Spheres aren't. Most human constructions aren't.

Can you choose what to have for breakfast? Yes. Does the same effect usually follow cause in controlled conditions? Yes. So some form of compatibilism is true.

That's precisely what I intended. All of those words denote things which, at face value, exist in different ways.

That's ok, Ray Brassier is the usual answer. ;)

I think it's appropriate to imagine things into existence so long as there's not a unique sense of what it means to exist. EG, Gods are myths, chairs are actual, the abstract concept of golf is something else...

Predicate logic fails to capture the full content of the Ontological argument. The Ontological argument moves from existence in imagination to existence in reality. So, the TOP depends on the distinction between existence in imagination and existence in reality. This crucial distinction can't be made in predicate logic. The existential quantifier, (Ex), is restricted to reality.

Honestly I think the ambiguity in the domain of quantification is the essential feature of the argument. You want there to be a single way in which something can be said to exist, collapsing the distinction between an imagined perfect entity and a real perfect entity - and if the argument is right, that's exactly what happens.

The distinction you draw absolutely is not in the argument's favour, since 'God exists in the imagination' is supposed to show through some property that 'God exists in reality', it may be that predication of 'is the greatest being imaginable' or like sentences is only a predicate of strictly imaginary entities... And really that's what's supposed to be so strong about the argument - you can go from supposedly minimal assumptions about imaginary entities and properties and deduce the existence of an entity. But if you would like to predicate God in a manner that makes God strictly imaginary that makes the existence of God a moot point.

@Fuji

What got you interested in accelerationism?

Thanks for the comments. I struggled a bit with the use of the word "truth." I know people involved in this process, well-meaning folks, and I think it is fair to describe their collective perspectives (facts) as their "truth." As the four wire rod companies -- which also have wire product manufacturing plants -- employ far fewer people than do the overall wire manufacturers, is it fair to decide a case based on the greater potential loss of employment?

I doubt there's any way to unambiguously evaluate which is the better choice. I've tried to come up with two 'recipes for deciding in an unambiguous way' - so the result may be uncertain but at least the methods' fairnesses can be evaluated.

I think if the bolded bit were the only criterion by which the case was decided it's not quite right. Imagine that account 1 predicts collection of horrible circumstances X, and account 2 predicts collection of horrible circumstances Y; the only deciding factor for whether we act in the manner suggested by account 1 or (xor, exclusively) 2 is the one which makes the most horrible predictions. In the absence of trying to evaluate which is the most likely or well reasoned the biggest crier-of-wolf would win.
____________________________________________________________________________________________
Method A
____________________________________________________________________________________________
If you must assume that X and Y (the predictions from account 1 and account 2 respectively) are equally likely, then going by the one which generates the most unfavourable circumstances would probably be the best (in the sense that it mitigates the worst things that are expected to happen).

So:
(1) comparative study: evaluate whether X or Y (the sets of horrible predictions) is more likely to happen
(2) comparative study: evaluate whether X or Y is more horrible.
(3) if both (1) and (2) agree OR account 1 is more likely then account 2 but both are equally horrible OR account 1 is as likely as account 2 but account 1 is more horrible OR the converse two statements (replacing account 1 with account 2 in between the ORs), go with the most likely one with the most horrible consequences.
(4) if they do not agree: comparative study: find areas of disagreement between account 1 and account 2, which parts of account 1 and account 2 are the most horrible and how likely are they to occur?
____________________________________________________________________________________________Method B
____________________________________________________________________________________________

Another perspective - if both accounts are compatible (like in @Srap Tasmaner's story) - assume one is true (that X will happen or that Y will happen), how likely is the other to happen?, find cites of contradiction and disagreement between them to evaluate how likely you are to be fucked in the other way and in both ways given both accounts and implementing all mitigating strategies for one.

If both seem unfair, back to the drawing board.

Have you considered starting a blog?

My favourite response to the ontological argument is an attack against the uniqueness of the entity it conjures into being.

(1) An entity X has property P.
(2) If X did not exist, then it would not have property P.
(3) X exists. (1,2, modus tollens)
P is typically 'greater than that which cannot be conceived' and X is god.

You can adjoin any set of properties to X, so long as P is among them X exists if the argument is valid. The argument is invalid anyway since (1) quantifies over the space of entities existentially. So (1) is equivalent to: there is a being Y such that Y = X and Y has property P, and a sub statement (there is a being Y such that Y=X) is exactly what the argument seeks to demonstrate, so it is circular.

edit: you don't even NEED 2 to conclude 3 from 1 with the quantifier.

