02/11/2024

Codebreaking without Command of the Language

Suppose we have a ciphertext in a letter in, say, Spanish. Can we break it without command of Spanish? The answer is yes. Back in 2019, I gave a quotation from 1737 to this effect in "Is Codebreaking Possible without Knowing the Language?".
I saw similar observations in two books I recently read.

Elonka Dunin and Klaus Schmeh, Codebreaking: A Practical Guide, which I reviewed yesterday:
"Experience shows that breaking an encrypted message does not necessarily require command of the language used."
(p.76 in Japanese, p.74 in English)

Liza Mundy, Code Girls: The Untold Story of the American Women Code Breakers of World War II, which I mentioned last month:
"What Friedman had also taught them during their training is that you can break a foreign cipher without understanding the language, as long as you know how the letters in that language behave."
(p.138 in Japanese)

01/11/2024

Japanese Edition of Codebreaking: A Practical Guide by Elonka Dunin and Klaus Schmeh

Codebreaking: A Practical Guide is now available in Japanese! The book is by Elonka Dunin and Klaus Schmeh and was first published in UK in 2020 (as I reported before). In US, a new expanded edition was published in September 2023 (announced here). The translation is based on the 2020 version, with additional appendices for reprint: Appendix D for the solution of the famous Zodiac killer cryptogram Z340, released right after the first publication plus Appendix E showing the Morse table. According to the translator's preface, the translation project started soon after the initial publication.

Although I'm mentioned as one of contributors to the original English version (p.16 of the translation), I believe the following is not too partial.

Use of the polite style in Japanese (cf. Hiragana Times), as well as presenting personal names in the original roman alphabet, is, I believe, the style commonly adopted in textbooks/manuals in the field of computing and cybersecurity where the expertise of the translator belongs.

The translator, who accepted the offer of this project after seeing the high rating of the original edition abroad and endorsements by experts, got absorbed in classical cryptography through reading the book. The numerous translator's notes inserted between paragraphs attest to the zeal to help beginners understand the content. Sometimes, the translator's note runs half of a page (e.g., p.229). Given the evident zeal, it could have been better but for the occasional typos and a few relevant errors.
The translator's notes are more than just a guide for beginners. I myself found useful the explanation of the very first steps of CrypTool 2 on p.7-8! At times, additional information is given, like the solution of the Silk Dress cryptogram in 2024 (p.153). I found interesting the episode of a Lorenz machine being sold on eBay for 9.5 pounds in 2016! (The news received media coverage at the time: NPR, BBC, The Guardian).

With so many topics about paper-and-pencil ciphers with many worked-out examples, numbering the sections adopted in the Japanese verion (including in the running head) is convenient for referencing and navigating. I never thought such a voluminous work (nearly 500 pages) can be published in Japanese at such an affordable price (3,480 yen before tax).
I enjoyed re-reading this book in Japanese, with occasional reference to the original in English.
Congratulations to the authors on the publication!

26/10/2024

A Great Cipher Left Undeciphered by William Blencowe

An intercepted letter from the French foreign minister Torcy to French ministers sent to the peace talk at Geertruidenberg, dated 3 April 1710, is endorsed "M Blencow cannot decypher them."
Filed with it is another letter from Marshal Villars to the Abbé de Polignac, dated 1 June 1710. This also seems to be undeciphered.
I uploaded a new article "A Great Cipher Left Undeciphered by William Blencowe" describing BL Add MS 61575 including these. I also added references to these undeciphered pieces in "Unsolved Historical Ciphers"

25/10/2024

Frederick I of Prussia's Transposition Cipher

I noticed three transposition ciphertexts (two of which are attributed to King Frederick I of Prussia) in DECODE. Two appeared to be non-decrypted, and I could solve them without much difficulty.
See my new article "Frederick I of Prussia's Transposition Cipher". I also listed them in "Unsolved Historical Ciphers."

24/10/2024

Cardinal Richelieu Used a Porta-like Pairing Cipher?

Just as I wrote about the outage of the British Library's online services the other day, another important source, Internet Archive, went down about 10 October because of a DDoS attack. It's a sad realization that online services we take for grated may be attacked anytime and the attackers may have success sometimes.

