 4 Reg. 8"). This is quoted as "ASVe, CX Cifre, chiavi e scontri di cifra, busta 4, reg.8" in Figure 6 of Bonavoglia (2021).
I noted three specimens in this cipher in DECODE.
R1844 ("Busta 11 f.114-115") enciphered with the key Franza (or Francia). Hieronimo Sippomano[?], Paris, 13 October 1578.
R1848 ("Busta 13 f.142-143") enciphered with the key Germania. Prague, 11 October 1578. This is the same as the specimen in Figure 6 of Bonavoglia (2021) cited as "ASVe, CCX, Dispacci degli ambasciatori, busta 13, c.142." [CX=the Council of Ten, CCX=Chiefs of the Council of Ten]
R1863 ("Busta 28 f.83-85") enciphered with the key Francia. 22[?] January 1580.
Figure 2 of Paolo Bonavoglia (2022), "The Enigma of Franceschi's Falso Scontro", HistoCrypt2022, presents another specimen with the Germania key, dated Prague, 23 January 1581, by Venetian ambassador A. Badoer, cited as "ASVe CCX Lettere degli Ambasciatori in Germania, b.12 c.155."
Related materials are in the DECODE database.
The four grids used for superencryption are in R1788 (State Archives of Venice [ASVe], "IT ASVe 0045 010 (Serie) Busta 4 Reg. 8"). This is quoted as "ASVe, CX Cifre, chiavi e scontri di cifra, busta 4, reg.8" in Figure 6 of Bonavoglia (2021).
I noted three specimens in this cipher in DECODE.
R1844 ("Busta 11 f.114-115") enciphered with the key Franza (or Francia). Hieronimo Sippomano[?], Paris, 13 October 1578.
R1848 ("Busta 13 f.142-143") enciphered with the key Germania. Prague, 11 October 1578. This is the same as the specimen in Figure 6 of Bonavoglia (2021) cited as "ASVe, CCX, Dispacci degli ambasciatori, busta 13, c.142." [CX=the Council of Ten, CCX=Chiefs of the Council of Ten]
R1863 ("Busta 28 f.83-85") enciphered with the key Francia. 22[?] January 1580.
Figure 2 of Paolo Bonavoglia (2022), "The Enigma of Franceschi's Falso Scontro", HistoCrypt2022, presents another specimen with the Germania key, dated Prague, 23 January 1581, by Venetian ambassador A. Badoer, cited as "ASVe CCX Lettere degli Ambasciatori in Germania, b.12 c.155."
14/01/2024
Double Reading Caused by Omission of Breaks
Two pre-war Japanese telegram cipher tables are presented in a blog (痩田肥利太衛門残日録その二). They are simple substitution ciphers for kana from 1923 and 1938 and are for internal use by the staff of Sendai Communication Office (serving both as a post office and a telegraph office).
The example message in the illustration of the blog is a famous double-reading message in Japanese.
The message:
kaneokuretanomu
is to be parsed as
Kane Okure. Tanomu. (Send money, please.)
but may also be parsed as
Kane Okureta. Nomu. (Money delayed. I'll drink.)
Whether enciphered or not, telegrams without punctuation are prone to this kind of parsing errors.
This reminded me of another example of such double reading I read when I was in elementary school.
A note in a clinic:
kokodehakimonowonuidekudasai.
should be read as:
Kokode Hakimono-wo Nuide Kudasai. (Take off your shoes here.)
but may also be parsed as
Kokode-ha Kimono-wo Nuide Kudasai. (Take off your clothes here.)
Such multiple possibilities of reading tend to occur in deciphering. One such example is given in my coauthored paper, "Deciphering Mary Stuart's lost letters from 1578-1584" (Cryptologia) (note 344). When introducing a codename for a secret messenger, Mary, Queen of Scots, writes:
"Le porteur s'appellera cy a present Renous Banque" (The bearer will now be called "Renous Banque").
But this passage, known from contemporary decipherment and printed in Labanff, v, p.479, seems to be a decipherment error. It would make better sense if parsed as
"Le porteur s'appellera cy-apres entre nous Banque" (The bearer from now on will be called between us "Banque")
There is always a possibility of this kind of double reading when deciphering a ciphertext without word breaks.
The example message in the illustration of the blog is a famous double-reading message in Japanese.
