Chapter 16

Anesthesia, Anoxia, and Drug-Induced Hallucinations

Why the pharmacological hypotheses fall short of the phenomenon they claim to explain

A Tale of Two Losses of Consciousness

In a hangar at Brooks Air Force Base in Texas, a fighter pilot strapped into a centrifuge spun until the blood drained from his head. The G-forces stripped oxygen from his brain. He went limp. His arms twitched. A few seconds later he came back. The instructor asked him what he had just experienced.¹

The pilot frowned. Bits and pieces, he said. Something about his kitchen. A face he could not place. A scrap of music. He could not put it together. The whole thing had lasted maybe ten seconds. He felt confused for another minute. Then he was fine.

That same week, in a hospital a thousand miles away, a man went into cardiac arrest. His heart stopped. Within fifteen seconds his brain's electrical activity collapsed. The monitors went flat. The team began compressions. Four minutes passed before they got a rhythm back. Six before he was breathing on his own. He was unconscious — by every medical measure unconscious — for those long minutes.

When he came around, he told the cardiologist a story. He had been floating above his own body. He had watched the team work. He saw the senior nurse step away to take a phone call. He heard his wife in the hallway, weeping, telling someone he had always wanted to be cremated. He had moved through what he later called a corridor of soft light and met his mother, who had died nine years before. She told him it was not his time. He came back. He could describe the hallway he had never seen, the nurse's call he had never heard, his wife's words he had never been told.²

Two episodes of unconsciousness. Two losses of brain function. One produced fragments. The other produced a story. The pilot's brain had been starved of oxygen for a few seconds, and he came back with mental confetti. The cardiac arrest patient's brain had been silent for several minutes, and he came back with a coherent narrative containing information he could not have known.

Same general category — unconscious from oxygen deprivation. Wildly different phenomena. That difference is the puzzle this chapter is built to address.

The Question Behind the Hypothesis

One of the most popular physicalist responses to the near-death experience literature goes like this. NDEs happen during cardiac arrest. Cardiac arrest deprives the brain of oxygen. Oxygen deprivation makes brains hallucinate. So NDEs are hallucinations. Drugs given during resuscitation might add to the effect. The dying brain may also dump its own chemicals — endogenous psychedelics like dimethyltryptamine, or molecules that act on the brain the way ketamine does. Combine these mechanisms, and you have a pharmacological cocktail that mimics dying. The experience is real to the experiencer. It is also, on this view, just chemistry.³

That argument has the virtue of being testable. We know what oxygen-starved brains do. We know what DMT does. We know what ketamine does. We know what awareness during anesthesia looks like. None of this is hidden science. We can place each phenomenology beside the NDE phenomenology and look. If the match is good, the hypothesis lives. If the match is bad, the hypothesis dies.

The match is bad. It is bad in specific, documentable ways. That is the burden of this chapter. The previous chapter examined the dying-brain hypothesis as Borjigin and her colleagues have developed it.⁴ Here we take up the related cluster — anoxia, anesthesia awareness, endogenous DMT, ketamine-like dissociation — and we ask whether any of them, alone or together, explain what we actually find in the cases.

The Anoxia Hypothesis

The anoxia argument is the simplest one. Anoxia means an absence of oxygen reaching tissue; hypoxia means oxygen below normal levels. When neurons run short of oxygen, they misfire. Misfiring neurons produce strange experiences. Cardiac arrest produces hypoxia. So NDEs are hypoxic hallucinations.

We know a great deal about hypoxic phenomenology because we have studied it deliberately for almost a century. Pilot training programs put aviators in low-oxygen chambers and watch what happens. Mountaineers above twenty-five thousand feet have left careful accounts. The U.S. Air Force researcher James Whinnery spent years documenting more than seven hundred episodes of gravity-induced loss of consciousness — what aviators call G-LOC — in centrifuge-trained pilots.⁵ Whinnery thought at one point that G-LOC dreamlets might be a window into the NDE. The more carefully he and other researchers looked, the less the two looked alike.