Thankfully, it seems Internet Archive (at least archive.org that I need) appears to be working now.
Thus, I could access Wits Interpreter (1, 2), attributed to John Cotgrave (Wikipedia), which I learned printed Richelieu's cipher the other day. (Somehow, at that time, before the attack on Internet Archive, my search could not find this on Internet Archive.)
The cipher attributed to Richelieu (p.491) is a Porta-like pairing-based cipher. But the other ciphers following this all seem to be taken from Porta's book. So, a more reliable source is needed to confirm Richelieu really used this cipher. (After writing this, I noticed the section on Richelieu in "French Ciphers during the Reign of Louis XIII" already mentions Wits Interpreter from a secondary source.)

 


19/10/2024

Some Ciphers from Henry IV's Time

When I was searching about the French ambassador Bordeaux in writing the post the other day, I ran across Dupuy 63. I added several ciphers from this volume:
ciphers of Seguier, Buzanval, and an informant to Henry IV (1601) in "French Ciphers during the Reign of Henry IV of France"; and
Francesco Guicciardini's Cipher (1597) in "Florentine Ciphers of Guicciardini (ca.1590-1593)".
It is notable that the three ciphers used in 1601 are of different nature. Buzanval (ambassador in The Hague) employed a symbol cipher with a seemingly large nomenclature in symbols. Seguier (ambassador to Venice) employed a symbol cipher, but his nomenclature was in Arabic figures with diacritics. The cipher of the king's informant employs Arabic figures both in the substitution table and the nomenclature, the character reminiscent of a cipher used with the Landgrave of Hesse-Cassel.

14/10/2024

Two French Ciphers (ca.1643) in English Archives

French ciphers are sometimes found in English (and probably other) archives. I wrote about the Prince of Conde's cipher and one other the other day. Now, I added mention of two French ciphers (M. de Boisivon, Prince Harcourt) in "Ciphers Early in the Reign of Louis XIV". Although these were shared with English royalists, they are similar to contemporary French ciphers.

13/10/2024

A Bundle of Ciphers of Lord Digby

It's almost a year since the British Library's computer system was disrupted by a cyber-attack (ransomware) in October 2023, but my links to BL materials do not work yet. It seems rebuilding the whole system to replace the previous patchwork of many legacy systems is taking time to fully recover their services (BL, detailed report as of March 2024 (pdf)). At least, I'm hopeful for future recovery because they say they have secure copies of all the digital collections and related metadata.

I was reminded of this when I saw volumes I studied some years ago in the DECODE database.
One of them is BL Add MS 72438 (a majority of no.8619-no.8645 in DECODE), which I found at https://www.bl.uk/manuscripts/FullDisplay.aspx?index=3&ref=Add_MS_72438 back in 2021, but the URL does not work now.
This volume contains "Cipher-keys and intercepted royalist correspondence from the papers of Georg Rudolph Weckherlin, government official, 1625-1647."
Georg Rudolph Weckherlin served Charles I in drafting, deciphering, and translating official correspondence, but after the Civil War broke out, he worked for the Parliament (DNB). In particular, he deciphered intercepted royalist letters (cf. n.36 in Philip Beeley, "Breaking the Code. John Wallis and the Politics of Concealment"). When Lord Digby's papers were captured after his defeat at Sherburn in October 1645, the cipher letters were delivered to Weckherlin (and a Mr. Frost), as I quoted in "Charles I's Ciphers".
F.1-f.24 are intercepted royalist letters (some are endorsed by or are in the hand of Weckherlin).
F.25-f.99 are cipher keys captured from Lord Digby's papers in October 1645. This includes the original of what I call "Third Cipher between Charles I and Henrietta-Maria (and Ministers) in Paris (Summer 1644-June 1645)" (f.67-68, DECODE R8687). F.100-101 (R8723) is its partial key (apparently by a codebreaker).
F.110-150 are ciphers used by Weckherlin while he was in the service of Charles I (ca. 1623-1641), partly in Latin and French.
I added two from Add MS 72438 which I believe are undeciphered in "Unsolved Historical Ciphers".

