I know this has been discussed before, but I’m not convinced that complexity penalties should apply to anything involving human witnesses.
Suppose someone theorizes that the sun is made of a micro black hole covered in lightbulbs, and there is no obvious physics being broken.… this is an obvious place to use complexity penalties. Simpler models can explain the evidence.
With the Bible though, we have witnesses that presumably entangle the Bible with a divine being. Complexity penalty in this case shouldn’t penalize for extra details. (Considering complexity penalties may still point to “this story is made up for social reasons, and here are some prior sources” instead of “god did it”… but this isn’t due to the amount of detail provided.)
...What? As a technical matter, the laws of probability say that evidence (eyewitness or otherwise) tells us how to update a prior probability, and ultimately a complexity penalty seems like the only way to get sensible priors.
I take you to mean that in a real eyewitness account, we should expect details. That seems more or less right, but largely irrelevant to what I’m saying—even the idea of a human-like mind is more complicated than it appears. That’s before we get to the details of the story (which we might doubt to some degree, in more trustworthy cases, even while paradoxically taking those details as evidence for some core claim).
Even the bare claim that God was involved with certain historical figures is another logically distinct detail we need to penalize before we get to the specifics of any one Gospel or source for the Torah. So the evidence of witnesses would need to overcome this penalty. And of course, in order for them to justify the beliefs about God, we would need to understand what that word means and how someone could directly or indirectly observe its object.
I may have misread your initial comment. To paraphrase to check my reading: you are penalizing due to complexity of a ‘god’ prior but, on the balance, eyewitness details should increase your estimate of the claimed witnessed set being true. More details from eyewitnesses do not then penalize further. The complexity of the god models are just so complex in the first place, that eyewitness details don’t increase your estimate much.
What I’m not grasping is what this sentence meant:
And even an Abrahamic God (or a divine Gospel Jesus, if we treat that as overlapping rather than a proper subset) is pretty detailed if we combine historical claims with some meaningful traits of divinity.
Functionally, we’re talking about the set of vaguely Bible shaped gods… not all the details would need to be true. Eyewitness claims that this bible shaped god interacted with a historical figure should STILL increase your estimate of it happening.… even though that increase may still be infinitesimal.
Excepting things like “the following sentence is false”, eyewitness details should always increase the chance of something like the referenced object existing. It may in parallel also provide evidence that the ‘custody chain’ is faulty or faked… but that’s a different issue.
Pretty much. I’m saying that “vaguely Bible shaped,” rather than “touched down only in Jackson County, Missouri in 1978,” is itself a detail to be justified.
I know this has been discussed before, but I’m not convinced that complexity penalties should apply to anything involving human witnesses.
Suppose someone theorizes that the sun is made of a micro black hole covered in lightbulbs, and there is no obvious physics being broken.… this is an obvious place to use complexity penalties. Simpler models can explain the evidence.
With the Bible though, we have witnesses that presumably entangle the Bible with a divine being. Complexity penalty in this case shouldn’t penalize for extra details. (Considering complexity penalties may still point to “this story is made up for social reasons, and here are some prior sources” instead of “god did it”… but this isn’t due to the amount of detail provided.)
...What? As a technical matter, the laws of probability say that evidence (eyewitness or otherwise) tells us how to update a prior probability, and ultimately a complexity penalty seems like the only way to get sensible priors.
I take you to mean that in a real eyewitness account, we should expect details. That seems more or less right, but largely irrelevant to what I’m saying—even the idea of a human-like mind is more complicated than it appears. That’s before we get to the details of the story (which we might doubt to some degree, in more trustworthy cases, even while paradoxically taking those details as evidence for some core claim).
Even the bare claim that God was involved with certain historical figures is another logically distinct detail we need to penalize before we get to the specifics of any one Gospel or source for the Torah. So the evidence of witnesses would need to overcome this penalty. And of course, in order for them to justify the beliefs about God, we would need to understand what that word means and how someone could directly or indirectly observe its object.
I may have misread your initial comment. To paraphrase to check my reading: you are penalizing due to complexity of a ‘god’ prior but, on the balance, eyewitness details should increase your estimate of the claimed witnessed set being true. More details from eyewitnesses do not then penalize further. The complexity of the god models are just so complex in the first place, that eyewitness details don’t increase your estimate much.
What I’m not grasping is what this sentence meant:
Functionally, we’re talking about the set of vaguely Bible shaped gods… not all the details would need to be true. Eyewitness claims that this bible shaped god interacted with a historical figure should STILL increase your estimate of it happening.… even though that increase may still be infinitesimal.
Excepting things like “the following sentence is false”, eyewitness details should always increase the chance of something like the referenced object existing. It may in parallel also provide evidence that the ‘custody chain’ is faulty or faked… but that’s a different issue.
Pretty much. I’m saying that “vaguely Bible shaped,” rather than “touched down only in Jackson County, Missouri in 1978,” is itself a detail to be justified.