Author here. Please note that this is the fifth installment of what is planned to be a very long series of articles about the scientific method targeted at a general audience. Comments and constructive criticism are welcome, but please keep the context in mind.
Hi Ron. I've definitely gotten into arguments with you about quantum theory on HN before, so it's nice to see something about which I broadly agree with you.
I skimmed back over your "Scientific Method Part 4" post and noticed the following paragraph:
> The reason science is naturalistic and atheistic is not because these are prejudices built into the method by fiat, it is because it turns out that the best explanations -- the most parsimonious ones that account for all the known data and have the most predictive power -- are naturalistic. The supernatural is simply not needed to explain any known phenomena.
I wonder if there's some nuance missed here. The natural follow-up question in my head is: can the scientific method ever support a supernatural explanation? What could such an explanation look like? How could it have predictive power whilst maintaining its supernaturalness?
I wonder if, actually, the scientific method is inherently at odds with supernatural explanations because as soon as an explanation has genuine predictive power (edit: and is parsimonious), it becomes natural.
> I wonder if there's some nuance missed here. The natural follow-up question in my head is: can the scientific method ever support a supernatural explanation? What could such an explanation look like? How could it have predictive power whilst maintaining its supernaturalness?
In principle religious prophecy could fit the bill. You could imagine a surprising and unambiguous religious prophecy about a future event, such as that the Yellowstone Caldera will erupt in February of 2025. If a series of such prophecies were successfully made about various events spanning a variety of disciplines or topics, each attributing the knowledge to the same deity, it would be difficult for me to not attribute the predictions to the supernatural.
In practice though, religious prophecy tends to either fail in being surprising or in being unambiguous. And when it is not unambiguous, it is not falsifiable.
edit: I would also add that it is important that the prophecy be about something that is independently verifiable as well.
Religious prophecy is an interesting example. It highlights the distinction between "explanation" and "prediction". Here, the explanation takes the form of a hypothesis that there is an omniscient deity. Is this a falsifiable hypothesis? Yes we can point to the accuracy of the predictions as an argument in its favour, but that doesn't differentiate between the deity hypothesis and - for example - an alien species with advanced predictive power that lives secretly among us. And if there were other tests that could be used to distinguish the deity hypothesis from alternatives, then I feel that the deity is behaving within the laws of nature.
> can the scientific method ever support a supernatural explanation?
Of course it can, which is to say, it can support an explanation that would be considered "supernatural" by today's standards. If there were evidence of supernatural (by today's standards) phenomena, science could easily incorporate supernatural (by today's standards) explanations for those phenomena. But the word "supernatural" means what it means (today) for a reason. A lot of people have looked for evidence of phenomena beyond the Standard Model and failed to find it. This is not to say that there might not be a breakthrough tomorrow, but I'll give you very, very long odds against.
> as soon as an explanation has genuine predictive power (edit: and is parsimonious), it becomes natural.
Yes, but that's kind of like how any AI technology that actually works is no longer AI. Science can easily incorporate deities, demons, psychic phenomena, Bigfoot (heck, that's just a new species, happens all the time). In fact, all of these things started out as bona fide scientific hypotheses back in their day. The thing that makes them "supernatural" is simply that they are at odds with the current set of data. That could change any time. But the predictive power of current theories means the odds are not with you.
It sounds like you are using it to mean "a supernatural phenomena is one which cannot be explained by the current scientific models". In this case, by definition the scientific method cannot support supernatural explanations.
Some people might take "supernatural" to mean "beyond any possible scientific law of nature, full stop". But even here we have that by definition the scientific method cannot support supernatural explanations.
I would argue that the vast majority of working scientists are naturalists - they accept something as being scientific only if it can be observed, measured, tested etc... .
> It sounds like you are using it to mean "a supernatural phenomena is one which cannot be explained by the current scientific models".
Yes, that's close enough.
> In this case, by definition the scientific method cannot support supernatural explanations.
Wrong, because in the first case you included the word "current" and in the second you didn't.