Another volume familiar to me is BL Add MS 18982 (DECODE (R8428-R8454)), which was at https://www.bl.uk/manuscripts/FullDisplay.aspx?index=0&ref=Add_MS_18982 when I found it before. It contains "Royalist correspondence, primarily addressed to Prince Rupert, 1645-1658."
Most of the undeciphered ciphertexts can be read by using already deciphered letters or with known keys. I added one from Add MS 18982 in "Unsolved Historical Ciphers".

12/10/2024

Chinese Enciphered Code Analyzed in Arlington Hall

I wrote something about Chinese enciphered codes in "Chinese Cryptography: 1871-1945", but the materials available are relatively scanty. Now, I learned in Liza Mundy, Code Girls that a Chinese enciphered code was among those analyzed in Arlington Hall during WWII:
"There was a French code called Jellyfish, a Chinese enciphered code they called Jabberwocky, another they called Gryphon." (p.154 of the Japanese edition)
The source given is a weekly report (RG 0457, 9032 (A1), Box 1114, "Signal Security Agency Weekly Reports, Jan to Oct 1943", Weekly Report for Section B-III, July 9, 1943). I wonder whether the original materials are not preserved.

10/10/2024

How about a Puzzle in Latin?

The additions reported yesterday include two unsolved ciphers (Bishop of Worcester, Serno Gilino). Both are in letters written in Latin (probably to Cardinal Wolsey) in the 1520s and both employ superscript digits (but they are clearly different ciphers).
For Worcester's cipher, I could identify symbols for single letters and some syllables. But when I apply these to undeciphered letters, I can only see some short fragments (e.g., "hu-n-c me-a-m", "no-n", "o-m-ni?-a") and the few long sequences of known symbols are incoherent to me (e.g., "...ri-s-tu-m-ri-po-r-ta-re-s-pe-re-n-t"). (I can see "riportare" but cannot think of words ending in "ristum" or beginning with "sperent".) For those versed in Latin, filling the blanks may be an interesting puzzle.
The relevant sections are now added in "Unsolved Historical Ciphers". See Ciphers during the Reign of Henry VIII for details.

09/10/2024

An Early English Cipher Used by Cuthbert Tunstall (1517)

Cipher was already in routine use in the first years of the reign of Henry VIII. I now added one used by Cuthbert Tunstall in 1517 in "Earliest English Diplomatic Ciphers". It is used in the context of "the new treatie of Cambray". (In the treaty of Cambray signed on 11 March 1517, Francis I, Henry VIII, and the Emperor agreed on a crusade against the Turks (Sanderson Beck).)
I also added several specimens from 1526-1529 (Edward Lee, Bishop of Worcester, Francis Poynts, Sylvester Darius, Serno Gilino) in "Ciphers during the Reign of Henry VIII".

29/09/2024

Two Syllabic Challenges (1653, 1654)

The volume I mentioned yesterday (ADD MS 4200) includes two letters that appear to be undeciphered: a letter of Bordeaux, French ambassador in London (1653) and a letter of the Prince of Condé (1654). I included sections for these in "Unsolved Historical Ciphers".
Both employ Arabic figures with diacritics and probably involve symbols representing syllables.

The ambassador's cipher would be similar to other French diplomatic ciphers having fairly regular assignment of syllables to numbers.
The Prince of Condé was in the Spanish Netherlands and fought the French at that time. If his cipher in 1654 was similar to the known cipher from 1655, it would involve non-trivial assignment of syllables to numbers.

28/09/2024

The Prince of Condé's Cipher with his Agent in London

The cipher used in the correspondence from 1655 between the Prince of Condé and Barriere, his agent in London, (Add MS 4200) is better than contemporary French ciphers in that single letters, syllables, other words and names, and nulls share diacritics. For example, single letters may be represented by figures with a circumflex (^) or an umlaut (¨) (or other symbols based on Roman letters), while some syllables and words are also represented with figures with a circumflex or an umlaut.
At the time, Condé was in exile in the Spanish Netherlands and was fighting the French troops. It may be speculated that he might have had his source of cipher other than the king's court.
See the new section in the updated article, "Ciphers Early in the Reign of Louis XIV".

