In a recent question I asked if it was epistemologically sound to consider alternative theories of consciousness to explain the visual phenomena that people blind from birth experience during a near-death experience, insofar as the most widely accepted theories of vision and neuroscience fail to explain this observed phenomenon.

Some people may reject resorting to less established or “fringe” theories of consciousness, but without argumentation that engages with those theories in order to refute them; instead, they dismiss those theories by default, because they are not widely accepted.

If someone dismisses these data points by conjecturing, "They are probably just having hallucinations of some kind", but without offering further details that clarify exactly how they are hallucinating - e.g., a theory of hallucinations applied to near-death experiences of blind-from-birth individuals that clearly explains how they are hallucinating in precisely the way they report - why should we prefer an explanation that might adhere closer to mainstream knowledge but does not actually offer an explanation for the observed data?

Much more generally: what makes an explanation an explanation? Given any suggested “explanation” for something, what criteria should it fulfill, categorically, to determine if it is a sufficiently good “explanation”, or not?

  • "clarify exactly how they are hallucinating" - what do you mean "how"? Like exactly which neurons are firing? We can't do that for explanations of much of any mental process. Or to exactly explain how they could've known what they hallucinated? But it may be somewhere between difficult and impossible to know that, given that we don't have access to every part of someone's memory. All that's required for a hallucination to be a valid hypothesis is a reasonable hypothesis for how they could've known anything that's presented as independently verifiable, or why the verification is flawed.
    – NotThatGuy
    Commented Dec 18, 2023 at 13:00
  • 1
    Note that the authors of that article themselves say "in order to demonstrate that the perceptions described by our blind experiencers are something other than mere fantasies or even complex hallucinations, it will be necessary to provide some kind of confirming evidence for them". They themselves seem to acknowledge that hallucination would be the default position, but the argument and evidence they present against it is just very weak.
    – NotThatGuy
    Commented Dec 18, 2023 at 13:02
  • @NotThatGuy "All that's required for a hallucination to be a valid hypothesis is a reasonable hypothesis for how they could've known anything that's presented as independently verifiable, or why the verification is flawed." - But then this pushes the problem to now having to define what counts as a "reasonable" hypothesis. You are essentially saying "X is valid if X is reasonable". Okay, how do we know if X is reasonable?
    – Mark
    Commented Dec 18, 2023 at 13:12
  • 1
    Probably outdated, but I remember a hypothesis out there according to which the body releases a significant amount of endogenous DMT (or something similar) when faced with apparent death, and this would account for the extreme hallucinations experienced during NDEs. I think it was found later that the body doesn't have the capacity to synthesize that much DMT proper (if almost any at all), but I wouldn't be surprised if it can generate something else to similar effect. Commented Dec 18, 2023 at 13:33
  • 3
    This question reminds me about a famous experiment by a high school physics teacher. The teacher placed a ceramic plate next to a Bunsen burner, then had the students enter the classroom and touch the plate to see where it was hotter. Students observed that the half of the plate further away from the burner was hotter than the half closer to the burner. Teacher asked students for explanation. Students all came up with tons of explanations using big science words like "convection". Then teacher explained he just rotated the plate 180° before letting them in the room.
    – Stef
    Commented Dec 18, 2023 at 18:57

8 Answers 8


Is "explaining away" something without offering details ever justified? How detailed should an explanation be to be considered valid?

The term 'explanation' is a bit like the term 'truth'. It means very different things to very different people. Consider that the SEP has at least 4 articles that highlights the topic:

So, philosophers, mathematicians, scientists, and ordinary people explaining their action have different approaches to what constitutes 'explanation'. Thus, unless providing a very narrow context, it's tough to make broad claims about justification of explanation just as one struggles with notions of evidentialism.

That being said, we can contextualize using your arguments regarding NDE from the lens of rhetoric. If you offer a point about NDE, and your opponent dismisses it meaningless with a shallow explanation, and then refuses to move forward without acknowledging the applicability of your claim, or refuses to address your concerns, then, while not necessary a deficiency in judgment, it does represent a deficiency in commitment to the process of argumentation. Obviously, people who have had NDEs and make claims about supernatural experiences represent a legitimate philosophical point of contention. If your opponent is a closed minded "skeptic" (not in the broader philosophical sense, but rather as a habitual "naysayer"), then what is being communicated is a boundary of consideration.

