It was the Spartans, the most warlike of the Greeks, who established the first system of military cryptography. As early as the fifth century B.C., they employed a device called the 'skytale' [which] consists of a staff of wood around which a strip of papyrus or leather or parchment is wrapped close-packed. The secret message is written on the parchment down the length of the staff; the parchment is then unwound and sent on its way (@ Kahn 82).
Francis Bacon, in Book VI of his 1623 Latin edition of
The Advancement and Proficience of Learning, uses the text of a
"Spartan letter sent once in a Scytale or round cypher'd staffe" as an
example in his section on "cyphars" (1640
translation, 268). Specifically, the text is used in a demonstration
of his "Bi-literarie Alphabet" cipher method, a means of enciphering a
message and concealing its presence by using two alphabets, the corresponding
letters of which differ slightly in appearance. The missive sent by
"cypher'd staffe" serves as the secret message: Bacon calls this the
"interiour letter". It is concealed in a quote "taken out of the first
Epistle of Cicero, wherein a Spartan Letter is involved." Cicero's
text serves as the "exterior letter" by which Bacon's secret message
example is disguised or steganographically concealed [Greek: stegos
= "roof", "cover"].
Now, in Julius Caesar's Gallic Wars, Caesar
tells of a time he sent a messenger with a letter to be delivered to
Cicero. "The messenger was instructed, if he could not approach, to hurl a
spear, with the letter fastened to the thong, inside the entrenchment of
the camp" (v.48). Given what is to follow, it is rather interesting that
the example texts chosen by Bacon in the explication of his method of
steganographic "bi-literarie" ciphering are associated with Caesar,
"cypher'd staff", and delivery of a message by means of a Spear.
Julius Caesar "invented" the first substitution cipher,
which bears his name to this day. To use it, one need only shift or
cyclically displace an alphabet a certain number of positions with respect
to itself. The "plaintext" or secret message is then enciphered by
substituting for each letter the corresponding letter from the shifted
alphabet. An example of a 21 letter alphabet key for this simple
substitution Caesar cipher method is shown below (capital letters are
"ciphertext", lowercase the "plaintext" substitution to make for each when
deciphering). This is the key used to solve the first 25 letters of the
acrostic cypher found in the Sonnets frontmatter.
The Testament of Love, published in 1532 and previously attributed to Geoffrey Chaucer. On the basis of the presence of an acrostic "signature", Thomas Usk is now recognized as the true author. The initial letters of each chapter spell "Margarete of Vitrw, have merci on thin[e] Usk". The spelling is somewhat "irregular", yet it is considered sufficiently accurate.
Hypnerotomachia Poliphili, published anonymously in 1499. The first letters of the 38 chapters spell "Poliam frater Franciscus Columna peramuvit" or "Brother Francesco Colonna passionately loves Polia". There was a Dominican monk by the name of Francesco Colonna who was still alive at the time the book was published. His acrostically written name is considered proof of authorship. Kahn states: "It was thus perfectly possible for Francis Bacon to have used steganography to simultaneously conceal and reveal his authorship of the Shakespeare works" (874).This Arte of Cypheringe , hath for Relatiue, an Art of Discypheringe ; by supposition vnprofitable; but, as things are, of great vse. For suppose that Cyphars were well mannaged, there bee Multitudes of them which exclude the Discypherer . But in regarde of the rawnesse and Vnskilfulnesse of the handes, through which they passe, the greatest Matters, are many times carryed in the weakest CYPHARS .
Bacon's 1605 Advancement of Learning was rewritten in Latin, greatly expanded, and issued in 1623 (the year Shakespeares Comedies, Histories, & Tragedies was published) as De Dignitate et Augmentis Scientiarum . An english "translation" of Augmentis was published in 1640. At the end of the enlarged section on cryptology (now 7 pages rather than just one paragraph), Bacon elaborates on the necessity for using ciphers which
may be managed without suspition. For if Letters Missive fall into their hands, that have some command and authority over those that write; or over those to whom they were written; though the Cypher it selfe bee sure and impossible to be decypher'd, yet the matter is liable to examination and question; unless the Cypher be such, as may be voide of all suspition, or may elude all examination.
It should be noted that we find this
"cyphar" sub-section at the conclusion of section three, whose subject matter is
"the measure of words [which] hath brought us forth
an immense body of Art, namely Poesie ; not in
respect of the matter (of which we have spoken before) but in respect of stile
and the forme of words, as Metre or Verse..." (263). Thus, Bacon
categorized cryptography as a form of "Poesie".
He considered "POESIS" to be one of the
three fundamental "partitions" of Science. It is given a side of the triangle of
fundamentals found at the base of the Light pillar, "SCIENTIAE", on the cover of
his Advancement and Proficience of Learning .