18/09/2024

A Syllabic Cipher of Cardinal Gualterio Reconstructed Manually

I uploaded a new article "A Syllabic Cipher of Cardinal Gualterio Reconstructed Manually (ca.1715)", in which I demonstrate manual reconstruction of symbols representing syllables used in letters of Abbé D'Estées, given assignment of symbols for single letters.
The letters record real-time reactions to historical events such as James Edward's expedition to Scotland (December 1715), Prince Eugene's victory over the Turks (August 1716), Spanish invasion of Sardinia (August 1717), and the Triple Alliance (January 1717). Actually, these are mentioned in cleartext. Hopefully, the ciphertext contains even more interesting contents. (I don't know whether the plaintext is in the archives.)

02/09/2024

Parliament's Ban on Ciphers during the English Civil War

Use of cipher was prohibited during the English Civil War by an order passed by the Parliament in April 1643: "That all such should be punished as Spyes and Enemies of the State, who hereafter should send any Letters or Papers written with Cyphers, or any other unknowne Characters." (Nadine Akkerman, "Enigmatic Culture of Cryptology" in Daybel et al. (ed.) (2016), Cultures of Correspondence in Early Modern Britain).

But the ban was only selectively enforced, at least in the view of the royalists. The Mercurius Aulicus (21 October 1643), a weekly royalist news pamphlet, accused the parliamentarians of the partisan application of the rule and printed an intercepted cipher letter subscribed by a Parliamentarian, a Matthew Durbun, pointing out that the parliamentarians "when they please can practice it, without the least transgression of their order, which it seems was made only for the punishment of the Kings friends but not for such innocent Rebels as they are."

How long the ban was in effect is not known for sure. While it was natural for the royalists to continue using ciphers (see "King Charles I's Ciphers"), we know that under John Thurloe (head of intelligence from July 1653), informants used ciphers (see "Codes and Ciphers of Thurloe's Agents"). Akkerman points out John White's A Rich Cabinet with Variety of Inventions (1653) promoted use of cipher when writing love letters and John Cotgrave's The Wits Interpreter (1655; 1662; 1671) described one of Cardinal Richelieu's cipher keys and recipes for secret ink. By these days, ciphers as well as steganographic techniques such as secret ink became quite common even among ordinary people.

01/09/2024

Duke of Nevers' Variable-Length Figure Cipher (1571)

From my renewed interest in variable-length figure ciphers I mentioned the other day, I uploaded my transcription of a ciphertext from an undeciphered letter in Italian from Lodovico Birago to the Duke of Nevers (1571) in "Unsolved Historical Ciphers".
Although the ciphertext is relatively short, once the continuous stream of figures can be broken into tokens (cipher symbols), homophonic solvers may readily decipher this.

15/08/2024

A Second Copy of Music Cipher to Charles II Acquired by the British Library (ca.2018)

Sheet music concealed a secret message to Prince Charles (later Charles II) fleeing after the defeat in the Battle of Worcester: "Conceal yourself. Your foes look for you." Another version concealing the same message is said to be for another Prince Charles, Jacobites' Charles III known as the Young Pretender. Both of these have a plausibly sounding story of their provenance. I wrote about these in "Musical Notes Concealing A Warning to Fleeing Prince Charles" back in 2015.
What I didn't know then is that the cipher to Charles II quoted in my article is in the British Library (Add MS 45850, f.68), but its provenance through the Port family I found in googling is about another copy, which the British Library newly acquired (as of February 2018) (Add MS 89288). I learned of this in the British Library's blog article, "'Conceal yourself, your foes look for you': revealing a secret message in a piece of music" (20 February 2018).

I was reminded of this topic when reading Nadine Akkerman and Pete Langman (2024), Spycraft, which refers to Nadine Akkerman (2018), Invisible Agents (the BL's blog also mentions forthcoming publication of this book). Akkerman discusses the nineteenth century copy in BL Add MS 45850 and considers it a hoax. Her dismissal of Jane Lane as the author based on the latter's literacy level is convincing. Even if we assume other authorship, it is hard to think of circumstances in which this kind of cipher came into play ("it is not as if Charles Stuart did not know his enemies were searching for him" etc.).

