Structured programming forces us to recursively decompose a program into a set of small provable functions. We can then use tests to try to prove those small provable functions incorrect. If such tests fail to prove incorrectness, then we deem the functions to be correct enough for our purposes.
