Chapter 12
When the Brain Goes Silent and the Person Keeps Speaking
I want to start in a hospital. Picture an intensive care unit at three in the morning. A woman in her sixties — a retired schoolteacher, three days out from a heart procedure — has just gone into ventricular fibrillation. The bedside monitor shrieks. The on-call team is already in the room. One nurse calls the code; another starts compressions; a respiratory therapist drops in an airway. The cardiologist scans the rhythm strip and reaches for the defibrillator paddles.
One of the monitors in that room — because of an ongoing research protocol — is a continuous EEG, a recording of the electrical activity at the brain's surface. (EEG stands for electroencephalogram, which just means "an electrical recording of the brain.") When her heart fibrillates, that EEG line does what it always does in cardiac arrest. Within seconds, the organized brain waves break apart. Within twenty seconds, they are gone. Within thirty seconds, the line is essentially flat.1
She is now, in every measurable medical sense, brain-silent.
The team works for nine minutes. Two shocks. Epinephrine. More compressions. On the third shock, the heart returns. They extubate her the next day. Two days after that, when she is well enough to talk, a research nurse asks her — gently, without leading — what she remembers about her arrest.
She remembers everything. She watched the team from a position near the ceiling. She knew which nurse was at her left side. She remembers what the cardiologist said about the second shock. She remembers a young chaplain who came in to pray with her husband in the family waiting area on the floor below — a chaplain her husband had never described to her and whose name she had not been told. The chaplain's name and timing check out. So does everything else.
Stories like this one are not isolated. They have been collected in prospective hospital studies, in five or six countries, for the last twenty-five years. They are the heart of this chapter.
I want to be honest about why this chapter sits where it does. We have already walked through cases of accurate distant observation in Chapter 10 and the powerful evidence from blind experiencers in Chapter 11. Each of those chapters carries its own weight. But cardiac-arrest cases are different in one crucial way: in cardiac arrest, we can measure what the brain is doing while the experience happens.11 We do not have to argue from inference to brain shutdown. We have the strip on the monitor.
That changes the conversation. If a person describes a hallway from above her body, the physicalist can say, maybe the brain was not as offline as you think. If a blind person describes a relative's clothing, the physicalist can say, maybe there are sensory channels we don't yet understand. Those moves are weaker than they look — Chapters 10 and 11 explain why — but they have a kind of plausibility. In cardiac arrest with EEG monitoring, the plausibility runs out. We can say what the cortex was doing. The cortex was doing nothing. That is why almost every serious challenge to physicalism from NDE research returns, again and again, to the cardiac-arrest cases.
The first crack in the wall — if I can put it that way — came from Pim van Lommel.
Van Lommel was not a parapsychologist looking for spectacular cases. He was a working Dutch cardiologist who had grown curious about what some of his cardiac-arrest patients told him after they were resuscitated. He decided to do something simple and rigorous: take every cardiac-arrest patient who survived in ten Dutch hospitals over a defined period, interview them within a few days of their arrest using a structured protocol, and see what proportion of them reported a near-death experience. No selection for "interesting" cases. No advertising for experiencers. Just the consecutive arrest survivors.
The result was published in The Lancet in December 2001 — one of the most respected medical journals in the world.2 Of 344 consecutive cardiac-arrest survivors, 18% reported a near-death experience, with 12% having what van Lommel scored as a "core" NDE.
That is the headline. But the methods are what matter. Van Lommel and his team carefully checked several of the alternative explanations physicalists had been offering for decades. Were the experiences caused by oxygen deprivation? If so, longer arrests should produce more NDEs — they did not. Were they caused by medications? If so, certain drugs should correlate with NDE incidence — they did not. Were they confabulations after the fact? Van Lommel reinterviewed the same patients two and eight years later, and the reports were stable to the point of word-for-word.
When The Lancet chose to publish that paper, the editors knew what they were doing. They were taking a field the medical establishment had long treated as fringe and putting it on the front shelf. The paper has been cited thousands of times.