14/08/2024

Use of Diacritics/Exponents/Vowel Indicators in Milanese Ciphers

Some years ago, I noticed the vowel indicator system characteristic of Spanish ciphers in the reign of Philip II had been in use from Charles V's time and similar symbols were also used in Milanese and other ciphers ("Tracing the Origin of Vowel Indicators in Spanish Ciphers").
I now see Milanese ciphers used combined symbols with diacritics or exponents as early as the mid fifteenth century but syllables were not formed systematically as with vowel indicators. I now added a section about this: "Use of Exponents/Diacritics/Vowel Indicators in Milanese Ciphers (1450s-1530s)."
It seems systematic vowel indicators are a degenerated form of such ciphers, but dating of the ciphers in the archives is necessary to assess such a hypothesis.

13/08/2024

Constantijn Huygens Jr.'s Secret in Simplistic Concealment Cipher

Constantijn Huygens Jr. (1628-1697), a brother of the physicist Christiaan Huygens, used a cipher in some part of his journals. In late twentieth century, it was found out that encrypted words can be read by simply ignoring odd-numbered letters. For example, b.mregtvelennphnōdesr reads met een hoer ("with a prostitute") (there is one extra letter, which may be an error). The cipher typically concealed such embarrassing privacy of the diarist. I learned of this in Christopher Joby (2014), The Multilingualism of Constantijn Huygens (1596-1687) p.282.


Constantijn Jr. was secretary of Prince of Orange William III (my favoutie historical character) since the latter became stadtholder in 1672. He records his personal experience in participating in major events such as William's campaings to oppose the French invasion, the expedition to England (the Glorious Revolution), and the Irish campaign to prevent the return of James II. Joby (2014) includes some quotes from these (p.284 ff.), and more would be found in Rudolf M. Dekker (2013), Family, Culture and Society in the Diary of Constantijn Huygens Jr, Secretary to Stadholder-King William of Orange. His journals seem interesting in describing historical events from a personal point of view. For example, the day after William III's coronation, from which he was absent, he wrote, "In the early afternoon I was with the king, who asked me where I had watched the coronation. I said that I had been busy deciphering the resolution of the States General, received in cipher, about the alliance with the Emperor, because I thought that the king would want to read this quickly. He asked me if I had received a coronation badge, and I answered no, without receiving much of a response." (ibid. p.40)

12/08/2024

Codebreaker Constantijn Huygens

I've heard that Constantijn Huygens Sr. (1596-1687; the father of the famous physicist, Christiaan Huygens) did codebreaking, but it was only recently that I learned that he regularly served in that capacity in Chapter 2 of Nadine Akkerman and Pete Langman (2024), Spycraft (p.153-156). According to this book, he studied cryptanalysis at the University of Leiden in 1616 and even got a pay raise while serving as a secretary to Prince of Orange Frederick Henry since 1624. The authors translate his proud words about his achievements in his autobiography: "At every siege, I proved my skills, anticipating the tricks of the enemy by means of my own knowledge of deceit ...." Particular reference was made to his contribution to the siege of Breda (1637) when requesting a pay raise.

His first achievement in the field appears to have been during Frederick Henry's siege of 's-Hertogenboschin 1629, when he was asked to decipher intercepted Spanish letters in cipher by using his knowledge of Spanish (Christopher Joby (2014), The Multilingualism of Constantijn Huygens (1596-1687), p.78).

He was not always successful. At one time in 1634, he said ciphers of the king of Spain were "more difficult to conquer than" the king himself. (Akkerman and Langman, p.154)


While his library contained many books on cryptography, ciphers he designed for royalists during the English Civil War were simple homophonic substitution ciphers albeit with an extensive nomenclator (ibid. p.156). (I'm inclined to think such ciphers were the most practical after all. John Wallis also proposed a simple Caesar cipher when asked for an "easy cipher", as noted in "John Wallis and Cryptanalysis".)


(By the way, Joby (2014) discusses "code switching", which has nothing to do with cryptography and may be broadly understood as switching to different languages when quoting etc.)