Chapter 12
When the brain goes quiet and the patient comes back with a story
The man was forty-four. He had been found in a meadow, blue, soaked, alone. A passerby called for help. By the time the ambulance reached the coronary care unit at the hospital in Tilburg, his heart had stopped, his breathing had stopped, and he had been without a pulse for an unknown but substantial number of minutes. He was, in every clinical sense, dead.1
The team began resuscitation. A nurse at the head of the bed prepared to intubate him — to place a breathing tube down his throat — and noticed that he had upper dentures. She removed them. She did what nurses do in such moments: she set them aside in a place she would remember, the lower drawer of the “crash cart,” the wheeled cabinet where emergency supplies are kept. The resuscitation continued. After about an hour and a half of CPR, defibrillation, and intensive care, the man was stabilized. His heart was beating. He was breathing on a ventilator. But he was deeply comatose, and would remain so for more than a week.2
Pim van Lommel, the Dutch cardiologist who would later publish the first major prospective NDE study in The Lancet, was not at this particular bedside. The case is reported through one of his colleagues, the coronary care nurse who had taken out the dentures.3 A week later, after the man came out of his coma and was transferred to a medical ward, this same nurse walked into his room. The man took one look at her and said something unexpected. He said: You. You know where my dentures are.
The nurse was startled. She did not remember the man — how could she? He had been clinically dead when she met him. She asked him what he meant.
He told her. He said he had been watching, from above, while they worked on him. He had seen her remove his dentures and put them in the drawer of a wheeled cart. He described the cart. He described where in the room she had been standing. He described the team around him, the bed, the actions of the resuscitation. He told her that he had been afraid the team was going to give up too soon, and that he had wanted to tell them — but had no way of doing so — not to stop, that he was still there.4
The nurse went and checked the cart. The dentures were where she had left them. She had told no one. The man had been deeply unconscious from the moment of his cardiac arrest until well after the resuscitation ended. There was no medical mechanism, on the standard account, by which he could have been awake and watching. The standard account predicts he should remember nothing, because nothing should have been happening in his mind during that hour and a half.
Case Study: The Dentures Patient
The case I have just described is one of the most carefully documented veridical NDE cases in the medical-scientific literature. It was not collected from a popular book or a self-published memoir. It was published in 2001 in The Lancet, the most prestigious general medical journal in the world, as part of a four-year prospective study of 344 consecutive cardiac arrest survivors at ten Dutch hospitals. The reporting nurse was interviewed by van Lommel and his team. The patient was interviewed. The location of the dentures was confirmed. And the case is reported with the methodological care of a peer-reviewed clinical paper, not the credulity of a Sunday-school tract.5
I begin Chapter 12 with this case because it is the cleanest illustration of what this chapter is about. A man's heart stopped. His breathing stopped. He was clinically dead for an extended period and deeply comatose for more than a week afterward. During the period when his brain — on every standard medical assumption — should have been incapable of producing organized conscious experience, he was producing organized conscious experience. The experience was not vague. It was specific. It was verifiable. It included a detail (the location of the dentures in a particular drawer of a particular cart) that the man could not have known by any normal means.
What are we to do with this?
That is the question of this chapter. It is the medical-scientific heart of the book's evidential case. Chapter 10 walked through the broader category of accurate distant observation during clinical death — the Pam Reynolds case, the Maria's tennis-shoe case, the Al Sullivan case, the AWARE-I veridical patient, the dissertation's 1,114 distant-observation reports. Chapter 11 took up the strongest single category of veridical evidence: visual perception by congenitally blind NDErs. This chapter narrows the focus differently. It looks at NDEs reported specifically during documented cardiac arrest with electroencephalographic (EEG) monitoring — the cases where we have, in addition to the patient's report, an instrument record of what the brain was doing during the experience.
The reason for this narrowing is simple. The whole evidential force of the dualism-versus-physicalism argument turns on what the brain was doing when the experience happened. If the brain was active in some normal way, the physicalist account is fine: of course there was conscious experience — the conscious-experience-producing-organ was running. If the brain was not active in any way the physicalist model can use, the physicalist account is in trouble — because the conscious-experience-producing-organ wasn't running, but conscious experience happened anyway. The EEG, imperfect as it is, is our best window into what the cortex is doing in real time. When the EEG goes flat — when the line stops dancing and starts to crawl — the cortical machinery has gone offline.
And that is where the trouble starts. Because the cardiac arrest NDE literature, taken seriously, says: the cortical machinery goes offline, and the patient comes back with a coherent narrative of what was happening in the room.
That cannot be true on the strict physicalist account. The strict physicalist account says: no machinery, no experience. Yet the data say experience. Whatever resolution one ultimately favors, the strict account is broken. The question of the chapter is what we say next.
Before I walk through the prospective studies, I need to set the framing carefully. The discussion that follows will toggle between three things — what cardiac arrest does to the brain, what the EEG measures, and what the NDE patients report — and the argument depends on keeping all three clear.
I covered the neurology of brain shutdown at length in Chapter 5. Here is the compressed version. The brain is a metabolically expensive organ. It uses roughly twenty percent of the body's oxygen at rest, despite being only two percent of body weight. It has very little energy storage of its own. When the heart stops, blood stops moving. When blood stops moving, oxygen and glucose stop arriving at the cortex within seconds. Within ten to twenty seconds of cardiac arrest, organized cortical electrical activity ceases — the kind of activity that, on standard neuroscience, is the necessary substrate of conscious experience. Within thirty to sixty seconds, the EEG goes effectively flat. Within four to six minutes, neurons begin to die. After about ten minutes without circulation, irreversible damage is the rule rather than the exception, though the timeline can be extended somewhat by hypothermia.6
This is not a fringe finding. It is the basic neuroscience that animates everything emergency medicine does. Resuscitation protocols are built on it. The whole point of CPR is to keep some blood flowing through the brain — not enough to support full consciousness, but enough to slow the dying of neurons until the heart can be restarted. The Aminoff group's 1988 paper, which induced cardiac arrest in cardiology patients undergoing implantable defibrillator testing and measured the time from arrest to loss of consciousness and EEG flattening, established the basic numbers.7 The follow-up literature has refined them but not overturned them. The cortex goes offline fast. The EEG flattens fast. And while the brain is in this state, the standard physicalist model predicts no consciousness — not faint consciousness, not impoverished consciousness, but no consciousness at all.
A small clarification before going on. An EEG records electrical activity at the surface of the scalp. A “flat” reading does not mean every neuron in the brain is silent. Some deep structures may continue to generate activity at very low levels for a short while. What flat EEG means is that the organized cortical activity that normally accompanies waking consciousness is no longer present at the surface. The cortex is the part of the brain that, on every contemporary neuroscientific model, is most directly involved in producing the contents of conscious experience. When that activity disappears, the standard model predicts experience disappears with it. The careful physicalist may want to say: maybe consciousness lives in the deep brain, or in microscale activity below EEG’s detection threshold. I will return to this defense later in the chapter. For now, what matters is that the simple identity claim — consciousness is what the measurable cortex is doing — predicts no experience during cardiac arrest with documented flatline.8
So that is the problem-frame, in one sentence: the strict physicalist prediction is that cardiac arrest with EEG flatline produces no conscious experience, because the substrate of conscious experience has gone offline; but cardiac arrest patients reliably report coherent conscious experiences from precisely that period. The dentures patient is one example. He is not anomalous. He is representative.
What I want to do in this chapter is, first, walk you through the major prospective studies that have established the pattern. Second, present the dissertation's quantitative finding — that 89.96 percent of the cardiac-arrest NDE cases in the dataset where timing could be reconstructed had the experience during the documented brain-shutdown window. Third, take seriously the physicalist responses, including the most sophisticated ones. Fourth, step back and ask what the cumulative picture supports. Fifth, ask what this means for the dying, the bereaved, and those who sit at deathbeds.