A few years later, on the western side of the United Kingdom, a Welsh ICU nurse named Penny Sartori did something van Lommel could not: she watched the cases as they came in.
Sartori spent five years on a single intensive care unit, taking detailed notes on every cardiac-arrest patient who came through. She personally interviewed them after their resuscitations. She had access to the bedside data — what the team did, what was said, what was on the monitors. Her sample was much smaller than van Lommel's, but her depth of access was unusual.3
She found NDEs at roughly the rate van Lommel had found. She also found something else: when an NDEr described details of his resuscitation, those details could often be cross-checked against the team's contemporaneous records. In one case, a patient described a particular nurse leaning over him in a particular way during compressions, and he was right.
Sartori was the first researcher to do this kind of bedside, in-person verification systematically inside an ICU. Her cases are an important confirming line of evidence.
Now we come to the most ambitious project in the field: Sam Parnia's AWARE studies.
Parnia is a critical-care physician with appointments at the State University of New York and at NYU. He has spent most of his career trying to understand what happens to consciousness during cardiac arrest. AWARE — short for AWAreness during REsuscitation — is a multi-hospital prospective study, the first one ever attempted at this scale, designed to do something van Lommel could not do: build the verification check directly into the resuscitation room.
AWARE I, published in Resuscitation in 2014, ran in thirty-three hospitals across three countries over four years. Two thousand and sixty cardiac-arrest patients were enrolled. Of those, only 140 survived to be interviewed — cardiac arrest, as we saw in Chapter 5, is a brutal medical event. Of those 140 survivors, 9 reported what the protocol counted as a near-death experience.4
Two of those 9 had veridical content. One in particular has become well-known in the literature — a fifty-seven-year-old man who described, in striking detail, three minutes' worth of his own resuscitation, including the actions and words of specific medical staff, and who placed the events at a moment when the medical record confirms he was in cardiac arrest with no pulse and no spontaneous respiration. His descriptions matched what staff actually did. He should not have been able to know any of it.
Two veridical cases out of nine NDEs may sound small until you remember the design. AWARE was looking for cases verifiable prospectively, with documented timing, in real-time medical settings — and it found them.
One AWARE I participant, a 57-year-old man, described awareness during the third minute of his resuscitation. He watched the team work from a position near the ceiling, identified the nurse running the code, and reported a specific sound from the automatic defibrillator at a moment when the bedside monitor showed no spontaneous cardiac activity. The team's contemporaneous notes confirmed the sequence and the personnel. By any honest standard, this was a real-time observation made when his eyes were closed, his pupils were fixed, and his cortex was silent.
AWARE II, published in 2023 in the same journal, was bigger and more sophisticated.5 It used portable EEG devices and cerebral oximetry (a measurement of brain oxygen) on patients during their actual arrests, rather than relying only on the standard hospital monitors. That gave the team something extraordinary: data on what was happening inside the dying brain in real time, while the resuscitation team worked.
The AWARE II findings are still being absorbed by the field. Two of them matter for our conversation here.
First: about 40% of the survivors who could be interviewed reported some recall of their arrest, and within that group, a meaningful subset reported coherent NDE-like experiences with veridical content. Second: in a portion of the EEG recordings during CPR, the team detected episodes of high-frequency gamma activity — short bursts of rhythmic brain electricity — even after the heart had stopped and the standard monitor had gone flat. This was unexpected, and the physicalist community quickly seized on it. We will engage that gamma surge in detail in Chapter 15, where it belongs.12 For now, two things are worth saying: the bursts are sporadic, not sustained, and whatever they are, they cannot account for the content of the verified NDE reports — accurate descriptions of distant events the patient could not have witnessed even with a fully functional brain in the body's actual position.
AWARE II's gamma-burst finding shows that the dying brain is electrically more interesting than we used to think. It does not show that the dying brain is producing the verified veridical content. Brief gamma activity in a flat-line cortex, without organized cortical processing, is not "consciousness" in any sense neuroscience currently recognizes — and it is certainly not "an accurate visual perception of a chaplain praying with my husband on a different floor."