If:
(1) Both accounts predict exactly the same bad things to happen.
(2) All the predictions that each make will happen.
(3) At least one of the accounts is true.

then

(4) the same bad things happen irrelevant of which is true.
IE: if (1),(2),(3). then (4)

I think (1) is unlikely, (2) is very unlikely, and (3) is false (since they might both be wrong). So if I was in your shoes, I'd look at what contrasts there are in the bad things they predict and how they predict them, then decide which is most likely to be true based on the contrasting elements.

Thank you very much.

Say you're measuring something's length. You make 10 measurements from a ruler whose smallest division is 1 mm. The measurements might be slightly different. But you expect them to be very similar. This similarity is quantified as the variance, $\small \sigma^2$, of the measurements. For the ruler, you expect them to be about 1mm variability, since you can determine the length within 1 mm from the ruler.

In contrast, consider a ruler whose smallest division was the size of the sun. This would be really bad at measuring stuff. Someone using it would just eyeball the length. But you can expect the variability to be much larger.


Now imagine that someone give someones measurements of something's length, and the fuck-off big ruler or the usual ruler have been used. The person you give the measurements to doesn't know which one is used. You give them 1000 measurements. They appear to be within about 1 mm of each other.

You ask the person to estimate the variability of the measurements, they say 'they're within 1 mm of each other', in in saying that they rule out the idea that the ruler the size of the sun was used. You can construct the above mathematical argument in terms of the proportion of measurements that are made with the fuck-off big ruler. If the proportion is very small, there's little evidence to inform the person looking at the data that they could've been made with the big ruler.

But by construction there's always a chance, a very small chance, of the big one.

Help?

Nutshell:
If you have two variables - the mean of their sum is the sum of their means.
If you have two variables with no correlation - the variability of their sum is the sum of their variabilities.
If you introduce a variable which has infinite mean and infinite variance to the sum, then the mean is infinite and the variability is infinite.
It doesn't matter what probability you assign to this infinity inducing mess, it'll always make an infinity inducing mess.
With sufficiently low probability of occurring, an infinity inducing mess can be made to resemble a precise experiment.
Thus, the data cannot distinguish an infinity inducing mess from a precisely measured law.
Thus, there's no evidence for the law.

It is, I can't figure out how to make it display. Can you help?

Yeah I think so. I think it's a way of highlighting that principles of parsimony play a fundamental role in determining scientific hypothesis - it's also a constructive argument, the procedure I gave above could be done to any experiment.*

There's a bit of a wrinkle though, there are systems which arguably should be modelled like this - a 'business as usual' mode and a 'oh god wtf is happening' mode, where the 'oh god wtf is happening' mode changes the global properties of the model. This has been brought up by Mandlebrot and Taleb with regards to financial time series. A bastardised version of the argument would be that financial models typically place too low a value on the relevance of the pathological part and thus they are surprised by pathology. For example, there were probability estimates of a financial crisis beginning in 2008 (apparently, trusting Taleb) being 10^-12. If you applied a 'raw' principle of parsimony to this kind of equation (a GARCH model in financial time series analysis), you end up with the 'best' model being one which contains no possibility of radical surprise.

I think the consequence of this is that as well as there being principles of parsimony. there should be an analysis of the consequences of assuming too simple a model.

*edit: with real valued quantities.

I'd been thinking along similar lines. To learn something isn't to learn something in a vacuum, it's to learn within a context. So in terms of the pictures, nothing new would be learned (if that thought experiment is valid); but in terms of knowledge about the duck-rules and the rabbit-rules, we could.

I haven't got a scooby.

Another contrast I thought of is that in Bayesian methods, the data are considered as fixed quantities in the likelihood and hypothesis tests are done using the product of the likelihood (actually a conditional distribution of the model parameters given the data here) and the prior - the posterior. There isn't an asymptotic distribution of the test statistic, there's simply some ratio involving priors and likelihoods.

If I have a null hypothesis that a population parameter mu has value q and calculate that, conditional on the null hypothesis being true, the probability of statistic S from a randomly chosen sample of n elements having a value in set A is p, are you saying that making that statement is inconsistent with a Bayesian view?