Hypoxic phenomenology has a signature. The signature is confusion. Pilots in the chamber lose color vision first. Their visual field narrows — what aviators call "tunnel vision," though it is not at all the tunnel of NDE reports. They become euphoric in a slack, drunken way. Their judgment fails. They report fragmentary images, sometimes pleasant, often nonsensical. They cannot follow a thought. When they regain consciousness they are disoriented for minutes. They rarely remember the episode in any organized way. The experience does not feel realer than reality. It feels foggy.⁶

NDE phenomenology is the opposite. The reports are characteristically lucid. NDErs describe their thinking as faster and clearer than ordinary waking thought, not slower and more confused. The narrative is sequential. Memory of the experience is sharper, decades later, than memory of ordinary events from the same period. Pim van Lommel's eight-year follow-up of cardiac arrest survivors found that NDErs' recall of the experience remained unchanged across years; non-NDE memories from the same period faded normally.⁷ Hypoxic memories do not behave that way. Hypoxic memories are foggy when fresh and gone in a week.

There is also a timing problem. NDEs are not confined to oxygen-starved states. They occur in patients with normal blood-oxygen levels — in deep hypothermia, in non-arrest cardiac events where oxygenation was maintained, in some psychological crises and accidents where there was no medical compromise of brain perfusion at all. If anoxia caused NDEs, normoxic NDEs should not exist. They do, and in not insignificant numbers.⁸

And there is a deeper problem with the brain-physics. The hypoxia chamber produces gradually developing hypoxia in a brain that keeps working — disordered, slowed, glitchy, but working. Cardiac arrest produces something closer to a switch being thrown. Within ten to twenty seconds the cortical EEG goes flat. The brain is not running poorly; it is not running. The chamber model and the cardiac arrest model are not the same physiological event. The chamber model cannot be exported to explain what happens in arrest, because in arrest there is no chamber-like state of disordered-but-active brain to host the hallucination.⁹

The DMT and Ketamine Hypotheses

The pharmacological argument moves from oxygen to molecules. Maybe the dying brain releases its own psychedelic. Dimethyltryptamine — DMT — is the natural candidate. It is one of the most powerful psychedelics known. It produces brief, intense, visionary experiences. The psychiatrist Rick Strassman, who ran the first FDA-approved human DMT trials in the 1990s at the University of New Mexico, speculated openly that endogenous DMT might be involved in dying.¹⁰

The trouble is that DMT phenomenology and NDE phenomenology look quite different up close. DMT users report geometric kaleidoscopes. They report contact with bizarre entities — beings of light made of moving math, "machine elves," chittering fractal organisms. The world becomes hyperreal in a strange way: textures pulse, colors saturate beyond what eyes can see, language fragments. The experience is famous for being radically other. NDE phenomenology is not bizarre in that way. NDErs report meeting their own grandmother. They report a familiar-feeling place. They report warmth, a sense of homecoming, an encounter with a personal Being. Where DMT scrambles the categories of waking life, the NDE deepens them.¹¹

Ketamine fits a little better. Ketamine is a dissociative anesthetic. At certain doses it produces out-of-body sensations and experiences of leaving the world. The British psychiatrist Karl Jansen built a sustained case in the 1990s that NDEs were essentially ketamine experiences produced by endogenous compounds binding the same receptors.¹² The OBE component is the strongest overlap. Some ketamine users do float above their bodies. Some pass through dark passages. Some encounter beings.