The word is upside down. The corresponding word on the other triangle, "HVMANA"
or "human", is also upside down. Next to each triangle is a black-eyed owl,
holding a lit torch. Owls are symbolic of secrecy, since they hide during the
day and only move about in the cover of darkness. And what does an owl say?
But it may be, that in the enumeration, and, as it were, taxation of Arts, some may thinke that we goe about to make a great Muster-rowle of Sciences, that the multiplication of them may be more admired... Neither have we (in our opinion) touched these Arts perfunctorily, though cursorily; but with a piercing stile extracted the marrow and pith of them out of a masse of matter. The judgement hereof we referre to those who are most able to judge of these Arts. For seeing it is the fashion of many who would be thought to know much, that every where making ostentation of words and outward termes of Arts, they become a wonder to the ignorant, but a derision to those that are Masters of those Arts: we hope that our Labours shall have a contrarie successe, which is, that they may arrest the judgment of every one who is best vers'd in every particular Art; and be undervalued by the rest.
Acrostics were common among the Greeks of the Alexandrine period, as well as with the Latin writers Ennius and Plautus, many of the arguments of whose plays were written with acrostics on their respective titles. Medieval monks were also fond of acrostics, as were the poets of the Middle High German and Italian Renaissance periods (@ Britannica, emphasis added).
Acrostics were popular at the time the Sonnets were published and they
generally took a person's name as their subject. Commendatory or explanatory
verses which acrostically spelled out the author's name were occasionally placed
at the beginning of books.
When an author uses the last letters of
each line to acrostically spell words, it is called telestich. An old example of this can be seen in the
Preface to the "Aenigmata Aldhelmi", or "Aldhelm's Riddles". The "Aenigmata" are
found in Aldhelm's treatise on prose, the Epistola ad
Acircium de Metris, written around 695. The first and last letters of each
line of the "Præfatio" are capitalized and they form a sentence which includes
the author's name. Aldhelm also used both acrostic and telestich methods in the
prologue of his poetic Carmen de Virginitate, where
the first letters reading down and the last letters reading up form a sentence
describing the contents of the poem.
Joshua Sylvester wrote many acrostic poems using his own name as well as the
names of others. Sylvester was best known for his translations of du Bartas'
French Divine Weekes and Workes, first published as
a collection in 1605, then reissued in 1608, 1611, 1613, 1614, &c. The
following acrostic/telestich sonnet was published posthumously in a collection
of Sylvester's work (@ 322):
J f patience true could termine passion's war R--- O ur thankefull Harpe had tendred long-a g O--- S ave that, our Griefs, whose deep-gulfs never eb B--- U nto you sAcred, by the which you se E--- A h, muse not, then, if all our Muse-work favou R--- H eart sad, Art bad; yet pray you read the res T--- S o deare Mecaenas, if your patience daig N--- Y our praises due to publish farre and n I--- L ifting your Name, the glory of your Sto C--- V nthrall to Time, for, Time that tryeth s O--- E lse had th' old Hebrews and brave Worthies al L--- STones wear,steel wasts,too weak to bear their glorie S--- E ven so devout as wee are found to do O--- R ecording loftie though wee low begu N---
Sylvester's association with "Anthonie Bacone" is interesting in the context of the present discussion because Anthony was a professional "spy" and cryptologist. He was an agent of Sir Francis Walsingham's Secret Service and was trained at Walsingham's cryptology school. Anthony traveled extensively and sent enciphered dispatches containing news of import back to England. He and his brother Francis were close friends, as can be seen in the dedication of the first edition of Francis' 1597 Essayes to "Anthony Bacon his deare...louing and beloued Brother" (@). Francis further dedicates them,
such as they are, to our loue, in the depth whereof (I assure you) I sometimes wish your infirmities translated vppon my selfe, that her Maiestie mought haue the seruice of so actiue and able a mind, & I mought be with excuse confined to these contemplations & Studies for which I am fittest....
On the title page of the 1605 edition of Deuine Weekes & Workes, a motto may be seen encircling a scene: "Acceptam Refero Lvcem: Sine Lvce Silesco" or "Light takes Credit for Its Reflection: Without Light, Silence". The translator seems to be indicating his dependence on the original work for the Light or textual substance of which his words are only a reflection or reproduction in another language. A crescent moon appears at the center of the "emblem" (shown below). The Bacon Brothers' family crest included a crescent moon.
