Chapter 7

What the Children Saw

At three in the morning, a fever broke. A nine‑year‑old boy named Eddie Cuomo opened his eyes in his hospital bed and started talking. He had been to heaven, he told his parents. He had seen Grandpa Cuomo and Auntie Rosa and Uncle Lorenzo. His father, embarrassed in front of the family physician, Dr. K. M. Dale, tried to wave the whole thing off as feverish chatter.

Then Eddie added one more name. He had also seen his big sister, Teresa.

Teresa was nineteen. She was four states away at college in Vermont. Eddie's father had spoken to her on the phone two nights before. Now the father grew agitated and asked Dr. Dale to sedate the boy.

That morning Eddie's parents called the college. The officials there had been trying for hours to reach them. Teresa had been killed in a car accident just after midnight — about three hours before Eddie sat up in his hospital bed and started talking.¹

Stories like Eddie's are why this chapter exists. Children's near‑death experiences are sometimes treated as a curiosity — a side note to the adult literature. They should not be. They are a separate strand of evidence, and in some ways the strongest strand we have. An adult who almost dies has spent decades soaking up cultural pictures of the afterlife from movies, novels, sermons, and dinner‑table talk. A nine‑year‑old has barely had time to soak up anything.

Michael Marsh, the careful Oxford physician we keep returning to, takes a different view. In his Chapter 7 he argues children's NDEs are dream‑states tied to brain maturation. We need to see whether his account holds up.

The Critic's Argument

Marsh treats children's NDEs in a section he calls "The Paediatric ECE/Dream Problem Revisited."² (ECE is his preferred label for what most people call NDEs — it stands for "experiences of clinical endangerment.") His argument runs in three steps.

First, the data. Marsh leans on the studies of pediatrician Melvin Morse and his team at Seattle Children's Hospital — the most important pediatric NDE work in print.³ Marsh fairly notes that Morse pursued the topic at real personal cost. He calls it a "tribute to his personal courage" that Morse kept going under criticism from colleagues.⁴ So Marsh and I agree on Morse's integrity. The disagreement is about what Morse's data mean.

Second, the comparison. Marsh emphasizes that children's NDE narratives are, in his words, far less composed and far less elaborate than the typical adult report.⁵ Children describe light and rainbows and "people in white" who they call God or Jesus, but they rarely give the adult‑style picture of an organized paradise. Only one child in Morse's main study saw a deceased relative; only one had a life review.⁶ This shorter, simpler quality, Marsh argues, is exactly what we would expect if NDEs are a kind of dream.

Third, the developmental link. Here is the heart of his case. Children's dream recall and dream content match their brain maturation. Under age eight, kids show "a paucity of substantial dream recall."⁷ After age eight, dream recall improves alongside what Marsh calls "visuo‑spatial competence" — the brain's growing ability to imagine layouts and bodies in space. The pediatric NDE corpus, he says, follows the same pattern: the older the child, the more elaborate the NDE. This parallel — dream development in lockstep with NDE elaboration — is for Marsh strong evidence that NDEs are essentially dream‑states with all that that entails.⁸

The philosophers John Martin Fischer and Benjamin Mitchell‑Yellin, Marsh's secondary partners in this critique, take a slightly different angle. In their Chapter 7 they argue that the similarity between adult and child NDEs does not, by itself, force a nonphysical conclusion.⁹ Children in hospitals, they note, are not blank slates. They have parents. They have grandparents who tell stories about heaven. Many of them have been to Sunday school. Some have just lost a family member and have heard adults talk about where she "went." A child's NDE may sound spontaneous, they argue, but the cultural soil was tilled long before.¹⁰

So we have two arguments to answer. Marsh says: children's NDEs are dreams scaled to brain maturity. Fischer and Mitchell‑Yellin say: children's NDEs aren't as culturally innocent as defenders claim. Both are serious arguments. Both are made in good faith. Both fail to account for what children's NDEs actually contain.

Where the Argument Cracks

Three problems sink Marsh's case before we even reach the strongest evidence.

The first problem is a confusion between what an experience is like and how a child describes it. Marsh notes correctly that pediatric narratives are simpler than adult narratives. From this he infers that the experiences themselves are simpler — closer to dreams. But a child's verbal range and a child's experience are not the same thing. A four‑year‑old who eats her first slice of birthday cake will not describe the flavor the way a pastry chef would. That doesn't mean she had a less rich taste. Children describe everything more simply than adults — the ocean, divorce, a thunderstorm, joy. To read the simplicity of the description as evidence the experience itself was thin is to make a category error. We will see the same problem in a different guise with the "vivid is not veridical" argument in Chapter 21.

The second problem is more decisive. The dream‑state hypothesis has nothing to say about veridical perception — the cases where a child accurately reports something he could not have learned by any normal means. Dreams do not give a child accurate knowledge of what his mother is cooking in another room. Dreams do not produce the details of a resuscitation that the dreamer never witnessed. Dreams certainly do not produce surprise meetings with a sister who died, hours earlier, in a state the child has never visited. Whatever else the dream‑state model explains — and it does explain some peripheral features — it stands silent before the cases that matter most. (Chapter 15 takes up Marsh's broader dream and REM‑intrusion arguments.)

