A number of years ago I posted on Facebook a response to something that Sam Harris had said. He said:
"we know [there are very good reasons for doubting an afterlife] from 150 years of neurology where you damage areas of the brain and faculties are lost and they're clearly lost, it's not that everyone with brain damage has their soul perfectly intact and they just can't get the words out, everything about your mind can be damaged by damaging the brain".
My response back then in italics:
--------------------------------------------------------------------------------------------------------------
Sam mentions, for example, that brain damage can result in not being able to recognise faces. So it's *not* the case that we really *do* recognise faces, but find we're cannot communicate to other people our knowledge. Our minds are actually *adversely* affected.
The problem with his argument, and the mistake that skeptics always make, is that he *conflates* the soul with the mind. In fact, even with brain damage, our souls can still be perfectly intact. Suppose someone has perfect unaided vision and puts on a random pair of eyeglasses. Wearing them she cannot read the registration plate of a car 25 metres away. It's not as if she *can* really see the plate but is unable to communicate this information. She really *can't* see the registration plate.
But it nevertheless remains the case that her unaided vision is still perfectly intact. Her bespectacled vision is poor, but not her unaided vision. And she merely needs to take the eyeglasses off to restore her vision. Likewise, with brain damage, our *minds* might be affected, but not our souls. Our souls merely need to detach from our bodies to restore full cognitive abilities (or indeed enhanced cognitive abilities).
-----------------------------------------------------------------------------------------------------------
I pasted in to Claude what Sam Harris said together with my response. It seems to me that it's not possible to dispute my point, so my expectation was that it would simply agree with me. It took absolutely ages to respond, perhaps about 15 to 20 minutes or so. However, it didn't agree with me. I reproduce its response in full below. The numbers in square brackets are what I myself added so that I can more easily refer to the relevant parts of its response.
[1] The eyeglasses analogy doesn't do the work you need it to, and the reason is instructive.
[2] When someone puts on the wrong glasses, we already have independent, overwhelming evidence that unaided vision exists and works: billions of people see fine without glasses, we understand the optics of why a bad lens distorts light, and the same person can immediately demonstrate intact vision the moment the glasses come off. The "intact vision underneath" isn't posited to explain away the bad-glasses case — it's independently verified.
[3] The soul hypothesis has none of that supporting structure. There's no independent case demonstrating that a soul exists, sees, thinks, or remembers apart from a brain. So the move isn't "we already know unaided vision exists, and this case doesn't threaten that" — it's "let's posit an entity for which we have zero independent evidence, give it exactly the property needed to absorb this counterexample, and declare the mind/soul distinction whenever brain damage threatens it." That's not analogous reasoning; that's a rescue device. Philosophers call this an unfalsifiable auxiliary hypothesis — you can save any theory from any evidence if you're willing to keep adding custom-fitted patches.
[4] There's a sharper problem too. With the glasses, we can check the underlying capacity separately from the impaired condition — take the glasses off and look. With the soul, there is no analogous check. You can't take the "brain glasses" off a living person and test whether the soul underneath sees fine. The claim is only ever cashed out as a promissory note about what happens after death, which is precisely the one moment no one can report back and verify it. A claim that is in-principle unverifiable in exactly the way that would let it be tested isn't a rival explanation of the neurology — it's structured so it can never be checked against the neurology at all.
[5] And the pattern of the damage is itself evidence against the hypothesis, not just a fact the hypothesis has to accommodate. It's not random what breaks: damage to the fusiform gyrus specifically knocks out face recognition while leaving object recognition intact; damage to Broca's area knocks out speech production while comprehension survives; damage to the hippocampus prevents new memories while old ones remain. The deficits track the brain's functional anatomy with remarkable precision — which is exactly what you'd expect if the brain is doing the cognitive work, and not what you'd expect if an immaterial soul is doing the work and the brain is merely a "radio" or "eyeglasses" the soul operates through. If the soul is the real seat of face-recognition, why would damaging one specific patch of cortex — and not the pineal gland, not "wherever the soul interfaces with the body," but the fusiform gyrus, reliably, across patients — knock out that one specific faculty? The transmission model has never explained why the correlations are so anatomically fine-grained rather than global or scattered.
[6] None of this proves no soul exists. But it does mean the eyeglasses analogy isn't evidence for the mind/soul split — it's a template for how to make any theory unfalsifiable, applied here without the independent support that makes it legitimate in the vision case.
