**Starting Fresh**I'd like to try to make this discussion much simpler. Let's forget all of the methods people have come up with...forget the "scientific method", "zetetic method", or any other method. Forget how science or other thinking was or is done in practice. None of that matters. Let's throw all that out.

We start fresh. The only thing we come into this with is the rules of logic.

*If anyone disagrees with any aspect of logic, then please share specifically what you disagree with, and why.***Some Rules, Perhaps?**In fact, let's put down some rules, just for this thread:

1. No discussing the scientific method, zetetic method or any other method that was created outside this thread.

2. No discussing how theory making and testing has historically been done in practice.

3. Every statement of disagreement should come with an explanation of why, and should follow logic correctly.

Can we stick to these rules?

**The Goal**Our goal is to find the correct set of theories on a given topic (a random example being, say, "the shape of the earth"

).

I say "set of theories", because we have to leave open the possibility that more than one theory could be correct.

**Thought Experiment and Unavoidable Steps**For the graphically minded, let's start with a thought experiment. Imagine a rectangle. Every coordinate position inside that rectangle represents a distinct theory on the topic. Thus, the rectangle contains a set of theories on that topic.

Forget how we arrived at these theories - it doesn't matter. All we require is that we have a set of theories which, at this point, have not been proven incorrect.

An observation is made (within the scope of that topic, of course, which means that each theory in the rectangle will make a prediction). Each of those theories (i.e. coordinate positions) will either predict that observation correctly, or incorrectly. Now we draw a shape within the rectangle that encloses only those theories that correctly predict the observation, and those that incorrectly predict the observation are outside the shape.

Now another observation is made, and the process is repeated. Some theories will lie within both shapes, because they predicted both observations correctly. Some will fall inside one shape and outside the other, because the theory was correct with one observation, but not the other. Other theories will lie outside both shapes, because they failed to predict both observations.

We can continue this process with each new observation.

Notice that I haven't said anything about how we arrive at the set of correct theories. I haven't chosen any method. I've only said that, given a starting set of theories, at some point we make observations, and we compare those to predictions made by each theory in the set. Any method we conceive will need to do this.

*If you disagree with this, please illustrate how this is possible by creating and describing an example method that does not do these steps in some manner.***In Search of a Method**Now, how can we reach our goal, that is, to find the set of theories that are correct?

First we need to define what a correct theory is. A correct theory is one whose predictions match the corresponding observations FOR ALL OBSERVATIONS THAT ARE POSSIBLE.

*If you disagree with that definition, then share your revised definition.*This has some implications. To demonstrate that a theory is correct, i.e. to prove it is correct, one must demonstrate that there are no observations that can ever be made that the theory will predict incorrectly. This amounts to drawing a shape for every observation that could ever be made, and KNOWING THAT YOU HAVE DONE SO. Until you have done that, you have not proven the theory correct.

Exactly how can you know that you have checked every possible observation? I say you can't. Hence,

you can never prove a theory correct. *If you disagree with that, I ask that you explain specifically and in detail how you would check every possible observation, and know that you have done so in practice.*Even if one could somehow succeed at this, there would remain the possibility that other theories exist that ALSO predict everything correctly, in which case more than one theory is correct. So even if you could prove a theory correct in this manner, you have said nothing about the correctness of any other theory.

Alternatively, proving incorrectness is far easier, because only one incorrect prediction is required, at which point that theory can be rejected and no longer needs to be considered. This amounts to drawing the shape for the first observation, rejecting every theory that doesn't lie within that shape, and then repeating with other observations on only the theories that remain within ALL of the shapes from preceeding observations. By doing so, the set of theories under consideration is reduced with each observation to a smaller and smaller set. In this way, you converge to the set of correct theories.

Put a different way, if I take a set of theories, and I observe that they all predict an observation correctly, what have I learned? Not much. The entire set of theories are still possibly correct, just as they were when I started.

If, however, I see that some theories in that set fail to predict an observation correctly, I have learned that those theories are incorrect, and I've made progress in that I've reduced the set of theories that I need to continue to consider.

So I ask, what would you do? Attempt to prove correctness, and thus meet all of the requirements it implies? Or attempt to prove incorrectness to converge to the subset of correct theories?

Or would you do something else?

*If so, explain exactly how you would proceed, using the same definitions I've used above to help with clarity.*Finally, if you feel the urge to comment about how you disagree with something above or think something is wrong, but you aren't willing/able to explain why with specifics, please don't bother responding. Your energy would be better spent cheering for your favorite sports team, or getting into a "yes-no" argument with a 4-year old.