An axiomatic method for proving the correctness of computer programs, using logical rules to reason about how statements change program variables.
Hoare's Oxford inaugural lecture, arguing that programming is a mathematical activity and that mathematical methods can ensure programs are correct and reliable.