Chapter 6

NDEs in the Blind

Vicki Umipeg has never seen anything. Not the face of her mother. Not a sunset. Not the color blue. She was born nearly four months early in 1950 and given pure oxygen to keep her alive. The oxygen saved her life and destroyed her optic nerves. From her first breath, her world was sound, scent, and touch.

Then in 1973, when she was twenty-two, she was thrown from a car after a crash on a Seattle road. Her skull was fractured. Her body was failing. As doctors and nurses raced to keep her alive in the emergency room, Vicki later said something extraordinary happened.

She saw.

She saw her own body lying on the table. She saw the gold wedding band on her finger. She saw the long brown hair she had never seen — and she saw a nurse cutting some of it away, which made her angry. She watched the medical team work. Then she rose through the ceiling and saw trees, and birds, and light.¹

If you are a physicalist — if you believe the brain is the whole story — Vicki's experience is a problem. A serious one.

A. The Critic's Argument

Michael Marsh saw the problem too. And he came prepared.

Marsh devotes several pages of his book to the research published by Kenneth Ring and Sharon Cooper in their 1999 book Mindsight. Ring is a respected psychologist and one of the founders of the field. Cooper is a careful researcher. Their book reports interviews with thirty-one blind people who had near-death experiences or out-of-body experiences. Twenty-one had NDEs proper. Of those, fifteen described some kind of seeing during their experience. Fourteen of the thirty-one were blind from birth. From among the congenitally blind, nine described visual perception during an NDE or OBE.²

Marsh attacks the research on three fronts.

First, methodology. He argues that Ring and Cooper used leading questions and accepted vague answers without pressing for clarification. When a blind subject described her vision during the NDE as fine or perfect or normal, Marsh wants to know what those words could possibly mean to someone who has never seen.³

Second, conceptual content. What does a tree mean to someone born blind? What does “gold” mean? Marsh suggests that blind subjects build internal pictures from what sighted people have told them — borrowed images, not real perceptions.⁴

Third, retraction. Marsh quotes follow-up comments from Vicki and from another blind subject named Brad. On a second interview, Vicki said it was hard to describe because it had not been visual in the ordinary sense. Brad said his experience was something like a tactile sense, since he had never known anything to compare it to. Marsh treats this as a near-retraction. He concludes that the alleged spiritualized form of sight is “entirely misconceived.”⁵

Then Marsh adds a body blow. He brings up the case of “Sarah” — the blind patient described in Larry Dossey's 1989 book Recovering the Soul. Sarah, supposedly blind from birth, allegedly saw the operating room in stunning detail during cardiac arrest. Sarah did not exist. Dossey eventually admitted he had constructed her from other cases. Marsh uses the hoax to suggest a general loss of confidence in the blind-NDE research enterprise.⁶

Fischer and Mitchell-Yellin take a different approach. They concede that blind NDEs have compelling appeal. But they offer a clever distinction. They split blindness into two kinds. They call the first no-input blindness, where the eyes do not work but the brain's visual processing might still be intact. They call the second processing blindness, where the brain itself cannot generate visual impressions. If a no-input blind person has visual experiences in an NDE, they argue, that might be explainable in physical terms — perhaps something like the visual hallucinations seen in Charles Bonnet syndrome, where some blind people experience strange visual scenes. Only processing blindness with veridical sight, they grant, would be hard for physicalism to explain.⁷

These are the arguments. Now we test them.

B. Identifying Weaknesses

Three weaknesses break the critics' case before it can land.

Start with the Dossey hoax. The fabrication is real, and it is shameful. The lesson is that a careless writer made up a story to dramatize a point. But notice what Marsh does with this. He pivots from a single fabrication by Dossey to general suspicion of all blind-NDE research. That is guilt by association. It is not an argument; it is a smear by adjacency. Dossey's hoax has nothing to do with Ring and Cooper's interview transcripts. It has nothing to do with Vicki Umipeg's case file. It has nothing to do with the medical records that backed up other cases. Saying it does is like saying that because a famous fraud once forged Vermeer paintings, no painting attributed to Vermeer can be trusted.

Worse, Ring and Cooper themselves caught a fabricated case before Mindsight went to print, added a published advisory disowning it, and named the researcher who had been deceived.⁸ That is what careful researchers do. The fact that they spotted and corrected fraud is evidence of integrity, not an indictment of the rest of their work. Marsh does not mention the advisory.

Second weakness: the “retraction” reading is selective. Vicki's actual statement, in full context, is not a denial of seeing. It is a struggle to describe seeing in a vocabulary she has never owned. Imagine being asked to describe the taste of wine when you have never tasted any food. You would say things like, “It's not really tasting — it's something more like color.” You would not be retracting. You would be reaching for analogies. Vicki had no internal library of visual concepts. The only language for sight she had ever owned was metaphors borrowed from other senses. When she said her experience was hard to describe because it was not visual, she meant: it was not visual the way you mean visual. It was something else. Brad's words follow the same pattern: he was contrasting his NDE with what he assumes normal vision must be like for sighted people.⁹ Marsh reads this as a near-confession. It is in fact a translation problem.