It would be one thing if all the cardiac-arrest NDE research were Western. It is not. An Iranian prospective cardiac-arrest study reported NDE incidence and core phenomenology consistent with the Western data — same out-of-body phase, same tunnel, same encounter with light, in a culture where Western NDE accounts had not been popularly disseminated. There are smaller confirming studies from India, Brazil, and parts of East Asia.6
What makes this important is what it rules out. The skeptic's first move when confronted with a Western NDE pattern is often, that's just American pop-religion working its way into people's hallucinations. The Iranian arrest survivors had no reason to be hallucinating American pop-religion. They reported what the AWARE survivors reported. The phenomenon does not appear to depend on which culture is hosting it.
When I sat down to do the quantitative analysis of cardiac-arrest cases for the dissertation, I expected a strong but limited result. What I found was stronger than I expected.7
Of the 5,278 cases in the full dataset, a substantial subset involved documented cardiac arrest, verified by medical records, witnesses, or both. Within that subset, I separated out cases where the timing of the conscious experience could be reasonably established — that is, whether the experience occurred during the period of measurable brain shutdown or only during peri-arrest periods (seconds before arrest or after return of spontaneous circulation). Of those timing-establishable cases, 89.96% had their conscious experiences during the documented no-measurable-brain-activity window.
Let me translate that. The strict physicalist prediction is that during this window, conscious experience should be impossible — or at most fragmentary, disordered, and incoherent. The data say the opposite. Conscious experience is not only reported during the window; it is the dominant location of the experience. When cardiac-arrest survivors say I had an experience, the overwhelming likelihood is that they are describing the period when their brain was offline.
A statistic like 89.96% can sound clinical. It is not. Behind that number are people. A Vietnam veteran whose heart stopped on an operating table and who, after his bypass, described a particular conversation in the surgeons' lounge that had taken place during his arrest. A young mother whose cardiac arrest came on the floor of her kitchen, who afterward told her sister exactly which neighbor came running across the lawn, what he was wearing, and what he said to her dying husband — none of which she could possibly have heard with her ears, and all of which her sister later confirmed.
There is one more finding from the dissertation worth pausing over here, because of how cleanly it cuts against physicalist expectation.
If consciousness is identical to brain function, then the worse the brain trauma, the worse the conscious report should be. People with severe traumatic brain injury show poorer recall, more confabulation, more confusion, more incoherence. The pattern is well-established in non-NDE neurology.
In NDE cases, the pattern reverses. When I sorted the cardiac-arrest cases by medical severity — duration of arrest, depth of physiological compromise, EEG documentation, multiple-organ involvement — and then sorted them by clarity and verifiability of the reported experience, the two scales correlated positively. The more severe the medical event, the clearer and more verifiable the report.
This is not what physicalism predicts. It is what dualism predicts. If the brain is the generator of consciousness, severe brain compromise should produce degraded experience. If the brain is the instrument of consciousness — the antenna, not the broadcaster — then severe brain compromise might affect the bandwidth of what the soul can communicate back to the body, but it would not necessarily affect the soul's own clarity. I think the data fit the second picture. The inversion is there in the numbers.
Because so much of this chapter rests on the EEG, I want to be careful about what it shows and what it does not.
Surface EEG measures the summed electrical activity of large populations of cortical neurons close enough to the scalp electrodes to register. When that signal goes flat — what clinicians call isoelectric — it means that organized cortical electrical activity at the cortical surface has ceased. It does not mean every electrical phenomenon in every neuron has ceased. There can be small, deep, brief activity that surface EEG would not pick up.8
The physicalist seizes on that gap: maybe consciousness is in some non-surface activity we are not detecting. The honest answer is: maybe it is, but the burden is now on the physicalist to identify what it is, generate a model that predicts what it would and would not produce, and explain why the cardiac-arrest NDE pattern looks the way it does. None of that has been done. "Maybe there is something we cannot detect" is not a model; it is a placeholder. And placeholders do not predict 89.96%, and they do not predict accurate observations of distant events the patient never witnessed.
