>If you define science in terms of the scientific method, you make it a fundamentally useless hobby that can never produce anything of value. If falsifying hypotheses is all you can do, you can never create the kind of knowledge others can build on. The part of science that creates value is building models and explanations that can make reliable predictions. And that can be used as building blocks of more elaborate models and explanations for more complex phenomena.
You don't define it this way in common parlance or even in articles. But you do need to heavily define it this way when categorizing things. If you don't do this you have a categorical mishmash where nothing is ever truly defined properly and filled with fuzzy meanings. See the word "entropy" there are at least 3 definitions for that word and these definitions confuse people and lead most people to not completely understand the concept of what entropy is EVEN at just an intuitive and fuzzy level.
I realize it appears that I'm being overly pedantic and inflexible but this is NOT the case.
To be flexible means you need to understand why language SHOULD NOT have rigid definitions in certain contexts but also WHY it should be extremely rigid and concrete in other contexts. You need BOTH. When it comes to categorization on what is "theoretical computer science" and other very technical topics we Need to have well defined categories and definitions.
The fuzziness of literature and poetry which remains open to interpretation has significantly lower utility in technical fields.
