All science is about building a model to help us explain how something in nature works and to predict how it will work in the future.
In the beginning the model is called an hypothesis
. After a few experiments that support the hypothesis we usually then describe it as a theory
. After many researchers have completed many experiments that support the theory and no experiments have been shown to refute the theory, it then is usually described as a law
There are several things you need to keep in mind throughout this process - usually called "Scientific Method":
- The original hypothesis must be stated in such a way that researchers can set up experiments consistent with the statement of the theory.
- All experiments must be conducted and documented to demonstrate this consistency.
- All experiments must be documented in such ways that any competent researcher with the right equpiment can replicate the experiment and achieve the same result.
- Any limitations known at the time must be stated within the hypothesis - more about this below.
In general, one experiment that does not support the hypothesis invalidates the hypothesis.
What often happens, however, is that a limitation not noted in the original statement of the hypothesis appears in one of the experiments. This, then, does not invalidate the hypothesis, it simply means this limitation must be built into the hypothesis.
For example, Newton's Laws of Motion work perfectly well if we are playing a game of pool by moving colored balls around a pool table. They don't work quite so well when we're talking about either things with huge masses such as stars or things moving very fast such as light. This doesn't make Newton's Laws "wrong" or "incorrect", it simply means we have to keep these limitations in mind.
If you keep these comments in mind, your questions about Methodological Naturalism
or the Demarcation Problem
can be put in proper perspective.
That's probably about enough for this post. I could talk about it all day.
Thanks for opening a very interesting thread.
All the best