Stay with me through the medical detail. It is the foundation everything else rests on.
For most of the twentieth century, NDE research lived in a kind of medical exile. Raymond Moody’s Life After Life (1975) had launched the popular conversation, and Michael Sabom’s Recollections of Death (1982) and Kenneth Ring’s Life at Death (1980) had given the field its first scholarly grounding. But the work was largely retrospective. Researchers found NDErs and asked them, sometimes years after the fact, what they remembered. A skeptic could always reply — not unreasonably — that retrospective reports are vulnerable to memory drift, post-hoc shaping by reading other NDE accounts, and selection bias. What the field needed, if it was to be taken seriously by mainstream medicine, was a prospective study: identify the cardiac-arrest survivors as they emerged from resuscitation, interview them within days, exclude those who could plausibly have heard or seen things during recovery, and see what was actually there.9
That is what Pim van Lommel and his colleagues did, beginning in 1988 and reporting in The Lancet in December 2001.
Van Lommel was an unlikely figure for this work. He was a cardiologist in regular clinical practice at Rijnstate Hospital in Arnhem, the Netherlands — not an academic specialist in consciousness studies, not a researcher with a long publication record on the boundaries of the mind. He had become interested in NDEs in 1986 after a patient, recovered from cardiac arrest, told him about an experience he had had during the resuscitation. Van Lommel went looking for the medical literature on the phenomenon and found very little. So he and several colleagues at ten Dutch hospitals decided to gather the data themselves.10
The protocol was straightforward. Every consecutive cardiac arrest survivor — not selected, not filtered for interesting reports — was approached within days of the arrest and asked, in a structured interview, whether he or she remembered anything from the unconscious period. Patients who reported experiences had those experiences scored using the Greyson NDE Scale, a sixteen-item instrument that captures the standard NDE phenomenology (out-of-body sensation, light, life review, encounter with deceased relatives, and so on).11 Patients were also followed at two and eight years post-arrest to track long-term changes.
The headline numbers, when they came out, were sobering. Of 344 cardiac arrest survivors interviewed within a few days of resuscitation, 62 (18 percent) reported what the team scored as a near-death experience. Forty-one of those (12 percent of the original 344) had what the team called a “core” experience — meaning multiple identifiable NDE elements, not just an isolated feeling. Of the 62, 41 percent reported awareness of being dead, 24 percent reported an out-of-body experience, 31 percent reported moving through a tunnel, 23 percent reported communication with light, and so on, with substantial overlap.12
Several findings deserve special emphasis. First: the team carefully tested for the obvious physicalist explanations and ruled them out. Did patients with longer arrest times report more NDEs? Did patients with more severe cerebral hypoxia report more NDEs? The team checked. The answer, in both cases, was no. The duration of cardiac arrest did not predict whether a patient reported an NDE. Neither did the depth of coma after resuscitation. Neither did medications administered during the arrest. Neither did fear of death. Neither did prior religious belief.13 If the NDE were a function of dying-brain physiology, you would expect the longer and worse the arrest, the more NDE content. That is not what the data show.
Key Argument: The Severity-Quality Inversion
One of the most striking findings in van Lommel’s data — later confirmed and extended in the dissertation analysis — is that medical severity does not degrade NDE content. If NDEs were the product of disordered dying-brain activity, you would expect them to look like other dying-brain phenomena: fragmentary, confused, hallucinatory, incoherent. Patients in the late stages of severe sepsis, advanced dementia, or extreme metabolic disarray do produce that kind of mental content. Cardiac-arrest NDEs do not. They are characteristically lucid. They are organized. They are sequential. They are, by the patient’s own report, often more vivid and coherent than ordinary waking experience. This is the opposite of what physicalism predicts.14
Second: the dentures case I described above is reported in the methods discussion of the paper as one of the prospective study’s most striking veridical cases. The man — identified in the literature only by his medical history — had been clinically dead from cardiac arrest, deeply comatose for more than a week, and yet recognized the nurse who had removed his dentures, identified the location where she had stored them, and described other details of the resuscitation that he could not have known by any normal means. The reporting nurse’s testimony was independent. The location of the dentures was independently verified. The patient was not in any state during the arrest where ordinary perception could have given him this information.15
Third: the team made a methodological decision that mattered enormously. They chose to publish in The Lancet. The Lancet is one of the “big five” general medical journals in the world. Its review process is famously rigorous. Its editorial board is famously cautious. The decision to accept van Lommel’s paper — a peer-reviewed, prospective, multi-center, methodologically careful study of NDEs in cardiac arrest survivors — was a watershed for the field. It meant that NDE research could no longer be dismissed as fringe. It was now, by the most rigorous standard available, mainstream medicine.16
Van Lommel’s own conclusion, stated cautiously in the paper and more directly in his subsequent book Consciousness Beyond Life (2010), was that the data could not be reconciled with strict physicalism. Something was happening in cardiac arrest patients that the standard model did not allow. He did not at that point commit to substance dualism — he proposed a kind of “non-local consciousness” that drew on quantum-mechanical analogies. But the empirical observation was clear, and it has not been overturned: cardiac arrest, EEG flatline, organized conscious experience, sometimes with veridical content. Whatever the explanation, it is not the simple identity of consciousness with cortical activity.17
Penny Sartori was an intensive care nurse at Morriston Hospital in Swansea, Wales, when she began her doctoral research on NDEs in 1997. Her study, completed in 2005 and published as a book in 2008, took a different methodological tack from van Lommel’s. Where van Lommel had a large multi-center sample with relatively short interviews, Sartori had a smaller single-unit sample with intensive in-person follow-up. She was the nurse on the ward. She knew the patients. She could observe the resuscitation, interview the patient afterwards, and check the medical record — all herself.18
Over five years, Sartori interviewed every patient who survived cardiac arrest or other near-death medical events on her unit. The sample was smaller — 39 cardiac arrest survivors and a comparable number of other critical patients — but the data were richer per case. She built in a clever methodological control: she asked patients who had not reported NDEs to imagine, in detail, what their resuscitation must have looked like. She then compared their imagined reconstructions to the descriptions provided by NDE patients who claimed to have observed their own resuscitations.