The message:
kaneokuretanomu
is to be parsed as
Kane Okure. Tanomu. (Send money, please.)
but may also be parsed as
Kane Okureta. Nomu. (Money delayed. I'll drink.)
Whether enciphered or not, telegrams without punctuation are prone to this kind of parsing errors.
This reminded me of another example of such double reading I read when I was in elementary school.
A note in a clinic:
kokodehakimonowonuidekudasai.
should be read as:
Kokode Hakimono-wo Nuide Kudasai. (Take off your shoes here.)
but may also be parsed as
Kokode-ha Kimono-wo Nuide Kudasai. (Take off your clothes here.)
Such multiple possibilities of reading tend to occur in deciphering. One such example is given in my coauthored paper, "Deciphering Mary Stuart's lost letters from 1578-1584" (Cryptologia) (note 344). When introducing a codename for a secret messenger, Mary, Queen of Scots, writes:
"Le porteur s'appellera cy a present Renous Banque" (The bearer will now be called "Renous Banque").
But this passage, known from contemporary decipherment and printed in Labanff, v, p.479, seems to be a decipherment error. It would make better sense if parsed as
"Le porteur s'appellera cy-apres entre nous Banque" (The bearer from now on will be called between us "Banque")
There is always a possibility of this kind of double reading when deciphering a ciphertext without word breaks.
08/01/2024
A Cipher of Margret of Austria?
A cipher used in the correspondence of Margaret of Austria, Governor of the Habsburg Netherlands in 1507-1515, 1519-1530, is preserved in Archives départementales du Nord (Un jour, une œuvre : un code secret aux Archives départementales!, 1 June 2022).
I believe the third image posted, if any, belongs to Margaret of Austria. It is a short list of symbols for names. Historians will be able to date it from entries such as
Soliciteur
de Zorn
de Bure
de fiennes
de Sempy
de Silstein
Brossart de fauquemont
Paule de Lichtein
The first image shows a simple substitution cipher alphabet which assigns numbers to the letters A-Z plus J at the end, with homophones for A, E, I, O, S, V. The syllable section assigns number to some (fa is enciphered as 35, fe is 36, etc.), but substitutes syllables for others (da is enciphered as fa, de is fe, etc.). Such mixed ciphers are known from Spanish ciphers in the 17th century (e.g., Cp.59, Cp.64, for which see my article).
The second image is part of a nomenclature, assigning numbers 266 to 614 to letters, syllables, words, names, and some functions. The entry of Queen Christina clearly shows it is no earlier than 1632. The following shows some entries of interest, from which I guess this is from 1682-1689.
269,313,357 chifre nul
270,314,358 annulle le chifre qui suit
271 Mr le Dauphin
276 Mr De Croissy *Colbert de Croissy, who assumed the title in 1662 and died in 1696
280 le Roy de Hongrie
299 Elect^r Palatin
300 la maison de B[r]aunswic et Lunebourg
303 la Reyne de Dannemark
328 Ill.[?] de Sreues[?]
365 La P^cesse d'Orange *Mary, Princess Royal (holding the title 1647-1650, died 1660)? or Amalia of Solms-Braunfels (holding the title 1625-1647, died in 1675 as dowager) or Mary of York (holding the title 1677-1694)
371 lEl^r de Mayences
373 le Card^l de furstenbourg
392 le Roy de Suede *Queen Christina (see below), not King, reigned 1632-1654.
396 Versailles *Court of Louis XIV from 1682.
411 Mr de Benting *William Bentinck (1649-1709, created Earl of Portland in 1689)
434 Republique
436 lEvesq. de Munster
439 le Mar^al d'humieres *Marechal d'Humieres, created in 1668 and died in 1694
440 le Card^l Cibo *Innocenzo Cybo was cardinal 1513-1550. Alderano Cibo was cardinal 1645-1700
443 M-r de Lauardin[?]
445 ce chifre est nul
455 M^r Diekvelt *Everard van Weede van Dijkvelt (1626 - 1702)?
484 le Car^dl D'Estree *Cesar d'Estrees (1628-1714)?
