[Paleopsych] New Scientist: Breaking Elgar's enigmatic code

[Premise Checker]

Breaking Elgar's enigmatic code
http://www.newscientist.com/article.ns?id=mg18424792.600
25 December 2004

    IT IS a story with all the makings of a blockbuster novel: a brilliant
    composer, an attractive woman, a secret letter and a mystery that has
    lasted 100 years. This story, though, is real. The composer was Edward
    Elgar, the renowned English musician who died 70 years ago. The young
    women was Dora Penny, a family friend. And the mystery? A short coded
    letter that, his music apart, remains one of Elgar's most enduring
    legacies.

    A study of the composer's papers reveals that for most of his life he
    was fascinated by cryptography. His letters and music scores, for
    example, are dotted with codes and anagrams. And the title of his
    Enigma Variations, first performed in 1899, hints at his delight in
    cryptic puzzles. He teasingly suggested that the melody on which his
    variations are based forms a counterpoint or matching voice to a
    well-known tune that is present in the piece only by implication. None
    of the many suggestions as to what this tune might be, including Auld
    Lang Syne and Rule Britannia, ring true, so the enigma remains.

    Yet Elgar left another, more intriguing, mystery. In 1896, while
    struggling to achieve recognition as a composer, he met Dora Penny, a
    young woman 20 years his junior. The daughter of a clergyman recently
    returned from Melanesia, she shared Elgar's interests in kites,
    cycling and football (they both supported Wolverhampton Wanderers).
    They exchanged letters and in July 1897, the halcyon summer of Queen
    Victoria's golden jubilee, Elgar sent her a letter in code. Its
    curious symbols, possibly inspired by Arabic script, seem to be based
    on the double-arched, cursive E in Elgar's signature (see page 58).
    Now known as the Dorabella cipher, after his nickname for Dora Penny,
    it remains unbroken. It has proved one of cryptography's most enduring
    puzzles.

    It is not surprising that Elgar was fascinated by ciphers.
    Code-breaking techniques have notable similarities to the process of
    composing formal harmony and counterpoint. Both activities involve
    sifting, shuffling and transposing parallel sequences of code or notes
    to find the best fit. For musicians, the challenge is to devise lines
    of music that sound pleasing on their own and also sound harmonious
    when played together. Take the round Frère Jacques, for instance. This
    is a simple example of repeated patterns of notes that overlap with
    each other. In a more complex manifestation it becomes a fugue.
    Experienced composers, like code-breakers, build up a repertoire of
    templates and patterns that can be tested and modified to suit.

    Links between music and ciphers go back centuries. One of the earliest
    known treatises on cryptography was written by Al-Kindi, an
    accomplished musician who was one of a group of Baghdad scholars
    working in the 9th century during the golden age of Islamic
    scholarship. Al-Kindi devised a revolutionary system for breaking
    substitution ciphers - messages encrypted by replacing, say, A with P,
    B with Q, and so on - based on an analysis of letter frequencies.

    Scholars had noticed that in the Koran, certain letters appeared with
    greater frequency than others, and they compiled a chart that ranked
    letters from most to least frequent. Al-Kindi realised he could use
    this chart to help crack substitution ciphers by replacing the most
    commonly used character in the ciphered text with the most common
    letter in Arabic, and then working through the chart to the least
    frequent.
    "Known as the Dorabella cipher, Elgar's code is one of cryptography's
    most enduring puzzles"

    Aware of this weakness, the Italian composer and architect Leon
    Alberti revolutionised cryptography in the 15th century with the
    invention of the cipher wheel. This is a rotating disc set within
    another disc, each with an alphabet inscribed around its rim. Match,
    say, A on one disc with C on the other and it becomes easy to create
    coded messages. Better still, it is simple to reset the position of
    the wheels at intervals to eliminate frequency patterns. This, Alberti
    thought, would make messages coded with his wheel impossible to break
    unless the settings were known. An electronic version of this device
    with multiple discs lay at the heart of the Enigma machine, a German
    cipher device used in the second world war and named after Elgar's
    variations by its German inventor, Arthur Scherbius.

    Cryptographers have also co-opted musical notation into their service.
    Everyone from 16th-century spies to illegal gamblers in 1950s New York
    have sent messages disguised as music. A typical cipher from the 18th
    century matches the first 12 letters of the alphabet to an ascending
    scale of 12 crotchets or quarter notes, and the next 12 letters to a
    descending scale of 12 quavers or eighth notes.