You can look at the probability of a random variable belonging to a set, you can't look at the probability of a fixed quantity belonging to a set (it's 0 or 1 depending on the interval and value). Also, the p-value assumes that the null hypothesis is true, it isn't a probability estimate of the null hypothesis. The important differences are what is considered random and how it's dealt with.

It's also possible to define p-value like objects for Bayesian analysis, but they are based on random parameters - so the likelihood actually is a probability distribution every time. This observation can also give an inconsistency between Bayesians and frequentists, and is termed a likelihood principle. Lots of hypothesis tests don't satisfy this principle.

It's a line at y=1 between x=0 and x=1, exactly. You can consider the infinite binary sequences (with trailing zeros) as numbers in [0,1]. Also, for the uniform distribution, each number is 'equally likely' (if you like measures the Radon-Nikodym density is 1 at all points in [0,1]). The fact that each number is equally likely in this sense can be shown to imply that if you 'throw a dart' into the interval, sampling a single number within it, its binary expansions' digits are fair coin flips with Heads=0 and Tails=1.

The procedure for going from a null hypothesis and an alternative hypothesis to a p-value with its usual interpretation is what being a strict Bayesian precludes. There are a couple of direct contradictions, firstly that the population parameter in the null is fixed - in Bayes it's random. Also the interpretation of a p value typically (as is the case for t,Z,F and Chi-square tests) relies on the frequentist interpretation of probability - long run frequency.

I feel the same about algorithmic information theory, had to Wiki to make sure the vague recollections I had of incompressibility and complexity weren't bull-crap.

Though I can think of an example tying my point and your point together. Imagine we're drawing a data vector $X$ from a standard uniform distribution. This is one that starts at 0 and ends at 1. It's a standard exercise in probability textbooks to show that this is equivalent to an infinite sequence of independent binary digits:

$.X_1 X_2 X_3 ...$

with $P(X_i = 1) = P(X_i = 0) = 0.5$ for all $i$. So you can get an algorithmically random sequence through a translation of the $X$ (minus its element-wise floor).

Edit: there is a helpful idea of a random variable, I've posted it here before... Will try and find it. Found it!

Let $(O,S,P)$ be a probability space where $O$ is a set of outcomes and $S$ a sigma algebra on the set of outcomes, then a random variable $X$ is defined as a measureable function on $(O,S,P)$ to some set of values $R$. A measurable function is a function $X$ such that the pre-image of every measureable set is measureable (element of the sigma algebra in their respective spaces).

I had in mind processes which are modelled using random variables rather than the conception used in algorithmic information theory. Most real numbers aren't computable and most real numbers' complexity is the same as having random bits for their decimal expansion, so in a sense the digits are patternless. This idea doesn't contradict random processes having patterns, just says that they don't necessarily have them.

The examples I had in mind were waiting times in queues and germination times in wild barley. Good to think of as random, but still contain patterns.

The typical method used for hypothesis testing using p-values is invalid if you are strictly Bayesian.

In which case predictions are made routinely.

Randomness isn't an absence of any pattern. Most things which can be considered as random have patterns. The basic example is a fair coin, flipping it gives you 50% chance of heads and 50% chance of tails. So it's a random outcome, but the generating process for the random outcomes has properties that can be analysed and accounted for - in principle anyway.

@sime

Change the thought experiment so that a person who's never seen it before and isn't compelled by the assumptions to see only a duck or only a rabbit. Assume they see the duck first, but then they see the rabbit. They will have learned something, namely that the picture can be seen as a duck or a rabbit, but there is no fact about the image which will allow them to distinguish duck from rabbit. If they've learned something, and it's not a fact about the image, what is it?

I'm not sure what you mean, regardless have some words.

Assuming Sally and Steve both see an identical image, it does not follow that they would interpret the image in the same way, unless they additionally share identical rules of judgement for identifying ducks and rabbits. And even then, they might still differ in their judgements if they each possessed different perceptual objectives, each seeing only what they wanted to see.

Let's refine it a little bit. Steve and Sally are asked to determine whether the image is of a duck or a rabbit. Assume Steve can only see the rabbit, whereas Sally can only see the duck. The domain of propositions that they're being asked to find evidence in is something like:

A ={those properties of the image which make it look like a duck}
B={those properties of the image which make it look like a rabbit}

however, each duck property of the image resembles a rabbit property, seen from a different view.
I believe the rules of judgement that they follow correspond to whether they see the duck or the rabbit. So Steve is following rabbit-highlighting-rules and and Sally is following duck-highlighting rules.