It should be understood that Francis Bacon used "many images drawn from light and darkness, the contrast between artificial and natural lights, and other 'light' effects, which constantly recur in his writings. [...] When thinking of mental activity, some picture of light seems nearly always to come before him" (@ Spurgeon 17). In the first of his Essayes, "Of Truth", we are told that
truth is a naked and open daylight that doth not show the masques and mummeries and triumphs of the world half so stately and daintily as candlelights. [...] The first creature of God in the works of the days was the light of the sense; the last was the light of reason; and his sabbath work ever since is the illumination of his Spirit. First he breathed light upon the face of the matter or chaos; then he breathed light into the face of man; and still he breathed and inspireth light into the face of his chosen. [...] Certainly it is heaven upon earth to have a man's mind move in charity, rest in providence, and turn upon the poles of truth (@).
In the 1608 second edition of Divine Weekes, a different "emblem" (shown below) was
used along with a shortened motto: "Acceptam refero." The words, like most
Latin, are laden with meaning. In addition to the above, "Acceptam refero" may
also mean "Acceptable to carry Back" or "Welcome Reference". Note that the "Sun circle" touches only the "F" and "B" crescents. Thus, the crescents with letters
corresponding to the initials of Francis Bacon are linked through a loop whose ends terminate at
the Light source.
It may also be
worth noting here that AA headpieces, one of
which is found on Shake-speares
Sonnets , are always drawn with relatively symmetrical halves mirrored
across a central axis. Absolute symmetry is broken, in every case, by features
and highlights which characterize a "light" side and a "dark" side. An axis
taken through the "F" and "B" spheres divides the "refero" emblem in a similar
way. Furthermore, the stylized "A"s in AA devices
are often drawn with two horizontal bars, so the letter "B" may be seen in the
upper half and an "F" in the lower half. And the "A"s are always drawn so that
they curl over, forming a crescent or "C". The AA
devices, like the "Acceptam refero" device, may be considered steganographic
"cipher images", concealing important semaphores in seemingly superfluous
"decorations".
Now, if it is "Acceptable to carry Back", the "F" crescent phase may be "carried back" in time, perhaps taking a short-cut along the bottom half of the Sun circle, and put back on its diametrically opposite and prior "B" phase. Thus, "F" = "B", as with a -4 Caesar alphabet shift. Or, the "B" may be carried up and back to its opposite "F" so that "B" = "F", a +4 Caesar shift. Francis Bacon used a -4 Caesar shift for the first 25 letters of the Sonnets cipher, then switched to a +4 Caesar shift for the following 29 letters. He used the +4 shift thereafter, enciphering his name in other "Shake-speare" works in certain key places, signaled by textual clues, as elaborated by Penn Leary in his Second Cryptographic Shakespeare . Whether or not the Divine diagram was expressly designed to serve as a "Welcome Reference", it may be seen as a succinct and beautiful iconic representation of Bacon's cipher system.
Philip van Marnix's alphabet key
solution for a cipher used by Don Juan de Austria in 1577 is shown to the left.
It is a 22 letter alphabet which does not include the letters "J", "K", "U", and
"W". While Trithemius used 24 letters in his cipher system, he excluded the
letters "J" and "V" and placed "W" at the end of his alphabet, after "Z".
The "exclusion" of letters from cipher
alphabets had several consequences. It made deciphering more difficult by
forcing the cryptanalist to try different alphabet keys. But more importantly,
it helped to defeat frequency analysis techniques, as de Crema clearly
understood back in 1401. De Crema's alphabet key did not include the letters
"J", "V" and "W" and the letters "A", "E", "O" and "U" were each assigned 4
substitutes.
While some excluded
letters were to be "translated" into their homophonic (similar sounding or
appearing) substitutes, others (say, "X") were considered "nulls" and skipped
over. It was also the case that "regular" letters in the middle of a solution
string could be considered nulls. Null letters "mean nothing and are intended to
confuse interceptors" (@ Kahn xiv). The use of "nulls" in
cryptograms was a common practice, known to the earliest cryptographers. In his
1605 Advancement treatise, Bacon noted the use of
"intermixtures of NVLLES, and NONSIGNIFICANTS" in ciphers. Again, in the revised
edition, he listed as one of the kinds of "cyphars" those which are "intermixt with Nulloes, or non-significant Characters"
(264).
All that said, the best reason
for using the particular 21 letter alphabet described here is that it works --
the bottom line in any cryptologic analysis. Strictly speaking, the only
requirement is that "the letters which make up the cipher alphabet cannot be
chosen at random; the key must be of such a nature that any one of the several
correspondents, desiring to make use of it, will have it at his disposal" (@
Gaines 69). That Bacon was so kind as to leave clues and hints is merely icing
on the cake, though they do serve as confirmation that the particular key and
cipher techniques are correct.