If you want to explore evidence, and your opponent refuses to, then that kills the friendly argumentation process. When this happens in a court of law, a judge simply compels the argument in the appropriate direction. In a debate contest, it should count against the reluctant party. But in normal, friendly argumentation, it hampers or kills discourse. Thus, you simply have learned something about your opponent's position and intent, and you must go from there.


Clearly, it depends upon the circumstances. A vague general hypothesis would not be accepted in a court of law or at a meeting in CERN to allocate a $1billion research budget; however, it might well be accepted during a drunken discussion at your local pub or at a cabinet meeting of HM Government.

If you ask me to take NDE's seriously, I would have a hard time squaring the idea with all my other ingrained beliefs, so I would probably dismiss your suggestion without feeling I needed to justify my attitude. That does not mean that I am right. Humans give the impression of being creatures of habit, so if we have come to accept an established 'truth' we tend to require a good deal of intellectual impetus to move from our customary position. We used to think the Sun circled the Earth, and a good deal of resistance was encountered when that view was challenged even though with hindsight it was nonsense.

That said, we cannot afford to assess and evaluate in a rigorous way every new idea that we encounter, or we would have no time for anything else. Some initial degree of surface plausibility is needed to engage our interest, and whether we follow it up depends on what is at stake and how much effort is required.

To return to NDEs, I would take the view that the plausibility test isn't satisfied. I don't have the time, knowledge, motivation or patience to evaluate the question. There are plenty of people who are deeply interested in NDEs, so I assume firstly that I should leave them to it because they are likely to make more and better progress than I would, and secondly that if they had achieved the plausibility test then we would have known about it by now.


Is "explaining away“ something without offering details ever justified? How detailed should an explanation be before it is considered valid?

The question assumes that one side has hard facts, or data, while the other has only a vague general hypothesis. Let’s look at that assumption first.

Regarding the facts, what is the procedure that has produced the hardness? In short, why is the data any good? Inability to replicate experimental results has become a terrible problem, regardless of the area of study. Refuting the conclusion of a poorly-designed experiment is easy, and does not require any particular level of detail. For example, in the humanities, is the study‘s nominally random sample of people actually self-selected? If so, the results are worth a lot less. Dismissing the value of such a study would not require the opponent to offer a set of competing details.

But let us assume a factual conclusion which is supported by procedures which are generally agreed to produce a fair and objective result. Now the burden has truly shifted. The opponent must offer details which show how the statement is suspect. For example, the statement contradicts other statements believed, with good reason, to be true; or the individual procedures within the larger set of procedures that support the challenged conclusion are actually weak or useless.

The challenge to the underlying facts is different from the challenge to the overall conclusion. The basic formula is F => C, or Facts thus Conclusion. To refute the conclusion is to disprove the entire line of reasoning; it is to deny the consequent.

By contrast, to refute the facts is to deny the antecedent; from such an argument, nothing follows. An antecedent might be false and the conclusion still could be true. The opponent who challenges the facts has the heavier burden of showing that, given the true facts, the conclusion either can not follow or is actually impossible.

Now, let us look at your specific question about near-death experiences.

If someone dismisses these data points [about NDE’s] by conjecturing that "they are probably just having hallucinations of some kind", but without offering further details that clarify exactly how they are hallucinating …, would that "explanation" be valid?

In my opinion, this explanation would not be valid. This challenge addresses only the facts. The conclusion is, "Reports of NDE’s represent a real phenomenon." The challenge attacks only the antecedent, the collection of reports from those who have had an NDE experience. People can hallucinate all sorts of things. If there is no explanation of why someone sees the particular set of images which are said accompany an NDE, then such arguments leave untouched the notion that NDE’s are real events.


I don’t have enough rep to comment. Can you discuss this with me here?

The question is hard to provide a definite answer to in its current form, but we can flesh it out. I think it’s a much more general question about how one can know if a theory or reasoning process or explanation is valid or not. That is a question that used to fascinate me, and I’m sure it’s a detailed philosophical topic, but in some ways I currently got used to assuming that nobody knows. But I would like to take this as an opportunity to look at the question from new angles.


In formal argument, other than premises, assertions without argument are fallacies.

Outside of formal argument, assertions without argument (like this one right here!) are just the normal way of communicating facts and opinions.

If you feel like it, you can ask them whether their statement is just an intuition, or if they can articulate an argument for why you should believe it too.

