In 1840, Edgar Allan Poe wrote an article in the
Alexander's Weekly Messenger, a Philadelphia newspaper,
where he challenged the readers to submit their own
substitution ciphers which he would decrypt. Initially,
he received cryptograms from around Philadelphia, but
soon after, they came in from all over the United States.
He published many of the cryptograms and their solutions
in fifteen numbers of the Alexander's Weekly Messenger.
The next year, Poe published his essay
called "A Few Words on Secret Writing" in
Graham's Magazine, in which he commented on the response
to his cipher challenge (see download below). The essay
also gave birth to the famous quote that "human
ingenuity cannot concoct a cipher which human ingenuity
cannot resolve".
In the 19th century, most people
considered secret writing and cryptography as a
mysterious esoteric art, and Poe had sparked a great
interest in cryptography with the general public. Thanks
to Poe's publications, cryptogram puzzles became popular
in newspapers and magazines. Inspired by the success of
the cryptograms and the interest in his essay, he decided
to write a short story that involved cryptography.
After writing The Gold-Bug, he
submitted the story to a writing contest, winning the
grand prize and $100. The story was published on June 21,
1843, in Philadelphia's Dollar Newspaper. It is regarded
as the first important publication in popular
non-technical literature that incorporated cryptography
in its story line. The Gold-Bug contains a detailed
description of how to solve a cryptogram using letter
frequency analysis. The story was an instant success and
helped popularize cryptography in the 19th century.
Edgar Allan Poe's The Gold-Bug is as
iconic to cryptography in literature as David Kahns
Codebreakers is to historical publications on the
subject. It became one of his most read and best known
stories. Many readers have set their first steps in
cryptology after reading Poes story - some even
became important codebreakers - and more than a few
writers were inspired by Poe, to write their own story
with secret writing and encrypted messages in it.
The Story
The main character in the story is
William Legrand, a man who lives at Sullivan's Island,
near Charleston, South Carolina, to escape from his
misfortunes. Legrand discovers a brilliant gold-colored
bug, but lends it out to someone else. When his friend,
the narrator of the story, visits Legrand, he is told
about the rare bug with a death's-head on its back, and
Legrand draws him a picture of the bug on a piece of
paper. A short while later, Legrand asks his friend to
come visit him immediately. Upon the friend's arrival,
the strangely behaving Legrand asks his friend to follow
him on an expedition into the woods near some rocks, to
search for a treasure. Afraid that Legrand has lost his
mind, the friend decides to accompany him out of concern
for Legrand's health. As it turns out, Legrand
accidentally had discovered a secret message in invisible
writing on the paper he used to draw the bug. Legrand
later explains his friend how he found the message and
how he was able to decrypt the message that started his
quest for a hidden treasure.
The Gold-Bug is not only an exciting
story about the discovery of an old treasure, but also a
great introduction to cryptography and codebreaking. It
tickles the reader's curiosity and Poe gives a detailed
description of how to decipher the cryptogram. While
doing so, he also provides the solution. However,
deciphering the message yourself is even more exciting
than reading how Legrand did it in the story. Can I
challenge you, just as Poe did, to decrypt Legrands
message, composed more than 160 years ago?
Decrypting the Message
Rather than just reading Poes
story, I will show you the technique and give you the
chance to do it all by yourself. It might be usefull to
read The Gold-Bug first, as the story might provide information
that will help to solve the cryptogram, but read
only to where the cryptgram appears. Don't cheat
by reading or peeking any further! Don't search the
Internet for Poe or The Gold-Bug, as this will also spoil
the fun.
The message is encrypted by
mono-alphabetic substitution, a cipher where the letters
of the alphabet are replaced by other letters or symbols.
We can calculate all possible combinations for 26 letter
of the alphabet, replaced by 26 symbols: the first letter
is substituted by one of 26 symbols, the second by one of
the 25 remaining symbols, and so on (26 x 25 x 24 x
3 x 2 x 1). In total, this gives
403,291,461,126,605,635,584,000,000 different ways to
allocate 26 symbols to 26 letters. How on earth could we
possibly decipher such a cryptogram? For centuries,
substitution ciphers were regarded as unbreakable...but
it is easier than it looks.
Although there are trillions of ways to
allocate a set of symbols to letters, there are only a
few ways to combine vowels and consonants in a natural
language. Strict rules determine which letter
combinations are possible and which are forbidden. The
syntax prescribes in what order words should be written
and which conjugations should be used. When we substitute
letters with symbols, those symbols still follow all
these rules and thus create patterns that we can detect.
Just as certain letter combinations are impossible (ZLG,
XOJ,...), so will certain symbols avoid one another. Just
as it is evident that the same vowels will always fit
within a given set of consonants (THR?ST, G??D...) so
will attract certain symbols each other. But where do we
start?
The mystery weapon to solve our message
is letter frequency analysis, the basis of all
codebreaking. Each language has its own typical
distribution of letters in a text. In English, the letter
e is by far the dominant letter, with an average of 12.7
percent. If we locate some of the most frequent vowels or
consonants in the ciphertext, or find recurring symbol
combinations, then the rules of the language will give us
strong leads to the words they are used in or the letters
they represent. Below, you'll find the letters of the
alphabet, ordered from most frequent at the left to least
frequent at the right (Poe used an older and slightly
different frequency table).
E T A O I N S R H L D C U M F P
G W Y B V K X J Q Z
For Legrands message, start by
taking a sheet with squares and write down the secret
message with a pen. Leave some blank rows between each
row of the message, to write your solution underneath the
symbols with a pencil (easily corrected with a gum).
Count how many times each of the symbols appears in the
cryptogram and write down the results in a table, ordered
from most to least frequent. You will see that one of the
symbols clearly stands out. This is the first major clue.
That most frequent symbol represents without doubt the
most frequent letter of the alphabet. Write your first
results underneath the according symbols on your message
sheet.
Next, you try to spot recurring
combinations of symbols. The most commonly used words in
English are, in order of frequency: THE, OF, AND, TO and
IN. Thus, you have to search for identical combinations
of symbols that contain the most frequent letter you
already found. You should spot each THE quite easily. If
so, you have discovered the solution for two more letters
that are used frequently. Make yourself a second table
with all the symbols and their corresponding letters you
already found.
By now, you should be able to find more
and more letters by completing fragments. If, for
instance, you find a fragment T?EE, it is not hard to
imagine what should follow the letter T. Vowels twins
(AA, EE, OO ) are common but not that many different
words contain such pairs. Try to find those words. If you
cant see it immediately, try all letters of the
alphabet until you get something readable. Each new
letter will help you to reconstruct more and more
fragments. Be patient. It could take some time before a
word appears in front of you, but once you have four or
five letters, youre in a straight line to the
finish.
Good luck...and make sure you don't get
bitten by the bug!
Note: in the original edition, one
symbol "(" was not printed near the end of the
message, right after 9;48; although Legrand
describes just that missing symbol to assist in finding a
word. Since the story refers to that symbol, it is
unlikely that it was omitted on purpose and probably got
lost during the publishing.