Third weakness: Fischer and Mitchell-Yellin's clever distinction does not survive contact with the data. Charles Bonnet syndrome produces hallucinations — random scenes, faces, animals, often in miniature — that have nothing to do with the patient's actual environment.¹⁰ They are not veridical. They do not match the room you are in. They do not include checking the gold ring on the finger or the make of the car involved in the crash. The blind NDEs we are discussing involve perceptions that match what was actually happening in the operating room or on the road. Charles Bonnet does not deliver veridical content. Random visual noise cannot accurately describe a real, specific event.

And the no-input/processing distinction misses something deeper. Even if Vicki had what Fischer and Mitchell-Yellin call no-input blindness, her brain has never received visual input — not once, in twenty-two years. There is no learned visual code. There is no “way of seeing” the brain has practiced. Even if her visual cortex were intact, what would it draw upon to generate accurate perceptions of a wedding ring it has never seen?

C. The Pro-NDE Response

Now to the cases themselves. Two will carry most of the weight: Vicki Umipeg and Brad Barrows. Both were blind from birth. Both saw during their experiences. And the details they reported track real events with embarrassing accuracy.

Start with Vicki. The 1973 accident left her in a hospital with a fractured skull and a damaged neck and back. She told Ring and Cooper that she suddenly found herself looking at her own body from above. She did not recognize the body at first. Then she saw the rings — a gold wedding band she knew by touch, and a plain gold band her father had given her, also identifiable by feel. She saw her hair. She watched a nurse cut some of it away as part of treatment. She was annoyed by this. She watched the medical team work on her body, hearing their conversations, but unable to make them respond to her.¹¹

Then she went up.

She passed through the ceiling, she said, and found herself outside. She saw trees. She saw birds and people. She saw light — overwhelming, beautiful, like nothing she had any frame of reference for. She entered that light. She met deceased childhood friends, including two girls from her school for the blind who had died young; in this place, Vicki said, they could see. She met a being she described as Jesus, who told her she had to go back. She came back into her body, and into pain.¹²

Now hear Vicki on her ordinary inner life. She told the researchers that she had never seen anything — no light, no shadows, nothing. And her dreams, she said, contained no visual impressions of any kind. Only taste, touch, sound, and smell.¹³

This matters more than people realize. Most adults blind from birth do not have visual dreams. Their dreams are auditory, tactile, and olfactory.¹⁴ Their brains do not generate visual content because their brains have never processed visual content. Vicki's dreams confirmed this for her own case. Sighted people sometimes assume the blind have a hidden visual life — that the brain is “really” seeing somewhere. The dream evidence shows otherwise.

So how did Vicki's brain produce, on this one day, a visual world that matched the actual emergency room? That correctly identified rings she could only know by touch? That correctly noted hair being cut? Pim van Lommel, the Dutch cardiologist who pioneered prospective NDE research, summed up the puzzle by noting that her observations could not have come from sensory perception, from a functioning visual cortex, or from imagination — given that there were verifiable parts.¹⁵

Now Brad Barrows. Brad was born premature, and like Vicki, was given oxygen that destroyed his sight. He has been blind from birth. In January 1968, at age eight, Brad developed pneumonia. His breathing stopped. As medical staff worked to revive him, Brad later said, he felt himself rise from his body, pass through the ceiling and the roof of the building, and stop above the street outside the Boston Center for Blind Children.

He looked down at the street. He saw fresh, deep snow. He saw a snowplow working its way up the road. He recognized a hill nearby that he used to climb. The scene was, he said, perfectly clear.¹⁶

Researchers later checked the weather records for that day. There had been a heavy snowstorm in Boston. Snowplows had been out. Brad's description matched the weather and the geography.¹⁷

Set this beside Marsh's reading of Brad's later interview comment. Brad did not say he had not seen. He said he had never had ordinary visual experience to compare it to. Of course he had not. He had been blind since birth. What he was reporting was a perception that, in his life, had no precedent — and which, when he later tried to describe it to a sighted interviewer, defied the borrowed vocabulary he had to use. The veridical content of what he described — snow, snowplow, hill, location — was checked. It was accurate.

Now step back and consider what these cases ask physicalism to explain.

Begin with the basic problem. The brain of a person blind from birth has not constructed a visual model of the world. The visual cortex, especially the primary visual cortex, gets recruited by other senses — touch processing, language, spatial reasoning — in those who are blind from birth.¹⁸ There is no library of visual experience to draw on. There is no template of what a wedding ring looks like, what a snowplow looks like, what a tree looks like.

The standard physicalist response to NDEs leans hard on the brain's ability to generate experience under stress. The dying brain hallucinates. The hypoxic brain dreams. A misfiring brain region produces an out-of-body feeling. In every one of those cases, the proposed mechanism uses materials the brain already has — sights it has seen, sounds it has heard, body schemas it has built. But for Vicki and Brad, the materials are not there.

An analogy may help. Imagine asking someone to describe a color they have never been told the name of, and which they have never seen, in a language they do not speak. The output, if it occurred at all, would be noise. Random words. Disconnected images.

That is not what we see in these cases. We see structured perception. Veridical perception. A wedding ring — identified, correctly, by sight, in a person who has only ever known rings by touch. A snowplow — described accurately, on the right street, on a day when snowplows were actually out. The precision is the problem.

Marsh tries to answer this by suggesting that blind people build internal pictures from sighted people's descriptions, then “see” those mental constructs during the NDE. But this misses the point twice over. First, those pictures, even if they exist, would be borrowed and approximate, not detailed. They would be a blind person's imagination of a sighted world, not the sighted world itself. Second, those imagined pictures would not happen to match the actual operating room, the actual rings on the actual fingers, or the actual snowstorm out the actual window. Veridicality is the wall.

Vicki and Brad are not isolated. Ring and Cooper documented twenty more cases, plus ten OBEs unrelated to a medical crisis. Around eighty percent of their respondents who had an NDE reported some form of seeing.¹⁹ These cases include patients who described medical equipment, conversations, the appearance of rooms, and the appearance of family members — perceptions they could not have built from prior knowledge or current sensation.

A fair-minded reader can ask, “But how strong are the verifications?” In Vicki's case, the medical record confirmed the emergency-room context, the hair cutting, and the procedures she described. In Brad's case, the snowstorm was on the public weather record, and the geography of the street and hill is fixed and checkable. These are not perfect, courtroom-grade chains. But they are real, multi-source confirmations that line up with the experiencer's description in a way that mere imagination cannot account for.

What we have, then, is a class of cases where (a) the experiencer's brain has never processed the kind of information being reported, and yet (b) the report contains accurate information about real, verifiable events. That is a serious empirical problem for any view that says the brain produces all experience.

The cumulative case stretches further. Beyond Vicki and Brad and the broader Mindsight sample, there are reports from across the field — Sabom, Long, van Lommel — of patients with damaged or impaired vision who described accurate visual content during their NDEs.²⁰ No single case settles the question alone. Together they point in one direction.

Notice what is being claimed and what is not. We are not claiming that the blind permanently see, or that NDEs cure blindness, or that consciousness “leaves” the body in some Cartesian science-fiction sense. The claim is more modest, and more interesting: during certain dying or near-dying states, conscious perception appears to operate without — and beyond — the sensory machinery of the body. That is the empirical signal. And it is consistent with substance dualism, the view that human beings are made of body and soul, which we develop in Chapter 24.

D. Counter-Objections

The strongest counter-objection runs like this. Maybe the blind, during an NDE, are constructing a pseudo-visual experience from non-visual cues — from sounds, vibrations, smells, and perhaps from sighted bystanders' commentary they overheard while unconscious — and then later remembering that construction as if it had been visual.

This is a real possibility worth taking seriously. But it does not fit the data.

Sound and vibration are not enough to identify which ring is on which finger by sight, or to specify the gold color of that ring. A sighted person standing at the bedside cannot generate a snow scene out the window for a child too sick to look. Auditory commentary — even if Vicki overheard it — would carry words about her treatment, not the visual fact of which color her ring was. The reconstruction theory works only if we already grant that the blind person has, somehow, the visual concepts that allow translation. They do not.

There is also a memory wrinkle. Several studies have asked whether NDE memories degrade or get embellished over time, the way ordinary memories often do. The answer, surprisingly, is no. NDEs are recalled with stable accuracy, in some cases over decades.²¹ If Vicki and Brad were merely confabulating, we would expect drift. We do not see drift.

A second counter-objection: Charles Bonnet syndrome. Some philosophers point to it as the natural-cause analogue. But Charles Bonnet hallucinations are well documented and strikingly different from veridical NDE perception. They are random. They are often Lilliputian — tiny figures, repeated geometric patterns. They do not match the patient's surroundings.²² Vicki and Brad reported the surroundings.

What we have established here is narrow but firm. People blind from birth, whose brains have never processed visual information, sometimes report visual perception during near-death experiences — and those perceptions sometimes match real events that can be checked. The brain-based explanations on offer either confuse fraud with research, mistake translation for retraction, or apply mechanisms (Charles Bonnet, no-input blindness) that do not deliver veridical content. The full timing question — whether such perceptions could occur during a brief window of recovering brain function — is taken up in Chapter 17. For now, the blind have shown us something worth pausing over. They have seen.