The history and purpose of cryptography

Every technique in this catalog — no matter how modern or mathematical — exists to deliver some combination of four properties. Losing sight of these is how organizations end up encrypting data that also needed to be signed, or signing data that also needed to stay secret.

The rest of this module walks through this arc in more detail — but the shape of the whole story is visible at a glance: millennia of clever secret-keeping followed by a few decades of applied mathematics that changed the rules entirely.

The earliest ciphers didn't hide that a message existed — they hid what it said, using methods simple enough to execute by hand in the field. The scytale rearranged letters (transposition); the Caesar cipher replaced each letter with another (substitution). Both assumed the method itself, not just a key, needed to stay secret — an assumption that held only as long as nobody studied the method carefully.

Around 850 CE, the Arab polymath Al-Kindi wrote the oldest surviving manuscript on breaking ciphers. His insight was statistical: in any given language, some letters occur far more often than others (E, in English), so a simple substitution cipher preserves those frequencies — count the letters in the ciphertext, match the most common one to the language's most common letter, and the rest unravels.

This is the moment cryptography stopped being purely a craft of clever concealment and became a contest with a countermeasure — every cipher design from this point on had to survive someone actively trying to break it, not just someone who happened not to notice it.

The response to frequency analysis was to stop using one substitution and use many: a polyalphabetic cipher that shifts by a different amount for each letter, following a repeating keyword, so no single letter frequency stays fixed. This scheme was first described by Giovan Battista Bellaso in 1553 — though it's almost universally known today as the Vigenère cipher, after Blaise de Vigenère, who in 1586 published a related but different autokey cipher and was credited with Bellaso's work by a 19th-century historian's mistake.

Whoever gets the credit, the cipher earned its nickname "le chiffre indéchiffrable" (the indecipherable cipher) and held that reputation for roughly three centuries. Charles Babbage privately broke it around 1854 using a technique based on finding repeated sequences in the ciphertext to estimate the keyword's length, but never published the result; Friedrich Kasiski independently rediscovered and published the same method in 1863, and the cipher's reputation never recovered.

By the early 20th century, encrypting by hand couldn't keep pace with the volume of military and diplomatic traffic. Arthur Scherbius patented the Enigma machine in 1918 — a typewriter-like device using rotating wired disks (rotors) to implement a substitution cipher that changed with every keystroke, reaching a huge number of possible configurations.

Enigma was first broken not by the famous Bletchley Park effort but earlier, by Polish cryptologists (Marian Rejewski and colleagues) in the early 1930s, who reconstructed its internal wiring mathematically. Their work was passed to Britain shortly before WWII, where a team at Bletchley Park — including Alan Turing — industrialized codebreaking at a scale that, as a side effect, helped establish the foundations of programmable computing.

In 1949, Claude Shannon published "Communication Theory of Secrecy Systems," applying the information theory he'd developed to formally define what a cipher can and can't guarantee. Shannon proved that a one-time pad — a key as long as the message, truly random, used exactly once — achieves perfect secrecy: a ciphertext that reveals mathematically zero information about the plaintext, regardless of an attacker's computing power.

This is a genuinely different kind of security guarantee from everything else in this catalog. AES, RSA, and ECC are all only computationally secure — breakable in principle given enough computing power, just not in any practical amount of time. Shannon's work drew that distinction precisely, and gave cryptography its first rigorous mathematical foundation.

Within about a year of each other, two developments ended the era covered by this module and started the one covered by the rest of the site. In 1977, the U.S. government standardized DES (the Data Encryption Standard) as the first publicly available, thoroughly analyzed symmetric cipher — the direct ancestor of AES. And in 1976–1977, Diffie, Hellman, Rivest, Shamir, and Adleman published the key exchange and encryption schemes covered in the Diffie-Hellman and RSA modules, solving the key-distribution problem that had limited cryptography for 2,500 years: two parties with no prior shared secret, communicating over a channel an adversary can watch, could now agree on one anyway.

Everything covered elsewhere in this catalog — AES, RSA, ECC, hashing, TLS — descends directly from that 1970s pivot. It's exactly what this site calls "classical" cryptography: not ancient, but the specific body of algorithms built between 1976 and roughly 2015, before the prospect of large-scale quantum computers put RSA and ECC's underlying hard problems at risk. The rest of this catalog is a tour of what that classical era actually built, and how it works.