The term shouldn't care whether it's physical or not. Since it'd severely limit it.
Ok. We should however then distinguish from 'demonstrated physics/science' (in the lab) from theoretical physics. I'd propose we put supernatural constructs of religion in the same category as theoretical physics in terms of defining the ground rules for 'evidence'.
Why wouldn't we compare religious and scientific evidence?
Well, I see no real point in comparing supernatural constructs of religion to something like EM fields. One shows up in the lab each time, every time. The other, presumably not so much. It does make sense however to compare *hypothetical* physics to religious supernatural constructs. The often both lack 'evidence' from the lab in terms of demonstrating cause/effect relationships.
The definition I provided with is all about determining what world you're in. How you do that is of less interest.
Well, it gets "messy" IMO when the goal posts start to move, and the number of supernatural constructs increases beyond one. If we add the potential for *multiple* supernatural constructs, that could easily result in multiple supernatural deities in a religion or multiple supernatural constructs in some specific scientific claim that makes such a claim unfalsifiable.
If it predicts nothing, then it's no hypothesis.
We run into areas of science (like inflation) where the actual "predictions" made by the originator of the supernatural construct do *not* match the actual observations. Guth for instance claimed that the universe was homogenous on the largest scales, whereas the Planck data shows hemispheric differences in the data set.
If we allow for *multiple* supernatural constructs *and* we allow for 'goalpost movement', then david has the potential to add a category called 'inflation+curvatons', and the original inflation claim can now *never* be falsified because we can add/or subtract more supernatural constructs as we see fit.
Likewise a religion might try to assign one force of nature to one supernatural construct, say EM fields with Zeus, and another force of nature with another supernatural construct, like associating floods with some other deity.
I'm therefore having a *very hard* time seeing the logic in allowing for an unlimited sized graph, with an unlimited number of supernatural constructs. It seems as though if we do not limit ourselves to *one* and only one, the falsification process goes flying out the window, and anything goes.
Since it doesn't provide with a proposition of a set of worlds.
Or rather, it's not a useful hypothesis in any way (since their purpose is distinguishing worlds).
Like I said, the "theory" of the concept gets a lot more messy when the "worlds" might include multiple 'gods', or multiple supernatural hypothetical scientific entities. I see no falsification possibility in such a model, and the concept of 'evidence' could mean anything to anyone.
Note also that hypotheses can be allowing for things they do not deal with.
For example a completely valid hypothesis would be one that proposes something about the human average height, but allows for everything in all other parameters.
IMO your model works great in *non theoretical physics* but not so great in the theoretical realm where I'm most interested in "evidence".
The following question is a good representation:
"If this hypothesis/theory was true or false, what would the difference be?"
In Guth's case, he *originally* said that the difference would be that the universe would be the same on the largest scales.
Thanks to Planck, that question can finally be 'put to the test'. It's pretty clear now that the *original* claim failed. The ability to add a new supernatural construct and *change* the claim (move the goalposts) ultimately precludes us from falsifying inflation theory. What's the point of having 'tests' that by design can only lead to *more supernatural constructs* every time the original claims are falsified?
I don't see why. Each modification is a falsification.
In a "religious" setting, each "modification' simply adds a new "god", and the assumption of at least one "god" is a given. In the realm of theoretical physics, each modification simply adds a new supernatural construct, and the existence of at least one supernatural construct is *assumed*. There no way to falsify the original claim in such a scenario.
Yup. Could.
Because what if that is the world we're living in?
By the way, if the set of worlds are equal for any two hypotheses/theories they are equally valid (though why someone would add a part that doesn't distinguish between any world is beyond me).
The only way I can imagine that someone could add things that doesn't help with distinguishing worlds is multiplications of propositions.
That doesn't change the set of worlds for the hypothesis/theory (i.e. doesn't help with anything). If you know of another way to add (and in extension subtract) without changing, let me know.
I'll have to chew on it for a bit. Ultimately I'm suggesting that A) God exists and B) connects to humans via the Holy Spirit. I'm ultimately proposing the existence of any additional 'physical thing' (which I assume has real physical substance), and potentially a new (or current) form of energy that allows God to interact with humans. In essence I'm adding multiple constructs (God and energy), but God is simply the physical universe and the energy could be an ordinary EM field. Let me think about it a bit.
Control in this case is just helping to extract additional information from the same event (i.e. evidence).
So consider the Planck/inflation scenario. We extract information from Planck that doesn't jive with the *original* claim about homogeneity on the largest scales. Do we use that 'evidence' to falsify the original claim, or do we allow the original claim to change to something new, and allow for another supernatural construct? Is it "fair" to then claim that there is actually evidence for inflation+curvatons, or is it fair to claim that the original inflation claim was falsified by a "test"?
Nope. I'd like to see a proposed hypothesis and a evidence analysis of that before I make up my mind.
Interesting. You seem pretty squeamish about applying this graph to a "religious" claim. Human are 'real' things and they report having experiences with something they associate with God, often through some device/thing they call a 'Holy Spirit'. It's a crude claim akin to your 'gravity makes things fall down', but it's actually a physically demonstrable claim in terms of even ancient human writings. It's harder however to physically define 'Holy Spirit", define the physical connection between Holy Spirit, and *test* it in a lab in controlled experimentation.
In *theory* at least however, the 'controlled experimentation' wasn't supposed to be a 'requirement' or we must by necessity rule out scientific claims like inflation and dark energy.
Now, whether I make all the work for that or not is irrelevant of course (with the small exception that other parts in the discussion might not agree with my proposition and/or analysis).
I'm willing to concede that it could be, but not that it is.
Well, the 'could be' part is encouraging. Admittedly associating a real and demonstrated form of energy with the term 'Holy Spirit' would provide more "detail', and more ability to 'test' the idea. As I mentioned however, if we start making controlled experimental results the minimum requirement of "evidence", that also automatically rules out inflation theory and string theory, probably large parts of QM. That seems overly restrictive, even by my standards.