Baconian Cipher

IPhenomenon

Assign every letter a five-position pattern of two symbols. Bacon used A and B; the modern bench uses 0 and 1 for clarity. Here is the classical 24-letter version — I and J share a pattern, U and V share a pattern, because five binary positions give 32 slots and the Elizabethan alphabet needed only 24:

Now the trick. The five marks need not be letters at all. They can be two typefaces in a printed book: ordinary roman versus italic, large versus small, one font versus another. Bacon's own example was a printed text in which that distinction carried a hidden message that a reader would never notice, because the visible sentence was perfectly ordinary prose.

The bench writes the patterns out as literal A and B, which makes them visible for study but destroys the whole point. In real use, the marks are the invisible part and the prose is the visible part.

IIKey space

The mapping is fixed and public; Bacon published it. There is no key in the cryptographic sense at all.

The key is the cover text: whatever prose you choose to hide the marks in. That makes the space unbounded in the same unhelpful way as the null cipher earlier on this loom.

What Bacon actually contributed is the discovery of encoding capacity. Five binary positions carry 32 values, which is exactly enough for an alphabet plus a little slack, and the slack is where the I/J and U/V mergers come from. Every digital character encoding since — five-bit Baudot, seven-bit ASCII, eight-bit bytes, sixteen-bit Unicode — makes the same trade between capacity and redundancy.

The Baconian is the first cipher in this collection whose weakness is not in the algorithm but in the carrier.

IIIEnciphering

• Write the plaintext in capitals, merging I with J and U with V if you are using the classical 24-letter table.
• Convert each letter to its five-mark pattern of A and B (or 0 and 1).
• Choose a cover text — a letter, a page of prose, a printed paragraph — long enough to supply five marks per plaintext letter.
• Assign a physical distinction to the two symbols. In print: two typefaces. In handwriting: two letterform variants, or two inks invisible to the naked eye.
• Write the cover text, applying the distinction in the pattern sequence. Work straight through; do not restart the pattern at word boundaries.
• Send the cover. The receiver reads the distinction, not the words.

IVDeciphering

• Obtain the cover text in a form where the distinction survives — the original printing, not a reset or a photograph.
• Read the text straight through, recording only the distinction at each character position as A or B.
• Group the marks into runs of five, starting from the first character.
• Look up each group in the table.
• If the message runs off the end of the cover, the missing marks are lost — there is no way to recover a truncated Baconian, which is why senders padded their cover text generously.
• Watch for the transmitter inserting a neutral character to adjust alignment. If the first groups decode to nonsense, try shifting the start position by one to four marks.

VWeakness

Typography is a fragile carrier. A book reset by a different printer loses the distinction entirely. A photocopy that renders italic as roman destroys the message silently and completely.

The capacity is tiny. Five marks per letter, with no compression, means a 500-word cover carries perhaps 200 characters of secret. Long messages need absurd quantities of cover text.

Multiplicity. Bacon devised a bilateral variant that also chooses between two near-synonymous ciphertext alphabets, which doubles the marks per character and halves the capacity again. Its complexity made it impractical, and Bacon knew it.

Modern detection is easy. Given a digital text, distinguishing two typefaces is a programmatic check, and any document that mixes fonts without typographic reason is immediately suspicious. Bacon's own era had no automated detection — which is precisely why the technique worked then and would not now.

VIPractice

·History note

Francis Bacon set out the biliteral cipher in The Advancement of Learning (1605) and expanded it in De augmentis scientiarum (1623). He proposed hiding the marks in two typefaces of a printed text, and he devised a second, more elaborate bilateral version that also chose between two parallel alphabets of words. Bacon's motive was partly cryptographic and partly philosophical: he was interested in a notation general enough to express anything, and the biliteral alphabet was his demonstration that any symbol set can be reduced to two. Five-bit encodings did not come into practical use until the Baudot telegraph code of 1870, and binary representation of text did not become routine until the middle of the twentieth century.

·Further reading

The Baconian is the ancestor of every system that hides data in a carrier's form: whitespace steganography, invisible Unicode characters appended to text, and the least-significant-bit methods used to embed data in images and audio.

It also gives the cleanest available answer to a question beginners ask constantly — what is the difference between a cipher and a code? A cipher transforms a message; a code replaces a message with a symbol from a pre-agreed list; a steganographic scheme hides the fact that a message exists. Bacon's contribution is a cipher and a steganographic scheme, published together, which is why it reads as modern despite being four hundred years old.

Next on the loom: Bifid Cipher — a hand cipher from 1901 that gets surprising strength out of a pencil.

+Weave into