"Do you have an explanation for X given Y?" can be a perfectly good question when you're really asking. On the other hand, if you really mean "if you don't have an explanation for X given Y, then Y is false", that is also a fallacy.


There is no general answer to your question. The degree of justification needed to legitimately dismiss possible evidence for a controversial POV depends on the degree of support provided for the controversial POV, and the degree of effort put into a rationale to dismiss it. There are, however, some other general principles that apply as well.

We humans are intrinsically Bayesian thinkers. We operate off a worldview, and approach all new information with that worldview as a presupposition. This is called a “prior” in Bayesian statistics. We humans also are notoriously bad at judging the validity of our prior judgments, such that we drastically overestimate the certainty by which we should hold them. This is why, when new innovative discoveries are made in a science, it often can take a generation, for the invalidly close-minded old guard to die off, before the useful new idea is adopted. The consequence of these two realities, that we think Bayesianly, and we also drastically overstate the justification for our priors, are that we humans in general are illegitimately close-minded against innovative views that challenge anything we have embraced as “true”. It is VERY VERY common that evidences are dismissed for illegitimate rationalized non-justified reasons.

There are several more ideas that are useful to consider. One is Lakatos’ Research Programmes. Imre Lakatos proposed the best model of how science should be done, and that is to treat an approach to a science problem as a “Research Programme” which consists of a mutually supporting collection of propositions along with an associated set of research methodologies. Lakatos’ approach can also be applied to philosophical worldviews. When a conflicting set of evidence is found to a Research Programme, the Programme is not immediately discarded, instead its advocates take this data as a challenge, and they endeavor to find some way to reconcile this data to the Programme. For example, the evidence that our stellar age data showed stars older than the Big Bang model said the entire universe was, was true for decades, yet was not treated as refuting the BB, but as an open problem. Eventually the BB was modified with a growing Cosmological Constant, and the two ages now are in sync. A serious philosopher or scientist SEEKS OUT data that challenges a preferred Programme, as the way to show the Programme is useful/progressive in solving problems in the field.

The other idea to consider is Apologism. Apologism is the branch of theology whose goal is to come up with cursory plausible dismissals of contrary data, so that the minds of the faithful can bask in the Faith in calm repose, without being disturbed by considering that there may be open data challenges that their Faith does not have an answer to. Apologism was named by theological organizations as a practice, but anyone who has debated people on almost any subject, will have encountered apologists for nearly any viewpoint that humans debate. Political apologism is particularly difficult to avoid, but one will find lots of other viewpoints suffer from this too. Creationism, Anti-Global Warming, anti-vax apologism, etc – apologism is everywhere. Many of these apologists form their own groups to hone their apologism skills, and share the best apologist rhetoric in newsletters and blog sites. As a general rule, apologists do not treat their POV as a research Programme, but as unquestionable. And the purpose of their rhetoric is to prevent serious consideration of challenging data, rather than proper investigation. When an apologist calls for evidence to be dismissed, such a call is almost universally unjustified.

Note that reductive materialism is also a POV, and it also has enough apologists to also form organizations, and create newsletters and blogs to spread their apologism, calling themselves Skeptic societies and sites.

With this set of ideas and observations providing a framework to consider your question, I think we can fairly easily see the dismissal of your evidence is a case of unjustified apologism.

However, stepping back to Lakatos’ Research Programmes, one set of challenging observations is NOT enough to overturn a useful Research Programme, such as Reductive Physicalism. To do that, and justify a different paradigm that is compatible with the “less popular consciousness models” requires that there be a growing list of seemingly intractable challenges for reductive materialism, and that there be another paradigm that provides an overall more useful model for solving this suite of problems, and does not face a new different and larger suite of problems of its own.

I believe this is currently the case for reductive physicalism, and even most physicalists agree, and have become NON-reductive physicalists. Non-reductive physicalism has major coherency problems, and it also faces many challenges. For some references on the suite of problems, I can recommend:

• Susan Blackmore’s “A Very Short Introduction to Consciousness” which spells out how physicalism cannot explain consciousness. Review:https://www.amazon.com/gp/customer-reviews/R1C1TJFIWBZ8ZQ?ref=pf_ov_at_pdctrvw_srp

• Jaegwon Kim’s Physicalism or Something Near Enough, which admits that the last half century of debate within philosophy of mind has shown that qualia are not physical. Review:https://www.amazon.com/gp/customer-reviews/R1LFTMUSP8VEWB?ref=pf_vv_at_pdctrvw_srp

