There are two types of science: (a) observational, (b) experimental.
The latter is basically heuristic (trial and error) and has been formalized into the 'scientific method' in which the trial is preceded by a hypothesis, and at the end there is some kind of a theory that has powerful predictive properties.
There is nothing here at all that disputes the roots of observational science in the observations of the natural world by the ancients, some of which are exquisitely detailed, well-preserved even today, and beautiful.
There is nothing here to say the ancients were incapable of doing experiments. Why would they be incapable? But the tools available for experimentation in those days were exceptionally simple. Even Galileo simply invoked a hammer and a feather.
Also there is nothing inherently inaccessible about today's science. There is really nothing to stop anyone studying it. It takes time, and the payoff is not that great. You have to have a genuine curiosity about how things work in the material world, beyond the5 senses, which is a bit of an isolating path.