The cortical literature itself is clear. The Aminoff and colleagues research from the late 1980s — induced cardiac arrest in patients undergoing internal-defibrillator implantation — showed that organized cortical activity stops within seconds of cardiac arrest, that loss of consciousness follows promptly, and that there is no measurable cortical activity during the arrest itself.9 The standard physicalist baseline is that there should be no consciousness in this window. That is the prediction the cardiac-arrest NDE data falsifies.
I should address the shelf targets, because skeptics raise them often.
AWARE I included a famous prospective design feature: visual targets — pictures placed face-up on high shelves in resuscitation rooms — visible only to a viewer up near the ceiling. The reasoning was straightforward. If NDErs are really observing from a position above the body, some of them should report having seen the targets.
In AWARE I, none of the survivors who reported NDEs had been resuscitated in rooms with the targets installed. Not zero hits — the design itself didn't have the chance.
This is sometimes reported as "AWARE found no evidence for OBE perception." That is a misreading. AWARE found two veridical cardiac-arrest cases with documented timing, including the fifty-seven-year-old man whose three-minute report we have already discussed. The shelf-target experiment did not have the sample size to test what it was designed to test. The veridical cases, on the other hand, were not in the experimental design — they came from the broader interview protocol, and they were positive. AWARE II is repeating the shelf design with a larger sample.
I want to set the two competing predictions side by side.
The strict physicalist prediction: consciousness is what the brain does. During cardiac arrest with documented flat EEG, the brain is not doing it. So no patient should report a coherent conscious experience from this window, and no patient should report verifiable observations from it.
The dualist prediction: the soul is a real, distinct constituent of the human person, and the brain is its primary instrument during embodied life. When the brain shuts down, the body becomes incapable of receiving the soul's signal back — but the soul itself continues. During cardiac arrest, then, the soul may continue to be conscious and to perceive while the body lies silent. After resuscitation, the soul re-engages the body, and the experience may or may not be communicated back to long-term memory depending on the integrity of the recovering brain.
The dualist prediction matches the data. The strict physicalist prediction does not. We will return to the larger metaphysical implications in Chapter 23; the point here is just the empirical fit.
Sophisticated physicalists know all of this. Most no longer hold the strict identity claim in its 1980s form. They hold what is sometimes called non-reductive physicalism: the view that mental states are realized by but not identical to brain states, and that there may be mental phenomena we do not yet have the neuroscience to explain.
Peter van Inwagen and Nancey Murphy are among the most thoughtful Christian advocates of this position. Murphy, in Bodies and Souls, or Spirited Bodies?, argues for a Christian physicalism that takes mental phenomena seriously without positing an immaterial soul.10 Their move is, in effect: yes, the strict prediction failed; perhaps consciousness is realized by mechanisms we have not yet identified. I want to grant this what it deserves. It is intellectually responsible. It is the position the strongest Christian physicalists actually hold.
"Maybe the brain is doing something during arrest that we just can't measure yet — and that something supports consciousness." This is the non-reductive physicalist's appeal. It deserves a hearing. But notice what it concedes: the original physicalist prediction has failed empirically, and the rescue depends on a future model that has not been developed and may not exist. Until that model arrives, the data look the way the dualist account predicts they will look.
Notice what has happened, though. The strict physicalist prediction has failed. The non-reductive response is not a positive prediction; it is an acknowledgment of that failure paired with an expectation that some future model will rescue physicalism. Whether that model is forthcoming is, at this point, a matter of faith — faith that the brain must be doing what physicalism needs it to do, even though we cannot currently see how.
The dualist account does not require that kind of faith. It rests on the older Christian conviction that the human person is more than the body — and it predicts what the data show.
Let me be careful at the end of this chapter not to claim too much.