The Eyeglasses example
Claude's very opening sentence [1] immediately suggested to me that it had not understood my argument at all, and this was vindicated when I read part [2] of its response. I want in this paragraph to clear something up. In my response to Sam Harris, I gave no evidence nor reasons to suppose that the filter thesis is a superior one compared to the production thesis, nor did I imagine I had. For even though I do think it's superior, that was not the purpose of my response. My purpose was to simply point out that there is an alternative thesis to explain why mind states are adversely affected by brain states. And should this thesis be the correct one, then negatively impacted mind states resulting from impaired brain function should not by themselves give reasons to doubt a soul, nor indeed have any consequences for one's self or soul once one is disembodied and existing within some afterlife environment. In short, I was questioning Sam Harris' assertion that we have very good reasons to doubt an afterlife since his point that one's mind can be adversely affected by brain damage doesn't, by itself, appear to achieve this. He would have needed to argue that this negative impact on the mind is far more plausibly accounted for by the production thesis than the filter thesis. However, he didn't even seem aware that there were any alternative theses to explain this fact, an ignorance seemingly shared by the vast majority of the skeptics of an afterlife.
Does Claude's explanation for how eyeglasses work and how they differ from the mind and body relationship, refute or at least cast considerable doubt on the filter thesis? Absolutely not at all. First of all, it is worth mentioning I could have mentioned any other number of examples where X adversely influences the output or functioning of Y, yet without X causing or producing Y. Examples of this are the picture displayed by a TV set, a song outputted by a radio, a mobile phone displaying a webpage, a prism splitting white light into its component colours, and so on. In each of these cases we can provide a full explanation of how Y is affected by X, just as Claude attempts with eyeglasses in paragraph [2]. The question here is, why cannot the brain and mind share a similar relationship to all these examples even though in this case we are ignorant of the causal mechanism? If one is unconvinced then consider this: Would providing full explanations for the various instances where X produces Y cast doubt on the production thesis? For example, someone defending the production thesis might point out that despite the fact that smoke is very unlike fire, fire nevertheless produces smoke. So even though consciousness is very different from brain processes, he might say that perhaps we shouldn't be too surprised that brains can produce consciousness. Clearly, it would be besides the point if I were to point out we have overwhelming evidence for how fire produces smoke, whereas we do not know how brains produce consciousness. For this gives us no reason to suppose brains do not in some currently unknown manner or other produce consciousness. But then, equally, the fact we do not know how the brain changes, modulates, and attenuates conscious states does not in itself provide any reasons or evidence that it doesn't do this.
Allow me to make what I said in the previous paragraph even clearer. It might be useful to compare my eyeglasses analogy to the balloon analogy of the Universe. Dots are painted on the surface of a balloon that are meant to represent galaxies. Blowing up the balloon represents the expanding Universe with the surface of the balloon representing the entire Universe. We can see that it makes no sense to talk about any of the dots being at the centre of the surface of the balloon, and we can see that none of them are located in any privileged position. One of the aims of this analogy is to facilitate the understanding that there need not be any privileged centre of the Universe, least of all that the centre is where we are. This is despite the fact that all the other galaxies are receding from us. Now, suppose I were to say this balloon analogy fails to provide any evidence that the Universe doesn't have a centre. Suppose I then proceeded to provide an explanation for why the dots move apart on the surface of the balloon in the way they do. This clearly would provide no reason or evidence whatsoever that the Universe does have a centre and isn't in fact expanding. It would be to completely miss the point of the analogy and the person giving the balloon analogy would rightfully flag me as either being stupid, or being deliberately obtuse.
So what Claude says in [1] is simply wrong and [2] is irrelevant and its assertion in part [6] that the eyeglasses analogy is not evidence is attacking a straw man.
I will advance reasons why I think we should favour the filter thesis over the production thesis in the My Positive Arguments section below.
The fine-grained correlations between brain and mind states
To be fair on Claude, it does in the latter part of its response attempt to put some meat on the bones of its contention. It mentions that the mind-brain correlations are not scattered or random but rather they are "anatomically fine-grained", and that the thesis that I have advanced, the filter thesis, does not account for this. Unfortunately, it never specified why it thinks the filter thesis does not account for this, and it is far from obvious to me. Consider our eyeglasses. Fogged up lenses affect our vision in one way, warped lenses affect our vision in another way, scratches on the lenses in yet another way, painting the lenses black would even make us go blind. Even if such correlations between the defects in our vision and damage or alteration to the lenses are not deserving of the label "fine-grained", it is arguably getting there.