So A=B, since each duck-highlighting-rule is also a rabbit-highlighting-rule. But when the image is seen as a duck, it is not seen as a rabbit. There are only two relevant interpretations of the image for the question: 'the image is of a rabbit' and 'the image is of a duck', supported by A and B respectively. But since A and B are the same, we cannot learn anything about whether the image is a rabbit or a duck.

Since you offered a summary post I'll give one too. Saying 'it's BS' wouldn't've let me wrestle with some relevant ideas I recently encountered (causal statistics), also a reasoned argument demands a reasoned response. :)

My view is that the testimonials as you have presented them are not sufficiently strong to conclude that NDE experiences generate veridical perceptions; in the sense that it cannot be said that the person perceived X because of an NDE. This is because the testimonials as you have presented them don't meet the evidential standards I outlined. IE they are possibly confounded, possibly riddled with priming effects and cannot control for unobserved factors. The probabilistic argument for their non-veridicality I outlined I believe is supported by the variability of the reports and their inability to meet the standards of experiences we already know to be veridical.

It is true that to try to infer causal properties from testimonial evidence, a form of observational study, places a lot of constraints on the content of the observations. This isn't a flaw in my argument, rather it speaks to the inappropriateness of testimonial evidence to establish that NDE perceptions are veridical. Further you did not address the the problems of inferring veridicality given sampling effects, or the thought experiment involving the mansion.

It was good fun debating with you though.

Whenever someone looks at an ambiguous figure, like the duck-rabbit, their perceptions are in such a state of undecidability:

1) Person sees a duck.
2) Person sees a rabbit.
(Google duck rabbit for images)

Imagine a person - Steve, could see only the duck. Imagine a person, Sally, could see only the rabbit. Steve could gain no more information about the hidden rabbit status of the duck, nor could Sally gain information about the hidden duck status of the rabbit. If you compose and conjoin what they know, there is no more attainable evidence. This is because Steve and Sally together have all information about the duck status and the rabbit status of the duck rabbit; evidence is consistent with the duck and the rabbit status, and so no further knowledge of the duck rabbit is attainable. However, if Steve and Sally were invited to look very closely at the duck-rabbit picture they were shown, they may be able to discover certain things they did not know before about how it was drawn. But this is impossible, since no further knowledge of the duck rabbit is attainable!

Where does this go wrong? I'm not completely sure.

@SophistiCat

Well, isn't the entire thrust of the Bayesian (aka epistemic) interpretation to psychologize probability?

I think so. But I don't think this accounts for whether Bayesian approaches to AI and the mind are correct or not. In my view AI questions about Bayesian methods are 'does this statistical model learn in the same way humans do?' or 'is this statistical model something like what a conscious mind would do?', but epistemic questions are 'does this interpretation of probability make sense of how probability is used?' and 'does (list of properties of Bayesian inference) give a good normative account of how we ought to reason?'. I can certainly see why AI questions would influence epistemic questions and vice versa, but there are definitely significant problems that affect one and not the other. For example, arguments about the likelihood principle (evidential claims must depend on a likelihood in some manner) are largely epistemic, but arguments about the framing problem (problems of parametrisation in learning algorithms) largely concern AI.

@Posty McPostface

I think the discrepancy in interpretations lends itself due to the possibility of hidden variables. Is this something that is considered in probability theory because it goes to the heart of the issue in my opinion?

I don't know what you mean, can you throw some more words at me please?

I look at testimony in terms of the whole, and I think it's important to take into account how people with no medical knowledge might describe things they see. If someone without medical knowledge was in the room, it's quite probable that they would describe things imprecisely. Just as people would do in their everyday lives, so I don't think this would necessarily mitigate the testimony. Although it might if we didn't have as many testimonials as we have (literally millions). What I find interesting in Pam's video, is that the doctors were confounded by her description of the experience. The fact that the doctors were baffled doesn't necessarily mean that she had an OBE, but it does suggest that her description was very unusual given her state.

I don't think we can check if she literally saw the implement, descriptions of experience are all we have to go on. There are plenty of other things she could've noticed and provided a precise description of - I gave a couple of examples in my post. I'm not saying for certain that we can rule out that Pam somehow observed the events in the room via an OBE, I'm saying that there are enough mitigating or problematic points in the video to render it not evidence of the veridicality of NDEs. I attempted to portray what actually would be strong evidence that an NDE was veridical in my post.