The third problem applies to Fischer and Mitchell‑Yellin's cultural‑soil argument. It is an argument that could be true, but it predicts the wrong content. If children's NDEs were mostly drawn from absorbed cultural input, we would expect them to contain that cultural input. We would expect Sunday‑school angels with feathered wings, golden gates, harps, halos. What we find instead is something quite different. Children describing angels without wings.¹¹ Children encountering deceased relatives they had never been told had died. Children seeing details of medical procedures their parents could never have prepped them for. The cultural‑absorption account predicts a Hallmark‑card heaven. The actual data predicts something else.

There is also a methodological mismatch. Fischer and Mitchell‑Yellin, careful philosophers, work largely from a small set of famous cases and from Long's broader claims.¹² Marsh, careful clinician, works largely from Morse's books and a few critical reviews. Neither engages the full body of confirmed pediatric cases catalogued by Rivas, Dirven, and Smit in The Self Does Not Die, where the children's cases are not isolated anecdotes but a recurring evidential pattern.¹³ The cultural‑conditioning hypothesis sounds reasonable in the abstract. Held up against the actual cases, it doesn't fit the shape of the data.

These three problems clear the ground. Now the actual evidence.

What the Children Saw

I want to walk through two cases in detail and then say why, taken together, they tell us something the dream‑state model cannot accommodate.

Katie

Begin with Katie. She was seven years old, and she nearly drowned. Found floating face‑down in a swimming pool, she was rushed to the emergency room and resuscitated. Her CT scan showed massive brain swelling. Doctors gave her about a 10 percent chance of recovery. She lay in a deep coma for three days, hooked to an artificial lung, before recovering — astonishingly — with no permanent damage.¹⁴

The pediatrician who treated her, Dr. Melvin Morse, asked Katie a few days later to tell him what had happened. He expected to hear how she had fallen in. Instead, she told him about meeting Jesus and the heavenly Father. She told him about a guardian named Elizabeth. She told him about two boys she met in the experience — boys named Mark and Andy. And she told him she had been allowed, while out of her body, to look in on her family at home.

Three details made Morse stop and listen carefully.

First: when Morse interviewed the intensive‑care nurses, they reported that Katie's first words on regaining consciousness were, "Where are Mark and Andy?"¹⁵ She kept asking for them. The nurses had no idea who Mark and Andy were. Katie had encountered them, named them, and remembered them — before she had any chance to talk to her parents about the experience.

Second: Katie described, in detail, what her family had been doing while she was unconscious in the hospital. She described what her mother was cooking. She described the clothes various family members were wearing. She described her brother playing in his room with a GI Joe in a Jeep.¹⁶ Katie's mother — at home, miles from the hospital, with no contact during these hours — confirmed that all of these details were exactly right.

Third: Katie's family was Mormon. They had not exposed her to traditional Christian iconography. Her mother had once explained the death of Katie's grandfather using a gentle metaphor — that he had stepped onto a boat while the family stayed on the shore.¹⁷ No tunnels, no Jesus, no heavenly Father, no winged Elizabeth had been part of Katie's instruction. The cultural‑conditioning hypothesis predicts that Katie would dream Mormon imagery. She didn't. She reported something else.

Morse pursued the case carefully. He published it in the American Journal of Diseases of Children, then assembled a team of eight researchers — anesthesiologists, pediatric neurologists, hospital staff — and ran a multi‑year study at Seattle Children's Hospital.¹⁸ What he concluded after three years was as plain as it was incompatible with the dream‑state account. Morse said that medical training teaches doctors to look for the simplest explanation, and that after working through every other proposed account he had come to think the simplest explanation for NDEs is that they are glimpses into a world beyond.¹⁹

That is a doctor who stayed close to the data.

Eddie Cuomo

Now back to the boy with whom we opened. Eddie was nine years old when his fever finally broke and he reported a journey. His list of relatives in heaven would have raised no eyebrows except for one name — Teresa. Teresa, the nineteen‑year‑old college student in Vermont, had been killed in a car accident around midnight. The college had been trying to reach the parents at home. Eddie woke at three a.m. and reported seeing her. By the time the family received the call at sunrise, Eddie's experience had already been spoken aloud — first to his parents, then to Dr. Dale.

Researchers call cases like Eddie's Peak in Darien experiences — encounters with deceased persons whom the experiencer did not yet know had died. The phrase comes from a sonnet by John Keats and refers to the shock of cresting a ridge and seeing an entire ocean you had not expected. Here it stands for the same shock in reverse: the dying see someone on the other side, and the family later learns, with a chill, that the person has actually died. Chapter 9 handles the Peak in Darien pattern in deathbed visions and adult cases at length. Children's cases form a distinctive subset.

Bruce Greyson — Chester F. Carlson Professor Emeritus of Psychiatry at the University of Virginia — has catalogued more than two dozen Peak in Darien cases.²⁰ Some are historical. Among the most striking is a Victorian‑era case Greyson recovers: two childhood friends, Edith and Jennie, both eight years old, both with diphtheria. Jennie died on a Wednesday. Her family, knowing how close the girls were, kept the news from Edith. On Saturday at noon, Edith — still believing Jennie alive — picked out two of her own photographs to be sent to her friend. Hours later she lapsed into unconsciousness. That evening she briefly woke and said to her father, "Why, papa, I am going to take Jennie with me!" Then: "O, Jennie, I'm so glad you are here." She slipped away before morning.²¹

These cases do something the cultural‑conditioning argument cannot survive. A child cannot absorb cultural ideas about people who have died if no one has told her they have died. The information was not, in any normal sense, available.

Why these cases together tell us something

I'm not claiming Katie and Eddie alone settle the matter. Chapter 4 handles the broader veridical‑perception case. Chapter 31 assembles the cumulative argument. What pediatric cases add — and add powerfully — is the elimination of confounders that haunt adult cases.

Adult patients have lived long enough to read books, watch documentaries, or hear NDE reports from someone in their network. A skeptic can always say, "Yes, but they could have heard about it." Children, especially young children, cannot. Their possible cultural input is small enough that researchers can practically map it. When a Mormon seven‑year‑old reports an Elizabeth‑figure outside Mormon iconography, when a nine‑year‑old names a sister who has died three hours and several states away, when a Victorian eight‑year‑old greets a friend whose death has been concealed — these are signals strong enough to register against a small noise floor.

Researcher William Serdahely directly compared adult retrospective accounts of childhood NDEs against contemporary pediatric NDE accounts. He found them indistinguishable in core content, and found that pediatric NDEs were not affected by gender, religiosity, age within the sample, or other demographic variables.²² Children encounter the same realm as adults — they just describe it more simply. That finding alone is hard to square with Marsh's developmental dream‑state model. If pediatric NDE content tracked dream development the way Marsh proposes, the content itself should change with maturity. It doesn't.

There is one more piece. Children encountering NDEs are also not typically told they are dying. Hospital staff and parents go to lengths to keep that information away. Yet in deathbed‑vision research close to NDE territory, dying children frequently know — and quietly tell the staff — that they are about to leave.²³ The Yale pediatric oncologist Diane Komp, who began her career a religious skeptic, ended it convinced that something real was happening to the children she cared for, after watching enough dying four‑year‑olds calmly say goodbye to nurses they had no business knowing they would never see again.²⁴ Chapter 9 develops the deathbed‑vision evidence at length.

Two children, decades apart, both knew things they had no normal way of knowing. That is the place where the dream‑state model begins to look less like an explanation and more like a story we tell ourselves to keep comfortable.

Two Counter-Objections

The strongest counter‑objection runs like this. Children may have absorbed more cultural content than we think. Death is everywhere — in cartoons, in storybooks, on television. A child who has heard a relative pray, "May Grandma rest in peace," has been given a story about heaven. The cultural‑conditioning argument does not require Sunday school; it just requires that a child have some concept of an afterlife. And almost every child does.

There are two responses, one general and one specific. Generally: Chapter 20 takes up the cultural‑conditioning argument as it applies to NDEs across the board. The short version is that NDE content frequently contradicts what we would expect from cultural absorption. The specific response is what we have already laid out: in the cases that matter most — the Peak in Darien cases, the verifiable‑perception cases, the cases where the content ran against the child's instruction — there is no plausible cultural source for the information conveyed. The argument doesn't fail because cultural exposure is impossible. It fails because cultural exposure cannot deliver the specific items being reported.

The second counter‑objection involves children's suggestibility. Researchers like Stephen Ceci and Maggie Bruck have shown that small children can be led, by careless interviewers, to "remember" events that never happened.²⁵ Could pediatric NDE reports be confabulated — the products of leading questions or unintentional coaching? Could parents and doctors be seeding the very stories they later report?

Two facts cut against this. First, in the strongest cases — Katie, Eddie, the eleven‑year‑old boy whose detailed resuscitation account Morse documented²⁶ — the children's reports were given immediately upon regaining consciousness, before parents and staff had reason to lead the witness. Second, the reports include details no adult was in a position to suggest. A nurse cannot lead a seven‑year‑old to ask, "Where are Mark and Andy?" if the nurse has never heard of Mark and Andy. The verifiable elements work as an objective check that suggestibility cannot dissolve. (Chapter 18 handles the memory‑and‑confabulation objection in full.)

Children's NDEs do not stand alone. They sit beside the dentures case, the AWARE study findings, Pam Reynolds, Vicki Umipeg's blind NDE — all the strands of a cumulative case made in Chapter 31. But pediatric NDEs are unusual in this: they remove the very confounders that make adult cases easy to dismiss. They are the part of the evidence that ought to make any fair‑minded skeptic stop and reconsider. Marsh's dream‑state model is elegant. It deserves the careful treatment we have given it. But when a nine‑year‑old wakes up and names a sister he could not have known was dead, "dream" is no longer a sufficient word.