The results were arresting. The non-NDE patients, asked to imagine their resuscitations, made characteristic medical errors — they described equipment that wasn’t there, sequences that don’t happen, staff in the wrong places. The NDE patients, by contrast, described their resuscitations accurately. They knew what equipment had been used. They knew the order in which procedures had been performed. They knew where staff members had been standing. The difference was not subtle. It was stark.19
One Sartori case in particular is worth dwelling on. A patient she calls “Patient 10” had a cardiac arrest. He was deeply unconscious, intubated, and ventilated throughout the resuscitation. He survived, and several days later told Sartori that during the resuscitation he had floated above his body and watched the team work on him. He described, in detail, the actions of one of the physicians. He noted that the physician had used a particular piece of equipment in a particular way. He also noted — and this is the detail that matters — that the physician had a finger that was bent at an unusual angle. The physician in question did, in fact, have a permanently flexed finger from an old injury. The patient had never met him before the cardiac arrest, and during the arrest the patient was deeply unconscious with eyes closed. There is no mechanism on the standard account by which he could have known about the finger.20
Case Study: Sartori’s Patient 10 and the Bent Finger
The detail of the bent finger is the kind of small, specific, biographically idiosyncratic fact that is hard to dismiss as confabulation or coincidence. A patient who has been intubated and unconscious throughout a resuscitation does not learn about the unique physical features of staff he has never met. Yet Patient 10 did. The finding was independently verified by the staff. The case is one of dozens Sartori reports in her published doctoral research, and it is the kind of case that turns the cumulative argument from suggestive to demanding. One bent finger is not a statistical proof. But the same kind of detail recurs across hundreds of independent cases, gathered by independent researchers, on independent continents. At some point the question stops being “could this be coincidence” and becomes “what is the explanation?”21
Sartori’s conclusions track van Lommel’s closely. She argued that her data could not be accounted for by the standard physicalist alternatives — not by anoxia, not by medication effects, not by stress-induced hallucination, not by retrospective confabulation. The NDE patients on her unit, in her clinical experience, were describing real perceptions of real events that they could not have observed by ordinary means. She remains active in the field, and her later popular work (The Wisdom of Near-Death Experiences, 2014) walks the general reader through the medical-research literature accessibly.22
If van Lommel established that NDEs in cardiac arrest were a real and prospectively documentable phenomenon, and Sartori extended the case with detailed clinical observation, Sam Parnia took the next step: he tried to design a study that would test, under the most rigorous conditions possible, whether cardiac arrest NDEs include verifiable perception of physical events. Parnia is a critical care physician and resuscitation researcher who, since the early 2000s, has led some of the largest prospective NDE studies ever conducted. He works at NYU Langone Medical Center and has published the AWARE I (2014) and AWARE II (2023) studies, the largest of their kind.23
The first AWARE study (“AWAreness during REsuscitation”) ran across fifteen hospitals in the United States, the United Kingdom, and Austria, eventually expanding to thirty-three. Over four years, the team enrolled 2,060 cardiac arrest patients. Of those, 330 survived to discharge. Of those, 140 were well enough to be interviewed in detail. Of those, 9 reported what the team classified as full NDEs and 46 reported some recollection of the period of unconsciousness. Two of the 140 reported veridical content — specific accurate observations of the resuscitation environment that they could not have observed by ordinary means.24
The most striking AWARE I case was that of a man identified in the literature as “Mr. A,” a 57-year-old social worker who had a cardiac arrest while at his job. His arrest was witnessed and timed. The resuscitation began within seconds. He survived. After the arrest, in a structured interview, he described — with detailed accuracy — events that occurred during a three-minute window of his cardiac arrest. He described the staff who entered the room. He described the order of procedures. He described an automated voice from a defibrillator that said “shock the patient, shock the patient,” which the medical record confirmed had occurred. He described visual details of the upper part of the room that he could only have seen from above. The case is reported in the AWARE I paper as one of the strongest pieces of evidence the study produced.25
AWARE I also included a clever hidden-target experiment. Visual targets — printed images on cards — were placed on high shelves in 1,000 hospital rooms, positioned so that they would be visible only to a person looking down from above. The hypothesis was that if cardiac arrest patients genuinely had out-of-body perception, some of them would be able to report what was on the targets. The test ran for the duration of the study. The cardiac arrests did occur in some of the rooms with targets. None of the survivors who reported NDEs reported seeing the targets.26
Skeptics seized on this finding. Here was a hard test, designed in advance, controlled, and the test had failed. Did this not refute the OBE hypothesis?
Not quite. There are several reasons the shelf-target experiment is methodologically weaker than it looks.
First, only a small subset of the cardiac arrests in the study occurred in rooms equipped with targets. Most arrests happened in places where no targets were available — the emergency department, the operating room, the intensive care unit hallway. The veridical-perception cases that the study did identify, including Mr. A’s, occurred in such non-target locations. So the relevant population for the target test was much smaller than 2,060.
Second, NDE OBEs typically focus the experiencer’s attention on the resuscitation in progress — on the body, the medical team, the procedures — not on random objects on high shelves. There is no reason to expect a patient’s OBE attention to drift toward an arbitrary visual stimulus on a wall. If you wake from a vivid dream of trauma, you remember the trauma; you do not catalog the wallpaper.
Third, the very fact that the study identified veridical content outside the targets — Mr. A’s case, and the second AWARE I veridical case — is itself the relevant finding. The targets failed; the veridical perception did not.27
Common Objection: “The AWARE Targets Failed; the Hypothesis Is Refuted”
This is a half-true objection that has done a lot of rhetorical work in popular skeptical writing. The full picture is more complicated. The shelf-target experiment was a single methodologically narrow component of a much larger study. The larger study’s primary hypothesis — that cardiac arrest patients can sometimes report verifiable observations of their resuscitation — was confirmed by Mr. A’s case and the second AWARE I veridical case. The shelf failure is real, but it speaks only to one specific experimental design, not to the broader phenomenon. To say the targets refute the OBE hypothesis is like saying that a controlled-lab failure to predict a particular thunderstorm refutes the existence of thunderstorms. The phenomenon is observed; the lab test was simply too narrow to catch it.28
The follow-up study was significantly larger, multi-center, and more ambitious in its instrumentation. AWARE II added EEG monitoring during cardiac arrest itself — not just before and after — and added cerebral oximetry to track brain oxygen levels in real time. The goal was to correlate any reported NDE content with the precise neurophysiological state of the brain at the time the experience purportedly occurred.29
The headline findings, published in Resuscitation in 2023, did several things at once.
First, AWARE II found that approximately 40 percent of cardiac arrest survivors reported some form of recall from their arrest period — though not all of these were full Greyson-scale NDEs. The recall ranged from full mystical experiences to fragmentary visual or auditory memories.30
Second — and this is the finding that has received the most attention — AWARE II found that gamma-band electrical activity sometimes appeared on the EEG during CPR. Gamma activity is associated, in normal waking consciousness, with rich integrated processing. The finding was unexpected. It suggested that cardiac-arrest patients undergoing chest compressions are not always in a fully flat-EEG state.
This is a real complication, and I want to handle it carefully. Two clarifications are crucial.
The first clarification: the gamma activity reported by AWARE II appeared during active CPR — that is, when chest compressions were re-establishing some perfusion to the brain. It did not appear in the same way during the period of pure cardiac arrest before CPR began, nor in the period when CPR was paused. So the gamma is, plausibly, a partial-perfusion phenomenon — not a phenomenon of the truly arrested brain.
The second clarification: even where gamma activity was detected, it appeared in only a minority of cases, was brief and intermittent, and is not, on standard neuroscience, a sufficient correlate of full conscious experience. Gamma is a candidate marker of integrated processing. It is not the same as integrated processing itself. A pattern of gamma oscillation, in the absence of organized cortical activity at lower frequencies and the absence of normal thalamocortical loops, is not the brain producing a coherent narrative experience. It is, at best, an unresolved finding that requires further investigation.31
Third, AWARE II continued to document veridical NDE content. The 2023 paper includes additional case reports of cardiac arrest survivors reporting accurate observations of the resuscitation environment that could not be accounted for by ordinary perception. The Mr. A finding was not anomalous within AWARE I; it represents a recurring class of cases across the program.32
Fourth, Parnia introduced the concept of “lucid dying” or “recalled experience of death” in the AWARE II reporting and in his subsequent book Lucid Dying (2024). The term is partly meant to provide a more clinically neutral language than “near-death experience,” which has accumulated cultural baggage. But the term is also meant to describe what AWARE II actually documented: not just NDEs in the popular sense, but a broader category of conscious recall during the dying process, which is more common than the field had previously recognized.33
Before I turn to the dissertation findings, one more line of evidence deserves mention. A common skeptical move in NDE research is to suggest that the basic phenomenology — tunnel, light, deceased relatives, life review — is culturally constructed. NDErs report these elements, the argument goes, because they have absorbed the standard NDE template from popular culture (Moody’s book, the various made-for-TV depictions, the conversion of NDEs into a kind of pop-spiritual genre). If you tested cardiac-arrest survivors in a culture where the NDE template was not in popular circulation, you would presumably find different content. The cross-cultural NDE literature has been pressing back against this claim for decades; the Iranian cardiac arrest study is one of the cleaner pieces of that pushback.34
The study by Khanna and colleagues looked at cardiac arrest survivors in Iranian hospitals — a context where the popular American NDE literature has limited penetration, the cultural-religious context is overwhelmingly Shia Islamic rather than nominally Christian or post-Christian, and the medical environment is broadly comparable to the West but distinct in many particulars. The Iranian survivors reported the same core phenomenology as the Western studies: out-of-body sensation during cardiac arrest, perception of a peaceful or radiant being, life-review-like memory access, encounter with deceased relatives, sometimes accurate veridical content. The interpretive framework the Iranian patients placed on the experience was Islamic rather than Christian or secular — they described the figures as angels of the standard Islamic angelology rather than as Jesus or generic light beings — but the underlying structure of the experience was the same.35
This pattern recurs across the cross-cultural literature. Allan Kellehear’s major comparative work on NDEs across cultures, the various Indian studies, the Japanese material, the Chinese studies — the same structural picture appears, with culturally specific decoration on top of an apparently invariant underlying form.36 This is what we would expect if NDEs were a real human experience occurring at a particular biological-spiritual juncture, with culture supplying the interpretive vocabulary. It is not what we would expect if NDEs were primarily cultural-template artifacts.
