Dorabella Cipher

In 1897, the composer Edward Elgar sent a short enciphered letter to his (much younger) lady-friend Dora Penny – because his nickname for her was “Dorabella”, this note has acquired the name “The Dorabella Cipher“. What it actually says is doubtless merely a trifle from one close friend to another (it’s hardly a Zimmerman Telegram, let’s be honest), but its inability to be deciphered has led to its status as an enduring and popular cipher mystery.

Elgar was fascinated by secret writing, even cracking a supposedly ‘uncrackable’ cipher published in Pall Mall Magazine: and part of his enduring fame arises from the way he concealed identities of various friends (as well as a well-known melody in counterpoint) in his famous Enigma Variations (which the German Enigma machine was named after, in homage to Elgar). Musicologists have managed to decrypt most of the secrets of the Enigma Variation: but what of his short enciphered letter?


Curiously, precisely the same cipherbet (‘cipher alphabet’) used here appears elsewhere in Elgar’s notes (which are riddled with cryptograms, puns, and deliberate misspellings). The letter-shapes are formed from a simple 24-letter symmetrical key (a pigpen cipher variant using 3 versions each of 8 rotated E-shapes, almost certainly a visual pun on Edward Elgar’s initials) – exactly the kind of thing a cryptologist would expect of a self-constructed cipher. The problem is that applying this (logical and apparently correct) key to the Dorabella Cipher produces an apparently nonsensical cleartext:-


The mystery of the Dorabella Cipher, then, is neither a whodunnit (because Elgar signed and dated it), nor even a howdunnit (because it seems that we already have the key to the cipher), but more like a “whodunnwhat” – though we can apparently decipher its text, we don’t know what it means, or even how to try to read it.

Currently, perhaps the most persuasive reading is that of Tony Gaffney (A.K.A. “Jean Palmer”), who proposes that the not-so-cleartext (above) was a written version of Elgar’s and Dora Penny’s shared private language, and so could only be read as a tricky combination of backslang, abbreviations, contractions, in-jokes, puns etc. Having said that, Tony’s attempt to reconstruct what it says remains somewhat tortuous: and so the mystery continues.

All of which talk of private language brings to mind the distinction the Annales historian Marc Bloch drew (in his posthumous book “The Historian’s Craft”) between intentional evidence (intended to influence others) and unintentional evidence (intended for an audience of one or less). I think Bloch’s idea was that the reliability of the evidence differed according to the use language was put to; and hence that some famous ciphertexts probably remain uncracked because the text they contain is unintentional evidence, too personalized a shorthand for anyone but an audience of one to read.

Will the Dorabella Cipher ever be cracked? Right now, I’d say probably no, simply for the reason that it is too short a text to do significant statistical analysis on, if (as I think likely) the plaintext was written in Elgar’s and Dora Penny’s shared private language. And that’s the difficult challenge I believe posed by many cipher mysteries: while enciphered intentional evidence can be too trivial, enciphered unintentional evidence can be simply too hard… not unlike trying to achieve triple-jump distance with a single leap.

Dorabella Cipher Links:-

There is also much more of interest in the moderated Yahoo group Elgar-Cipher.


  1. avatar mark stahley November 10, 2012 4:33 pm

    I think the symbols are sounds not letters. It’s difficult for me 2012 American, to know the exact pronunciation but I believe the first line starts ” Am I your Enigma?”
    and to make it harder, Enigma is spelled out in a musical way
    Dora stuttered so possibly that is how she said it without stuttering(?)
    am eI uer Iy-E,N-e, aA-e, ge-e, em-me,a-I ?
    Am I your enigma? in p-
    riory Y B I,” E E” eh. A Rene is PA Y Y in P-
    apa’s P-As? PS a P-A-I- M> Pa I? oui
    PA is Public Assylum where he gave weekly concerts before he was famous.

  2. avatar Narissa Andrews November 29, 2012 1:22 pm

    This has been solved by that Australian guy Tim Roberts and it makes total sense and Id accept it as the correct solution.

  3. avatar nickpelling November 29, 2012 1:46 pm

    Narissa: I wish it were so, but strongly suspect that it’s not –

  4. avatar Narissa Andrews November 29, 2012 1:57 pm

    Ill hold it as the best solution upto yet but ive got a uk composer looking at it at the moment with fresh eyes so see what he thinks.

  5. avatar PatiB December 22, 2012 7:00 pm

    I think it’s easy- they aren’t sounds, or even letters. They are 19th century icons… Each symbol represents one of them and/or an action: I met you, you turned away, you came back, we fell in love, it was wrong, we parted ways.