487 le Pro^ce de Groninge
489 chifre nul
495 M^r le Peletier
528 Les Provinces unies *1581-1795
530 La Reyne christine *reigned 1632-1654, died in 1689
533 annule le precedent
586 M^r Heinsius *Daniel Heinsius (1580-1655)?? or Anthonie Heinsius (1641-1720)
589 Le Con^el de Vienne
I believe the third image posted, if any, belongs to Margaret of Austria. It is a short list of symbols for names. Historians will be able to date it from entries such as
Soliciteur
de Zorn
de Bure
de fiennes
de Sempy
de Silstein
Brossart de fauquemont
Paule de Lichtein
The first image shows a simple substitution cipher alphabet which assigns numbers to the letters A-Z plus J at the end, with homophones for A, E, I, O, S, V. The syllable section assigns number to some (fa is enciphered as 35, fe is 36, etc.), but substitutes syllables for others (da is enciphered as fa, de is fe, etc.). Such mixed ciphers are known from Spanish ciphers in the 17th century (e.g., Cp.59, Cp.64, for which see my article).
The second image is part of a nomenclature, assigning numbers 266 to 614 to letters, syllables, words, names, and some functions. The entry of Queen Christina clearly shows it is no earlier than 1632. The following shows some entries of interest, from which I guess this is from 1682-1689.
269,313,357 chifre nul
270,314,358 annulle le chifre qui suit
271 Mr le Dauphin
276 Mr De Croissy *Colbert de Croissy, who assumed the title in 1662 and died in 1696
280 le Roy de Hongrie
299 Elect^r Palatin
300 la maison de B[r]aunswic et Lunebourg
303 la Reyne de Dannemark
328 Ill.[?] de Sreues[?]
365 La P^cesse d'Orange *Mary, Princess Royal (holding the title 1647-1650, died 1660)? or Amalia of Solms-Braunfels (holding the title 1625-1647, died in 1675 as dowager) or Mary of York (holding the title 1677-1694)
371 lEl^r de Mayences
373 le Card^l de furstenbourg
392 le Roy de Suede *Queen Christina (see below), not King, reigned 1632-1654.
396 Versailles *Court of Louis XIV from 1682.
411 Mr de Benting *William Bentinck (1649-1709, created Earl of Portland in 1689)
434 Republique
436 lEvesq. de Munster
439 le Mar^al d'humieres *Marechal d'Humieres, created in 1668 and died in 1694
440 le Card^l Cibo *Innocenzo Cybo was cardinal 1513-1550. Alderano Cibo was cardinal 1645-1700
443 M-r de Lauardin[?]
445 ce chifre est nul
455 M^r Diekvelt *Everard van Weede van Dijkvelt (1626 - 1702)?
484 le Car^dl D'Estree *Cesar d'Estrees (1628-1714)?
487 le Pro^ce de Groninge
489 chifre nul
495 M^r le Peletier
528 Les Provinces unies *1581-1795
530 La Reyne christine *reigned 1632-1654, died in 1689
533 annule le precedent
586 M^r Heinsius *Daniel Heinsius (1580-1655)?? or Anthonie Heinsius (1641-1720)
589 Le Con^el de Vienne
02/01/2024
British Codebreakers' Keys of French Ciphers during the War of American Revolution
Back in 2015, I reported use of four codes by the French during the American Revolutionary War in "Code Switching in French Diplomatic Correspondence and Intercepted Letter of Barbé-Marbois (1782)". (By the way, code switching, the theme of the article, was also prescribed during the reign of Louis XIV, as reported by Jörg Ulbert, quoted in "Code Switching in French Diplomatic Correspondence during the Peace Negotiation at Rijswijk (1696-1697)").
To facilitate identifying the four codes, labelled A-D in my article, the letter "E" is assigned the following numbers according to my reconstruction:
Code A: E=94, 395, 709, 865, 1043
Code B: E=302, 461, 590, 871
Code C: E=164
Code D: E=51, 475, 485
I noticed related keys are among the collection in BL, Add MS 32263:
f.220-225 (DECODE* R7681)
f.226-227 (DECODE R7682)
f.228-229 (DECODE R7683)
f.230-231 (DECODE R7684)
f.234-235 (DECODE R7687)
Of these, the "left" key in R7687 matches Code C and the "left" key in R7683 matches Code D. I was hoping the British codebreakers' keys might shed light on the indication system for code switching, but I have not been able to find a clue. (In particular, I thought "867" might be an indicator for Code D, but I find nothing to support it.)
As a starting point for future search, the following describes these items.
R7682, R7683, R7684, and R7687 are the British codebreakers' key worksheets, with printed numbers 1 - 1100. It seems to cover two different codes, one written left to the number and the other written right to the number.