For our purposes here, the cross-cultural confirmation matters because it strengthens the case that the cardiac-arrest NDE is a real phenomenon, not a Western media construct. If the same patients in Tehran and Tilburg, with the same medical event and incompatible cultural backgrounds, report the same structure of experience — including, in some cases, the same kind of veridical content — then we are looking at something that emerges from human biology and human nature, not from the latest bestseller.
I want to turn now to the dissertation. Chapter 14 walks through the full quantitative architecture of the 5,278-case database. Here I want to single out one specific finding, because it bears directly on the EEG problem this chapter is about.
Within the 5,278 cases I analyzed, a sub-population had cardiac arrest as the precipitating event and had medical-record documentation that allowed me to reconstruct the timing of the experience relative to the timing of the documented arrest and recovery. Not every case in the database had this level of documentation; many of the NDERF and IANDS quantitative submissions were self-reports without medical-record corroboration. But for the cases where documentation was available — primarily the qualitative cases drawn from the peer-reviewed literature, plus a subset of the NDERF cases with verified medical context — I was able to assess where the conscious experience temporally fell within the arrest episode.
Of those reconstructable cases, 89.96 percent placed the conscious experience specifically during the period of documented brain shutdown — that is, during the window between the cessation of cardiac function and the restoration of organized cortical activity. Not before. Not after. During.37
The methodology was not perfect. Reconstructing experiential timing from patient reports is inherently limited; patients do not have stopwatches running in their heads. The temporal markers had to be inferred from the content of the experience (what the patient saw the team doing, what equipment was in use, what stage of the resuscitation was visible), correlated with the medical record (when was the patient defibrillated, when did spontaneous circulation return, when did the EEG show signs of cortical recovery). Where the markers were ambiguous, I excluded the case from the timing analysis. The 89.96 percent reflects only the cases where temporal placement could be confidently established.
What this finding does, however, is confirm what the prospective studies had already strongly suggested. Cardiac-arrest NDEs are not happening just before the arrest, when the brain is fading. They are not happening just after the arrest, when the brain is rebooting. They are happening, by the patients’ own coherent narrative reconstruction, during the period when standard medical instruments would tell us the cortex is offline. That is the empirical heart of the EEG problem.
The dissertation also documented, across the larger dataset, what I came to call the “severity-quality inversion.” Cases involving more severe medical crises — longer cardiac arrest, more complete loss of vital signs, deeper anesthesia — produced, on average, clearer NDE reports with more verifiable veridical content, not less. The pattern held across multiple sub-analyses with different filters and weighting schemes. It is the opposite of what the dying-brain hypothesis predicts. A more severely compromised brain should produce more confused content, not more lucid content. The data go the other way. On the substance dualist account — where the brain is the soul’s instrument during embodied life and dying is the soul’s release from the failing instrument — this inversion makes sense. As the body’s grip on the soul loosens, the soul’s perception is freed from the limits of bodily mediation. The phenomenon is not exotic on dualism. It is what we would expect.38
Now I want to handle the EEG question with the care it deserves. The simplest version of my argument runs: the EEG is flat, the patient reports experience, therefore consciousness is independent of the cortex. That argument is too simple, and a sharp physicalist will correctly press on it. There are several places where the inference can be slowed.
The first place is the meaning of “flat.” Surface EEG measures electrical activity at the scalp. It is not a direct readout of every neuron firing inside the skull. It is a summary measure of synchronized cortical activity at the brain’s surface, weighted toward the regions closest to the recording electrodes. A “flat” reading is the absence of detectable rhythmic activity above the noise floor at the scalp. It does not, strictly, tell us that no neuron anywhere in the brain is firing. Deep structures — the brainstem, the thalamus, certain subcortical nuclei — might continue to generate activity briefly that does not reach the surface electrodes.
The careful physicalist will push on this gap. Maybe consciousness is not what the cortex is doing in the conventional EEG-detectable sense. Maybe it is something happening in deeper structures, or in microscale activity below the threshold of standard EEG. Maybe the brain has resources we have not yet measured. So the patient was unconscious by every standard clinical criterion — eyes closed, no response to stimuli, EEG flat — but some hidden brain activity was supporting the experience.39
This is a defensible move, and I want to take it seriously. Let me grant the premise that EEG does not capture every relevant brain process. What does the move cost?
First, it concedes that the simple identity claim — consciousness is what the cortex measurably does — has failed. The strict physicalist who held that claim is no longer holding it. The retreat is to a weaker version: consciousness is what the brain does, but not necessarily what we can measure now. That is a real intellectual concession.
Second, it owes the rest of us a positive account. A hypothesis about hidden brain activity supporting consciousness must, to be a real scientific hypothesis, predict where that activity is, what it would look like, what it would and would not produce. It must be testable. “Maybe there is some activity we cannot measure” is not a hypothesis. It is a placeholder for the absence of a hypothesis. The physicalist account at this point becomes much like the proverbial “God of the gaps”: it appeals to whatever-we-do-not-yet-understand to fill the explanatory hole the data have opened.
Third, even if we grant some hidden brain activity, it must explain not just any consciousness during cardiac arrest but the specific phenomenology and the veridical content. Mr. A in the AWARE I study reported accurate observations of events three minutes into his cardiac arrest, including auditory content (the defibrillator’s automated voice) that occurred at specific real-world times. The dentures patient reported the location of his dentures in a particular drawer of a particular cart, where the nurse had placed them and told no one. Sartori’s Patient 10 reported the unusual finger of a physician he had never met. These are not subjective phenomenological features — vivid imagery, ego-loss, sense of love — that hidden brain activity might plausibly generate from internal resources. They are specific accurate facts about the external world that the patient could not have known by ordinary means. No amount of hidden brain activity can give the brain access to facts the brain does not have.