What I mean by taking her experience as a whole, as opposed to picking out one or two things that might be explained in other ways, is the following: Does her description of the events in question match what others have described in terms of her metaphysical experiences. So, not only are we considering what she describes while in surgery, but does what she said match what others claim to have seen in their experiences. We are also considering things that some might find unimportant, but are important in terms of the truthfulness of their statements. For example, remarks that some might skip over like feeling very light while outside the body, which would be in keeping with not having gravitational effects. Also, how they describe their communication experience, viz., mind-to-mind. These are small things that you might find unimportant, but can tell you something about the veracity of their testimony in terms of how it lines up with other testimonials.

The fact that there are consistent groups of themes within NDEs is interesting, but is not evidence that the NDEs are observing something 'out there' which is real. You can see the same thing with the many religions which include 'tree of life' imagery; they are incompatible accounts of purportedly real phenomena with common mythopoetic structures. I think it's also plausible that the consistencies observed in NDE and other mystical experiences can be generated by there being an encultured, primordial mythopoetic structure with broad parameters - life/death imagery, revelation, calm/home feelings, out of body experiences, otherworldly visitations...

Again, as I've said many times I find the consistency of these testimonials remarkable. Whenever you're looking at the testimony of a large number of people, even in normal testimonials we find inconsistencies. This is why testimonial evidence is generally weak, but as in an inductive argument, the conclusion is either strong or weak based on the kind of evidence. This is why my early evaluation of what makes testimonial evidence strong is important.

The consistency is interesting, but as above it is not sufficient to get to the veridicality of NDEs. Further, there is a causal element in the argument you've made. Namely, that when descriptions of NDEs contain some independent information the NDE effected person could not have had access to without the NDE, that suggests the NDE is veridical. My point is that enough and the right kind of NDE descriptions with these properties would be evidence that NDE experiences are veridical.

Enough in the sense that people exposed to the NDE are likely to produce the right kind of testimony. The right kind being precise, non-confounded, non-primed, sequentially accurate descriptions being necessary to establish NDEs veridicality in general. I don't think it's appropriate to attempt to establish the veridicality of X using X.

This plays into my previous probabilistic argument, and I'll address your response to it here.

Imagine we have a huge mansion and 50 people are given a photograph of different rooms in the mansion. Each person is then tasked to describe the contents of the photograph. I think this is roughly equivalent to your example of the 100km^2 area. If a person describes elements of the photograph accurately and doesn't describe things which aren't in the photograph, we'll call their description accurate. Now imagine we hide the mansion and the photographs, and give the written descriptions to a third party.

The third party's job is to provide an account of the observations of the people. We'll call this third party Steve: I think it's likely that Steve would describe a large house with decor of a certain colour, descriptions of and numerosity of rooms - a very good description of the real mansion's rooms and general decor. Can Steve conclude from the descriptions alone that the 50 participants were shown photographs of a real mansion or paintings of a fictitious one? I don't think so.

Can Steve derive any evidence from the descriptions that they're descriptions of photographs or paintings? Can Steve claim this on the basis of the consistency of the reports?

Edit: let's further imagine that half of the rooms had a blue decor and half had a red decor. Steve may conclude based on the inconsistency of the decor that the house was unreal. Or alternatively, he could say that the decor changes and derive the 'blue' theme and the 'red' theme from the descriptions - grouping them based on which theme was present.

In this scenario, the veridicality of the experiences of the people was assumed, but Steve still cannot derive any evidence that the descriptions refer to a painting of a mansion with red and blue decor, photographs of a mansion with red and blue decor, or that they come from separate things all together.

________________________

As another experiment, imagine that 100 people are given the same photograph and are tasked to describe it as completely as they can. It is unlikely that pairs of people produce descriptions containing exactly the same propositions, but there will be a high degree of agreement and no contradictions. Steve, again serving as a third party, could conclude that their perceptions were likely to be veridical based on the consistency. Imagine instead that 50 of the people were shown one photograph and 50 were shown another which had the same fixtures but different decor. Steve could not conclude that their perceptions were veridical due to the split. Alternatively, Steve could decide that 50 of the people described a house with blue decor, and 50 of the people described a house with red decor. I think your argument from testimony is essentially this: there are blue and red thematic subgroups, therefore Steve can conclude the descriptions are of photographs.