Endorsed "De La Luzerne a Montmorin, Nov^r 1780" In a different hand, "Vergennes" is written above the name, and "Luz-ne a Vergennes Dec^re[?] 1783.
The left key corresponds to Code C.
This includes many substitution alphabets (but Codes A-D are not among them). The following lists the working title (?) and the "E" section. (It should be noted that sometimes it is not clear which lines belong to the section. In the first place, the same number occurs multiple times, and the meaning of the list is not completely clear.)
"587 la" 272 270 -- 135 187
705 718 -- 24 17
"361 de" 17 634 -- 267 270
817 533 -- 402 250 116 339 -- 616 522
267 87 -- 211 185
"250 l" 361 402 -- 740 135
"607 le" 587 240 -- 305 801
211 185 -- 817 303
"794 de" 961 632 -- 951 449
"816 C" [None. Though "e" is the most frequent letter, it often occurs in syllables and so it is possible that "e" does not occur in a short specimen.]
"1125 de" 28 386 -- 451
"144 Ameriq" 985 401 -- 285 751
"377 B" [None.]
"670 de la" 119_ 767 -- 430 486 342 139 -- 207 49
672 900 -- 588 869
"223" 369_ 1021 -- 290 119_
"342" 143 1070 -- 315 439
903 869 -- 139 670
"103 de" 1 767 -- 516 426
293 591 -- 903 869
"767 gs[?]" 162 1133 -- 802 63
290 119_ -- 670 430
481 1 -- 103 516
799 898 -- 253 369
"297" [None]
"812 des" [None]
*DECODE
Héder, M ; Megyesi, B. The DECODE Database of Historical Ciphers and Keys: Version 2. In: Dahlke, C; Megyesi, B (eds.) Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022. Linkoping, Sweden : LiU E-Press (2022) pp. 111-114. , 4 p. [pdf] Megyesi Beáta, Esslinger Bernhard, Fornés Alicia, Kopal Nils, Láng Benedek, Lasry George, Leeuw Karl de, Pettersson Eva, Wacker Arno, Waldispühl Michelle. Decryption of historical manuscripts: the DECRYPT project. CRYPTOLOGIA 44 : 6 pp. 545-559. , 15 p. (2020) [link] Megyesi, B., Blomqvist, N., and Pettersson, E. (2019) The DECODE Database: Collection of Historical Ciphers and Keys. In Proceedings of the 2nd International Conference on Historical Cryptology. HistoCrypt 2019, June 23-25, 2019, Mons, Belgium. NEALT Proceedings Series 37, Linköping Electronic Press. [pdf]
To facilitate identifying the four codes, labelled A-D in my article, the letter "E" is assigned the following numbers according to my reconstruction:
Code A: E=94, 395, 709, 865, 1043
Code B: E=302, 461, 590, 871
Code C: E=164
Code D: E=51, 475, 485
I noticed related keys are among the collection in BL, Add MS 32263:
f.220-225 (DECODE* R7681)
f.226-227 (DECODE R7682)
f.228-229 (DECODE R7683)
f.230-231 (DECODE R7684)
f.234-235 (DECODE R7687)
Of these, the "left" key in R7687 matches Code C and the "left" key in R7683 matches Code D. I was hoping the British codebreakers' keys might shed light on the indication system for code switching, but I have not been able to find a clue. (In particular, I thought "867" might be an indicator for Code D, but I find nothing to support it.)
As a starting point for future search, the following describes these items.
R7682, R7683, R7684, and R7687 are the British codebreakers' key worksheets, with printed numbers 1 - 1100. It seems to cover two different codes, one written left to the number and the other written right to the number.
R7682
Endorsed "... Luzerne a Verg[enne]s April 1780 ..."R7683
The "left" key matches Code D.Endorsed "De La Luzerne a Montmorin, Nov^r 1780" In a different hand, "Vergennes" is written above the name, and "Luz-ne a Vergennes Dec^re[?] 1783.
R7684
Endorsed "C^te de Rochambeau avec la Chevalier de la Luzerne, July 1781"R7687
Endorsed "Luzerne et Marbois au Marquis de Castiris 1782"The left key corresponds to Code C.
