The Code that Failed: Testing a Bacon-Shakespeare Cipher

by Terry Ross



Although the most popular current strain of antistratfordianism is Oxfordian, for many decades Francis Bacon was the favorite candidate of those who doubted that William Shakespeare wrote the works attributed to him. Many of the objections Oxfordians raise against Shakespeare were developed originally by Baconians (something Oxfordians are not always eager to admit), and Baconians had the advantage over Oxfordians that their candidate had lived long enough not only to write the works but indeed, they claim, to have overseen the publishing of the First Folio in 1623: Oxford died in 1604, Shakespeare in 1616, but Bacon lived until 1626.

Since so many of the same points are raised by both Baconians and Oxfordians, the case for Shakespeare is ably presented elsewhere on the Shakespeare Authorship page. One major difference between the Oxfordians and the Baconians is that the latter are much more likely to display a fondness for cryptography. There have been many ingenious, complicated (and sometimes nutty) efforts to show that Bacon hid the evidence of his authorship within the works themselves, in code or cipher. The most comprehensive response to these efforts is The Shakespeare Ciphers Examined (1957), by William F. and Elizebeth S. Friedman, which is thought to have pretty much demolished cryptographical claims for Bacon.

The temptation to find new Baconian ciphers is a powerful one, however, and Penn Leary, who greatly admires the Friedmans and who is determined to avoid the errors that they found in earlier Baconian works, believes that he has uncovered the actual cipher Bacon used to sign "Shakespeare's" works. He explains his methods in great detail in The Cryptographic Shakespeare [1987]; much of this work is available online at the Penn Leary site. I decided to test Leary's methods for myself, and in this essay I will explain his methods, and I will give the results of my tests--results that completely invalidate those methods.

To Table of Contents

Penn Leary's Baconian Ciphers

The principal cipher Bacon used, according to Leary, is based on a 21-letter alphabet: "ABCDEFGHIKLMNOPQRSTVY." This alphabet omits the letters "JUWXZ," but, as Leary explains, in Elizabethan times the letters "I" and "J" were often used interchangeably, as were the letters "U" and "V," and "W" was often printed as "VV." Since the Roman alphabet did not use "X" and "Z," Bacon, according to Leary, omitted those letters. In addition, the numerals from "1" to "9" could be expressed as the letters from "A" to "K" (remember, there is no "J" in this alphabet), while the numeral "0" may be omitted, since the Romans had no zero. Given this 21-letter alphabet, Leary believes Bacon then replaced each letter with the one that comes four places later. Thus "A" becomes "E," "B" becomes "F," and so on. Here is the 21-letter alphabet, and beneath it in lowercase is the cipher equivalent of each letter.

Here is a string of letters one might find in Shakespeare's works: "TSVAI." Applying Penn Leary's methods to this string produces "bacen." Remembering that spelling had not yet become standardized in Bacon's day (there are dozens of ways to spell "Shakespeare"), one might well be inclined to count "bacen" as a spelling of "bacon." Penn Leary has found a great many such Baconian signatures in Shakespeare's works--as of 1989 there were 113, and I'm sure the number must be significantly larger by now. Leary has also found instances in cipher of the word "name" and of something resembling the word "cipher" itself, and he does not confine himself strictly to the method I have described, but since the overwhelming majority of his findings is based on his application of those methods to find what he believes is Bacon's hidden name, my analysis is confined to his search for "Bacon," though a similar analysis could easily be applied to "name" or "cipher."

To Table of Contents

Did Bacon Write the Funeral Elegy?

I decided to apply Penn Leary's methods to the Funeral Elegy that has recently been attributed to Shakespeare, but that has not, to my knowledge, been subjected to a search for ciphers. If the Funeral Elegy was indeed written by the author of Shakespeare's works, and if Francis Bacon wrote those works and "signed" them by incorporating ciphers of his name, we should expect to find such ciphers hidden in the Funeral Elegy. On the other hand, if we failed to find such signatures, then perhaps the Funeral Elegy was not written by the author of The Tempest and Coriolanus.

