Chapter 15

The Dying-Brain Hypothesis

Borjigin, the Pig-Brain Studies, and Why a Thirty-Second Surge of Gamma Activity Is Not the Explanation the Headlines Said It Was

A Headline Cycle, Three Times Over

In August 2013, the Proceedings of the National Academy of Sciences published a paper from Jimo Borjigin’s laboratory at the University of Michigan. The title was technical: “Surge of Neurophysiological Coherence and Connectivity in the Dying Brain.” The findings were striking. Nine rats had been put through experimentally induced cardiac arrest while their brains were continuously monitored on EEG — the device, common in hospitals, that measures electrical activity at the surface of the brain through small sensors placed on the scalp. Within roughly thirty seconds of cardiac arrest, before the EEG line went flat, the rats’ brains showed a brief, intense burst of gamma-band activity — the very high-frequency rhythms (above twenty-five cycles per second) that neuroscientists associate with active conscious processing in awake animals.¹

The press response was immediate, and unsubtle. Outlets ran articles claiming that scientists had finally found, in dying rat brains, the neural mechanism behind near-death experiences. The mystery, finally, had been solved. It was the language of breakthrough.

It wasn’t, but the headlines mostly stuck.

Nine years later, in 2022, a German team reported a striking single case — an eighty-seven-year-old patient who unexpectedly went into cardiac arrest while his brain was being monitored on EEG following surgery for a subdural hematoma (a serious bleed underneath the membrane that surrounds the brain). His EEG showed elevated gamma activity in the moments around death. A case report. Interesting, suggestive, but a sample size of one.²

A year after that, Borjigin returned. The 2023 PNAS paper described four comatose ICU patients whose ventilators had been withdrawn after their cases were determined medically futile. Two of the four — not all four — showed surges of gamma activity in the dying interval, particularly in regions of the back of the brain implicated in conscious processing.³ The headlines repeated the cycle. Coverage spoke of hints of fleeting consciousness, the science behind why people see lights at death, and signs of awareness at the moment of dying.

This chapter is the slow correction the careful literature has been writing all along.

What Is the Hypothesis, and What Is at Stake?

The dying-brain hypothesis is, in its current form, the strongest contemporary physicalist response to NDE research. The claim is straightforward: as the brain dies, it briefly produces heightened consciousness, and NDEs are this brief heightened consciousness. The proposed mechanism is gamma synchrony — a final, paradoxical surge of organized high-frequency electrical activity, normally associated with rich conscious states. Borjigin’s rat studies, the 2022 single-case report, the 2023 four-patient study, and (so the popular framing claims) the pig-brain research from Yale together form what some commentators describe as the empirical foundation for an entirely physicalist account of near-death phenomenology.

I take the hypothesis seriously. Of all the physicalist explanations I have examined in my doctoral work, this one comes closest to engaging the actual NDE evidence. It is not the temporal-lobe hand-waving of the 1970s. It is not the now-failed REM-intrusion proposal of Kevin Nelson. It is real laboratory work, carefully conducted, peer-reviewed, partially replicated across species. The question is whether it does what the headlines claim it does.

Three questions structure the chapter. First — what does the research actually show, when read carefully and not through the press release? Second — does the proposed mechanism explain the actual phenomenology of NDEs, including the parts that matter most evidentially? Third — where does the hypothesis genuinely succeed, and where does it fail?

The short answer, given here for those who want the destination before the journey: the hypothesis succeeds in showing that some dying brains do something unexpected before they go quiet. It fails — comprehensively — at explaining the central evidential category in NDE research, which is veridical perception during clinical death. The careful literature has been making this case for a decade. The chapter makes it again, with the data updated.

The Borjigin Findings, Carefully Read

The 2013 PNAS paper deserves a careful read. The experimental design was elegant. Nine rats received continuous EEG monitoring before, during, and after experimentally induced cardiac arrest. The researchers tracked changes in power across frequency bands, coherence between brain regions, the directional flow of activity, and the coupling between slow and fast rhythms.

The finding: within roughly thirty seconds of cardiac arrest, before the EEG went isoelectric (the technical term for “flat”), a transient surge of synchronous gamma oscillations appeared. The gamma activity was global — present across cortical regions. It was highly coherent — meaning the regions oscillated together. Front-to-back directed connectivity increased. Phase-coupling with slower theta and alpha rhythms grew tighter. In short: just before the brain shut down for good, there was a brief, organized burst of activity that looked, in some respects, like the activity associated with conscious processing in awake animals.

What the 2013 paper claimed was modest. The authors interpreted their findings as evidence that the dying mammalian brain is not necessarily as electrically quiet as had been assumed — that some neural correlates of conscious processing might persist briefly during the perimortem (around-death) interval. They did not claim the rats had near-death experiences. Rats cannot tell us. They did not claim that their findings produce accurate distant perception, lucid narrative recall, or any of the specific features that make human NDE evidence probative.

The bridge to humans came in two steps. In 2022, Vicente and colleagues published a single case report — an eighty-seven-year-old man whose EEG was being monitored for seizure activity after neurosurgery when he unexpectedly died. The recording captured elevated gamma activity in the perimortem window.² A case report — interesting, suggestive, but again a sample size of one and complicated by the patient’s underlying neurological injury.

The 2023 Borjigin paper used a more controlled design. Four comatose ICU patients, all of whom were going to die, all of whom had family members consenting to continued EEG monitoring as life support was withdrawn. EEG was recorded continuously through the dying process. Of the four patients, two showed surges of gamma power, gamma-band coupling with slower rhythms, and increased connectivity in the posterior temporo-parieto-occipital region — an area sometimes called the posterior “hot zone” for conscious processing. The other two showed no such surges.³

The 2023 finding is interesting, and limited. Two of four under specific circumstances is a minority pattern, not a universal one. All four patients had pre-existing brain damage from prior cardiac arrests; the two with gamma surges had documented seizure histories. Peter Fenwick, the British neuropsychiatrist and longtime NDE researcher, observed soon after publication that the gamma activity in these two patients might reflect terminal seizure-like activity rather than near-death consciousness.⁴

The Pig-Brain Studies: A Useful Correction

The chapter title mentions “the pig EEG studies.” This phrase deserves careful unpacking, because it is one of the places the popular framing has run furthest from the data.

The Yale BrainEx study (Vrselja and colleagues, 2019, published in Nature) restored cellular activity to pig brains four hours after death. The Andrijevic et al. follow-up in 2022, called OrganEx, extended the perfusion technology to the whole pig body. Both papers are remarkable. Both are frequently invoked, especially in popular discussions of consciousness and dying, alongside Borjigin’s gamma-surge findings — as if they had shown something similar.

They did not. Stefano Daniele and Zvonimir Vrselja, the lead authors of the BrainEx work, were emphatic on this point in every interview and in the paper itself: at no point did they observe organized electrical activity associated with perception, awareness, or consciousness. In fact, the perfusion protocol was specifically designed to suppress any potential return of organized activity, on ethical grounds. What BrainEx and OrganEx demonstrated is genuinely remarkable but different from what the popular framing implies. They showed that brain tissue retains substantial cellular and metabolic potential after the heart stops — that the cascade of cellular death is more reversible than scientists had assumed. They are studies of tissue resilience. They are not studies of dying-brain consciousness.⁵

If anything, when we read the pig-brain literature carefully, it cuts against the dying-brain hypothesis rather than supporting it. These studies show that even when cellular function is restored hours after death, the brain does not spontaneously generate organized electrical activity. There is no evidence of any consciousness in those pig brains. The most thoroughly studied post-mortem mammalian brains in the scientific literature show, when carefully measured, the absence of consciousness, not its surprising presence. The pig is not on the side of the popular framing.

Why the Hypothesis Fails to Explain the Data

Now to the heart of the matter. Even granting everything Borjigin has demonstrated — and an additional charitable margin for what future research might discover — the dying-brain hypothesis fails to account for the actual NDE phenomenology in five distinct ways. Each one matters; the cumulative weight is what the next two chapters build on.

(1) Content. The Borjigin mechanism is local. Gamma synchrony in a person’s brain cannot, by any principle of neuroscience known or proposed, produce accurate information about events occurring in distant rooms. Yet the strongest NDE evidence (the cases I develop in Chapter 10) consists of precisely such information. Maria’s tennis shoe on the third-floor ledge. Pam Reynolds’s accurate description of the bone saw and the conversation in the operating room. The 1,114 cases of accurate distant observation in the dissertation database. None of this is in any way affected by gamma surges in the experiencer’s own dying brain. The brain is local; the information is non-local. The veridical content is what the dying-brain hypothesis simply does not address.

(2) Consistency. Random or chaotic dying-brain activity should produce random or chaotic phenomenology. The phenomenology found in well-documented NDE reports is anything but. It is structured and consistent — across cultures, across decades, across patient populations, across age groups. The nine-point core pattern (developed in Chapter 9) is robust and replicated. A neuropathological account predicts cross-cultural variation, idiosyncratic content, dream-like incoherence. The data show structure, not noise.