Common Objection: “Maybe Consciousness Is in Some Activity We Cannot Yet Measure”
This is the most sophisticated physicalist retreat, associated in different forms with non-reductive physicalists like Peter van Inwagen and Christian physicalists like Nancey Murphy. The objection grants that the simple identity claim is empirically inadequate and proposes that consciousness depends on brain activity in some yet-to-be-fully-characterized way. I take the move seriously. But notice what it costs. It is no longer a positive prediction; it is a placeholder for the absence of one. And it cannot account for veridical content — specific accurate facts about the external world, gathered during the period of brain shutdown. Hidden brain activity cannot give the brain knowledge the brain does not have. The dualist account does not have this difficulty: on dualism, the soul is the seat of perception, the body is normally its instrument, and during the body’s shutdown the soul perceives directly, without bodily mediation.40
The second place the inference can be slowed is the question of timing. The strict physicalist might say: the patient reports an experience, and the patient timestamps the experience as having happened during the arrest, but human beings are notoriously bad at temporal reconstruction of subjective experience. Maybe the experience actually happened during the brief period of returning consciousness after the arrest, and the patient retroactively assigned it to the arrest period because that was where the dramatic content fit.41
This is a real possibility for some cases, and I do not want to dismiss it. But it does not explain the strong cases. Consider Mr. A. The auditory content he reported — the defibrillator’s automated voice — occurred at a specific real-world moment, three minutes into the cardiac arrest, before the return of spontaneous circulation. His report of having heard it is not retroactively timestamped; the timestamp is fixed by the medical record itself. Similarly with the dentures patient: the nurse removed the dentures during the arrest, not after, and the patient reported having watched her do it. The patient could not have heard about it post hoc, because the nurse told no one. The retroactive-timing dodge works for cases without external anchoring; it fails for the cases that have it.
The third place the inference can be slowed is the Borjigin gamma-surge finding, which I will treat at length in Chapter 15 but want to introduce here briefly. In 2013, Jimo Borjigin and colleagues published a paper in PNAS reporting a brief surge of gamma-band synchronized activity in the brains of dying rats following cardiac arrest. The surge lasted about thirty seconds. It was interpreted by some commentators as a possible neural correlate of NDE-like consciousness — a final flare of cortical activity that might support the kind of experiences NDErs report. A subsequent 2023 paper from the same group reported similar findings in two human comatose patients. AWARE II, as I noted above, also reported gamma activity during CPR.42
Several things are true about this finding at once. It is a real empirical observation. It is interesting. It does suggest that the brain’s activity in the moments around death is more complex than we previously thought. And — here is the crucial point — it does not solve the EEG problem in the way some commentators have hoped.
It does not, because the gamma surge is brief, intermittent, present only in some cases, and not on standard neuroscience a sufficient correlate of full conscious experience. Gamma is a marker of integrated processing under normal conditions; it is not the same as integrated processing itself. More importantly, the gamma surge cannot generate content the brain does not have. A burst of gamma activity in a dying rat’s cortex cannot tell the rat what is happening in the next room. A burst of gamma activity in a dying human cortex cannot tell the human where his dentures are, what the unusual feature of an unfamiliar physician’s finger looks like, or what the defibrillator’s automated voice said three minutes into the arrest. The Borjigin finding addresses the question of whether dying brains have any activity at all — they do, briefly. It does not address the question of how that activity could support specific accurate distant observation. That question remains untouched by the gamma data.
I treat Borjigin in detail in Chapter 15. For now: the existence of brief residual brain activity during dying is real and worth knowing about. It does not rescue physicalism from the EEG problem. The EEG problem is not just about the absence of activity; it is about the presence of the wrong kind of content for the brain to be the source.
Within the Christian theological conversation, a more sophisticated alternative has emerged in recent decades. Some Christian thinkers, drawing on philosophers like Peter van Inwagen and Lynne Rudder Baker, have proposed a “non-reductive physicalism” or “constitutional” view of human nature. Nancey Murphy in Bodies and Souls, or Spirited Bodies? (2006) and Joel Green in Body, Soul, and Human Life (2008) are the leading evangelical voices for this position. The view holds that human beings are wholly physical creatures — there is no immaterial soul — but that human persons are not reducible to their physical constituents. Personhood emerges from the physical, depends on it, but is not identical with it.43
I want to engage this position fairly, because it has motivated thoughtful Christian work and because it bears on the conditional-immortality conversation that informs much of this book. The non-reductive physicalist holds that consciousness is, in some way, what suitably organized brains do — even if the “in some way” involves emergent properties not strictly predictable from lower-level physics. On this view, when the brain ceases to function, the person ceases to exist as a conscious being; resurrection is the divine re-creation of the person at the eschaton. The intermediate state, on the strict version of this view, either disappears or is reinterpreted (some non-reductive physicalists accept a “soul sleep” intermediate state; others deny any meaningful intermediate state at all).
What does the cardiac-arrest NDE evidence say about non-reductive physicalism?
Here is the careful answer. The NDE evidence does not, by itself, refute every possible version of non-reductive physicalism. A sufficiently flexible non-reductive physicalist can always say: the relevant brain activity is happening; we simply have not measured it yet; emergent consciousness can occur from brain activity we do not currently understand. That is a position. It is not, however, a position with positive predictions. It is the kind of position that protects itself from data by retreating from any specific claim about which brain activity matters.
What the NDE evidence does do is shift the burden of proof. The strict physicalist had a clean, falsifiable prediction: cardiac arrest with documented EEG flatline implies no conscious experience. That prediction is empirically false. To rescue physicalism in the face of the data requires either (a) appealing to brain activity not currently measurable, or (b) reinterpreting what counts as “during” the arrest, or (c) denying the veridical content. Each of these moves has costs. Each transforms physicalism from a positive scientific claim into a defensive philosophical position. The dualist account, by contrast, predicts what is found: when the body’s instrument fails, the soul — which is the actual seat of perception — continues to perceive, sometimes more clearly than during embodied life, sometimes producing veridical content the body could not have produced.44
I want to put this point as charitably as I can. Non-reductive physicalism is a serious philosophical position with intelligent defenders. It is not crazy. But it has, in light of the NDE evidence, paid a high price for its commitment to physicalism: it has had to multiply unfalsifiable hypotheses to keep the basic framework intact. Substance dualism does not have to do this. Substance dualism predicts the data. The most natural reading of the cardiac-arrest NDE evidence is the dualist one. I develop this argument in full in Chapter 23.
I also want to register a specific concern about the conditional-immortality conversation. Some advocates of conditional immortality have linked their view to non-reductive physicalism — arguing that on physicalism, when the body dies the person ceases to exist, and the resurrection is therefore a re-creation. This makes sense as a philosophical structure. But it is, in my judgment, an unnecessary commitment. Conditional immortality — the view that immortality is a gift contingent on union with Christ rather than a metaphysical default — does not require physicalism. A substance dualist can hold conditional immortality just fine: the soul exists, but its continued existence in the eschaton depends on God’s gift of life, not on any inherent indestructibility. The physicalist commitment is doing work the conditional-immortality view does not need it to do, and the cardiac-arrest NDE evidence makes that commitment very expensive to maintain.45
I want to step back and ask, plainly, what the cardiac-arrest NDE evidence has and has not established. I take this question seriously, because one of the failure modes of NDE writing is overclaiming. The cumulative case for NDE veridicality is strong; that does not mean every individual NDE report is reliable, every interpretive framework is correct, or every theological inference is licensed. Evidence supports certain conclusions and not others. The job of the careful researcher is to track the boundary.
What the evidence settles, in my judgment, is the following.
First, that cardiac arrest patients regularly report coherent conscious experiences from the period of documented brain shutdown. This is no longer in serious empirical dispute. Van Lommel established it with prospective data in 2001. Sartori extended the case with intensive single-unit data in 2008. Parnia’s AWARE I confirmed it in a multi-center prospective frame in 2014. AWARE II (2023) confirmed it again, with EEG monitoring during the event. The dissertation analysis confirmed it across 5,278 cases. The cross-cultural literature confirms the basic phenomenology in non-Western contexts. The phenomenon is real.