R7681
Endorsed "Oct. 1780 Luzerne à Verg[ennes] Philad[elphia]"This includes many substitution alphabets (but Codes A-D are not among them). The following lists the working title (?) and the "E" section. (It should be noted that sometimes it is not clear which lines belong to the section. In the first place, the same number occurs multiple times, and the meaning of the list is not completely clear.)
"587 la" 272 270 -- 135 187
705 718 -- 24 17
"361 de" 17 634 -- 267 270
817 533 -- 402 250 116 339 -- 616 522
267 87 -- 211 185
"250 l" 361 402 -- 740 135
"607 le" 587 240 -- 305 801
211 185 -- 817 303
"794 de" 961 632 -- 951 449
"816 C" [None. Though "e" is the most frequent letter, it often occurs in syllables and so it is possible that "e" does not occur in a short specimen.]
"1125 de" 28 386 -- 451
"144 Ameriq" 985 401 -- 285 751
"377 B" [None.]
"670 de la" 119_ 767 -- 430 486 342 139 -- 207 49
672 900 -- 588 869
"223" 369_ 1021 -- 290 119_
"342" 143 1070 -- 315 439
903 869 -- 139 670
"103 de" 1 767 -- 516 426
293 591 -- 903 869
"767 gs[?]" 162 1133 -- 802 63
290 119_ -- 670 430
481 1 -- 103 516
799 898 -- 253 369
"297" [None]
"812 des" [None]
*DECODE
Héder, M ; Megyesi, B. The DECODE Database of Historical Ciphers and Keys: Version 2. In: Dahlke, C; Megyesi, B (eds.) Proceedings of the 5th International Conference on Historical Cryptology HistoCrypt 2022. Linkoping, Sweden : LiU E-Press (2022) pp. 111-114. , 4 p. [pdf] Megyesi Beáta, Esslinger Bernhard, Fornés Alicia, Kopal Nils, Láng Benedek, Lasry George, Leeuw Karl de, Pettersson Eva, Wacker Arno, Waldispühl Michelle. Decryption of historical manuscripts: the DECRYPT project. CRYPTOLOGIA 44 : 6 pp. 545-559. , 15 p. (2020) [link] Megyesi, B., Blomqvist, N., and Pettersson, E. (2019) The DECODE Database: Collection of Historical Ciphers and Keys. In Proceedings of the 2nd International Conference on Historical Cryptology. HistoCrypt 2019, June 23-25, 2019, Mons, Belgium. NEALT Proceedings Series 37, Linköping Electronic Press. [pdf]
26/12/2023
Manifesto Printed in Secret for Uprising against the Shogunate
Righteous indignation against the corrupt shogunate made Oshio Heihachiro (Wikipedia), government official in Osaka himself, to organize an uprising in 1837. In the previous year, he had drafted a 2000-character manifesto and had it printed secretly. In order to prevent the woodcutters from reading the content, the woodcut was prepared in 32 divisions, so that each piece looked like nothing but a cipher. In the image below, divisions between woodcut blocks are clearly seen as horizontal spacing. (I learned of the episode in an epic manga by Minamoto Taro, Fuunji-tachi (Wikipedia, Wide edition, vol.14, p.212).) Although the prematurely started uprising was suppressed in a day, the manifesto of Oshio, who was also a scholar, was hand-copied and circulated widely.
References:
Here is an image of the manifesto preserved in Joshoji (成正寺), the temple also housing Oshio's tomb and a memorial, as well as the woodcut reproduced in 2006 (関西・大阪２１世紀協会「なにわ大坂をつくった100人 第20話 大塩平八郎（1793－1837年）」).
The copy shown in the photo on an auction site is in the form of a scroll and appears to be a different copy from the same woodcut (なにわ歴博ブログ ～等身大の学芸員生活～).
The image below is taken from the National Diet Library (image, catalogue, 大塩中斎先生天保救民告文).
The image at another auction site seems to be a contemporary handwritten copy (『委託HK◇大塩平八郎檄文 巻子装 天保八（1837）年（思想家 陽明学 大塩中斎 近世古文書 大坂 天保の大飢饉 書状 幕臣 大塩平八郎の乱）』).
Other copies (early printing? in different layouts) are found here (大塩平八郎檄文) and here (大塩平八郎檄文).
A transcription is found at 本山美彦「土佐堀川が育んだ改革者たちの言葉（6）」 (with modern translation and annotations) and also at 資料385 大塩平八郎の檄文.
Subscribe to:
Posts (Atom)