A summary of the argument: if the sampling behaviour of supposedly veridical NDE experiences matches the sampling behaviour without reference to NDE, there's no reason to conclude there is an NDE-veridicality mechanism from the data.

Edit: I wrote the previous post with reference to videos as independent observers, but other sources of independent verification would suffice.

In summary: I think the veridicality of NDEs turns on their aggregate properties, rather than the accounts of NDEs made by specific individuals. I don't have the data, so I don't want to make any quantitative conclusions - but I can give you the thrust of the argument I'd examine with the data.

I'm reminded of Gettier - what are the chances that someone's NDE contains information they 'could not have known without an out of body experience' by accident? IE, what are the chances that an NDE can be judged as veridical purely through sample size effects?

So I think it's appropriate to look at phenomena which would allow an NDE to be judged as veridical. Namely, if someone provides a description of the NDE and if this description matches a video record. What properties of a description make this likely, without reference to the causal mechanism of being out of body? Well, there are a few things that can increase the probability that a description matches some things in the videos.

(1) The NDE experience contains vague rather than precise descriptions of events. This is the case for the woman in the video's NDE-perception of a 'tooth-brush like device' in the room, which could fit any tool with a mechanical base and rotating upper part. This could describe lots of surgical tools. I'm not saying this is necessarily the case for every report as I'm sure there are very precisely described events in some NDE accounts that match the videos quite faithfully.

I think it would be fair to remove these cases from the veridical NDEs, since a trained liar could produce these statements. Such as 'there were at least 15 people in the room', 'there was a scalpel used by a woman to make an incision around my head". This may account for the woman in the video's 'there are 20 doctors' in the room statement, but we cannot obtain information one way or the other without access to her first description of the event [which was articulated with the doctor, so there is confounding].

(2) There is some prior knowledge about the procedure. This is usually the case, since participants will have a rough description of the procedure's goals, risks and benefits before deciding whether to have the procedure. I don't think testimony could distinguish a mental event stimulated by memory of the procedure's description causing a 'recollection' or 'NDE observation' during the procedure. This could account for the woman's perception of someone saying 'the veins are too small' [the video doesn't say that a specific woman actually said this, only that they had difficulty cutting some of the veins due to the size]. 'The veins are too small' might also be a likely phenomenon when doing vascular surgery on the small veins in the brain!

This should be used to rule out descriptions of events that can be accounted for through memory rather than through NDE since testimony cannot allow us to observe the causal mechanism (I think this point was made obtusely by Jeramiah earlier in the thread). The causal mechanism being 'this person is having a veridical NDE'. This wouldn't rule out all descriptions of things in the procedure - for example a doctor who entered's hairstyle could be a part of the perceptions in the NDE, and the hair-style is unlikely to be part of the priming. I know there are cases of things like this.

(3) A veridical NDE should be treated like a test for confusion/awareness. Someone is asked to recall various facts about themselves and their immediate environment. If someone, say, could tell there was a tree in the window but forgot their name they would be judged as experiencing confusion. Multiple correct answers to the right questions (a temporal order) are required for an NDE to be judged as veridical.

Applying these filters to your data should give you candidates for veridical NDE assessment. After applying these filters, I believe a very small number of cases will remain.

So, how about a positive criterion? For an NDE to be considered veridical, it should contain multiple correct observations of documented events in the correct order. I believe this is a fair criterion, since veridical perception usually occurs along with a temporal order of events. There are numerous things to control for:

(A) Whether an observation is correct or incorrect should be judged from a written version of the NDE experience made independently from discourse with another observer of the event. This prevents priming effects.

(B) The subject should be in deep anaesthesia to prevent non-NDE perceptual events.

When these have been removed, there should be a small proportion of people who have experienced things which can be used to make a case for veridicality. But - a small proportion is exactly what is expected purely through chance. I'd make a ballpark estimate (asspull) that 1 in 5000 NDE descriptions would have at least two correct precise ordered descriptive elements in them that cannot be accounted for by any of the above filters.

And we would be left with the conclusion that the vast majority of NDE events are not connected in some way post-filtering to the real world. By the previous argument I made, I believe this would show that NDE perceptions are not veridical.

It's true that most people who are near death, or have had some experience that brings them near death don't experience any of the listed testimonial reports. However, I don't see how this negates the millions of consistent reports of those that have. All of us have experiences that most people haven't had, that doesn't mean that my experiences are any less real, or that my reports of those experiences are doubtful in the same sense you're doubting NDEs. You seem to be concluding that because more people haven't had the experience that that somehow makes the millions of experiences of those who have had the experience doubtful. It does raise questions, I agree, but I don't see that it means that what people are experiencing in these NDEs is any less veridical.