Sure enough, Penn Leary's Baconian cipher identifies the Funeral Elegy as a genuine work by Shakespeare (that is to say, Bacon). I will quote the English version of the lines in which I have found these ciphers, and immediately beneath I will quote the lines as they appear when translated according to Penn Leary's Baconian rules (I call the result "Bakish" for short):

Here is line 151 of the Funeral Elegy in English and in Bakish:

       Which now that subject's merits doth rehearse
       Cmngm rsc bmeb acfnigb'a qiynba hsbm yimieyai
The letters "tsmeri" translate into "baqiyn," which counts as "bacon."

Here is line 264 in English and in Bakish:

       Of that same ignorance which makes
       Sk bmeb aeqi nlrsyergi cmngm qeoia
The letters "tsamei" translate into "baeqin," which also counts as "bacon."

Line 473 in English and in Bakish:

       Else what avails it in a goodly strife
       Ipai cmeb ecenpa nb nr e lsshpd abynki
"tavai" translates into "becen," which also counts as "bacon."

Sometimes the name Bacon appears in cipher and backwards. Here is line 68 in English and Bakish:

       When now his father's death had freed his will
       Cmir rsc mna kebmiy'a hiebm meh kyiih mna cnpp
The letters "isfat" translate into "nakeb," which is backwards for "bekan," which counts as "bacon."

Line 141 in English and Bakish:

       For if his fate and heaven had decreed
       Ksy nk mna kebi erh miecir meh higyiih
Again, the letters "isfat" count as "bacon."

Line 558 in English and Bakish:

       Immure those imputations I sustain
       Nqqcyi bmsai nqtcbebnsra N acabenr
The letters "isust" translate into "nacab," (remember that in the cipher's 21-letter alphabet "u" and "v" are the same letter) which is backwards for "bacan," which counts as "bacon."

Occasionally Bacon's signature may occurs over two lines of Shakespeare, as in the last three letters of line 285 and the first four letters of line 286:

       Whose illness is the necessary praise
       Must wait upon their actions only rare
       Cmsai nppriaa na bmi rigiaaeyd tyenai
       Qcab cenb ctsr bminy egbnsra srpd yeyi
The letters "isemust" translate into "naiqcab," which is backwards for "bacqian," which also counts as "bacon."

Penn Leary has also found Bacon's name by looking at the first letter of each line, or of each capitalized word, or at every other letter in a string, but I confined myself to simple backwards and forwards searches and still found that Francis Bacon seems to have proclaimed his authorship of the Funeral Elegy no fewer than seven times:

A Baconian might well take this as compelling evidence that the author of the Funeral Elegy was also the author of Shakespeare's works--and that the author's real name was Francis Bacon.

To Table of Contents

How Many Ways Are There to Spell "Bacon"?

Penn Leary's method found seven instances of Bacon's "signature" in the Funeral Elegy, but the reason for this is not that Bacon wrote the poem but that Leary's method is guaranteed to find Bacon's "signature" in virtually every text of any length in the Roman alphabet. What makes this phenomenon possible is the enormous number of character strings that count as an instance of the name "Bacon."

Below are three lines of characters. In the top line are the 26 letters and 10 numerals used in English; in the middle line is the transformation of those 36 characters into the 21-letter alphabet Leary thinks Bacon used; and in the bottom line are the equivalent letters as they would appear in the cipher that Leary thinks Bacon used (what I call "Bakish").

       English:   abcdefghijklmnopqrstuvwxyz0123456789
       Bakish:    efghiklmnnopqrstvyabccc-d--efghiklmn
Penn Leary's method is to translate English into Bakish and then look for the string "bacon" or its equivalent. Now, Leary's man had a very short name, and it should always be easier to find strings of letters that correspond to "bacon" than, say, to "McGillicuddy" or "Shakespeare," but Leary makes things even easier for himself by allowing a great many equivalents for "bacon."