But it might be more instructive in this context to think of the analogy of a television set rather than eyeglasses, although I go into more detail than others have done regarding this analogy. So the TV signal itself, which is invisible to our eyes, carries all the information about a programme being screened. It carries the information about which pixels should light up, in what colour, and when. And all this information constitutes the programme being screened, such as a film. However, damage to the TV sets internal components can adversely affect the picture being displayed in all manner of ways. Perhaps the set might develop dead pixels, or the information conveyed by the TV signal is misinterpreted and some pixels light up the wrong colour. Indeed, no end of other aberrations of the picture can manifest themselves. But crucially, note that no amount of damage could result in an actual change to the plot of a film being screened. In other words, damage to the internal components does absolutely nothing to the TV signal itself, it still conveys precisely the same information regardless of the state of the TV set receiving the signal.
In this analogy, the TV signal represents the soul or the self. The TV picture stands for what we normally might refer to as the mind, which is the result of the self or soul as it manifests itself and gets shaped by the brain. So damage to the brain might result in damage to the mind, but if this analogy is roughly accurate, it cannot damage the self or soul.
Note that any damage to the components can result in "fine-grained" aberrations in the picture displayed. This despite the fact that the programme being shown has its origin from outside of the TV set. So the question is, why would one expect the soul-brain relationship to be different? Claude never specifies why, and I have the suspicion there simply is no good answer to this question. Claude and people articulating this objection don't appear to have thought this through.
The Entity as a Rescue Device
Getting back to what Claude says, in paragraph [3] it states, "there's no independent case demonstrating that a soul exists, sees, thinks, or remembers apart from a brain". It also asserts that this soul, or "entity" to use Claude's terminology, is a "rescue device" to save the afterlife hypothesis. In the very next paragraph, paragraph [4], it says: "[t]here's a sharper problem too. With the glasses, we can check the underlying capacity separately from the impaired condition — take the glasses off and look. With the soul, there is no analogous check".
So what is this self? Quite simply, it's the common-sense conception of what the vast majority of us consider our selves to be. The self, or "entity", is the thinker that has thoughts, the experiencer that undergoes experiences. So experiences and thoughts don't just exist by themselves, they are not free-floating as it were. Rather the idea here is that experiences cannot exist without an experiencer, thoughts cannot exist without a thinker, and that this experiencer or thinker is simply what we mean by the self. Further, this self, together with its conscious states, would usually be thought of as being non-physical. After all, no self, nor even any of its conscious states, can be detected. At best, only the underlying neuronal activity of conscious states can be detected.
Should this conception of the self be correct, then it follows that we are the very same person throughout our lives. Our moods, intelligence, interests, general personality characteristics can change throughout our lives, and when it comes to our moods, they can change from one hour to the next, yet we are still the very same person, self, experiencer or thinker throughout these changes. And this aligns with how we feel. Most of us are convinced we are the very self-same individual throughout all these changes. It should also be added that this conception of the self entails that our sensory perceptions, thoughts, memories, general personality characteristics and so on, do not comprise our selfhood, rather they are merely attributes of this self. The position known as substance dualism is the most common position in the mind-body debate that incorporates such a conception of the self.
So if we accept this concept of a self, there are two possibilities: Either this self ends with the death of the body, or it survives the death of the body. If the former, then the self requires the brain to enable us to think, remember, undergo emotions and more generally to have experiences. If, on the other hand, the self survives the death of the body, then the brain plays no role in the production of these attributes; it merely changes, modulates, attenuates and frequently inhibits them. This implies that when a self is in its disembodied state it will experience its unchanged or unaltered attributes or conscious states.
Would it be reasonable to believe that such a self survives the death of its body? I suggest that even if we ignore all the evidence suggesting we survive, such as NDEs and so on, it might be reasonable to do so. We just need to consider that our bodies change throughout our lives. So how can there be an underlying self or essence that doesn't change when this self is produced by a changing body?
Is this Self or Entity Falsifiable?
