> It's called the "null hypothesis." It's the gold standard of scientific experimental design.
I think this definition is tying too much epistemic certainty on the part of the scientist towards the null hypothesis versus the alternative hypothesis being tested, hence my disagreement to what I consider a view lacking nuance. In short, if the scientist has no reason to doubt the null hypothesis then there is no reason to test. So the scientist must first be willing to allow the null to be proven unlikely / rejected.
I don't think we're going to see any further agreement this deep in the semantics, so let's move forward understanding each other to be in general agreement on the metaphysical construction of the scientific method.
> And in science, we assume such things are not real until positive evidence leads us to a different conclusion
We assume them to be untestable, not necessarily false. Just unable to be tested. The scientific method only has three states: untested, agrees with available evidence, or rejected by available evidence. If no evidence, then untested.
String theory is a great example of this. Wonderfully mathematical and logical, but we haven't figured out how to test major components of it specifically yet that would distinguish it from alternative theories.
> In short, if the scientist has no reason to doubt the null hypothesis then there is no reason to test.
On the contrary, that's the point at which positive evidence may contradict the null hypothesis, assuming it exists. But the null hypothesis must be the default initial assumption.
My point is that if the scientist assumes anything but the null hypothesis a priori, it's not science, it's marketing. This is why the null hypothesis is the default initial position in any legitimate scientific investigation.
The alternative is to assume the truth of a theory and seek falsifying evidence. But this may require proof of a negative, which is frequently impossible. I can't prove Bigfoot's nonexistence, but this failure doesn't support Bigfoot's existence.
I think this definition is tying too much epistemic certainty on the part of the scientist towards the null hypothesis versus the alternative hypothesis being tested, hence my disagreement to what I consider a view lacking nuance. In short, if the scientist has no reason to doubt the null hypothesis then there is no reason to test. So the scientist must first be willing to allow the null to be proven unlikely / rejected.
I don't think we're going to see any further agreement this deep in the semantics, so let's move forward understanding each other to be in general agreement on the metaphysical construction of the scientific method.
> And in science, we assume such things are not real until positive evidence leads us to a different conclusion
We assume them to be untestable, not necessarily false. Just unable to be tested. The scientific method only has three states: untested, agrees with available evidence, or rejected by available evidence. If no evidence, then untested.
String theory is a great example of this. Wonderfully mathematical and logical, but we haven't figured out how to test major components of it specifically yet that would distinguish it from alternative theories.