(3) Quality. Dying brains, in every other context where we have careful data, do not produce lucid, organized, sequential, narrative experience. Patients revived from near-drowning, near-suffocation, anoxic emergencies, or partial cerebral ischemia rarely describe coherent stories. They describe fragments, confusion, paranoia, brief flashes. By contrast, NDE reports are characteristically clear, organized, sequential, and morally significant. Nigel Shaw’s 2024 critique of the gamma-band model in F1000Research presses precisely this point: gamma activity in known clinical contexts — terminal seizures, ketamine intoxication, DMT — does not generate the phenomenology actually reported by NDErs. The hypothesis predicts the wrong character of experience.⁶

(4) Blind NDEs. This is, in my judgment, the decisive failure. A brain whose visual cortex has never developed normal patterns of visual processing cannot, on any plausible mechanism, produce accurate visual perceptions through a brief surge of gamma synchrony. Yet congenitally blind NDErs (the cases in Chapter 11) consistently report visual content that matches independently verified events. Vicki Umipeg’s case alone is enough to falsify the strict version of the dying-brain hypothesis. The brain has no visual template to hallucinate from. Whatever Borjigin’s gamma surges may do in some dying brains, they cannot do this.⁷

(5) Anesthesia and barbiturate cases. Anesthetics suppress cortical activity, including gamma. Barbiturate sedation suppresses gamma. The hypothesis predicts that NDEs should be rare or absent under deep anesthesia and barbiturate-induced cortical silence. The evidence shows otherwise. NDEs occur in these contexts at rates broadly comparable to other medical settings — and in the most carefully documented cases, occur during periods when the EEG was confirmed flat by direct measurement.⁸ The hypothesis predicts the wrong outcome in precisely the cases where the medical record is most thoroughly documented.

What Borjigin Herself Says

It is worth noting — and this is rarely acknowledged in the popular framing — that Jimo Borjigin herself does not claim that her work explains NDEs. The 2023 paper concludes with admirable epistemic restraint: the authors note that “the mechanisms and physiological significance of these findings remain to be fully explored.”⁹ The data show that the dying brain can still be active under some conditions. That is not a claim that NDEs have been explained.

The pattern, then, is consistent: a striking experimental finding, then eager journalistic overstatement, then careful clinical caveats from the authors themselves, then slow correction in the peer-reviewed literature. The careful version is what we are tracking in this chapter.

What the Hypothesis Might Legitimately Explain

Now I want to be fair. The dualist account I am defending throughout this book does not require NDE phenomenology to be entirely independent of brain physiology. It requires the experiencer to be a soul, not a brain. These are different claims, and the difference matters.

Some basic phenomenological features of NDE experience may indeed be co-produced by dying-brain physiology. The sense of heightened vivid awareness — gamma synchrony is associated with rich conscious processing in awake humans, so a final burst of synchrony could plausibly intensify the subjective vividness of whatever the soul is then perceiving. The sense of ego-loss or boundary-dissolution — the brain’s normal suppression of certain self-models may relax under anoxic stress, contributing to the characteristic loss of body-anchored selfhood reported in NDEs. Some life-review-like memory access — limbic structures, the brain regions involved in memory and emotion, may briefly hyper-activate as the brain’s normal inhibitory systems shut down. Each of these contributions, if real, is a real contribution. The dying brain may be co-shaping the experiential textures.

This is fully compatible with substance dualism. The dualist position I argue throughout this book holds that the body is the soul’s instrument during embodied life. As the instrument fails, the experiences had through it change in character. Some phenomenological features may be partly artifacts of the failing instrument; others may be the soul’s own perceptions emerging into the foreground as the body’s ordinary input recedes. The dualist account does not require us to peel the experience apart into “purely brain” and “purely soul” components — that overstates what either substance dualism or NDE evidence can deliver. What dualism requires is that the experiencer be the soul, not the brain.¹⁰

Now to the deeper philosophical move. Some commentators — Susan Blackmore in her writings, Christopher French in interviews, the popular-science press almost reflexively — have framed the Borjigin findings as “NDEs explained.” This is a category error.¹¹ A partial mechanistic account of some phenomenological features is not an explanation of the full phenomenon.

Compare. Suppose we fully characterized the auditory cortex’s neural correlate of musical experience. That would tell us something about how the brain processes music. It would not tell us that music is “just brain activity” — music is also melody, structure, meaning, emotional resonance. The correlate is a piece of the explanation, not the explanation.

Borjigin may have characterized a neural correlate — possibly — of some perimortem subjective texture. He has not thereby explained the veridical content, the cross-cultural consistency, the blind NDEs, or the anesthesia cases. The phenomenon to be explained is much larger than the slice the hypothesis addresses.