Because the use of the 21 letter alphabet me and a -4 Caesar substitution
applied to acrostically selected letters yields the 25 letter solution given
below, we know it is the one Bacon used to encrypt his message. The great length of the solution and its ability to explain
anomalous features of the "ciphertext" dedication preclude the possibility that
the solution is just a function of "randomness". Further analysis supporting
the conclusion that the 25 letter string simply can not be random will be
discussed nonetheless, after the solution is demonstrated.
This is a sentence consisting of Four
words and 25 letters which ends at the first proper use of a period and the only
lowercase letter in the dedication ("Mr."). The last word in the solution string
is the name of our Author, spelled perfectly and correctly: "Bacon".
The
first word, "Nypir", refers to a mathematician about whose name "we do not know
the correct spelling...since many forms of the word are found, such as Napeir,
Nepair, Nepeir, Neper, Napare, Napar, Naipper" (@ Gibson). Indeed, when
"Nypir" is considered within the total context of the message, it is clear that
the word refers to the mathematician John Napier, the celebrated inventor of
logarithms.
Initially, it may seem problematic that
mention is made of Napier's revolutionary work prior to the first publication of
Mirifici Logarithmorum Canonis Descriptio in 1614.
However, "Napier tells us that he had been working on his invention of
logarithms for twenty years before he published his results, a statement that
would place the origin of his ideas about 1594" (@ Boyer 311).
Furthermore, Kepler wrote that in 1594, "Tycho Brahe was led by a Scotch
correspondent to entertain hopes of the publication of the Canon or Table of
Logarithms" (@ Macdonald xv). There is also evidence that
the 1619 Mirifici Logarithmorvm Canonis Constrvctio,
which explains the theoretical nature of logarithms, was written well before the
Descriptio.
Numerus Artificialis, or simply Artificialis, is used in the body of the
Constructio for Logarithm, the number corresponding to the logarithm being
called Numerus Naturalis.
Logarithmus, corresponding to which Numerus Vulgaris is used, is however employed in
the title-page and headings of the Constructio, and in the Appendix and
following papers. It is also used throughout the Descriptio published in
1614 ; and as the word was not invented till several years after the
completion of the Constructio (see the second page of the Preface, line
12), the latter must have been written some years prior to 1614 (@ 85).
Given Bacon's extensive contacts with the best minds of his time, it is not at all unlikely that he knew about Napier's invention prior to 1609. From the time of first conception, Napier seems to have openly discussed his invention with other leading-edge mathematicians and thinkers, such as Kepler and Brahe. Indeed, unlike the generally selfish and proprietary attitude of modern inventors, Napier was interested in the common good. He noted in an "Authors Preface" to the 1616 English translation of Descriptio that his
inuention, being (as all other good things are) so much the better as it shall be the more common, I thought good heretofore to set forth in Latine for the publique vse of Mathematicians. But now some of our Countreymen in this Island well affected to these studies, and the more publique good, procured a most learned Mathematician to translate the same into our vulgar English tongue... (@).
It is to be expected that a great intellect (such as that responsible for the Sonnets ) would have some knowledge of mathematics, since it is "so necessary for man, that (as I thinke nowe) so much as a man lacketh of yt, so muche hee lacketh of hys sence and wytte" (@ Record). Thus, it should not be too surprising to find that the author of Shake-speares Sonnets knew about logarithms, as can be seen in Sonnet 136:
In things of great receit with ease we prooue,
Among a number one is reckon'd none.
Then in the number let me pass vntold,
Though in thy stores account I one must be,
Fourthly, that the foote of the piller is Orchematicall, yt is to say, founded by transilitien or ouerskipping of number by rule and order, as from 1 to 3, 5, 7, &9 : the secret vertue whereof may be learned in *Trithemius, as namely by tables of transilitien to decypher any thing that is written by secret transposition of letters, bee it neuer so cunningly conveighed. (e.a.)
A side note, "* Polygraphiæ suæ lib 5," indicates
that Book 5 of Trithemius' Polygraphiae should
be consulted. In Book 5, Trithemius presents his polyalphabetic "tabula recta"
which "uses the normal alphabet in various positions as the cipher alphabets. Each cipher alphabet produces, in other words, a Caesar
substitution" (@ Kahn 136, e.a.).
Watson's "Pasquine Piller" and "preaty
obseruations" contain the main elements of the Sonnets cipher:
The Fourth enumerated "principall" feature includes a reference to the section of a cryptography book which deals specifically with Caesar substitution; Bacon's Sonnets cipher is solved by using a 4 letter shift Caesar substitution method.