  6. avatar nickpelling December 22, 2012 8:25 pm

    Pati: Any evidence for this apart from it sounding right to you?

  7. avatar Andrej December 23, 2012 4:05 am

    Hello, Im currently working on a solution which is based on secret cryptic system I had as a kid. Back in 1897, the old alphabet was used, which only used 24 characters (J & W are not included). That suggests three sets of 8, which correlates with the 1, 2 and 3 squiggle symbols. Each of these than point in a certain direction, North, South, East, West, NE, NW, SE and SW. If we organize the letters in three squares of 3 by 3 leaving the center square empty. Depending on the number of squiggles we select the square, and according to the direction we choose the letter. The Question that remains, is in what order were the letters arranged in the squares?


  8. avatar nickpelling December 23, 2012 3:42 pm

    Andrej: what you’re describing is very similar to what is known as “pigpen cipher” or “masonic cipher”, and people have pointed out the similarity between this and the Dorabella Cipher’s alphabet many times before. Having said that, for me the biggest mystery about the Dorabella Cipher is that there’s no obvious reason it should have been non-trivial – as I recall, in July 1897 Elgar had only just met Dorabella, so why would he send her enciphered in a ridiculously arcane way? Similarly, the two had only just got to know each other, so all the so-called “decrypts” I’ve seen containing bizarre phrases in some alleged ‘private language’ make no sense to me.

    In my opinion, the answer is likely to be simple, playful, and obvious once you see it: really, we’re probably all looking straight through it!

  9. avatar Soroush Hoseinpour February 13, 2013 11:35 am

    He has sent a letter to his love and why it shouldn’t be a poem?
    Underlined words (love&love) in the original letter are same.
    These two (T)s are same too.

    what do you think?

  10. avatar nickpelling February 13, 2013 1:55 pm

    Soroush: the letter that the Dorabella Cipher was folded into was from Alice Elgar (Edward Elgar’s mother) to the Penny family, but I’ve never seen a copy of it, let alone seen any underlined words in it. Do you have a copy or a scan of this that I can see? Otherwise I can’t really make any comment on what you propose, sorry. :-(

  11. avatar soroush February 14, 2013 12:38 pm
  12. avatar Thomas Upton February 26, 2013 5:46 pm

    This is very close to J S Bach’s shorthand for musical notation. Elgar uses curves where Bach uses zigzags. It is a code, in the sense that longer work is hidden, but this is a quicker way than staff and notes. Look up “shorthand for musical notation.

  13. avatar nickpelling February 26, 2013 6:30 pm

    Tom: do you mean basso continuo?

  14. avatar Decrypter X March 6, 2013 5:44 am

    Guys: I think I just got it (or something similar) the first 3 “squigglies” mean “why”. Not saying anything else right now SUPER PUMPED!!!!!

    Decoder X

  15. avatar Decrypter X March 6, 2013 6:01 am

    Wait nevermind. My “solution has multiple outcomes. Sorry for false alarm. But I did find a way to write like this that I believe is almost virtually impossible to decode( unless you know how, of course). Not sure yet if I should release it. Let me know.

    PS when I write a cipher, I sign it “Encoder X” and my buddy who helps me “Encoder Y”.

    Pps when we decode ciphers, we sign them “Decoder X” and Decoder Y.

    Decoder X

  16. avatar Cat Darensbourg July 14, 2013 6:44 am

    Thank you for your article pointing out the cipher was pig-pen based, most likely. I did some sketches for how the alphabet might have been constructed and posted it on the “Ancient Cryptography” website if you would like a look. (My other, earlier theory that it might be a pig-pen, then a rail-fence, i have set aside for now, and the other cipher stuff is merely discussion at this point.)

  17. avatar Narissa Andrews July 16, 2013 2:15 pm

    The last 11 letters only have 7 different symbols.
    Say you give each symbol a letter, starting at the NE Triple symbol, we have ABCDBAECBFC

    I dont think thats ever going to spell a real word? A tune maybe?

  18. avatar nickpelling July 16, 2013 2:20 pm

    Narissa: “not long to it”? 😉

  19. avatar Narissa Andrews July 16, 2013 2:58 pm

    Yeah that works :) back to the drawing board.

  20. avatar Allan G. August 25, 2013 7:29 pm

    Posted this earlier but under wrong heading. Should be this one – The Dorabella Cipher.

    My decrypt is (with my punct.):

    Forli, Malvern.

