A while back I was having a conversation with someone who professed to work with undead spirits and the supernatural. As you know, I don't think that appealing to the supernatural is a good explanation for whatever phenomenon we are attempting to describe. As a matter of fact, from a Bayesian point of view, it's much more likely that a naturalistic explanation that has an intrinsically higher prior probability, that can also explain the phenomena, is the better candidate for explanation.
I didn't actually get into a full-blown arguement with this person, I just sort of asked some Socratic-dialog like questions to see if she could quell my skepticism. One of the things I asked was how it was that these ghosts and other supernatural beings intake useful energy and release non-useful energy. She didn't seem to understand what I was asking, so I used a car and the human body as an example. Gasoline/food goes in, heat/methane/CO2 goes out. She simply asserted that supernatural beings consume "electromagnetic energy". This was an obvious attempt at using the literary genre of science to make it sound like she was doing actual science, but I obviously saw through that (i.e. I asked "what do you mean by that?" but she couldn't answer, since there are multiple types of E-M radiation).
What she then said was that science was wrong in the past, so it will probably be wrong in the future, and used the shift from a geocentric view of the universe to a heliocentric view of the universe as an example. "Laws of physics" she said, "were completely overwritten in the past, so they will be overwritten in the future". Therefore, she will be vindicated in some future physics that will show that her beliefs were always compatible with science. I actually do think that will happen, but not from the physical sciences. The part of science that will shed light on why she believes what she does will probably come up in neurology and psychology, or other areas of cognitive science, not from the physical sciences. Well, that part is already available, it's just that this scientific knowledge hasn't become common knowledge. Yet.
Anyway, if I had been thinking faster, I would have corrected her misrepresentation of the progression of science with the analogy I'm about to write. But all I could say was that previous theories were never really overturned, it was just that newer theories were less wrong than previous ones.
So yeah, here is the analogy: Imagine you live in an apartment building Harlem and you attempt to draw a map of all of New York City. Let's say you've never been outside of the block that your apartment building is on. This map might be accurate for say, your immediate neighborhood, but it won't be accurate for all of NYC. Let's say, then, that you decide to walk a 10 block radius around your apartment. Again, this map might be more accurate than your previous map; it will be accurate for the 10 block radius around your apartment and a bit beyond, but it still won't be accurate for all of NYC. Next, let's say that eventually, you decide to walk all around Manhattan island and then attempt another map of NYC. Again, this map will be more accurate than your previous maps, but it won't be accurate for the entirety of NYC.
Each update to this map, as you explore more and more of NYC with your own eyes, will be more accurate than your previous maps until you've been to all of NYC. Moreover, you will be able to see the logical evolution of these maps and why the previous maps got what they got right and got what they got wrong. No successive map will be completely unrelated to a previous map. Furthermore, each map follows a Bayesian updating scheme, where new information is included into previous information.
The scientific model of knowledge progression that the psi-ist person was promulgating was a wholly unBaeysian progression of science. As Isaac Azimov wrote, they think that one day science will say that the Earth is a sphere, and then the next day science will say that the Earth is a rectangle, the next day a trapezoid, the next day a donut, etc. None of those models has any logical progression between them, and none of those models follows any sort of Bayesian updating. They are just random.
So the next time I encounter someone who attempts to appeal to some future physics that will one day vindicate their psi beliefs, I'll have to remember to use that map analogy to set them on the correct path. Hopefully, any reader(s) of my blog will do the same 😉