The cardiac-arrest cases do not settle every question we will ask in this book. They do not, by themselves, tell us what kind of soul human beings have, what the soul encounters when it leaves the body, or what becomes of the unredeemed at the moment of death. Those are theological questions, and we will come to them in their own chapters.
What the cardiac-arrest cases do settle is more limited and, I think, more important than people often realize. Decades of prospective, peer-reviewed, multi-center research have established that human consciousness can persist during periods when the brain — by every measurable standard contemporary medicine can apply — is doing nothing. This is a reproducible pattern across decades, countries, and survivors who have nothing in common but the experience of having died and come back. Whatever the explanation turns out to be, it cannot be the simple identity claim that consciousness is what the brain does — because consciousness is reliably reported when the brain is doing nothing measurable.
For families standing at a bedside, watching a monitor flatten, this evidence is not abstract. It speaks to the question every grieving daughter and every anxious husband asks at some point: is she still in there? The medical answer is one thing. The deeper answer, the one the cardiac-arrest cases gesture toward, is another. The body's silence is not necessarily the soul's silence. The schoolteacher whose story opened this chapter was somewhere — not nowhere — during the nine minutes her EEG ran flat.
The chapters that follow build on this foundation. Chapter 13 takes us to encounters with unknown deceased relatives — cases that, like the cardiac-arrest cases, cannot easily be explained on physicalist principles. Chapter 14 brings the full cumulative case together. Chapter 15 engages the most sophisticated physicalist counterargument — Borjigin's gamma-surge research — at length.
But the heart of the empirical case is here. When the EEG goes flat and the patient still tells you what happened in the room, the body's silence has not been the soul's silence.
↑ 1. The standard timeline is well-documented in the cardiac-resuscitation literature: organized cortical electrical activity ceases within roughly ten to twenty seconds of cardiac arrest, and the EEG goes effectively flat within thirty to sixty seconds. For the empirical baseline, see Chapter 5 of this book and the references there to Michael J. Aminoff and colleagues' classic 1988 work on EEG during induced cardiac arrest.
↑ 2. Pim van Lommel, Ruud van Wees, Vincent Meyers, and Ingrid Elfferich, "Near-Death Experience in Survivors of Cardiac Arrest: A Prospective Study in the Netherlands," The Lancet 358, no. 9298 (December 15, 2001): 2039–2045. The longer popular treatment is van Lommel's Consciousness Beyond Life: The Science of the Near-Death Experience (New York: HarperOne, 2010), which engages the dualist implications more directly than the original journal paper could.
↑ 3. Penny Sartori, The Near-Death Experiences of Hospitalized Intensive Care Patients: A Five Year Clinical Study (Lewiston, NY: Edwin Mellen Press, 2008). Sartori's later popular synthesis is The Wisdom of Near-Death Experiences: How Understanding NDEs Can Help Us Live More Fully (London: Watkins, 2014).
↑ 4. Sam Parnia et al., "AWARE—AWAreness during REsuscitation—A Prospective Study," Resuscitation 85, no. 12 (December 2014): 1799–1805. For Parnia's longer popular treatments, see Erasing Death: The Science That Is Rewriting the Boundaries between Life and Death (New York: HarperOne, 2013) and his more recent Lucid Dying: The New Science Revolutionizing How We Understand Life and Death (New York: Hachette, 2024).
↑ 5. Sam Parnia et al., "AWAreness during REsuscitation—II: A Multi-Center Study of Consciousness and Awareness in Cardiac Arrest," Resuscitation 191 (October 2023): 109903. The portable EEG and cerebral oximetry data are a major methodological advance over AWARE I and have shifted the conversation about what the dying brain is and is not doing during CPR. The gamma-burst finding is engaged in detail in Chapter 15 below; the responsible reading is that the bursts are real but cannot account for the verified veridical content.