    Bon ? (16th char is unresolved),

    A and Dai (=Alice and I) opin’ (=opine)
    met St Swithin eighty six.; wed at Brompton Oratory but owed takc …….. (final 8 chars dodgy due to ambiguous cipher chars)

  21. avatar Allan G. September 25, 2013 3:24 pm

    Revised decrypt (chars 1-39) now reads:

    ForlE [=pronounce as Forlee], Malvern Link.

    A. and Dai’s [=Alice and I’s] qk [=quick] o [?opinion?] –
    met St Stephen eighty six etc as above

  22. avatar Allan G. September 25, 2013 9:40 pm

    Final 8 chars (after ‘but owed takc’):

    Mogul ob’d[=obliged]

    Full decrypt

    Forle, Malvern Link.

    A and Dai’s qk o. Met St Stephen (e)ighty six. Wed at Brompton Oratory but owed takc. Mogul ob’d. (87 chars)

  23. avatar Šuruppag November 20, 2013 2:47 am

    After 30 years he had probably forgotten which signs he had assigned to each letter and so he simply reassigned them differently.

    Considering we found a version of it that is a simple substitution cipher in his notes, and he thought Dora would be able to solve it, I really doubt there is a complex stystem at work.

    So why is it that expert cryptographers cant easily solve it? I guess its brevity just really makes it hard to analyze. Too bad he didn’t write her a sappy wall of text.

  24. avatar nickpelling November 20, 2013 7:57 am

    Šuruppag: actually, most simple substitution ciphers can be cracked with a 30-letter ciphertext (or even less with a couple of really good guesses), so the Dorabella Cipher should be more than long enough to be cracked. So, the mystery is why it can’t be cracked despite its length, not because of it. :-)

  25. avatar Allan G. November 29, 2013 8:02 am

    Re my decrypt of 25 Sept.
    In a letter to his friend Kilburn, dated Aug 3 1897 (3 weeks after Dorabella), Elgar signed off in his usual way then, unusually, added the following:
    “Somebody has given me a ‘fountain pen’ & it’s not a complete joy – yet.”
    Suggest chars 24-27 (recovered as SQKO) = (abbr.) squeako (the sound of his new pen).
    Oh – and chars 78-79 prob = BS (giving TABS, not TAKC).
    Hey ho.

  26. avatar Bryan February 4, 2014 4:25 am

    I keep seeing people stating that Elgar’s signature is at the bottom of the Dorabella Cipher, yet if you compare it to his actual signature the two are nothing alike. I think it is actually a clue to decipher the text. I’m gonna play around with it a bit :)

  27. avatar , Rick A. Roberts February 22, 2014 6:14 am

    I believe that I have solved the Dorabella Cipher using Elgar’s Cipher Code mentioned in the article. The message reads,” BE IT CLEAR NOW YOU FINALLY KNOW WHY ON WEDNESDAY IN OUR QUAINT IN HOME SERMONS YOU DON’T SHOOT THE GOOSE THAT LAYS A GOLDEN EGG.”
    Edward Elgar was very witty and liked to use puns or play on words. He referred many times about, “Don’t Shoot the Goose That Lays A Golden Egg”. This I believe was mentioned in the, “Unitariun Register” also.
    A man and a woman found a special goose that laid golden eggs. They thought that they would become very rich. However, by cutting the goose, they thought that they could get all of the gold coins inside the goose at once. When they cut open the goose, there were no more golden eggs.
    Elgar was a famous compose who composed, “Pomp And Circumstance”, famous marching songs and other patriotic songs that are famous allover.
    Rick A. Roberts

  28. avatar Anton Alipov August 16, 2014 10:59 pm

    A different reproduction (allegedly photographic) of the Dorabella cipher is presented at the Elgar Museum’s website.
    Here’s my brief report:

    Has this photo (the Museum’s) been previously known? Is it original or faked (for whatever reason)?

  29. avatar nickpelling August 17, 2014 11:27 am

    Anton: thanks very much for leaving a comment and passing a link to your page. When I tried to track down Dora Penny’s Elgar donation at the London School of Music, an archivist there told me that several boxes of her early Elgar material were somehow lost (in Leeds? I can’t recall fully) en route to London, which presumably included the Dorabella Cipher. So – unless there has been some magical discover of this hitherto-lost material – the reproduction in her book is all we have to work with.

    Hence I suspect that what has happened here is that, in the absence of the real thing, a fake version of the cipher was mocked up to put on display as part of an exhibition… but I shall ask the curators about this, just to be sure. Will let you know what they say! :-)

  30. avatar Anton Alipov August 17, 2014 2:42 pm

    Thank you in advance!