But the comparison breaks down on the details. Ketamine experiences are characteristically dysphoric — laced with fear, dissociation, panic. NDEs are characteristically peaceful, with most NDErs reporting profound calm. Ketamine experiences are visually drug-like — geometric, patterned, distorted. NDEs are visually environmental — the room, the relatives, the landscape — and where features are unusual the unusual features are described as more real, not less. Ketamine experiences fade quickly and rarely transform a life. NDE aftereffects last decades and reshape entire personalities.¹³

The most careful empirical comparison to date is the 2019 study by Charlotte Martial, Vanessa Charland-Verville, and colleagues at the University of Liège. They used semantic similarity analysis to compare 625 NDE reports with 15,000 written reports of experiences induced by 165 different psychoactive substances. Of all the drug classes, ketamine produced the closest match to NDEs. But "closest" is not "same." Even the ketamine-NDE similarity was systematically incomplete. The features ketamine reproduces best — out-of-body sensation, perceptual unusualness — are precisely the features the dualist tradition would expect to be partly co-produced by the dying body's neurochemistry. The features ketamine does not reproduce — life review, encounter with deceased loved ones, lasting moral transformation, accurate distant observation — are precisely the features that mark the NDE as more than the brain.¹⁴

And the empirical question of whether the dying brain actually releases enough DMT or DMT-like molecules to drive an experience has not been settled in the affirmative. The 2019 Borjigin-lab study found small amounts of endogenous DMT in dying rat brains. The amounts were not large. Whether the human dying brain does anything similar at scale, and whether such a release would actually generate the lived NDE, remains undetermined.¹⁵ The endogenous-DMT hypothesis is a hypothesis. It is not a finding.

Anesthesia Awareness Is Not What NDEs Are

One last move in the pharmacological cluster: maybe NDEs are episodes of anesthesia awareness. Anesthesia awareness — sometimes called intraoperative awareness — happens when a surgical patient is paralyzed and unable to move but is partially conscious because the anesthetic is insufficient. It is rare. The best estimates put the rate at roughly one or two cases per thousand surgeries.¹⁶

It is also one of the most traumatic experiences in medicine. Patients who suffer it report exactly what the situation suggests: pain, paralysis, panic, the helpless awareness that they are pinned to a table while their body is being cut open. Many develop severe post-traumatic stress disorder. The clinical literature on anesthesia awareness is harrowing reading. It is not the literature of NDEs.

NDEs do occur during surgical anesthesia in some documented cases. The cardiothoracic surgeon Michael Sabom and the neurosurgeon Robert Spetzler each documented NDEs in patients undergoing procedures with deep anesthesia and EEG flatline.¹⁷ The most famous of these — Pam Reynolds's case, treated in Chapter 12 — involved deliberate hypothermic cardiac standstill, taped eyes, ear-canal speakers playing white noise to monitor brainstem function, a documented EEG flatline, and an NDE during which Reynolds described the bone saw used on her skull and the surgical conversation. None of this fits the anesthesia-awareness pattern. Reynolds did not report pain. She did not report panic. She reported floating above her body, a tunnel, a meeting with deceased relatives, a peaceful return. The pharmacology of anesthesia awareness predicts the opposite at every point.

The Compound Hypothesis and the Cumulative Failure

The thoughtful physicalist will not stake everything on a single mechanism. The thoughtful physicalist will argue that the dying body produces a unique pharmacological cocktail — hypoxia plus endogenous DMT plus ketamine-like dissociation plus residual anesthetics plus the dying-brain gamma surge — and that the cocktail, not any single ingredient, produces the NDE. The compound hypothesis is harder to refute because it is harder to specify. It is also, for the same reason, weaker as a scientific account. A theory that gets stronger by becoming more vague is not actually getting stronger.¹⁸

The deeper problem with any pharmacological account — whether single-ingredient or compound — is that pharmacology produces local content. A chemical in the brain affects what the brain does. A chemical in the brain cannot let the brain perceive what is happening in another room or another building or another state. The strongest NDE cases are precisely the ones in which the experiencer perceives accurate information about events far from the body: the conversation in the waiting room, the family member arriving by car at the hospital, the object on the windowsill on the floor above. No drug in the medical-cabinet, and no endogenous compound the human body is known to produce, can transmit information about distant events to the user. That is not how chemistry works.