Assessment

Step back. What can we conclude?

What Borjigin has shown, carefully read: some dying brains do something unexpected during the perimortem window. The 2013 rat data is reliable. The 2022 single-case human data is suggestive. The 2023 four-patient study found gamma surges in two of four cases — under specific circumstances of prior brain damage and seizure history. The gamma activity is real. The interpretation is contested.

What Borjigin has not shown — and does not claim to have shown — is that these surges produce the actual reported NDEs; that they explain the veridical content; that they explain the cross-cultural consistency; that they account for blind NDEs; that they account for the lucid, narrative quality; or that they predict NDEs in the medical contexts (anesthesia, barbiturates) where they actually occur. None of that has been demonstrated.

The Yale pig-brain studies, properly read, are even more limited as evidence for the dying-brain hypothesis. They show cellular and metabolic resilience hours after death — a remarkable finding, with potential applications in stroke and trauma medicine — but they explicitly find no organized electrical activity associated with consciousness. They are tissue-resilience studies. They do not support the popular framing in which they are sometimes invoked.

The hypothesis at its best is a partial mechanism for some neuro-correlates of NDE phenomenology, fully compatible with the dualist instrument-account I have been developing across the book. The hypothesis at its worst — in headlines and confident dismissals — is a category error parading as a paradigm shift.

The cumulative empirical case for substance dualism (developed across Chapter 10 through Chapter 14) is unaffected by Borjigin’s findings. Its center of gravity — accurate veridical perception during clinical death, accurate visual content from congenitally blind NDErs, accurate identification of deceased relatives the patient did not know had died — lies entirely outside what the dying-brain hypothesis can address. The strongest evidence has not been touched.

Two things to note before closing. The pattern of striking finding → eager headline → slow correction is methodologically instructive: the careful version of the story is recoverable for any reader willing to read past the press release.¹² And the deeper point — the dying-brain hypothesis, even at its strongest, does not eliminate the dualist account. It at most suggests that the dying instrument may produce some experiential textures of its own. The dualist account predicts and accommodates this. The strict physicalist account, by contrast, predicts that the strongest NDE evidence should not exist. It does. That asymmetry is the ongoing argument.¹³

Why This Matters

Why does any of this matter, beyond the seminar room?

For the pastor sitting with a family at a deathbed, the news cycle will at some point claim that scientists have explained near-death experiences as a final brain surge. Care is required. The careful literature does not say this. The headlines do, and the headlines are wrong — or at least vastly outrunning their evidence. A pastor who has read the actual papers is in a position to honor a family’s reports of a loved one’s deathbed experience without pretending it has been refuted by the latest PNAS publication.

For the family member who has heard about a loved one’s NDE — the grandmother who described seeing her deceased husband in the room, the father who reported hovering above the operating table — this matters too. The experience is real. The veridical content is unaffected by the latest gamma-band paper. Whatever grandma saw is not explained away by Borjigin’s rats.

For the dying person who reads a popular article about brain surges and wonders whether the rich Christian theology of dying — Christ meeting the soul at the moment of departure, the conscious intermediate state, the encounter with the Light — has now been rendered obsolete: it has not. The science has not done what the headlines claim it has. The classical Christian theology of dying has, if anything, gained empirical confirmation across the past five decades of NDE research. The Borjigin findings do not change that.

For the Christian physicalist in the conditional-immortality movement — the position I have engaged with respect throughout this book — this matters most directly. The dying-brain hypothesis cannot rescue physicalism from the NDE evidence. The veridical cases stand. The blind cases stand. The anesthesia cases stand. The dualist account remains, by a substantial margin, the most empirically adequate.

The next chapter takes up the cluster of pharmacological hypotheses — anoxia, anesthesia awareness, endogenous DMT, ketamine-like dissociation. Chapter 17 is the cumulative refutation, gathering the residual physicalist hypotheses (REM intrusion, temporal-lobe stimulation, cultural conditioning) and showing that the same evidential pattern repeats across all of them. The pattern, in the end, is the argument: every physicalist hypothesis fails on the same evidential point, and the failure is not an accidental gap but a structural feature of trying to explain a non-local, soul-level phenomenon with local, brain-level mechanisms.

The brain matters. The brain is a magnificent, intricate, biologically wondrous instrument. Its dying is a real loss. But the brain is the soul’s instrument, not its source — and Borjigin’s rats, his comatose patients, and the Yale pig-brain perfusions have not, in the end, shown otherwise.