The reason that the supernatural is what it is today is because no observations require it as an explanation. But that could change at any time.
> I would argue that the vast majority of working scientists are naturalists - they accept something as being scientific only if it can be observed, measured, tested etc... .
The scientific method is to come up with the best explanation that accounts for all observations [1]. Things that are not observed need not be accounted for.
It's interesting to see the author answer with "yes." As I understand, two of the underlying principles presupposed by science as an endeavor are methodological naturalism and the principle of the uniformity of nature. The second only really needs to be locally true, i.e. maybe fundamental physical laws or constants don't necessarily hold across all time and space, but they hold at least enough to be able to infer them from data, make predictions, and confirm or disconfirm those predictions. Methodological naturalism seems to be more necessary, though.
There are at least two forms of supernaturalism I can think of. First is the unmoved mover that lays down fundamental laws and constants, creates the basic ontology of physical substrate, and sets everything in motion, but from that point on, never actively intervenes. This is the God of Deism and the clockwork universe of the early Englightenment. Second is one or more personal forces that are not constrained by the physics of our universe, who actively watch and care what happens, can potentially be pleaded with, and may intervene to cause events that have no physical explanation. This is the more common God or gods of theistic traditions, or spirits of animistic traditions.
Each of these would be impossible to investigate scientifically because their actions would need to formulated purely in terms of internal state, or intent and teleological goals. They could not be predicted. These supernatural beings could potentially act in any manner they please, making them impossible to understand purely by observation. Instead, you could at best hope they might explain to you their actions and motivations and that these would not change over time. This is why religion relies upon prophesy and received knowledge or revelation, not observation.
On the other hand, it seems the author is here using "supernatural" not to necessarily refer to things that are actually supernatural, but to encompass things like ghosts and remote viewing that may be postulated as supernatural, but assuming they were real, could also have naturalistic explanations we just don't currently have the means to figure out. Science can and even currently tries to investigate things like this. It might be niche, but it happens. Even T1 research universities have departments dedicated to paranomal research, Duke most prominently that I'm aware of among US schools.
1 point by AnimalMuppet on Dec 3, 2022 | parent | context | un‑favorite | on: The Truth Matters and Secular Humanists Should Def...
To some degree. (Warning: long rant follows.)
I define "the scientific method" as a process with four steps. 1: Systematic observation. 2: Looking for regularities in the observations. 3: Forming hypotheses to explain the regularities. 4: Experimentally testing the hypotheses.
(This isn't just me. Wikipedia describes the scientific method very similarly. This isn't me trying to cook the definition to make a point.)
Now assume, for purposes of argument, that God exists. And since lots of people use the word "God" to mean lots of different things, I'm going to define what I mean. By "God" I mean a being with personality - someone, not just something. Someone who existed before the physical universe, and exists independently of it - "outside" it in some sense, though I don't mean in a geometric way. Someone who created the physical universe, and who can, at his sole discretion, reach into the physical universe and change things. (You can think of this kind of like using a debugger - you can stop the program and change the value of a variable, with no antecedent in the flow of execution of the program.) And if and when he does, things actually change - it's not just a change in our perception; things actually physically change.
All this I ask you to assume for the sake of this argument. You don't have to believe it, but it's there for the argument to make sense.
And one more assumption: Assume that this God actually does change something, and that science observes it at step 1. The question is, what's science going to do with it?
They're going to throw it out at step 2, because there is no regularity. There's no pattern, unless God does this a lot.
But if they don't throw it out at step 2, the next problem comes at step 3 - forming a hypothesis. The way science currently is, "God" is very much out of favor as a hypothesis. But in our thought experiment, God actually is the explanation. Science is never going to propose the correct explanation for such an event.
But even if science proposes God as the hypothesis at step 3, the next problem comes at step 4: how are you going to test it? "Um, God, could you do that again? And, um, sign it this time?" You can't run the experiment. I don't see how you could run the experiment even in principle.
So in that sense, the supernatural exists outside of science. This doesn't tell us anything about the supernatural; it merely tells us that science is not a useful tool for examining it.