I agree that a raw proportion of people not having NDEs isn't good evidence that NDEs content is non-veridical. I'm glad of your example because it picked up on several ambiguities in my response.

For example, less take our experiences in everyday life, if I'm at a party with 100 other people and 15 of these people claim that X happened, and that those closest to the experience or who claim to have had the same experience agree that X happened, then your claim is that X probably didn't happen. Your claim is based not on counter-evidence, but on the fact that the 85 other people at the party didn't report those same experiences, or that they have no recollection of X happening. But all they are saying is that they didn't have the experience. If they had the experience and reported completely different reports, then I would say that you have an argument, but that's not what their saying. Their saying, I was at the party but didn't experience what those 15 people experienced.

I'm not referring to the raw proportion of people who have had NDEs vs those who have not had NDEs, I'm referring specifically to the people who have had NDEs and then self report content given the Gallup poll results. If you look at the numbers in my response, you'll see that the % reports from the subjects who have had NDEs (otherworldly experiences etc) sum to more than the % of people in the general population that have had NDEs in the first place. This is 28% vs 15% from the same poll.

The relevant situation to consider is the conditional event: what are the people reporting given that they self identify as having had an NDE. In your party example, 15 people claiming that X happened is ok - since if we want to learn about the event X, we are considering self report of people who have already seen X, not the 85 people in the party who didn't.

My argument turns on this idea of conditional reports. I can state it more precisely now, with reference to the previous 'door seeing' example. Say there are 200 people in my door seeing study. I only expose 100 of them to the door. I then record self reports of whether they see the door or not. I think if this experiment was conducted, close to 100% of the people who saw the door would say that there was a door then, this makes the condition probability of seeing the door given being exposed to the door close to 100%. I am claiming this is good evidence because of high consistency in the reports of people who have been exposed to the door. It is irrelevant that 50% of the people in the study didn't see the door.

So let's apply this to NDEs, do we observe very consistent regularity in the reported content of people who have experienced NDEs? No, there are many different stories. This means that exposure to an NDE doesn't (probabilistically) entail having a particular experience. Whereas exposure to the door does (probabilistically) entail seeing the door.

This is where my analogy of 'the dress' comes in. Exposure to the dress generates disjoint perceptions - the dress is really blue vs the dress is really white. What we can say about the evidence here is that of the people seeing the dress, people will see a blue OR white dress - a disjunctive event -
with high probability. Similarly, of those who have self reported mystical content of NDEs, we can say there is a high probability of a disjunctive event - namely at least one of the things on your list.

As an aside, that the % of people who responded who had NDEs with self reported content categorised by the poll was only 28%, this is evidence that being exposed to a near death scenario will not necessarily generate a near death experience. This would be similar to being exposed to 'the dress' and seeing no colour, but that is absurd.

So what we can conclude from the testimony is that people who identify as having NDEs and who self report mystical experiences will have at least one of 15 mystical themes in it. Evidence for the disjunctive event isn't evidence for the veridicality of experiencing any disjunct, rather like in the dress example. This leads me to conclude that:

1) There are general themes within NDEs for those who self report content.
2) There is no evidence that exposure to near death generates a particular NDE.
3) From 2), there is no evidence that the perception of a given NDE type is veridical.

You can also conclude from the similarity of the experiences of NDEs and certain psychoactive drugs that there are unobserved factors mediating the relationship between the mystical experience being near death. We are in the same situation with regards to the dress, there is an unobserved perceptual/mental/neuro-chemical event that gives rise to the disjunction blue/white rather than any of the specific disjuncts. IE it is more likely that an unobserved factor drives the appearance of the disjunction than the truth of any disjunct. In stark contrast to the door situation.

There's usually a Bayesian or frequentist method to do anything. The major reasons people choose to use Bayes or frequentist afaik is pragmatic, more to do with the availability of software and the speed of algorithms than anything philosophically fundamental.

I can say I wouldn't use frequentist estimates for problems with spatial dependence though - they take a long time algorithmically -, whereas there are very efficient Bayesian methods for it.