    I also thought about this possibility, but the strange side of it is that, in our computer age, the simplest way to “fake” (or “reproduct”) something is just to make a scan/photo of it. In this case they could just print the image from Wikipedia, frame it beautifully and present to the audience. Why introduce additional labour to “mock” the cipher, moreover – with discrepancies?! One should expect more respect to the relics from a museum.

  31. avatar nickpelling August 17, 2014 3:04 pm

    Anton: well, people do all kinds of things for all kinds of reasons. My guess would simply be that the print-out didn’t look hand-written enough. :-)

    But I’ve already emailed the curators to ask, so hopefully we’ll get a definitive answer before long…

  32. avatar Vedran October 3, 2014 8:30 pm

    Hi Nick.
    So have you ever gotten a reply from them (regarding the photo)?

  33. avatar Anton Alipov October 21, 2014 5:35 pm

    Yes, very interesting to know if the museum explained what’s actually the case with this alleged photo…

  34. avatar Rick A. Roberts November 19, 2014 2:47 am

    Has anyone read my deciphering of the Dorabella Cipher that I posted on 22FEB2014 ? I believe that it is very accurate, and that it makes logical sense. Check it out. Thank you.

  35. avatar Mark Pitt March 4, 2015 7:02 pm

    The key to the Dorabella Cipher is ‘DORA X’.

    Insp. M Pitt

  36. avatar Mark Pitt March 11, 2015 9:54 pm

    Using the DORA X key provides the following:

    The first line of the Dorabella Cipher – Having lazy tranquil warm rest and recuperation LC crowd,ed kites…..

  37. avatar Allan Gillespie May 29, 2015 7:52 am

    I’ve posted a solution showing Running Key with stutter and indicator, explanation of the fullstop at char 65, reference to a mystery news ban etc. Also the name of the author. Link is

  38. avatar Anton Alipov May 29, 2015 10:21 am

    Of course I’m not a native English speaker, but, sorry, “Cry rane pfog ufeletharge” does not convince me. Or maybe this is something Irish?

    Considering the story in its entirety, I think that this “cipher” is nothing more than a mystification introduced by Mrs Powell herself.

  39. avatar mark pitt July 16, 2015 8:20 pm

    The solution to the Dorabella cipher is based on the dial cipher set at AO and the courage card set sequence. This is encoded into the cipher date. The initial 14 symbols decipher ‘to fix set ao set weekly’, minus most of the vowels. The remainder of the cipher then rotates through the dial cipher every 7 and translates via the music cipher. My full explanation is available if anyone is interested I have posted it on the Elgar cipher yahoo group. I explain in the article my breakdown of the cipher date and the indication of the dial setting AO and show how I made my decipherment. This is all supported by the 1896 Pall Mall Magazine articles regarding ciphers as read by Elgar, which hugely support my decipherment.

  40. avatar mark pitt July 16, 2015 8:35 pm

    In short the date contains a pi sign at start, indication of dial cipher. There is Elgorod encoded into the date as explained in my article which indicates the ingenious first set of the dial I.e. AO so to complete both the names Elgar and Dora (backwards). Then a music note (the no. 9) and then the 7(repeat). The main lines in the courage card set match the main lines in the date. AO is also the dial setting in the pall mall magazine (1896). Read my article as posted to the Elgar cipher yahoo group on 14/7/15.

  41. avatar Dan Biddle July 22, 2015 9:25 pm

    Place a set of musical staves over the top of each line, every symbol is either a crotchet, quaver or whatever (I haven’t the musical knowledge to work it out) and it will form a composition.
    Elgar was a composer, from my hometown no less, so this seems like the obvious thing that it will be.

  42. avatar Gunnar Johannson September 8, 2015 5:20 am

    Hi I don’t know music either but they are the finger placements on the keys of 3 different instruments. I’d love to hear the song. He made a little diddy to that young lady she likely asked him I would love to hear your new stuff and he sent the keys order for her to figure out til he came out with his new music.

  43. avatar Aymii September 13, 2015 10:24 am

    I can’t help but remember back to when my daughter was learning to read music, she would make little “scribbles ” like this to remember which notes and fingers to place .. he was in love with the woman , of course he wrote music for her that she would understand .. Great website you have here by the way

  44. avatar nickpelling September 13, 2015 11:53 am

    Aymii: well, there you have the paradox. On one fragment, the lettershapes are definitely used as a cipher, while on another (the Liszt fragment), they appear to be used as music. And on the third (the Dorabella Cipher itself)… we just don’t know either way.

    Thanks for the compliment about the website, though it’s not really as neatly organized as I would like. 😮

Leave a Reply