The same problem hits the congenitally blind cases. A surge of any neurochemical through a visual cortex that never wired for ordinary sight cannot produce accurate visual perception of the room. There is no scaffolding for the chemical to act on. And the same problem hits the cases of meeting recently deceased relatives whose deaths the patient did not know about. Pharmacology does not deliver news.

What the Pharmacological Hypotheses Really Show

The honest assessment is this. The pharmacological hypotheses do explain something. Some sub-features of NDE phenomenology — the OBE sensation in particular, the timeless quality of experience, the sense of meaning — likely share neurochemistry with anesthesia, anoxia, and psychedelic states. The dying brain is doing things, and what it does shapes some of the experience.

Anesthesia awareness explains very little, because the pharmacology and the phenomenology are simply different — anesthesia awareness is a story of pain and panic; the NDE is a story of peace and structure. The anoxia hypothesis explains less than was once thought, because hypoxic phenomenology is the opposite of NDE phenomenology in nearly every documented respect. The DMT hypothesis explains some of the visionary texture of certain NDEs and almost none of the rest, because DMT experiences are characteristically bizarre and NDEs are characteristically familiar. The ketamine hypothesis is the strongest of the four because the OBE component matches; even there, the Martial study showed ketamine to be similar to but systematically distinct from the NDE pattern.

None of these accounts — and no plausible compound of them — explains the veridical core. None explains the cross-cultural consistency the cumulative case brings to bear. None explains why NDErs are transformed for life while ketamine users go home and have lunch. The pharmacological cluster is not a competitor explanation for the strongest NDE evidence. It is a partial gloss on some of the texture.

And the cluster strategy itself is telling. To rescue physicalism from the NDE evidence, the physicalist has to multiply mechanisms — anoxia for some cases, DMT for others, the dying-brain gamma surge for cardiac arrest, ketamine-like dissociation for the OBE, residual anesthetics for surgical NDEs, REM intrusion for the borderline cases. Each ingredient is offered for a different feature. None of them is offered for the whole. The dualist account, by contrast, is one account: the experiencer is the soul; the body is the instrument; the body's varying neurochemistry partly co-shapes the experience in each context. One framework. One ontology. The proliferation of partial physicalist explanations is a sign of empirical strain, not strength.¹⁹

Why This Matters at the Bedside

If you have sat with a dying loved one — or if you have wondered what your own dying will be like — the pharmacological hypotheses promise a strange comfort. They say: do not worry too much about the stories. The grandmother your father saw in the hospital was a chemical. The light your mother described was a chemical. The peace was a chemical. None of it was real. The real you ends with the heartbeat. Whatever comes after is a kindness the brain plays for itself on the way out.

Some Christians find that comfort tolerable because they trust the resurrection to come later. Many do not. And the evidence does not support the comfort.

The drugs we know about do not produce what NDEs produce. The body's own pharmacology, as best we can measure it, does not either. Whatever your father met when he died, it was not a chemical. Whatever your mother saw was not a chemical. The grandmother was not a chemical. The peace was not a chemical. The empirical evidence does not let us say with certainty what they encountered. But it does let us say, with growing confidence, what they did not encounter — and a brain-bound hallucination is what they did not encounter. The phenomena are larger than that.

This matters because the dying themselves often need to know. Dying patients ask. They ask their nurses and their chaplains and their pastors and their children, sometimes in full lucidity in the last days, what is happening to them. The honest answer is not, "It is just chemistry; do not worry." The honest answer — supported by the best contemporary research — is, "Something more than chemistry is going on, and the historic Christian tradition has been describing it for two thousand years."

The next chapter pulls all of this together — the dying-brain hypothesis from Chapter 15, the pharmacological cluster from this chapter, and the residual physicalist accounts (REM intrusion, temporal-lobe stimulation, cultural conditioning) — and asks, cumulatively, whether physicalism has the resources to explain what the NDE evidence actually shows. The answer, the dissertation found, is no. The next chapter is the case for that answer.