How many ways are there to spell "bacon"? Consider first just the consonants "bcn." In Leary's cipher method, the English letter "T" becomes the Bakish "b." I have counted 14 ways used by Leary to spell the Bakish "c" (though no doubt there are more than 14 that he would accept), because Leary allows any Bakish letter or string of letters that may have a hard "c" or "k" sound: "c" in Bakish = "U," "V," or "W" in English; "ch"="UD," "VD," or "WD"; "k"="F"; "q"="M." The Bakish "n" occurs whenever "I" or "J" occurs in English. Thus with 1 way to get "t," 14 ways for "c," and 2 ways for "n," there are 28 different English combinations of letters that produce the Bakish consonants in "bacon" forwards, and another 28 ways backwards.

But the real trick is in the spelling of the "a" and "o." As Leary says, "for the spelling of 'Bacon' in deciphered plaintext [i.e., Bakish], conceivably almost any vowel or combination of vowels might be substituted for the 'a' and the 'o' in Bacon's name" (Leary, p. 217). If we allow all 6 vowels (Bakish "aeiouy" = English "SAEKQR") but also allow any two or three consecutive vowels, as Leary does, then there are 258 different ways just to spell the "a" and 258 different ways to spell the "o" in Bacon. With the 28 different ways to spell the consonants, there are now 1,863,792 (28 times 258 times 258) different ways in English to generate the Bakish "bacon" forwards and another 1,863,792 ways backwards. Thus, there are a total of 3,727,584 strings of characters that will count as ciphers for "bacon." While it's true that Elizabethan spelling was somewhat variable, it wasn't THAT variable.

We may look for any of these 3,727,584 magic strings in consecutive letters in Shakespeare's works, or we may look at every other letter, or at the first letters of lines, or at the first letter of each capitalized word, or at every letter in each capitalized word. Occurrences of the letters "x" and "z" in a string are not a problem, as we simply ignore them. If we come across numerals, we may use "1," "4," "5," "6," and "9" to help spell "bacon," because in Bakish they become "e," "h," "i," "k," and "n." The only surprise in Leary's finding 113 instances of Bakish "bacon" in Shakespeare is that the number is so tiny.

Some English words are automatic Bakish "bacon" generators -- they each contain a string of letters that will count as "bacon" in Bakish. Here is a table of such words as they occur in the works of Shakespeare (a more complete list is also available). In the first column is the number of times the word occurs in Shakespeare, in the second column is the word itself in English, in the third column is the word translated into Bakish, and in the fourth column is the string within the Bakish word that Leary would consider the signature of Francis Bacon (some "bacons" are forwards and some are backwards). The numbers are taken from the Harvard Concordance, which does not include stage directions, speech prefixes, lists of characters, and the like, so the actual numbers of occurrences should be somewhat higher.

          NO.|  SHAKESPEARE   |    BAKISH      | BACON
            1| affirmation    | ekknyqebnsr    | nyqeb
            1| affirmatives   | ekknyqebncia   | nyqeb
            2| alder-liefest  | ephiy-pnikiab  | nikiab
            1| believest      | fipniciab      | niciab
            5| boatswain      | fsebacenr      | bacen
            1| briefest       | fynikiab       | nikiab
           15| chiefest       | gmnikiab       | nikiab
            7| confirmation   | gsrknyqebnsr   | nyqeb
            2| confirmations  | gsrknyqebnsra  | nyqeb
           54| counterfeit    | gscrbiykinb    | biykin
            5| counterfeited  | gscrbiykinbih  | biykin
            4| counterfeiting | gscrbiykinbnrl | biykin
            1| counterfeitly  | gscrbiykinbpd  | biykin
            5| counterfeits   | gscrbiykinba   | biykin
            2| counterfeit'st | gscrbiykinb'ab | biykin
            2| countervail    | gscrbiycenp    | biycen
            1| diameter       | hneqibiy       | neqib
            1| grievest       | lyniciab       | niciab
            4| intrusion      | nrbycansr      | bycan
            1| liefest        | pnikiab        | nikiab
            1| relieveth      | yipnicibm      | nicib
            1| reviewest      | yicniciab      | niciab
           12| travail        | byecenp        | byecen
            1| travail'd      | byecenp'h      | byecen
            2| travails       | byecenpa       | byecen
            1| treasuries     | byieacynia     | byieacyn
            1| untrussing     | crbycaanrl     | bycaan
            2| viewest        | cniciab        | niciab
            1| vieweth        | cnicibm        | nicib
According to Leary's method, every time Shakespeare uses any of these words, what he's really doing is saying, "Hi, I'm Bacon."