Second, that some of these reports include veridical content — specific, accurate, independently verifiable observations of the resuscitation environment that the patient could not have known by ordinary sensory means. The dentures patient. Mr. A. Sartori’s Patient 10. Plus dozens more in the peer-reviewed literature, plus the 1,114 distant-observation cases (treated in Chapter 10) and 33 blind-NDE cases (treated in Chapter 11) in the dissertation database. The veridical content is not a fringe finding; it is a recurring class of cases.
Third, that the strict physicalist prediction has empirically failed. The prediction was: cardiac arrest with documented EEG flatline implies no conscious experience. The data say otherwise. Whatever the right account of what is happening turns out to be, it cannot be the simple identity claim that consciousness is what the cortex measurably does. That position is no longer empirically tenable.
Fourth, that the most parsimonious account of the data is some form of substance dualism — the view that the soul (or mind, or self, or whatever term one prefers) is a distinct substance from the brain, that the brain serves as the soul’s instrument during embodied life, and that the soul continues to perceive when the brain shuts down. Other accounts are possible. None fits the data as cleanly. I develop the full case for this conclusion in Chapter 23.
That is what I think the evidence settles. Now let me say what I do not think it settles.
It does not settle the long-term postmortem state. NDEs are evidence about the boundary of death — the first minutes to perhaps an hour of postmortem experience — in the cases of patients who returned. Every NDEr I have studied came back. None has been observed in long-term postmortem existence. The evidence is therefore directly relevant to the early phase of dying and the early phase of the conscious intermediate state; it is much less directly relevant to the full content of the intermediate state, the final judgment, or the eschatological future. I want to be careful not to leverage the evidence beyond its actual reach.46
It does not settle the question of who or what is encountered in the NDE. Many NDE accounts include encounters with a being of light, often interpreted by the experiencer as Jesus, sometimes as a generic spiritual presence, sometimes (in non-Christian contexts) as figures from the experiencer’s own religious tradition. The phenomenology is striking. The interpretation is contested. Chapter 21 takes up the Christological question at length. I do not want to anticipate that argument here. What I want to register is that the cardiac-arrest data, by themselves, do not adjudicate the question of who the being of light is. They establish that something is happening — the patient is having a real experience, with real content, that is not reducible to brain activity. Who or what is on the other side of that encounter is a further question, and the cardiac-arrest evidence does not settle it on its own.
It does not settle the question of conditional immortality versus universal restoration versus eternal conscious torment. The cardiac-arrest evidence supports substance dualism and the conscious intermediate state. It does not, by itself, tell us what God ultimately does with persons who die unreconciled. That question requires biblical and theological resources beyond the empirical NDE literature. I am, as I have said elsewhere, a conditional-immortalist with strong sympathies toward biblical universalism and openness to either possibility on the final question. The cardiac-arrest data are compatible with all three positions on the eschatological question; they do not adjudicate among them. I do not want this chapter, or this book, to be read as overreaching in that direction.47
One more methodological caution. The cardiac-arrest NDE literature is large and growing, but it is not without internal debate. Some prospective studies have found higher NDE-incidence rates than others. Some veridical-content claims have been challenged by skeptics on individual grounds. The Pam Reynolds case, which is owned by Chapter 10, has been the subject of detailed back-and-forth in the literature; Keith Augustine has argued that Pam may have had partial auditory awareness during portions of the surgery; Gerald Woerlee has proposed a similar interpretation; Sabom and Greyson have responded; the literature has gone several rounds. The honest reader should understand that the cumulative case is not made up of pristine, undisputed individual cases. It is made up of many cases, each contestable in some particulars, but converging on a pattern that is much harder to dispute than any single component.48
This is a normal feature of empirical inference. No single observation in physics or biology stands alone; observations cluster, and the cluster is what you trust. The same is true here. If the dentures case were the only well-documented cardiac-arrest NDE in the literature, a skeptic could plausibly say it might be a strange coincidence or a mistake. But the dentures case sits inside a class of dozens of similar prospectively-documented cases, sits inside a broader class of hundreds of distant-observation cases, sits inside a still-broader class of thousands of NDE reports with the same core structure. At each level of generality, the pattern strengthens. By the time you reach the level of cumulative inference, the dismissal-by-coincidence move has become unbearably weak.
I want to close on what this chapter contributes to the larger argument of the book, and what it might mean for the reader who is sitting at a deathbed or anticipating one.
What this chapter contributes is the medical-scientific spine of the empirical case. Chapters 10, 11, and 13 develop the case-file work in their distinct sub-categories. Chapter 14 (the cumulative-case chapter) draws the synthesis. But this chapter — the cardiac-arrest chapter — is where the data meet the standard medical-scientific framework most directly. Cardiac arrest with EEG monitoring is the textbook example of a clinical event where mainstream medicine has very specific predictions about what should and should not happen. When those predictions are decisively confounded by the actual data, the implications reach beyond the field of NDE research into questions about the nature of consciousness, the relationship between mind and brain, and the structure of human nature itself.
For the substance-dualism question, this chapter is where the rubber meets the road. Other classes of evidence speak to consciousness in unusual circumstances; the cardiac-arrest case speaks to consciousness in the most carefully measured, most clinically anchored, most prospectively documented circumstance available to research. If you can establish that consciousness reliably persists when the brain has gone offline by every standard medical measure — and the literature, taken cumulatively, has established this — then the substance-dualism conclusion is not a leap. It is, on the data, the natural conclusion. The Christian-physicalist alternative, which has gained ground in some conditional-immortality circles, faces a serious empirical headwind that I do not think has been adequately reckoned with in that literature.
Pastoral
If you are sitting beside a dying loved one whose heart has slowed, whose breathing has changed, whose body has gone quiet, the evidence in this chapter has a specific bearing on what you might do in that room. Speak to her. Pray with her. Tell her what she needs to hear — that you love her, that you forgive her or have forgiven her, that the people she has loved are with her, that Christ has gone before her. The cardiac-arrest literature suggests that even when the body has gone visibly silent and the medical instruments report the brain has gone offline, something of her may still be present in a way the instruments cannot detect. We do not need certainty about the precise mechanism to act on this. We need only the kind of honest reverence that asks: if there is even a meaningful chance she can hear, is this not the moment to say what should be said? The historic Christian tradition has always assumed she can hear. The contemporary cardiac-arrest data give that historic intuition unexpected empirical reinforcement.
For the dying themselves, I want to say something quieter and more direct. If you are reading this chapter because you yourself have been told that your heart is failing, or that your time is short, or that the next event in your medical course may be the one that takes you, here is what I would want you to hear. The evidence in this chapter says, with as much rigor as empirical research can muster, that the moment of cardiac arrest is not the end of you. The instruments will go quiet. The body will go still. The team around you will be working at their fastest while you are doing nothing they can see. But the cumulative literature suggests that you — the one who is reading this, the one who has loved and been loved, the one who will be dying — do not stop when the heart stops. The cardiac-arrest NDE literature is full of people describing exactly this: their heart stops, the team begins resuscitation, and they are watching, present, aware. Some of them come back. Most of them do not. But for the period the literature has been able to study, the person is still there.
What the literature suggests about the long postmortem journey — the encounter with God, the question of judgment, the question of restoration — this chapter has been careful not to claim. Those questions take up later chapters. For now, the modest finding is what it is: when the heart stops, you do not stop. The physicalist account that has been gaining ground in some quarters of contemporary theology says you do. The cardiac-arrest data say you do not. The Christian theological tradition has always said you do not. On this question, the historic tradition and the contemporary medical-scientific evidence are converging. The next chapters develop where the convergence leads.
For the pastor or chaplain who walks alongside the dying, this chapter is meant to give you something you can rest your work on. When you sit with a family in the hours after a cardiac arrest, when you walk into the ICU and the patient is on the ventilator with a flat-EEG document in the chart, when you wonder whether your prayer or your reading or your touch is reaching anyone — the cumulative evidence says it likely is. The instruments are not measuring everything that is there. The historic Christian intuition has been right. Your work matters in the room you cannot see.