• Daniel Stoljar’s Physicalism, which spells out how physicalism cannot deal with Hempel’s Dilemma, and the ontology of physicalism makes no sense in light of modern (quanta and relativity) physics. Review: https://www.amazon.com/gp/customer-reviews/R13R2OUNXMIN6H?ref=pf_vv_at_pdctrvw_srp

• The data compiled by parapsychology – see the Parapsych Association’s page summarizing what is demonstrated beyond reasonable doubt: https://parapsych.org/articles/36/55/what_is_the_stateoftheevidence.aspx

This is a large and diverse suite of challenges, which physicalism as a research programme has made no progress on for a century. When to shift to a different paradigm is a judgement call, but increasing numbers of philosophers and scientists have realized that it is past time to move on from physicalism.


Knowledge is essentially a set of judgements that are logically consistent. For example, if I tell you that a new bird species --without entering into details-- has been found, you will link the fact with your knowledge. That is, such bird will not have three feet or eat lions. You will think in a common bird within some standard deviation features.

Knowledge is based on logic and serves survival. When knowledge is logically flawed it will decrease the survival probabilities of the holder; for example, if you hold contradictions about relationships (for example, that you should be honest with friends, but that you can anyway steal their wallets), you will essentially destroy relationships. In the same way, you can destroy your house or your own body. From the opposite point of view, knowledge that is logically consistent is a competitive advantage in natural selection.

Now, what is an explanation?

An explanation is an attempt to make a logical connection between some new judgement with the whole knowledge. If I say that 1+1=3, you might doubt it, which means that you can't find a logical connection between such judgement and your whole knowledge. Well, there's an explanation for such judgement, which can allow its connection with your knowledge: a weighing balance that rounds to the next integer weighs a watermelon of 1.4kg as "1". But two watermelons of the same weight will round up to "3" (2.8kg). This is not a trick: this is a real fact. So, now, you have connected the fact that 1+1=3 can be true in certain circumstances.

Notice that in general, acquiring knowledge does not change previous knowledge, knowledge just "grows". But in certain cases, a new judgement can change a big subset of previous knowledge, that is known as a "paradigm shift".

They are probably just having hallucinations of some kind", but without offering further details that clarify exactly how they are hallucinating [...]

There, you are asking an explanation of a small judgement, while you have proposed a big change in knowledge (a paradigm shift) when "consider[ing] alternative theories of consciousness to explain the visual phenomena". The burden of proof in this case is in your hands, you should propose, write, submit and get you theory accepted, before disqualifying current theories due to lack of explanations about secondary facts.

consider alternative theories [...] to explain [...] insofar as the most widely accepted theories [...] fail to explain this [...] phenomenon.

In addition, such is appears to be a false dichotomy and a hasty generalization. A small flaw in accepted theories does not justify changing them by an unaccepted theory with a small strong point.


Yoda said, “Do, or do not. There is no try.” I find this loosely analogous to the idea that there are only explanations. There are not “sufficiently good” explanations, and there is no difference between “explaining” and “explaining away”. So, “Explain, or do not explain. There is nothing really that important in-between.”

I will suggest that an explanation is a model. This is an extremely commonplace idea, but maybe it can help you if you haven’t considered it before.

When we perceive phenomena, and we ask “why”, we are simply asking for a model - a representation, a copy, of the structure, the information - of reality, of the world. As one dissolves one’s understanding of the world into more and more abstract categories, one may shift from see the world as physical - a collection of atoms and particles - to virtual and informations - values in quantum fields, specifying parameters like mass, force, position, etc. Ultimately, it appears that all causation, from this point of video, is explainable by a sequence of physical events, with laws specifying exactly why and how one thing leads to the next, such as the principle of least action, and the previously held belief that there was no action at a distance, and current knowledge about how subatomic particles interact, and how the exchange of I think bosons or something is how fundamental forces work.

(Maybe I am wrong, I’d like to learn about shortcomings to this perspective.)

So, there is no need to distinguish between a socially unpopular explanation that you think has merit, vs. a mainstream one; or, an explanation that tries to dismiss some alternative hypothesis, vs. proposing one. The only relevant matter is providing the correct model of the phenomenon. And I guess there is more to be said on ways if verifying that it is the correct model, but obviously, the idea of scientific testability and the revising of hypotheses comes to mind.

Scientific models.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .