"Science" is, loosely speaking, a process for formulating and evaluating predictive models of the behavior of systems with unknown characteristics, based on emp

"Science" is, loosely speaking, a process for formulating and evaluating predictive models of the behavior of systems with unknown characteristics, based on empirical observation. Theoretical "computer science" is, most of the time, a matter of building abstractions from smaller concepts and making analytical, not empirical, observations of the characteristics of those abstractions. This is much closer to what is generally thought of as "abstract mathematics". The closest you're likely to get to scientific methodology when programming is when you have to write test code to figure out what the crap this freaking third-party library you're stuck with is actually doing because clearly it isn't what the documentation says it should and... at which point we've moved away from "computer science" and into why "software engineering" is such an embarrassment compared to real engineering disciplines. If you want to hear someone smarter than myself expound similar ideas, look up some of E. W. Dijkstra's work--the man was a genius, a "computer scientist" par excellence, and he expected software engineers to act like engineers.