↑ 6. See the dissertation's literature review at Near-Death Experiences as Evidence for Substance Dualism within the Conditional Immortality Debate (Th.D. diss., Trinity College of the Bible and Trinity Theological Seminary, 2025), Chapter 2, for the cross-cultural cardiac-arrest research. The Iranian prospective study and additional smaller studies from India, Brazil, and East Asia consistently find the core phenomenology that van Lommel and Parnia found in Western samples. Surbhi Khanna's collaborative work with Bruce Greyson on cardiac-arrest NDE prevalence is also relevant here. For the broader cross-cultural picture, see Allan Kellehear, Experiences Near Death: Beyond Medicine and Religion (New York: Oxford University Press, 1996).
↑ 7. The methodology and full data are presented in dissertation Chapter 4. Briefly: cardiac-arrest cases were drawn from both the qualitative (832 cases from peer-reviewed scholarly sources) and quantitative (4,446 cases from NDERF and IANDS) sub-datasets. Timing was coded based on the available medical-record description, witness account, or experiencer's own placement of the experience within the documented arrest interval. Cases in which timing could not be reasonably established were excluded from the percentage calculation. The 89.96% figure is the percentage of timing-establishable cardiac-arrest cases whose conscious experiences fell within the documented no-measurable-brain-activity window rather than peri-arrest periods.
↑ 8. The standard caution about surface EEG is well-known to neurophysiologists and is acknowledged in van Lommel's Lancet discussion section. The point is not that surface EEG is uninformative — it is highly informative for cortical function — but that it is not the totality of brain activity. The physicalist appeal to undetected deep activity is logically possible; the question is whether such activity, were it present, would be of a character to support coherent conscious experience and accurate distant perception. There is currently no neuroscientific model that says it would.
↑ 9. Michael J. Aminoff and colleagues' work in the late 1980s on EEG during induced cardiac arrest in patients receiving internal-defibrillator devices remains the standard reference for the rapid loss of organized cortical activity following circulatory arrest. The key finding for our purposes: organized EEG activity disappears within seconds, and there is no measurable cortical electrical activity during the arrest itself. See also Bruce Greyson's review articles in The Lancet Psychiatry and elsewhere for surveys of the medical-NDE literature against this baseline.
↑ 10. Nancey Murphy, Bodies and Souls, or Spirited Bodies? (Cambridge: Cambridge University Press, 2006). For the broader Christian-physicalist position engaged constructively, see also Joel B. Green, Body, Soul, and Human Life: The Nature of Humanity in the Bible (Grand Rapids: Baker Academic, 2008). For the critical engagement these positions deserve, see Appendix D of this book and J. P. Moreland, The Soul: How We Know It's Real and Why It Matters (Chicago: Moody, 2014); John W. Cooper, Body, Soul, and Life Everlasting: Biblical Anthropology and the Monism-Dualism Debate, rev. ed. (Grand Rapids: Eerdmans, 2000).
↑ 11. For the methodological standard of "real-time veridical observation during documented brain shutdown" as the strongest test case for the dualism-physicalism debate, see Janice Miner Holden, "Veridical Perception in Near-Death Experiences," in The Handbook of Near-Death Experiences: Thirty Years of Investigation, ed. Janice Miner Holden, Bruce Greyson, and Debbie James (Santa Barbara: Praeger, 2009), 185–211. Holden's careful taxonomy of veridical-perception cases is the standard the AWARE design was attempting to operationalize prospectively.
↑ 12. The Borjigin gamma-surge research — Jimo Borjigin et al., "Surge of Neurophysiological Coherence and Connectivity in the Dying Brain," Proceedings of the National Academy of Sciences 110, no. 35 (August 27, 2013): 14432–14437, and subsequent pig-brain follow-ups — is the most-cited contemporary physicalist counter-evidence and deserves a chapter of its own. That chapter is Chapter 15. The brief note here is only that even granting the empirical reality of dying-brain gamma activity, such activity has not been shown to constitute consciousness in any well-developed sense and cannot account for the verified content of cardiac-arrest NDE reports.