A known word in an unknown message, and the leverage it creates
A crib is a guess about what a fragment of plaintext says — a word, a phrase, a standard opening — placed against an enciphered text in the hope that the alignment reveals something useful. The name is old, probably borrowed from the students' sense of a hidden answer, but the technique itself is as old as organised cryptanalysis.
The logic is straightforward. If you know — or can reasonably assume — that a message begins with a particular phrase, you have a second piece of text whose relationship to the ciphertext is fixed. Every letter of your guessed plaintext maps to the letter above or below it in the cipher. In a monoalphabetic substitution, a reliable crib of a dozen letters hands the analyst a dozen entries in the substitution table for free, and frequency analysis supplies the rest. Even a short, confident crib is often enough to crack the alphabet open.

The technique grew more powerful, not less, as cipher machinery became more sophisticated. Against the Enigma, cribs were indispensable. German operators followed rigid procedures — weather reports began with the word for weather, U-boat grid references followed predictable formats, officers' messages often opened with the same courtesies. The British analysts at Bletchley Park called these known or suspected fragments cribs and built their work around them. The bombes — the electromechanical machines designed around Alan Turing's theoretical approach — did not work by exhaustive search alone; they worked by testing whether a specific crib, placed at a specific position in a message, produced a consistent set of Enigma settings. The crib turned an impossibly large search into a much smaller one.
Cribs could also be manufactured. When analysts needed a plaintext fragment and none was available from habit or routine, stations occasionally induced the enemy to generate one. Laying mines in a known location and then monitoring the resulting radio traffic — knowing the position would be reported — was one documented method. The expected grid reference became the crib.
What made a crib dangerous was not only that it unlocked a single message. A confirmed crib meant confirmed settings, and confirmed settings meant every other message sent on the same key that day could be read. The leverage was enormous and asymmetric: one reliable phrase, correctly placed, could open an entire network's traffic for twenty-four hours.
The vulnerability was procedural rather than mathematical. A cipher system that allows an adversary to predict any portion of plaintext — through formula, habit, or sheer repetition — has already conceded a foothold. Cribs are what operator habit produces: not a flaw in the wiring, but in the people using it.