Monoalphabetic

A cipher is called monoalphabetic when it uses a single substitution table for the entire message: each plaintext letter is mapped to one ciphertext letter, always the same. If A → D at the start of the message, then A → D at the end too. No evolution over time. It’s the oldest and simplest family, and also the most pedagogical for understanding how a cipher works — and how it falls.

Canonical examples

• Caesar (50 BCE): fixed alphabet shift. Key is an integer between 1 and 25. The shift-3 (Julius’s version) gives A→D, B→E, C→F… Key space: 25.
• ROT13 (Internet, 1980s): Caesar of 13. Special property: it’s its own inverse (ROT13 twice yields the plaintext). Heavily used on Usenet to hide spoilers and corny jokes.
• Atbash (Hebrew alphabet, ~600 BCE): alphabet mirror. A↔Z, B↔Y, C↔X… No key: it’s a single deterministic transformation. Appears in biblical texts (the word “Sheshach” in Jeremiah 25:26 = “Babel” in Atbash).
• General monoalphabetic substitution: the table is an arbitrary permutation of the 26 letters. The key is that permutation. Key space: 26! ≈ 4 × 10²⁶ — astronomical in theory, yet general substitution falls in minutes to frequency analysis.
• Affine: c = a × p + b mod 26. Combines shift (b) and multiplication (a). 12 valid values for a × 26 for b = 312 keys. More subtle than Caesar, still monoalphabetic.

The golden rule of monoalphabetic ciphers

All these ciphers share one fundamental weakness: they preserve letter frequencies. If E is 12.7 % of plaintext, the letter replacing it is exactly 12.7 % of ciphertext. Cryptanalysis just has to count, sort, and match.

This weakness isn’t designer carelessness; it’s a mathematical consequence of the architecture. A fixed table cannot make a frequent letter “disappear” — it just moves it. To break that rule, you must either swap tables along the way (polyalphabetic ciphers) or encipher by groups instead of letter-by-letter (polygraphic ciphers).

Cryptanalysis walkthrough

Take the ciphertext (English, general monoalphabetic substitution):

WK XIIW MV WKZJ WD VKWAB G XV B BWAK GZ CFKZ

1. Counting: list frequencies. W shows up 6 times, K 5 times, Z 3 times.
2. Frequency hypothesis: the most frequent letter is probably E. So W = E.
3. Short words: spot single letters → likely A or I. We see G alone, then B alone — try G = I, B = A.
4. Common digrams: KZ → maybe EN, RE, ON. We saw K is high, so try K = T (THE pattern).
5. Common word: WKZ (three letters with no apostrophe) → likely THE. With W = E? No, that’s XHE. Try W = T, K = H, Z = E. Recheck — THE everywhere coherent.

We iterate. On a longer text (200+ characters), about ten iterations suffice. On isolated short words you sometimes get stuck. The whole beauty of manual cryptanalysis is in this dance between counting and hypotheses.

Index of coincidence as a signature

The index of coincidence of a monoalphabetic ciphertext stays the same as the plaintext (≈ 0.067 in English, 0.074 in French). It’s a recognizable signature at first glance for a cryptanalyst: faced with an unknown ciphertext whose IC sits at ~0.067, you know you’re in monoalphabetic territory before you’ve tested any key. Conversely, a ciphertext with IC ~0.04 is probably polyalphabetic.

Stepping up the security

If monoalphabetic isn’t enough:

• Polyalphabetic (Vigenère, Beaufort, Autokey, Enigma): you switch substitution tables based on position. Flattens frequency. The key becomes a word, a phrase, or even a long random sequence. Crackable by Kasiski + IC, but vastly more resistant.
• Polygraphic (Playfair, Hill): you encipher groups of letters (pairs, triples) rather than single letters. Preserves frequencies at the n-gram level — harder to attack.
• Homophonic: you assign multiple ciphertext symbols to each frequent plaintext letter, picked at random. Also flattens frequency. The Beale ciphers (19th century) are an example.

None of these are mathematically unbreakable — you have to switch to AES or the one-time pad for that. But they raise the bar enough for serious pedagogical or recreational use.

Key takeaways:

• Monoalphabetic = one substitution table for the entire message. Oldest, simplest family.
• Includes Caesar, ROT13, Atbash, general substitution, Affine.
• Preserves letter frequency → crackable by frequency analysis in minutes.
• Index of coincidence unchanged from plaintext (~0.074 FR, 0.067 EN). Recognizable signature.
• To improve security: move to polyalphabetic (Vigenère), polygraphic (Playfair), or modern (AES).