It is only fair to point out that Leary is careful to use facsimiles of Shakespeare quartos and the First Folio, and I have not always done so. It is very significant to him that the first word of dialogue in the First Folio is the first spoken word of The Tempest: "Bote-swain" (page 215 ff.). However, he need not have bothered with the original spelling, because under his loose rules both "Bote-swain" and "Boatswain" contain strings that count in Bakish as "bacon." Similarly, although not all Elizabethan spellings of "counterfeit" will work, among those that do are "contrefaict," "contrefait," "counterfait," "counterfaite," "counterfeict," and "counterfeight." If anything, using original-spelling texts should increase the odds of finding one of the 3,727,584 magic strings.

In addition to the automatic "bacons," there are what we might call "semi-automatics": strings that may be found as parts of two words that frequently go together. "ISFAT," for instance, occurs whenever "is" "his" or "this" immediatedly precedes "fat," "fatal," or "father." When Falstaff says, "There live not three good men unhanged in England, and one of them is fat and grows old" (1H4:2.3.130-32 [Riverside]), Leary's method would see in the line only another Bakish "bacon." Another such string is "JESUST": according to Leary's methods, whenever a word beginning with a "T" follows "JESUS," the importance of the passage lies not in any religious reference but in the occurrence of the Bakish "bacon." One would expect that Leary's methods would provide plenty of similar evidence that Bacon translated the New Testament, but Leary never tests his method on anything other than Shakespeare.

This is a surprising omission on Leary's part, because in The Cryptographic Shakespeare, citing the analysis by the Friedmans, he says this of earlier attempts to find Baconian cryptograms in Shakespeare

	. . . hundreds of books have been written on the subject. . . . 
	These contend that cryptograms or ciphers in the works amount to
	concealed signatures of Francis Bacon, who himself had written a
	work on cryptography.  But these ciphers tend to cancel each other
	out or are so broad as to demonstrate that almost any works were
	written by Bacon. [page 92]
To Table of Contents

So What Else Did Bacon Write?

I determined to test Penn Leary's methods according to the standard he himself proposed, and if it turns out that his cipher "is so broad as to demonstrate that almost any works were written by Bacon," then it must be rejected. I have applied his methods to works that he does not claim Bacon (or Shakespeare) wrote. If these works may also be used to generate Bakish "bacons," then either Bacon wrote them and hid his name in cipher, or else their appearance has nothing to do with Francis Bacon, and is just the predictable result of the expected occurrence of some of the 3,727,584 magic strings of characters in English that Leary counts as occurrences of "bacon" in Bakish. I wrote a UNIX script to search for such strings, and ran it on several texts that are available online. Here's some of what I found (links are provided to the sources of all texts I tested):

I began by looking at Elizabethan works and found 14 Bakish "bacons" in Edmund Spenser's Amoretti and Epithalamion and another 64 in The Shepheardes Calender -- was Bacon really Spenser as well as Shakespeare? I then applied Leary's methods to Bacon himself and found 61 Bakish "bacons" in Bacon's own Essays -- what possible reason could Bacon have had for hiding cryptographical evidence that he wrote something that bore his name? Following up my earlier suggestion that the string "JESUST" would produce quite a few Bakish "bacons" in the gospels, I searched the King James version and found 83 Bakish "bacons" in Matthew, 38 in Mark, 44 in Luke, and 88 in John (of the 253 in the four gospels, 44 were instances where the name "Jesus" was followed by a word beginning with a "T," and 209 were produced by other strings). Clearly, if Penn Leary's method were reliable, the King James gospels must have been written by Bacon.