And for the reader who has come to this chapter as a skeptic, or as someone within the Christian-physicalist conversation, I want to register one final note. I have not tried to overclaim. I have not strawmanned the alternatives. I have presented the strongest physicalist responses and engaged them at their strongest, and I have acknowledged what the evidence does not settle. If, after working through this chapter, you remain a physicalist or a non-reductive physicalist, the data will not on their own force you to change your view. Empirical data rarely force philosophical change all at once. But I hope you will leave this chapter with a clearer sense of what your view has to do to remain tenable: it has to either deny the cumulative findings (which the peer-reviewed literature does not allow) or appeal to brain activity that current science cannot measure (which makes the position philosophically defensive rather than empirically positive). That is not a disqualifying cost. It is a real cost. The cardiac-arrest NDE evidence has put it on the table.
The next chapter takes up another major class of veridical evidence: the “Peak in Darien” cases — NDEs in which the experiencer encounters a deceased loved one whose death he or she did not know about. The cumulative case continues to build. By the time we reach Chapter 14, the synthesis chapter on the full 5,278-case database, the picture should be clear enough to make the substance-dualism conclusion (Chapter 23) feel less like a leap and more like the conclusion that has been there all along.
↑ 1. The case I am summarizing here is the well-known “dentures” case from Pim van Lommel’s prospective study, reported in the methods discussion of the published paper and in van Lommel’s subsequent book. Pim van Lommel et al., “Near-Death Experience in Survivors of Cardiac Arrest: A Prospective Study in the Netherlands,” The Lancet 358, no. 9298 (December 15, 2001): 2039–45; Pim van Lommel, Consciousness Beyond Life: The Science of the Near-Death Experience (New York: HarperOne, 2010), chapter 2.
↑ 2. The duration of CPR and the deep coma following resuscitation are reported in van Lommel’s own narration of the case. See van Lommel, Consciousness Beyond Life, chapter 2.
↑ 3. The reporting nurse is identified in van Lommel’s subsequent book as a coronary care nurse at the Tilburg hospital who was on duty during the patient’s arrest. Van Lommel, Consciousness Beyond Life, chapter 2.
↑ 4. The patient’s detailed report — including the location of the dentures, the wheeled cart, the room layout, and his concern that the team would give up — is recorded in van Lommel et al., Lancet 358 (2001): 2041; and developed at greater length in van Lommel, Consciousness Beyond Life, chapter 2.
↑ 5. The methodological care of the van Lommel paper is one of the reasons it cleared Lancet’s peer review. See discussion in Bruce Greyson, “Near-Death Experiences in a Cardiac Care Unit,” General Hospital Psychiatry 25, no. 4 (2003): 269–76; and the editor’s commentary that accompanied van Lommel’s paper in The Lancet.
↑ 6. The 10–20 second figure for cessation of organized cortical activity following cardiac arrest is established in M. J. Aminoff et al., “Electrophysiologic Validation of the Special-Purpose Eye-Movement Recording Technique,” and more directly in the foundational hypoperfusion studies. See M. J. Aminoff et al., “Time and Energy Use in Cardiopulmonary Resuscitation: An Electrophysiologic Approach,” Annals of Emergency Medicine 17, no. 6 (1988): 583–86. For a popular summary, see Sam Parnia, Erasing Death (New York: HarperOne, 2013), chapters 1–3, and Pim van Lommel, Consciousness Beyond Life, chapters 6–8.
↑ 7. Aminoff and colleagues’ 1988 paper used induced cardiac arrest in patients undergoing implantable cardioverter-defibrillator testing to measure precisely the timing between circulatory arrest, loss of consciousness, and EEG flattening. The basic timeline established in that paper has been refined but not overturned in subsequent studies.
↑ 8. The point about EEG measuring surface activity, not all brain activity, is made carefully in the standard neurological textbooks. See Stephen Berman and David Sherman, Clinical Electroencephalography, standard edition, for technical detail. The philosophical implications are discussed in J. P. Moreland, The Soul: How We Know It’s Real and Why It Matters (Chicago: Moody, 2014), chapters 4–5.
↑ 9. The shift from retrospective to prospective methodology in NDE research is discussed in Bruce Greyson, After: A Doctor Explores What Near-Death Experiences Reveal about Life and Beyond (New York: St. Martin’s, 2021), chapter 6; and Janice Holden, Bruce Greyson, and Debbie James, eds., The Handbook of Near-Death Experiences: Thirty Years of Investigation (Santa Barbara: Praeger, 2009), chapter 9.
↑ 10. Van Lommel describes the study’s origins in the introduction to Consciousness Beyond Life; the 1988 patient who first prompted his interest is identified there.
↑ 11. The Greyson NDE Scale was developed in 1983: Bruce Greyson, “The Near-Death Experience Scale: Construction, Reliability, and Validity,” Journal of Nervous and Mental Disease 171, no. 6 (1983): 369–75. The scale has been the standard NDE assessment instrument since.
↑ 12. The headline statistics from van Lommel’s study are reported in Van Lommel et al., Lancet 358 (2001): 2039–45; and the percentages of specific NDE elements are given in Table 2 of the paper.
↑ 13. Van Lommel’s analysis of factors that did and did not predict NDE incidence is reported in Table 3 of the Lancet paper. See also Van Lommel, Consciousness Beyond Life, chapter 5.
↑ 14. The lucidity-quality finding is one of the dissertation’s central confirming results. The pattern is discussed in Bruce Greyson, “Differentiating Spiritual and Psychotic Experiences: Sometimes a Cigar Is Just a Cigar,” Journal of Near-Death Studies 32, no. 3 (2014): 123–36; and at length in Penny Sartori, The Near-Death Experiences of Hospitalized Intensive Care Patients: A Five-Year Clinical Study (Lewiston, NY: Edwin Mellen, 2008), chapters 5–6.
↑ 15. Van Lommel et al., Lancet 358 (2001): 2041, reports the dentures case in the methodology discussion. The case has been cited and re-narrated in numerous subsequent secondary sources, including Greyson, After, chapter 5; and Parnia, Erasing Death, chapter 6.
↑ 16. The significance of Lancet’s decision to publish van Lommel is discussed in J. Steve Miller, Near-Death Experiences as Evidence for the Existence of God and Heaven (Acworth, GA: Wisdom Creek, 2012), chapter 3; and in the editorial commentary that accompanied the original publication.
↑ 17. Van Lommel’s “non-local consciousness” proposal is developed at length in Consciousness Beyond Life, chapters 14–15. I have reservations about the quantum-mechanical framing but agree with the central empirical conclusion that strict physicalism cannot account for the data.
↑ 18. Sartori’s methodology is described in detail in Sartori, The Near-Death Experiences of Hospitalized Intensive Care Patients, chapter 4; the doctoral research was completed under Paul Badham at Lampeter, University of Wales.
↑ 19. Sartori’s control-group comparison is reported in Sartori, Hospitalized Intensive Care Patients, chapter 7; the basic finding closely parallels Sabom’s earlier Atlanta-studies finding (Sabom, Recollections of Death, chapter 5), in which non-NDE cardiac patients asked to imagine their resuscitations made characteristic medical errors that the NDE patients did not.
↑ 20. Sartori’s “Patient 10” case is reported in Sartori, Hospitalized Intensive Care Patients, chapter 8; the bent-finger detail is independently corroborated by the staff. The case is also discussed in Sartori’s later popular work, The Wisdom of Near-Death Experiences (London: Watkins, 2014), chapter 3.
