An appropriate deduction might be "Cars have windows and can move, houses have windows and cannot move. The presence of windows alone is not what allows the car to move."
A fair point - here's a generalisation. W denotes windows, M the ability to move, a and b are two objects for which their possession of windows and ability to move (or otherwise) is known, and x is some other object.
That works better; the conclusion is practically useless due to the amount of combinations that wouldn't allow the car to move, but at least it's reliable.
Going past that would require messing around with things that don't move until they do, or vice versa. Also known as science.
That sounds plausible but is, in fact, kinda a made up distinction you just came up with. People up to and including Isaac Newton used the phrase 'natural philosophy' to describe what they were doing. 'Testing' in any meaningful sense of the word was a part of that more often than not. Even Pythagorean astronomy was implicitly testing things by making predictions of the movement of celestial bodies. So, no, but thanks.
Edit: also worthwhile, I feel, mentioning that a lot of good science is purely observational and involves no direct testing, even of theorems. E.g early paleontology would, I feel, fit into that theme
You're mincing words to not acknowledge the blatant qualitative difference between what those Greek philosophers were doing under the label "natural philosophy", versus what people past Bacon (exemplified in your comment by Newton) were doing.
And, as a result, your comment boils down to an "ackshyually" leading to a clearly idiotic conclusion.
Even Pythagorean astronomy was implicitly testing
Emphasis mine. Doing it "implicitly" doesn't cut it out; this shit needs to be explicit and systematic. You need to take the bloody window off the car and see if it still moves, then take off the lights, so goes on.
Even when direct experimentation is not possible due to the nature of the subject, you need to formulate a bunch of alternative explanations and find a way to sort them out. i.e. explicitly test shit.
And this is so fucking obvious that I'm not wasting my time further with you. If you're so blatantly ignorant on the scientific method, Wikipedia is a good start.
Mate, first of all: chill. Second of all: me, mincing words? You're the one who's decided the phrase 'natural philosophy' only applies to ancient Greeks. It's literally just what science used to be called. Being very very hung up on a specific definition of a word or phrase and excluding other common usages is not a good basis for an argument. There need have been no argument at all. 'We call it science now' seems to be what you meant, but it's not coming across well
Oh. And you can still do science without a theory. It's called data collection and it is absolutely vital.
Edit: a good example would be Rosalind Franklin's work on the structure of DNA. She did some incredible science with x-ray diffraction which was vital to Crick and Watson's theory of the structure.
Edit X2: also 'doing it implicitly' is totally fucking fine if the result of being invalidated is you create a more refined model of reality. Which is, like, pretty much what the whole thing about astronomy is...