I thought I could do better than Penn Leary's total of 113 for Shakespeare, and I found 1449 Bakish "bacons" throughout the works, or 1136 more than Leary himself found. That seems an impressive total, but the works comprise an enormous set of strings to search through. George Puttenham's Arte of English Poesie is about one tenth the size of Shakespeare's works, and searching the original-spelling text I found 192 Bakish "bacons." Spenser's Faerie Queene is something under a third the size of Shakespeare's works, and I found 548 Bakish "bacons" there. The 1449 for Shakespeare seems, if anything, a little low.

But since Bacon was alive when both The Arte of English Poesie and The Faerie Queene were written, perhaps all I had done was prove that they indeed had also been written by Bacon. I decided to check some texts written long before and long after Bacon could have had a hand in them.

The kind of cipher Leary uses to generate what I call Bakish "bacons" is known as a "Caesar cipher," after Julius Caesar, who is thought to have used a letter-substitution cipher. I decided to search the Latin text of Caesar's Gallic Wars, and I found 64 Bakish "bacons" (every time a word beginning with "i" follows the word "atque" in Latin, a "bacon" appears in Bakish). If Bacon wrote Caesar's Gallic Wars as well as the plays of Shakespeare, it would explain why Henry VI, like Gaul, is divided into three parts.

Leaping forward many centuries, I tried the method on Longfellow's Song of Hiawatha and found 327 Bakish "bacons." As it happens, the string "IAWAT" in English translates into a backwards Bakish "bacon," so the very name "Hiawatha" is another "automatic," like "boatswain" or "counterfeit." If every instance of the name "Hiawatha" is actually an instance of Bacon's name, does this mean we should read the poem as his autobiography?

Edgar Rice Burroughs's more recent Tarzan of the Apes generated a disappointingly low 135 Bakish "bacons" (perhaps Bacon wrote the work but wasn't proud of it), but I found a whale of a lot of them in Melville's Moby Dick: 347.

Worried that I was biasing my search by concentrating on literary works, I tried instead that classic of political prose, the Federalist Papers, and found 239 Bakish "bacons." In a more recent and less classic text, the Senate Majority Whitewater Report, I found 396 (what did Bacon really know about Vince Foster's death?)

Indeed, I have yet to find any substantial text in English or Latin that will NOT produce Bakish "bacons" using Leary's methods.

To Table of Contents


Unless Penn Leary is prepared to claim that Bacon wrote everything that ever has been and ever will be written in English (and Latin, for that matter), the inescapable conclusion is that his method is, when judged by his own standards, useless as a means of determining the authorship of Shakespeare's works. The words "boatswain" and "counterfeit" (and "Hiawatha") are not evidence for Bacon's authorship, nor are the words "his father," "Jesus therefore," and "is sweet." Leary never rejects any Bakish "bacon" that he finds, and his rules are so loose that one is guaranteed to find rashers of "bacons" anywhere one looks.

What is most unfortunate about such cryptographical exercises as Leary's is not the waste of time and energy -- it seems a pretty harmless hobby -- as what it does to one's sense of Shakespeare and Bacon. All the plays and poems are treated as if they are without meaning or significance (Leary continually refers to Shakespeare's words as "ciphertext") unless someone finds the magic key that unlocks the Baconian wisdom hidden behind all that poetry. The ideal result of Leary's methods would be to throw out all 38 plays, the sonnets, and poems, and gaze upon the 1449 "bacons" that may be extracted. If we accept Leary's methods, then the only thing Shakespeare says to us (and he says it 1449 times) is "Hi, I'm Bacon"; and Francis Bacon himself becomes little better than a dog marking its territory.


See Penn Leary's reply to this essay.
See Hiawatha's Cryptographing, my response to Leary.
See the UNIX and Perl scripts I have used to test Leary's methods.
Was Francis Bacon a welterweight? See a list of over 700 words that Leary would have you believe contain Francis Bacon's signature.
Go back to the Shakespeare Authorship home page.