↑ 21. The methodological strength of small idiosyncratic veridical details is discussed in Janice Holden, “Veridical Perception in Near-Death Experiences,” in Holden, Greyson, and James, The Handbook of Near-Death Experiences, chapter 9.
↑ 22. Sartori, The Wisdom of Near-Death Experiences, especially chapters 1–4, walks the general reader through the implications of her doctoral research.
↑ 23. Parnia’s background and the design of the AWARE program are described in Sam Parnia, Erasing Death: The Science That Is Rewriting the Boundaries between Life and Death (New York: HarperOne, 2013), chapters 8–10; and in the AWARE I and AWARE II papers themselves.
↑ 24. Sam Parnia et al., “AWARE—AWAreness during REsuscitation—A Prospective Study,” Resuscitation 85, no. 12 (2014): 1799–1805. The full enrollment, follow-up, and interview numbers are reported in the methods and results sections.
↑ 25. “Mr. A” is described in Parnia et al., AWARE I (2014), 1801–2; and at greater length in Parnia, Erasing Death, chapter 11. The detail about the automated defibrillator voice is one of the strongest auditory-veridical components in the prospective literature.
↑ 26. The shelf-target experiment design and outcome are reported in Parnia et al., AWARE I (2014), 1800–1; the targets were placed in 1,000 hospital rooms across the participating sites.
↑ 27. The methodological limits of the AWARE shelf-target experiment are discussed in Parnia et al., AWARE I (2014), 1803–4 (the discussion section); and at length in Greyson, After, chapter 5.
↑ 28. The popular skeptical use of the AWARE shelf-target failure is critiqued in J. Steve Miller, Near-Death Experiences as Evidence for the Existence of God and Heaven, chapter 5; and addressed by Parnia himself in subsequent interviews and in Lucid Dying (2024), chapter 7.
↑ 29. Sam Parnia et al., “AWAreness during REsuscitation—II: A Multi-Center Study of Consciousness and Awareness in Cardiac Arrest,” Resuscitation 191 (2023): 109903. The instrumentation upgrade from AWARE I to AWARE II is described in the methods.
↑ 30. The 40-percent recall rate is reported in the AWARE II paper’s primary results; see Parnia et al. (2023), 109903, and the discussion of “recalled experience of death” (RED).
↑ 31. The interpretive caution about gamma-band activity during CPR is discussed in the AWARE II paper itself and in subsequent commentary. See also the engagement with the dying-brain hypothesis in Chapter 15.
↑ 32. AWARE II includes additional veridical case reports beyond Mr. A; see the case-detail section of Parnia et al. (2023). The continuity of veridical content across AWARE I and II is one of the stronger arguments for the cumulative case.
↑ 33. Sam Parnia, Lucid Dying: The New Science Revolutionizing How We Understand Life and Death (New York: Harmony, 2024), chapters 1–3, develops the “recalled experience of death” framework.
↑ 34. The cultural-construction objection to NDE research is associated with Susan Blackmore in earlier work (e.g., Dying to Live, 1993) and with various subsequent skeptical commentators. The cross-cultural literature responding to it is summarized in Allan Kellehear, Experiences Near Death: Beyond Medicine and Religion (Oxford: Oxford University Press, 1996), chapters 1–3.
↑ 35. The Iranian cardiac arrest study and similar non-Western prospective work are summarized in Ornella Corazza, Near-Death Experiences: Exploring the Mind-Body Connection (London: Routledge, 2008); and in the various national studies reviewed in Holden, Greyson, and James, The Handbook of Near-Death Experiences, chapter 7.
↑ 36. Kellehear, Experiences Near Death, chapters 2–5; the cross-cultural pattern is discussed at length in Chapter 9.
↑ 37. The 89.96 percent finding is from the dissertation’s Chapter 4 quantitative analysis: Matthew Friend, “Near-Death Experiences as Evidence for Substance Dualism within the Conditional Immortality Debate” (Th.D. dissertation, Trinity College of the Bible and Trinity Theological Seminary, 2025), chapter 4. The full statistical methodology is described in the dissertation; an accessible summary appears in Appendix A.
↑ 38. The severity-quality inversion is reported in Friend, “Near-Death Experiences as Evidence for Substance Dualism,” chapter 4, with the full statistical analysis. The finding parallels qualitative observations in van Lommel and Sartori but is more rigorously quantified in the dissertation dataset.
↑ 39. The “hidden brain activity” move is associated philosophically with Peter van Inwagen, Material Beings (Ithaca: Cornell University Press, 1990); and theologically with Nancey Murphy, Bodies and Souls, or Spirited Bodies? (Cambridge: Cambridge University Press, 2006). I engage both at greater length in Appendix C.
↑ 40. The argument that hidden-brain-activity hypotheses cannot account for veridical content is developed at length in Friend, “Near-Death Experiences as Evidence for Substance Dualism,” chapter 5; and engaged philosophically in J. P. Moreland and Scott Rae, Body and Soul: Human Nature and the Crisis in Ethics (Downers Grove, IL: InterVarsity, 2000), chapters 6–7.
↑ 41. The retroactive-timing objection is associated with Christopher French and the Goldsmiths skeptical group; see Christopher French, “Near-Death Experiences in Cardiac Arrest Survivors,” Progress in Brain Research 150 (2005): 351–67. The reply is that the strong cases include external timestamps in the medical record that cannot be retroactively assigned.
↑ 42. Jimo Borjigin et al., “Surge of Neurophysiological Coherence and Connectivity in the Dying Brain,” PNAS 110, no. 35 (2013): 14432–37; and Jimo Borjigin et al., “Surge of Neurophysiological Activities in the Dying Human Brain,” PNAS 120, no. 19 (2023). The 2023 paper extends the rat-brain finding to two human comatose patients.
↑ 43. Nancey Murphy, Bodies and Souls, or Spirited Bodies? (Cambridge: Cambridge University Press, 2006); Joel B. Green, Body, Soul, and Human Life: The Nature of Humanity in the Bible (Grand Rapids: Baker Academic, 2008). For a thorough engagement, see Appendix C.
↑ 44. The argument that substance dualism predicts the data more cleanly than non-reductive physicalism is developed at length in Chapter 23; the philosophical case is made strongest in Richard Swinburne, Mind, Brain, and Free Will (Oxford: Oxford University Press, 2013); and J. P. Moreland, The Soul, chapters 4–6.
↑ 45. The argument that conditional immortality does not require physicalism — that a substance dualist can hold conditional immortality just as easily — is developed in John W. Cooper, Body, Soul, and Life Everlasting: Biblical Anthropology and the Monism-Dualism Debate, rev. ed. (Grand Rapids: Eerdmans, 2000), chapters 7–9; and is the standing position of the dissertation. I treat the question more fully in my forthcoming companion volume on conditional immortality and postmortem opportunity.
↑ 46. The methodological caution about the limits of NDE evidence — that it speaks to the boundary of death, not the long postmortem state — is developed in Friend, “Near-Death Experiences as Evidence for Substance Dualism,” chapter 6; and is consistent with the careful framing in Bruce Greyson, After, chapter 12.
↑ 47. The cardiac-arrest data are compatible with conditional immortality, conservative biblical universalism, and (less plausibly but not disprovingly) eternal conscious torment. The eschatological question is settled by biblical-theological argument, not by NDE empirical data. See further discussion in Chapter 30.
↑ 48. The Pam Reynolds debate — including the auditory-awareness counter-proposals by Keith Augustine and Gerald Woerlee, and the responses by Sabom, Greyson, and Holden — is reviewed in detail in Chapter 10; and in Janice Holden, “Veridical Perception,” in Holden, Greyson, and James, The Handbook of Near-Death Experiences, chapter 9. The relevant point for this chapter is methodological: the cumulative case does not depend on any single contestable case being beyond all challenge.