Tony Gaffney, a chess player / tournament organizer I knew back in the early 1980s when playing for Hackney Chess Club, made some fascinating comments to my recent blog post on The Subtelty of Witches and Eric Sams’ attempted solution to the Dorabella Cipher.
Firstly: having spent a looong time in the British Library looking at ciphers (you’ll see why shortly), Tony was happy to tell me that it in fact has three encrypted books, all using simple monoalphabetic ciphers:
(1) MS Add. 10035 “The Subtelty of Witches” (Latin plaintext),
(2) Shelfmark 4783.a.30. “Ebpob es byo Utlub, Umgjoml Nýflobjof, etc. (Order of the Altar, Ancient Mysteries to which females were alone admissible: being part the first of the Secrets preserved in the Association of Maiden Unity and Attachment.)” London, 1835. (English plaintext)
(3) Shelfmark 944.c.19. “Nyflobjof es Woflu” (Mysteries of Vesta), pp.61, London 1850 (?). (English plaintext).
Secondly: without realising it, I had already seen an early version of Tony’s own proposed Dorabella decipherment in the comments to the Elgar article on the BBC Proms website, attributed to one “Jean Palmer”. You see, back in 2006, this was the pseudonum Tony used to write (and POD publish through authorsonline) a book containing a thousand (!) furtively ciphered messages that were placed in (mainly Victorian) newspapers’ personal columns: I shall (of course) post a review of this “Agony Column Codes & Ciphers” here once my freshly-printed copy arrives.
It turns out that Tony is also a frequent poster (under the name “Tony Baloney”) to an online code/cipher cracking forum called Ancient Cryptography I was previously unaware of (probably because its definition of “ancient” seems to extend only as far back as 1450, Bible Codes [pah!] excepted). The forum has specific threads devoted to the d’Agapeyeff Cipher, the Beale Papers, Zodiac Killer Ciphers, and the Kryptos Sculpture (for example), as well as some delightful oddities such as a link to recordings of shortwave Numbers Station broadcasts (coded intelligence messaging). If you want a friendly online forum for discussing attempts to break these historical ciphers, this seems like a sensible place to go.
But back to Tony Gaffney: given that he deciphered a thousand (admittedly mainly monoalphabetic substitution) messages, it should be clear that he is no slouch on the decrypting front. Which is why it is interesting to lookat the latest version of his proposed solution to the Dorabella Cipher. As far as I can tell, this involves simply using exactly the same cipher crib as appears in Elgar’s notebook (?), but interpreting the text that comes out as having been written in a kind of phonetic-style backslang. Here are the two stages (note that the hyphens are inserted as part of the interpretation, not part of the transcription):-
Deciphered: B-ltac-ei-a-rw-unis-nf-nnellhs-yw-ydou
Anagrammed: B-lcat-ie-a-wr-usin-fn-nshllen-wy-youd
Plaintext: B hellcat i.e. a war using effin' henshells(en)? why your
Deciphered: inieyarqatn-nte-dminuneho-m-syrr-yuo
Anagrammed: intaqraycin-net-dminuenho-m-srry-you
Plaintext: antiquarian net diminuendo?? am sorry you
Deciphered: toeh-o-tsh-gdo-tneh-m-so-la-doe-ad-ya
Anagrammed: theo-o-ths-god-then-m-so-la-deo-da-ay
Plaintext: theo o' tis god then me so la deo da aye
On the one hand, I’d say it is more plausible than Eric Sams’ proposed solution: but on the (inevitably negative) other hand, it doesn’t quite manage to summon the kind of aha-ness (AKA “smoking-gunitude“) you’d generally hope for – as Tony’s book no doubt amply demonstrates, the point of a secret love note (which is surely what Elgar seems to have sent Dora Penny?) is to be both secret and to convey something which could not openly be said. But is this really it?
Some people like to say that the real point of tackling apparently unbreakable ciphers is to be found in the travelling rather than in the arriving – that the real prize is what we learn about ourselves from butting our horns against that which is impossible. To which I say: gvdl zpv, bttipmf.
Not a very nice thing to say 😉
Interesting article, though 🙂
True – but perhaps if everyone swore in cipher, the world would be a better place. 😉
Glad you enjoyed the post! 🙂
my my caesar, what a negative one (-1). ;0)
Nice writeup…
I looked at the Dorabella msg. Any chance it is a snippet of a musical score… a shared melody between the two of them? ….
That seems like a pretty complex puzzle otherwise for a man to expect his young lover to solve….
Though many (including me!) have suggested that it might possibly be some kind of enciphered music, on further reflection I really don’t think it is.
Given the symmetrical nature of Elgar’s “E/EE/EEE” cipherbet (formed from 8 x 3 rotated and repeated shapes), I do wonder whether there might be some kind of shape-driven cipher system going on – i.e. something like “rotate each row or column round after every letter”, so that the alphabet repeats every three or eight symbols. Note that there’s a four letter repeat within the message, and overall it looks to be just a little more than just monoalphabetic, so this would be just about the right level of complexity for a self-made cipher system. Basically, might the Dorabella cipher be a kind of rotating pigpen? I haven’t yet checked, but it’s a plausible hypothesis…
Cheers, ….Nick Pelling….
It reminds me of a game we played where you lay one piece of paper over another, the edge of one somewhat lower. Then write the message across the division. You can still read it if you have the top half, but working out the lower can be really difficult, especially if the top sheet is laid at an angle.
Diane: curiously enough, that’s not too far from the method the Knights Templar used to verify the first “cheques” to the Holy Land. In fact, these messages then mutated into simple ciphers split in two… but that’s another story entirely. Note that whereas most (Knights Templar + cipher) stories are simply novelistic fabrications, this one comes from a historian I met who spent a lot of time in the French Templar archives.
Alright Nick… solve this for me:
345-180-65-500-40-(73 1/3)-65-80-240-40-15-250-80-125-75-200-(31 1/4)-90?
44-208-64-32-68-64-128-(10 2/3)-68-130-64.
176-96-176-200-552-288-104-504-192, 168-192-64-920-320-(106 2/3)-90-24-304-136-760-320-80-170-24…
[8-4-5]
I looked at the English letters that have been derived from the code. When I ran a comparison with the standard frequency of letters in the English language, the correlation was almost perfect. The only outlier was the letter n, which appears most often in the cipher.(6th in a standard frequency dist.) Otherwise the concurrence suggests that the deciphered clear-text is in fact English. The question of course, is still what the heck does it mean.
Follow up to above comment. Since running original comparison of Dora to Standard English letter frequencies, I came upon a freq. table for old English. Turns out that my outlier “n” is the most frequent letter in Dora and old English! I don’t believe Dora qualifies as strictly old English, but likely is somewhere in between. It is interesting, however, that “n” is at the top of both frequencies.(6th in standard English.)
Caesar seems out of character.
Read in German:
Ich habe was anderes raus gefunden..eine verschlüsselte Botschaft.Kam aber noch nicht weiter..bisher dieses:
Erste reihe:Seit heut nutz Pice Saffee ihr Haus.
Zweite reihe:stuhl…
Fortsetzung folgt,sobald ich es fertig habe.
Ich nutze dass finger Knochen Alphabet..
Can you crack my code? Will post hints later. Everything is in Spanish and phonetic spelling of Standard Arabic.
Gato Zouga Reciber Primera Comer Morir Zougua Comer Bicicleta Morir Burt quaal Gato Salir Bicicleta Saltor Primera Gato Reciber U-M Th’man Th’man Beber Comer Comer Gato
http://linceicipher.com
http://linceicipher.com/2015/03/12/two/
No prizes awarded. Yet.