However, as I explained in the previous section, this accusation has no merit since the conception of the self or soul I outlined is the default belief of the vast majority of humankind. This includes many of those who emphatically reject an afterlife. Even Karl Popper, the very person who proposed the falsifiability criterion in the first place, held more or less this conception of the self, even though he was personally disposed to reject an afterlife. The crucial point being that my conception of the soul was not simply conjured up to save the afterlife hypothesis — indeed, nothing could be further from the truth. Hence, to label it an auxiliary hypothesis is simply inaccurate, and grossly so.
The first point to make is that even if, contrary to the facts, the conception of a soul I have advanced were purely an auxiliary hypothesis with no independent reasons to subscribe to it, it is interesting to note that auxiliary hypotheses can often be fruitful in rescuing the main theory. Indeed, this fact was instrumental in formulating more sophisticated conceptions of falsifiability and even questions whether falsifiability is always a suitable guidance to scientific progress.
My Positive Arguments
I think it might be useful to some people if I were at this juncture to give my own positive reasons for preferring the filter thesis over the production thesis. There is all the evidence of course, and I find all the anomalous phenomena near death persuasive. For example, near-death experiences including shared-death experiences, deathbed visions, terminal lucidity, apparitions of the dying, clocks stopping at the moment of death, and so on. I also find persuasive the alleged memories of young children who appear to remember a past life and where the details the child provides largely check out on being investigated (the main examples being the research by Ian Stevenson). But I won't go into details here since all this evidence is explored extensively elsewhere and by people more knowledgeable about this evidence than I am.
In addition to the evidence, there are also philosophical arguments for preferring the filter thesis to the production thesis. Perhaps the most compelling is that the production thesis is arguably cripplingly problematic in explaining how there could be any mechanism or causal chain whereby the brain produces consciousness. Consider that consciousness is supposed to come into being as the end consequence of physical chains of causes and effects. Such causes and effects are cashed out in the form of processes that we can measure; namely particles with physical properties such as charge, momentum, spin and so on, and their interactions. But at the end of such causal chains we get a sudden abrupt change from these measurable processes to subjective experiences such as, for example, the greenness of grass, the warmth of love, the smell of roses and so on. But here it seems we have an unbridgeable ontological gap between the termination of such physical causal chains, and such raw experiences. There is no appropriate mechanism, or conceivable causal chain, whereby such qualitative experiences could be created. This difficulty of how the brain could produce consciousness has been labelled the hard problem.
An Appraisal of Claude's Response
Part [1] of its response where it states that my eyeglasses analogy doesn't achieve what it needs to achieve, is not justified by [2] since [2] is irrelevant.
Parts [3] and [4] convey the idea I am simply conjuring up some "entity", as Claude refers to it, which will accommodate the fact that brains adversely affect mind states, hence allowing it to potentially survive death. This then makes such an entity an unfalsifiable auxiliary hypothesis. The unwary reader would be surprised indeed to discover that this mysterious sounding "entity" is simply the pretty much universal concept of the self that we all instinctively accept, even from those rejecting an afterlife. This means that the entity or self is not something I've conjured up out of thin air, and hence cannot be an auxiliary hypothesis. It is true that, strictly speaking, this self, and indeed the filter thesis, are unfalsifiable, but so too is the production thesis. However, this isn't relevant since neither of these theses are scientific hypotheses in the first place. Finally, regarding these two sections, why is Claude more or less repeating in part [4] of its response what it had already said in part [3], yet stating that [4] is an even more compelling problem for such a self or soul than [3] is?
Part [5] of its response concerning the fine-grained correlations between brain and mind states is potentially better since this does seem to make it plausible that the straightforward explanation for such fine-grained correlations is the obvious and simplest one that the brain produces consciousness. But in order for the production thesis to be a superior explanation here, such fine-grained correlations need to be absent where X adversely influences the functioning of Y, yet doesn't produce Y. But if X is very complex, such as in a brain, this seems questionable. Especially when something considerably less complex, such as a television set, has what might be labelled "fine-grained correlations" between the functional state of its internal components and the picture it displays.
So, in summary, Claude's arguments sound superficially compelling, but providing we have sufficient understanding of the underlying issues, we can see that they completely fall apart when we examine them. Further, it's adopted a partisan position where it appears to be more focussed on "winning" the argument, or in persuasion, rather than in a genuine search for the truth. Given that so many people find these LLMs as being authoritative and reliable and use them extensively in writing scholarly papers and essays, I do find it of concern that these LLMs are prone to employing fallacious arguments as well as adopting partisan positions.