The words in the given pillar arrangement must be re-arranged so the meaningful letters are aligned to yield the acrostic/telestich solution; the words in the Sonnets frontmatter must essentially be re-arranged one word to a line and aligned along the right side to yield the (enciphered) telestich solution. The author explains that "ALL such as are but of indifferet capacitie, and haue some skill in Arithmetike " may be able to appreciate the "Piller"; the full significance of the Sonnets cipher solution can only be understood when the "message" is considered in a mathematical context. The term "Pasquine Piller" is derived from the Italian "Pasquino", a "name given an antique Roman statue...which was annually decorated and posted with verses" (@ Webster's ); the Sonnets Dedication, in all capitals with periods separating the words, has the appearance of a Roman inscription. The first two letters in the Piller are AA and the author's first "obseruation" is that the "whole piller...is by relation of either halfe to the other Antithetical " or symmetrically contrastive; the AA headpiece above the first of the Sonnets is similar in style to the one found throughout the Hekatompathia and AA headpieces always feature a dark "A" and a light "A" mirrored across a central axis. A final point should be made regarding the title page and numerous appearances of an AA device in The Hekatompathia. The title page contains three noteworthy features: a man holding a type of Spear (halberd) and two important headpieces. The "headpiece" used at the bottom appears in a slightly different form on the 1623 First Folio of Shakespeares Comedies, Histories, & Tragedies and Bacon's 1620 Novum Organum -- the arrow on the right lacks the "message" hanging from the middle of the shaft. The headpiece is a steganographic pictorial cryptogram which contains (at least) two instances of the name "Bacon" in the form of rebuses.
Above is an image of a headpiece used throughout the Hekatompathia. The same block used to print it was used in the production of the counterfeit Furtivis cryptography book printed by John Wolfe in 1591. Again, it should be noted that the counterfeit was essentially perfect, including the 1563 original publication date, with the exception of the AA headpiece.
At the head of the first page of Shake-speares Sonnets there is an AA device which is very similar to that found in the Hekatompathia and counterfeit Furtivis. While there were (at least) 14 distinct AA designs used until a last appearance in the 1720
edition of Bacon's Essays, they differed widely. The
design shown above, which appeared on the Sonnets,
seems to have been an attempt to reproduce or imitate the Furtivis device.
The unusual AA headpiece at the top of the Hekatompathia title page consists of two overlapping
images of what will be refered to as the Furtivis AA
device. This double Furtivis AA headpiece also
appears in several places in Ben Jonson's 1616 Works. Ben Jonson was a good friend of Francis Bacon.
Several important references are made to Bacon/Shakespeare in the Works, including the Poet/Lawyer "Ovid the play-maker"
(@ 279) in Poetaster and Sogliardo in Euery Man out of his Humour, a lampoon of the Actor
Shakspere who was apparently paid to act as a decoy for Bacon after the trial of
Essex. Sogliardo purchases a crest, a "Boore without a head," and takes as his
motto "Not without mustard," a clear reference to Shakspere's "Not without right" motto.
Curiously, Carlo comments, "I commend the Heralds
wit, hee has decyphered him well : A swine without a
head, without braine, wit, any thing indeed, ramping to gentilitie" (124,
e.a.). Though much other Jonsonian evidence could be presented, the last piece
given here is the beginning of a poem published in Under-Woods (@ 225, e.a.) and
originally written for Francis Bacon's birthday:
Haile, happie Genius
of this antient pile !
According to the technical experts, the length of the Sonnets solution is sufficient to warrant the
conclusion that it is a valid decipherment.
In
order to further demonstrate that the 25 letter solution can not be random, an
acrostic analysis computer program was written which performs several types of
acrostic letter selection in addition to the one that yields the solution to the
Sonnets cryptogram. An interactive version is
available on this site. Several kinds of selection procedures were used in order
to generate a broader statistical sample. The program selects letters from words
of the following types:
Every word When the program determines that a
particular word in the text being analyzed is one of the 7 types, each letter in
the word is then selected and added to the appropriate acrostic string. Letters
are processed by choosing every letter from the left to the right end of the
word and every letter moving from the right to the left end of the word. That
is, an "offset" value is incrementally increased so that every letter of a word
is added to two acrostic strings of letters for that word type. [From the left end of
"Pig"] Short words will not "contribute"
letters to acrostic strings of letters chosen using a large offset -- they are
effectively "skipped". For example, a Capitalized 3 letter word will not
"contribute" to an acrostic string consisting of letters which are offset 5
letters to the right from the leftmost letter of Every Capitalized word (i.e.
the 6th letter of every Capitalized word). The 9 letter word "SHAKSPERE" was
chosen to serve as a "control" and present a contrast to the shorter word
"Bacon". Any words could have been used -- the intent of
the exercise was to demonstrate that a sufficiently long string of letters will
not "show up" with greater than expected frequency when text is searched using
acrostic selection and Caesar substitution. The choice of "Bacon"
effectively tested for two things: the frequency of appearance of the full 25
letter solution string (since it contains the short string) and the frequency of
appearance of a relatively short string. "Bacon": 21^5 = 4,084,101 to
1. Now, meaningful text is not random.
That is, the Sonnets can not be considered a
sequence of random letters. However, it should be pointed out that in selecting letters acrostically from meaningfully ordered
words, the chances of "finding" a meaningful word are relatively low. An
English text only yields its meaning when letters and words are processed (read)
in a linear and sequential order, along a horizontal axis. But the direction of
acrostic selection is oriented orthogonally with respect to the normal
horizontal "reading axis". Even when a letter is selected from every word of the
text, the procedural axis is vertical: it is as though all the words were
aligned vertically and the selection proceeds "across the rows". So, a random
acrostic consists of letters taken "out of order" with respect to the standard
profile of meaningfully arranged letters.
The point is that acrostically selected letters will not appear with the same
frequency as letters found in "regular" meaningful text, where every letter is
included in the analysis. It is thus less likely that a string of
acrostically selected letters will contain meaningful words or sentences. It is
extremely unlikely that a 25 letter meaningful English sentence like "00 Nypir
Cypphrs Bekaan Bacon" will appear in a string of letters resulting from a
randomly chosen type of acrostic selection. So, out of about 31 million strings
consisting of 690 million letters, not one instance of
an acrostically selected 9 letter string spelling "SHAKSPERE" was found. The
string "Bacon" occurs once for every 720,005 acrostic strings. That is about one
sixth of the number resulting from random selection of each letter: 4,084,101 to
1. Again, the relatively long string
"SHAKSPERE" did not occur, in keeping with the odds given above of
794,280,046,600 to 1. The Elizabethan English texts showed a lower number of
"Bacon" occurrences, relative to the number of strings checked: once for every
955,340 strings. That is about one fourth of the random probability and still
relatively close to the frequency of occurrence in the modernized text of
"Shake-speares" works.
concealment cipher goes by various names, as null cipher, open-letter cipher, conventional
writing, dissimulated writing, and so on... The name "null cipher"
derives from the fact that in any given cryptogram the greater portion of
the letters are null, a certain few being significant, and perhaps a few
others being significant only in that they act as indicators for finding
truly significant letters. [...] Significant letters may be concealed in
an infinite variety of ways. The key...may be their positions in words, or
in the text as a whole. It may be their distance from one another,
expressed in letters or in inches, or their distance to the left or right
of certain other letters (indicators) or of punctuation marks
(indicators)...
Note that the text of the Sonnets
dedication contains periods (or decimal points) at the end of every word. The anomalous punctuation and capitalization serves to draw
attention to the significant cipher letters, the last letters of each
capitalized word.
Concerning the decryptment of
concealment cipher, we regret to say that cryptanalysis has little help to
offer. Fortunately, most of these ciphers depend absolutely on the belief
that they will not be recognized as cipher, and once they are so
recognized, they present no resistance. In those few cases where the
secret message is not at once obvious, it is sometimes useful to arrange
the words (or sentences) in columns, or in rows, for a closer
inspection...and where the alignment from the left gives no results,
letters or words can be aligned from the right...
The author
recommends acrostic analysis, the very method used to solve the Sonnets cipher, "in those few cases where the secret
message is not at once obvious." Aligning capitalized words from the right,
then using a -4 Caesar substitution on the rightmost letters, yields the
solution to Bacon's cryptogram. Trying Caesar substitution is reasonable and
appropriate, as Mrs. Gaines explains:
Many of the classic
ciphers, fundamentally of the concealment type, are also substitution
ciphers, and their decryptment would follow substitution methods. Of
these, perhaps the best known is Bacon's biliteral cipher.... The subject
is fascinating, and the literature of cryptography is rich with examples.
[...] Concealment cipher has, of course, the unique virtue of being able
to convey messages under circumstances which make it seem that no
communication has passed... (e.a.)
This Arte of
Cypheringe , hath for Relatiue, an Art of Discypheringe ; by supposition vnprofitable; but,
as things are, of great vse. For suppose that Cyphars were well mannaged, there bee Multitudes
of them which exclude the Discypherer . But in
regarde of the rawnesse and Vnskilfulnesse of the handes, through which
they passe, the greatest Matters, are many times carryed in the weakest CYPHARS .
How comes it all things so
about the
The
fire, the wine, the men ! and in the midst,
Thou stand'st as if some Mysterie thou did'st
!
The second most extensive use of a Furtivis type device appears in a book published 2
years later in Scotland. There are 7 instances, with 1 inverted, in the 1584 Essayes of A Prentise, in the Divine Art of Poesie by
James I, a book which was widely circulated because of his political stature. In
the preface to the section "Ane Metaphoricall Invention of a Tragedie called
Phoenix" may be found a Pasquine Piller whose acrostic/telestich solution reads "Es Me Stewart D Wike" (@).
Now, The
Hekatompathia was printed in 1582 by "Iohn Wolfe" in London, 2 years before
Thomas Vautroullier printed The Essayes of A
Prentise in Edinburgh using the same AA block.
The block was used a second time in 1584 for the publication of Dauid Powel's The historie of Cambria, "Imprinted at London by Rafe
Newberie and Henrie Denham" (@) and notable for its apparently unique "Boar
Riders Spearing Dragons" headpiece. The use of the
AA block by printers was clearly not
"proprietary". Subsequent to its use by Vautroullier in Edinburgh and
Newberie in London, John Wolfe again used it in London in his 1591 counterfeit
printing of the cryptography manual De Furtivis Literarum
Notis . Why did Wolfe "ruin" what would have been a perfect counterfeit,
including the 1563 date, by using an AA headpiece
imprint?
Every Capitalized word
Every UN-capitalized word
The leftmost word in every line
The rightmost word in every line
The leftmost word in every line when the
word is Capitalized
The leftmost word
in every line when the word is UN-Capitalized
For example, when a three letter word is
of the type "every Capitalized word" (say, "Pig"), it will append letters to
acrostic strings consisting of:
0 letter offset: the leftmost
letter of Every Capitalized word (i.e., 1st letter of the word: "P");
1 letter offset to the right from the
leftmost letter of Every Capitalized word (2nd letter: "i");
2 letter offset to the right from the
leftmost letter of Every Capitalized word (3rd letter: "g");
[From the right end of
"Pig"]
0 letter offset: the rightmost
letter of Every Capitalized word (3rd letter: "g");
1 letter offset to the left from the
rightmost letter of Every Capitalized word (2nd letter: "i");
2 letter offset to the left from the
rightmost letter of Every Capitalized word (1st letter: "P").
Letters in the
acrostic strings resulting from every permutation of "type" and end-offset are
then substituted for corresponding letters in each possible Caesar alphabet
shift. Thus, each acrostic string will yield a total of 21 strings: the
original string of "unshifted" or "plaintext" letters and 20 strings generated
by consecutive Caesar shift and substitution. Applying every possible Caesar
substitution to each acrostically selected string of letters increases the
sample size 20 fold and helps demonstrate the uniqueness of the -4 Caesar
substitution solution.
All the
strings of letters resulting from this process are checked for the presence of
the target "plaintext message". Every string is also checked for the presence of
a backward instance of the target. Strings too short to contain the "message"
text are not checked and therefore don't add to the reported total number of
strings checked.
The original
spelling text of Shake-speares Sonnets was divided
into "blocks" of text, each consisting of a sonnet number and corresponding
sonnet. The Title page and Dedication comprised the first block of text. The
whole was searched for occurrences of the strings "Bacon" and "SHAKSPERE" with
the following results:
Found Bacon in block 1:
Select words from top to bottom, left to right;
rightmost letter of every word;
-4 Caesar shift (a=S):
OONAYPIRCYNPPYAKAARHRSBEKAANBaconQDGAYPAYRNCPDACNICDPP
Found Bacon in block 1:
Select words from top to bottom, left to right;
rightmost letter of every Capitalized word;
-4 Caesar shift (a=S):
OONYPIRCYPPHRSBEKAANBaconQDGAYPAYRNCPDACNICDPP
+++ Summary +++
Blocks of text: 155
Lines of text: 2328
Words in text: 17965
Letters in text: 75028
Found 2 occurrences of "Bacon".
Acrostic strings checked for "Bacon": 444192
Total letters in All Acrostic strings: 14483196
Found 0 occurrences of "SHAKSPERE".
Acrostic strings checked for "SHAKSPERE": 370524
Total letters in All Acrostic strings: 14014140
It is to be expected that a short string
will show up more often than a longer one. That is, the longer the "target"
string, the less likely it is that it will appear. The simple probability that a
particular word will result when each letter is chosen randomly is given by the
"product rule". Specifically, "the number of r
-permutations of a set of n objects with repetition
allowed is n^r " (@ Rosen 272). The
probabilities are:
"SHAKSPERE": 21^9 =
794,280,046,600 to 1.