@Sam26

I'm interpreting the OP an the other posts you have made as invitations to analyse whether the testimony of people who have NDEs is sufficient to conclude that they are experiencing something real or actual. IE: that the self reported content of the NDE is real in the same sense that 'I see the door to my house and it's really there' is real. I'll call this property 'veridicality' of the NDE.

So, how can the veridicality of the NDE be established? One necessary condition for their veridicality would be to see if the experiences within the NDE obtain in a majority of experiences. If you take people to the door of your house, most people would see the door - if most people didn't see the door, then I think it would be grounds to doubt that there was a door. If I saw the door, and most people didn't, then I would have grounds to doubt that my experiences containing the door were veridical.

IE, if most people do not experience a specific NDE event (like love or disembodiment, in the list of 15) given that they have had an NDE and that they have almost died, then there's grounds for doubting the veridicality of the experience. 15% of people have had an NDE in those conditions in America according to a Gallup poll, of these 9% reported the classic out of body experience, 11% said they had entered another realm and 8% had encountered otherworldly beings. Assuming these categories are independent and exhaustive, you can obtain that 72% of people who had an NDE experienced nothing contained in the list or nothing at all. I'd be surprised if they were exhaustive, but it still looks like the majority of people who had an NDE couldn't categorise it according to listed tropes in the poll OR alternatively they nearly died and experienced nothing.

However, this doesn't mean 'nothing is going on', I'm reminded of 'the dress' perceptual illusion that was commonplace on the internet a few years ago. There were consensus views that it was white and that it was dark blue. I think it would be quite silly to argue whether it was really white or really dark blue, the far more likely occurrence is that there is some mental or perceptual event that induces the different colours in different people.

I believe this is analogous - there is no majority consensus on 'what happens during an NDE', and no experiential consensus (a long list of alternatives) on what the contents of the NDE are. I therefore believe it's likely that NDEs are the result of some currently not understood mental or perceptual event that need not correspond with anything 'out there'.

I am sorry, my statistics and hypothesis testing background is too basic and rusty to fully appreciate your comments. I didn't mean to advocate likelyhoodism though - I only mentioned it as an example of Bayesians not being satisfied with prior probabilities and seeking ways to avoid them while still preserving what they think are Bayesianism's advantages. — Sophisticat

The thrust of the comments is that contemporary statistics uses plenty of methods and mathematical objects that are not consistent with contemporary philosophy of statistics' accounts of evidential content and the methods and objects used to analyse it. One response would be 'so much the worse for statistics', but I think it's so much the worse for philosophy of statistics since these methods observably work.

I think whether Bayesian models of the mind or of learning in general are accurate in principle is mostly orthogonal to interpretations of probability. Would be worth another thread though.

Human thought is consistent? Human mathematical thought isn't consistent. Within a specific formal system, sure, but there are intuitionist, constructionist and para-consistent versions of mathematical theories. For example in para-consistent mathematics, the real numbers turn out to be countable and not countable at the same time, whereas in standard mathematics they are simply not countable. In constructionist mathematics, the existence of non-measurable sets cannot be established whereas they are objects in classical mathematics. Para-consistent and intuitionist/constructivist mathematics disagree on proof by contradiction (fine in paraconsistent, not fine in intuitionist).

@Sophisticat

There's a big probability I'm being unfair based on unfamiliarity with the literature, just whenever I've read philosophy of statistics it is usually concerned with things very separate from the current practice of statistics - especially statistical modelling. Still, if you notice any ways I'm being unfair I'd like to hear them.

I forgot to reply to this:

But wasn't the very idea of "subjective" probability to take our psychological intuitions as the primary source of probability valuations? There seem to be conflicting agendas here. But on the other hand, if we give up the simplistic rationalism of Bayes, won't we then diverge from scientific (not to mention mathematical) probability, carving out a special theory that's only relevant to psychology? — Sophisticat

There's been an attempt to assess the consequences of giving up the rational utility maximisers/probabilistic rationality since the 70's, following behavioural economics and the experimental psychology behind it. A landmark paper in this regard is 'Prospect Theory' by Kahnneman and Tversky.

Our intuitions being a primary source of probability evaluations is quite vexed (as you put it), since our intuitions demonstrably contain no untrained competence in evaluating phenomena subject to regression to the mean and sample size effects, also not Bayes theorem. This isn't too surprising, as any field of science doubtlessly has many phenomena which will not have untrained competence regarding them.

In my view, if there is a conflict of the intuition with something that is already unambiguously formalised, go with the formalisation.