Unsolved Puzzles

Secret Writing: A History of Codes, Ciphers and Codebreaking

Course 1 · Chapter 12 · Unsolved and Hidden in Plain Sight

Every chapter so far ended with a solution: a method broken, a machine read, a lock opened. This last chapter is different. It collects the messages that, as far as the public record goes, nobody has yet been able to read, shows you how to make puzzles of your own, and then sums up the journey.

Why Some Messages Stay Unread

A cipher can resist for several different reasons, and it helps to tell them apart before looking at any particular case.

  • Too little text. Frequency analysis (Chapter 3) needs plenty of letters. A short message simply may not contain enough evidence, however clever the attacker.
  • A one-time key. If the key was random, as long as the message and used once (Chapter 10), there is nothing to find.
  • Not a cipher at all. Some texts may be an invented language, a code, a hoax, or meaningless. No amount of cleverness reads a message that was never there.
  • Lost context. A code needs its codebook (Chapter 7). If the book is gone and nothing else is known about the writer, the meaning can vanish with it.

Four Famous Open Cases

The Voynich manuscript

A hand-written book of curious plants, star charts and bathing figures, in an alphabet nobody recognises, now kept in a university library in the United States and dated by testing of its vellum to the early fifteenth century. Its text has been attacked by professional codebreakers and enthusiasts for over a century. Proposals include a cipher, a code, an invented language, a natural language in unusual script, and a hoax, and none has been accepted.

The Zodiac ciphers

In the late 1960s a killer in northern California sent newspapers letters containing coded sections. One of the four ciphers was solved within days by a schoolteacher and his wife. Another was reported solved by amateurs decades later. The remaining two are still unread, and the identity of the sender has never been settled.

The Beale ciphers

A nineteenth-century pamphlet tells of a treasure buried in Virginia and gives three ciphertexts, made of numbers, said to give the location, the contents and the names of the owners. The second was solved with a method that uses a famous public document as the key, and describes the treasure. The other two have not been read, and many historians suspect the whole story is a fiction.

Kryptos, passage four

Chapter 10 introduced the sculpture outside an American intelligence headquarters. Three of its four coded sections have been solved. The last, a short passage of 97 characters, has resisted every attempt, although the sculptor has released a few of its letters as hints over the years.

The open cases at a glance Voynich manuscript: a whole book in an unknown script; the text is long, but nothing is agreed.
Zodiac: four ciphers, two read, two still open.
Beale: three ciphers, one read, two open, and doubts about the whole tale.
Kryptos: four sections, three read, one open and very short.
Not Yet Verified Every date, count, place and status in this section is written from general knowledge and was not checked against a source for this chapter. Solutions are sometimes announced, and claims sometimes fail, so search for the current state of each case before repeating any of it.

Making Your Own Puzzle

You now know enough to build ciphers that are fun to solve, and that is a good way to understand why the old ones fell. A good puzzle fits its solver: easy enough to crack, hard enough to enjoy.

DifficultyMethodWhat the solver needs
EasyCaesar shift (Chapter 2)Try all 25 shifts, or spot the pattern
MediumKeyword substitution (Chapters 3 and 4)Letter frequencies and some guessing
HarderVigenere with a short key (Chapter 5)Repeated-pattern spotting, then frequencies
Very hardOne-time pad (Chapter 10)The key, because nothing else works

Three practical tips. Write a message long enough to contain real evidence (about fifty letters is a fair minimum for substitution puzzles). Give the solver a hint, such as the subject or a word that appears. And always solve your own puzzle once from the ciphertext alone, so you know it works.

If you like scripts, a shift cipher or a letter counter takes only a few lines in almost any language. They make a pleasant exercise, though nothing in this course requires them.

Try it yourself: the last message This is a Caesar shift. Find the shift and read it: YMJ XYTWD LTJX TS. (Hint: the commonest letter in English is E, and the ciphertext has a J that looks like an E.)
Show the worked answer

If J stands for E, the shift is 5 letters forward (E, F, G, H, I, J). So subtract 5 from every letter: Y becomes T, M becomes H, J becomes E, giving THE. Continuing: XYTWD becomes STORY, LTJX becomes GOES, TS becomes ON.

The message reads THE STORY GOES ON.

The Whole Story in One Table

Looking back, the same pattern repeats: a method is invented and trusted, someone finds the weakness, and a better method follows.

ChapterThemeThe lesson
1 to 2Hidden meanings and ancient ciphersHiding the message differs from hiding its meaning
3 to 4Frequency analysis and Renaissance officesStatistics beat simple substitution
5 to 6Vigenere and the spymastersA famous cipher can fall; people are the weak link
7 to 9Telegraph, Enigma and Bletchley ParkReused keys and habits give machines away
10 to 11Cold War and public keysKey delivery was the real problem
12Unsolved puzzlesSome messages keep their secret

Where to Go Next

This course told the story and let you try the old methods by hand. To learn how modern security actually works, continue with Cryptography Fundamentals and HTTPS/TLS Fundamentals, which pick up where Chapter 11 left off. For the wartime background, the history courses on the Second World War cover the events that codebreaking helped to change. And if the unsolved cases caught your interest, the best next step is to read a sourced account of one of them and try the puzzles yourself.

Not Yet Verified Course names and what they contain are cited from this site's course list and not rechecked in this chapter. Check the course index before relying on them.

Exercises

Exercise 1

Encipher the word HELLO with a Vigenere cipher, using the keyword CAT repeated as needed (C is shift 2, A is shift 0, T is shift 19). Write out the five letters of the key you use, then the five ciphertext letters.

Exercise 2

Three unknown ciphertexts, each a few hundred letters long. (a) One letter appears about 12 times in every 100, a few appear about 8 times, and many are rare. (b) Every letter appears about 4 times in every 100, with no pattern at all. (c) The commonest letters are flatter than in (a), but the same three-letter groups keep reappearing a fixed distance apart. For each, say which kind of cipher it most likely is (simple substitution, Vigenere with a short key, or a random one-time key) and why.

Exercise 3

Make a puzzle of your own. Encipher the two words OPEN SESAME with a Caesar shift of 7, keeping the word break. Then write one hint for the solver, and say which weakness from this course would let a stranger crack your puzzle without the hint.

Course complete: the short recap Chapters 1 and 2: codes versus ciphers, steganography, the scytale, Caesar and Atbash.
Chapters 3 to 6: Al-Kindi's frequency analysis, Renaissance cipher offices, Vigenere and its fall, and the spymasters.
Chapters 7 to 9: telegraph codebooks, Enigma, Bletchley Park and the code talkers.
Chapters 10 and 11: one-time pads, the Cold War, public keys and the crypto wars.
Chapter 12: the open puzzles and how to make your own.

Quick Reference

  • Reasons a message stays unread: too short, a random one-time key, not really a cipher, or lost context
  • Open cases: the Voynich manuscript, two Zodiac ciphers, two Beale ciphers, the last Kryptos passage
  • Good puzzle: long enough, a hint included, and solved once by you
  • Pattern of the whole course: invent, trust, break, improve
  • Check: the current status of each open case

The end of the course: thank you for reading. Keep your eyes open: ciphers hide in plain sight more often than you would think.