Verification vs. falsification
Ed Clarke has observed that this general area of improving confidence in a specification should probably be called falsification rather than verification
This is not so different from the shift in testing terminology
- Does a test case succeed or fail if it exposes a problem?