    Musicians, too, seem to enjoy adopting simple codes when composing
    melodies. Many pieces of music contain motifs based on initials, words
    or short phrases written using the seven musical notes A to G. German
    musical notation also allows the addition of S and H - equivalent to
    E-flat and B. The musical cipher B-A-C-H is common, and both the BBC
    and the composer Dmitri Shostakovich have used their initials as coded
    musical signatures. However, note-based codes can be used for more
    than just adding labels: Robert Schumann and Alban Berg both used
    short coded motifs in their compositions as references to illicit love
    affairs, and in the early 19th century John Field, the celebrated
    Irish composer of nocturnes, thanked some particularly generous dinner
    hosts with melodies based on B-E-E-F and C-A-B-B-A-G-E. Elgar made use
    of this technique too. In 1885 he composed a duet for two sisters
    based on their family name G-E-D-G-E, and 15 years later he
    mischievously ciphered the names of some of his critics into the
    demon's chorus in his oratorio The Dream of Gerontius.

    It is also possible to create a cipher using musical rhythm. Morse
    code has obvious potential. The Australian-born composer Barrington
    Pheloung used conspicuous Morse code rhythm patterns in his music for
    the UK television series Inspector Morse, and even encoded the
    murderer's identity into the incidental music for some episodes.

    It is known that Elgar attempted to learn Morse code and it is
    possible that he used it in his music. For example, it could explain
    his Enigma theme, which has a distinctive rhythmic structure that
    suggests calculated design: the first motif is followed by itself
    reversed, forming a rhythmic palindrome: two short notes, two long
    notes; two long notes, two short notes. This pattern is repeated three
    times in total.

    Its symmetry is striking, a feature that would do credit to a
    20th-century modernist but it is odd for its time. So what might this
    pattern represent?

    In Morse code two dots represents I and two dashes M. So the motif
    could be read as a repetition of "I am, am I?" This accords with
    Elgar's admission that the theme represented the sense of loneliness
    and inadequacy he felt at the time he wrote it, and the observations
    by others that Elgar had been deeply hurt by cruel put-downs from
    critics. It suggests a heartfelt but defiant response.

    Morse code may also crop up in a cryptic letter that Elgar sent to
    Dora in 1901. Within the message he inserts short, distinctive motifs
    from his Enigma Variations, in particular a fragment from the
    Dorabella variation and the opening of the initial theme. The segment
    reads: "Whether you are as nice as", three short notes, three short
    notes, "or only as unideal as", two short notes, two long notes.
    Interpreted as Morse code, these mysterious notes become SS and IM,
    inviting the interpretation: "Whether you are as nice as sugar and
    spice or only as unideal as I am".

    What clues do we have to the meaning of the Dorabella cipher itself?
    Analysis of the frequency distribution of the characters in the
    message reveals a pattern typical of a substitution cipher, but all
    attempts to break it based on this assumption have failed. Writing in
    the journal The Musical Times in February 1970, cryptographer and
    musicologist Eric Sams analysed the cipher for telltale patterns of
    letter groups, such as sequences of the form xyyx that have a limited
    number of possible vowel-consonant equivalents (S-E-E-S, for instance)
    and which could offer clues to the cipher. Sams did not get very far,
    however, and his results are unconvincing.
    "Elger mischievously ciphered the names of some of his critics into
    the demon's chorus"

    Elgar appears to have offered the key in an exercise book containing
    the address "Tiddington House", where he lived from 1927. He listed
    the symbols used in the Dorabella cipher matched against the letters
    of the alphabet. The cipher follows a simple pattern, with single,
    double and triple E-like characters, each in eight possible
    orientations - upright, rotated 45 degrees clockwise, 90 degrees
    clockwise and so on. This gives a total of 24 potential characters,
    and as with many ciphers, I and J share a single character, as do U
    and V. Samples on the page written using this code reveal the messages
    M-A-R-C-O E-L-G-A-R (Marco was his pet spaniel) and A V-E-R-Y O-L-D
    C-Y-P-H-E-R. But when applied to the Dorabella cipher this key does
    not generate anything that makes obvious sense.

    Since simple substitution fails, it seems likely that Elgar used a
    form of double-encipherment, such as letter substitution followed by
    letter shuffling, perhaps coding the message in alternate letters (for
    example, the 1st, 3rd, 5th and so on). But the appearance of a
    repeated four-letter group makes letter shuffling unlikely.

    This does not rule out the use of multiple substitution alphabets
    using a keyword, however, a sort of manual version of the Alberti
    cipher wheel. If the keyword is D-O-G, for example, you might carry
    out letter substitution by first coding A as D, B as E and so on, then
    coding A as O, B as P and so on, finally coding A as G, B as H and so
    on.

    The detective work is complicated by Elgar's eccentric spelling. For
    instance, he once wrote of Dora Penny "warbling wigorously in
    Worcester wunce a week". And what appears to be a potential cipher
    message in the Tiddington House exercise book reads: "DO YOU GO TO
    LONDON TOMORROW?" A message lacking E, the most frequently used letter
    in English, seems deliberately designed to confuse. Elgar even put a
    small mark below each letter O, perhaps hinting that they might be
    dropped from the message altogether.

    Did he use these kinds of tricks in the Dorabella cipher? It seems
    likely. Seventy years after his death, Elgar's most intriguing
    composition is yet to be cracked.

    From issue 2479 of New Scientist magazine, 25 December 2004, page 56