"OONYPIRCYPPHRSBEKAANBacon":
21^25 = (aprox.)
1,136,272,166,000,000,000,000,000,000,000,000 to 1 (or 1.13 x 10^33 to 1).
It may be easily
demonstrated that the letter frequencies which characterize meaningful text are
not the same as those of acrostically selected letter strings. A comparison of
the frequency characteristics of letters in text read normally (i.e., all
letters) and those in an acrostic string (e.g., last letters of words), will
quickly show that various acrostic selection procedures generate very different
"frequency profiles" than that which characterizes "normal" text. For example,
consider the letter frequency profile of the original spelling text of George
Puttenham's 1589 Arte of English Poesie, shown
below. Numbers below letters indicate the frequency of a letter's occurrence as
a percentage of total letters analyzed.
In order to further demonstrate that the
probability of a very long string occurring as the result of acrostic selection
and Caesar substitution is sufficiently small to warrant the claim that the
cipher must be intentional and valid, additional text was processed by the same
acrostic analysis computer program described above.
The remaining works of "Shake-speare"
were divided up into lines of from 1 to 8 words and blocks of from 10 to 13
lines (both numbers being determined by a random number generating function).
The text used was not the original spelling -- unfortunately a digital version
is not freely available. It was used only as a "random sample" of parsed
meaningful text, albeit ironically selected.
The complete list of "hits" is
available for perusal, along with a list which includes context. None of the acrostic instances of "Bacon" is accompanied by
additional meaningful text. The three longest words which show up next to
"Bacon" (in separate instances) are very short and hence random: "COOL", "TACT",
and "NOOON". Here is a summary of an Acrostic Analysis of the complete works of
"Shake-speare":
Total Words of text: 922239
Total Letters: 3891140
Bacon's found: 66
Acrostic strings checked: 47520312
Letters in all Acrostic strings: 792898680
SHAKSPERE's found: 0
Acrostic strings checked: 31537800
Letters in all Acrostic strings: 690145596
In order that the sample space might
include "olde English" text, the original spelling text of George Puttenham's
1589 Arte of English Poesie and Richard Mulcaster's
1582 Elementarie were processed with the following
results:
Total Words of text: 183767
Letters in text: 823493
Bacon's found: 10
Acrostic strings checked: 9553404
Letters in all Acrostic strings: 167434428
SHAKSPERE's found: 0
Acrostic strings checked: 6516636
Letters in all Acrostic strings: 148118292
A total of
1,106,006 words of text were analyzed. The 9 letter string "SHAKSPERE" was not
found at all in a search of 38,054,436 strings consisting of 838,263,888
letters. In a check of nearly a billion letters, there
was not a single instance of a 9 letter "target". And exactly one occurrence
was found of a string over 2 times as long as
"SHAKSPERE", the meaningful four word sentence
Linked Summary
I have...vast contemplative ends...for I have
taken all knowledge to be my province.... I hope I should bring
in...profitable inventions and discoveries.... This...is so fixed in my
mind as it cannot be removed. [...] I will [become] a true pioneer in that
mine of truth.
The solution contains self-references:
the word "cyphhrs" and initial "zerOes" (cyphers) refer to
both mathematics (echoing "Nypir") and the cryptogram itself.
The letters AA prior to "Bacon" and the "OO" at the head of
the solution echo the Sonnets AA headpiece, which imitates the one found in
the 1591 counterfeit publication of the
cryptography book De Furtivis
.
The AA
device which appears in Napier's Descriptio is the same as the one on Bacon's
1620 Novum Organum and the 1623 Comedies, Histories, & Tragedies of
"Shakespeare".
The first appearance of an AA headpiece was on a book of which Bacon
owned a copy containing proof sheets.
The final letter of the solution
corresponds to the only lowercase letter in the Dedication and the first
appropriate use of a period.
The odd punctuation of the Dedication
draws
attention toward the significant last letter of words.
The odd punctuation also corresponds
to mathematical decimal notation which first appeared in the
form used today in Napier's 1616 Descriptio.
Words in all caps in the Dedication
correspond to Capitalized word acrostic (telestich) selection.
The letters in Bacon's abbreviated
alphabet key correspond to those used by ancient Romans, whose
inscriptions were written using only capital letters, with periods between
every word.
Four unnecessary vertical spaces and the
final word "FORTH" correspond to the Caesar shift amount and number of words in the
first 25 letters of the solution, as well as the number of "Bacon"
homophones in the complete 54 letter solution.
Francis Bacon understood and wrote
about steganographic cryptography prior to the
publication of the Sonnets. He later expanded this work, focusing
almost exclusively on steganographic methods. It is appropriate to give
him the last word: