Chapter 13

Out-of-Body Experiences and the Temporo-Parietal Junction

A woman lies on a hospital bed. Sixty-four electrodes have been implanted under her skull. The neurologists are mapping her seizures before surgery. They send a small electrical pulse into a patch of brain tissue called the temporo-parietal junction, or TPJ — a junction where the temporal lobe meets the parietal lobe, just above and behind the right ear. The patient suddenly feels herself sinking. Then floating. Then she says, with her eyes wide open, that she sees herself from above. But only her legs and lower trunk. The image flickers and twists. Her legs seem to shrink. Then they rush toward her face and she ducks.¹

That experiment is the centerpiece of one of the most influential skeptical arguments against near-death experiences. The claim is bold and clear. Out-of-body experiences are not glimpses of a soul leaving the body. They are glitches — brief malfunctions of a brain region that normally tells us where our bodies are. If a neurosurgeon can switch on an OBE with an electrode, the argument runs, then OBEs are clearly products of the brain. Period.

I want to take this argument seriously. Olaf Blanke, the Swiss neuroscientist whose laboratory has done much of this work, is a careful scientist. Michael Marsh, who features Blanke's findings in chapter 6 of his book, is right to bring them into the conversation. But when we look closely — and ask the questions Blanke's lab did not ask — the argument breaks apart.

A. The Critic's Argument: The Brain Builds the Body

Marsh's case begins with a basic point from neuroscience that nobody disputes. Our sense of being located in our bodies is not a brute given. The brain builds it. Different streams of information — what our eyes see, what our muscles report, what the inner-ear balance organs tell us — have to be stitched together. That stitching happens largely in the multimodal association cortex of the parietal lobe and the nearby TPJ.² Pull on one of those threads, and the whole tapestry can shift. Marsh thinks the OBE is exactly that kind of shift. Pull the thread — through epilepsy, migraine, stroke, or a surgeon's electrode — and the patient feels herself rising, floating, looking down on her own body.

The flagship case is the 43-year-old epileptic woman I just described. Blanke's team published her in Nature in 2002. With low-current stimulation, she felt herself "sinking into the bed" or "floating" two meters above it. With higher current, she saw herself from above, though only the lower part of her body, and she reported that her body parts changed shape and size. When she opened her eyes, she even ducked away from her own legs, which seemed to fly toward her face.³ Marsh treats this as a striking demonstration that focal disturbance of the TPJ alone can produce something OBE-like.⁴

Two years later, Blanke's group expanded the study with six more patients. Four had complex partial seizures. One had a vascular lesion. One had a temporo-parietal lesion and a history of migraine. Across these very different cases, Marsh notes, one feature stood out. Their pathology clustered in or near the TPJ.⁵ One patient with right-sided ischemic damage felt himself rise up in his chair, see his "double" still seated below, and even watch his wife from both vantage points at once — a phenomenon called heautoscopy, the seeing of a second self. He felt unusually happy and weightless.⁶ Another patient, with left fronto-temporal-insular damage, slowly tipped backward in his chair until he believed he was now behind the nurse adjusting his IV.⁷

From these scattered cases, Marsh draws a sweeping conclusion. If a damaged or stimulated TPJ can make a perfectly conscious person feel out of her body, see her own body from a distance, or even view her surroundings from above, then we already have a sufficient brain-based explanation for the OBE that often accompanies the NDE. The body-schema is a cortical construction, period. Pathology can scramble that construction. NDE-OBEs, on this telling, are simply more dramatic versions of the same scrambling. Marsh even suggests that a "blinded" reader, set to compare these epileptic narratives with NDE narratives, might struggle to tell them apart.⁸ The case rests on three claims: TPJ activity correlates with body-self processing, TPJ stimulation can produce OBE-like experiences, and so OBEs are products of the TPJ.

The argument has been extended in two directions since Marsh wrote. Blanke's own group has shown that non-invasive transcranial magnetic stimulation over the right TPJ can briefly disrupt healthy volunteers' ability to imagine their bodies from a different perspective.⁹ And in 2007, two separate teams — Ehrsson in London, and Lenggenhager and Blanke in Lausanne — used virtual-reality goggles and synchronized touch to give healthy subjects a brief, mild illusion of being "in" a body other than their own.¹⁰ These were widely reported in the press as scientists having "induced" out-of-body experiences. Skeptics treat these studies as further support for Marsh's conclusion. We will see in a moment that they actually deepen its problems.

B. Identifying Weaknesses

Stack those three claims and the structure looks impressive. Pull them apart and it wobbles. Three problems matter most.

First, Blanke's induced experiences are not what NDE patients report. Read the descriptions carefully. The Swiss patient sees only her legs and lower trunk. Body parts shrink and balloon. Limbs fly at her face. Her perception of her own body is fragmented, distorted, cartoonish. Compare that to a typical NDE-OBE. The patient hovers near the ceiling. She sees the entire room from a stable, coherent vantage point. She watches the resuscitation unfold over many minutes. She hears conversations clearly. Her descriptions check out against medical records. Pim van Lommel, in the very Lancet paper that put NDE research on the global map, made the point sharply: experiences induced by drugs, hypoxia, or electrical stimulation "are not identical to NDE."¹¹ Bruce Greyson and other clinicians have said the same for decades.¹² A flickering image of your own legs is not the same phenomenon as a long, lucid out-of-body episode in an operating room. To call them the same thing requires blurring the very details that make NDEs evidentially interesting.

Second — and this is the key point — Blanke's lab never tested for veridical perception. They never asked whether their patient could accurately describe events she could not have seen by ordinary means. They never put a target above her in the room. They never asked her to read what was on the bedside chart while she "floated" two meters up. They never tested whether she could perceive what was happening in the next room, or down the hall, or in the parking lot below. Chris Carter has hammered this point in his work on NDE skepticism, and he is right to do so.¹³ The 2002 Nature paper is two pages long. It contains rich phenomenology and zero veridical verification. So even if we grant everything Blanke's lab observed — that the TPJ can produce a feeling of being elsewhere, with distorted self-imagery — we have learned exactly nothing about whether the TPJ can produce the thing that makes NDE-OBEs evidentially explosive: accurate perception of real-world events from the reported vantage point.

Third, even where Marsh's epileptic patients reported "veridical" content, the meaning of that word shrinks dramatically on inspection. The migraine patient who saw his wife while floating above his chair was already in a room with his wife. He could see her with normal vision before the aura started.¹⁴ Seeing her again from a slightly different mental angle is not a paranormal event. It is recombination of immediate sensory memory. None of these patients reports knowing things he could not have known by normal sight, hearing, or inference. NDE-OBEs routinely do.

C. The Pro-NDE Response

Here is where the argument turns. The thing the TPJ hypothesis cannot reach — cannot even approach — is the body of evidence that NDE patients perceive real events outside the field of any sense organ they own. We treat the strongest cases fully in Chapter 4 and Chapter 5. Here I want to draw the line that no purely TPJ-based account can cross.

Take the well-known dentures case from van Lommel's Dutch hospital. A man arrives in cardiac arrest, blue, unresponsive. The nurse removes his dentures and places them in a drawer of a crash cart. A week later, when the same nurse approaches him on a cardiac ward, he recognizes her instantly. He tells her where to find his dentures — in that drawer, in that cart, in the room where he was resuscitated. He describes the resuscitation team, the layout, what was said.¹⁵ Now ask the TPJ hypothesis to do its work. The patient was unconscious. His brain was in arrest. There were no eyes open, no working ears. Even if his TPJ had somehow produced a vivid sense of floating, by what mechanism could a misfiring TPJ tell him which drawer the dentures were in? The TPJ is a body-self processor. It does not have eyes that see through walls.

Or consider Maria's tennis shoe, the case that pulled hospital social worker Kimberly Clark Sharp into NDE research in 1977. A woman in cardiac arrest later reported floating outside the building and seeing a tennis shoe on a third-floor window ledge — with very specific details about the shoe, including a worn spot at the toe. Sharp, doubtful, went and looked. The shoe was there, exactly as described.¹⁶ Bring the TPJ hypothesis to bear on that. A scrambled body-schema is supposed to have located a particular shoe on a particular ledge of a particular floor of a hospital the patient had never been in before? The hypothesis collapses on contact with the data.

Pam Reynolds, treated fully in Chapter 5, gives us an even sharper example. While her body lay with eyes taped, ears blocked by speaker-emitting molded plugs, body cooled to the temperature of an ice bath, blood drained, brain stem unresponsive, EEG flat, she later described the bone saw used on her skull, its case, and a comment made by a cardiac surgeon — all confirmed by surgical records.¹⁷ A TPJ explanation for that case is not strained. It is impossible. The auditory cortex was being actively jammed. The eyes were taped shut. Whatever a TPJ does, it does not see through tape and hear through 100-decibel clicks while a brain stem is silent.

Or take a case from The Self Does Not Die: a Spanish patient cared for by Dr. Manuel Sans Segarra reported not only her own resuscitation but the surgery happening in an adjacent operating room — a femoral neck fracture being repaired by orthopedic surgeons she had never met, at the very minute her heart had stopped.¹⁸ Her TPJ, on the skeptical theory, somehow acquired information about a procedure happening in another room, performed by people she had never seen, while her own brain was in cardiac arrest. The hypothesis is asked to do work it was never built for.

And there is a feature of NDE-OBEs that the stimulation studies cannot match in principle. NDErs almost universally describe their experience as more real than ordinary waking life, not less. A 2013 study comparing memories of NDEs to memories of real events and imagined events found that NDE memories scored higher than real memories on phenomenological measures of clarity, detail, and self-referential content.¹⁹ Blanke's stimulated patient knew her experience was being triggered by an electrode. The VR-induced "OBEs" of Ehrsson and Lenggenhager were rated by the subjects themselves as mild, brief, partial illusions — nothing like the hyper-real, life-changing quality NDErs report. To equate the two requires not just a small leap but the deliberate erasure of the most distinctive psychological feature of the NDE.

Now consider the patient population. NDEs occur in roughly eleven to eighteen percent of survivors of cardiac arrest.²⁰ If the TPJ produces them by malfunctioning under physiological stress, why does the malfunction occur in only a small minority? Every cardiac-arrest patient experiences the same dramatic loss of cerebral blood flow. Every one of them suffers the same disruption of cortical function, including in the TPJ. Yet most report nothing. The cases that do occur are remarkably similar across patients, cultures, and centuries — we treat that consistency in Chapter 22. A random brain glitch should produce a wide spread of incoherent fragments, not a stable, recognizable phenomenology with veridical content. Selectivity plus consistency is not what dysfunction looks like.

And then there is the population of NDErs themselves. The patients in Blanke's studies have demonstrable TPJ pathology — tumors, lesions, ischemic damage, epileptic foci. NDE patients, by contrast, are almost always neurologically normal before the cardiac event that nearly kills them. Many of them go on to live for decades with no further OBE. Their brains scan as healthy. If the TPJ "produces" their OBEs, it does so in a structurally normal brain, under conditions where the cortex is in fact dramatically less active than in Blanke's awake epileptic patients. The skeptical model thus implicates a brain region in producing a coherent, accurate experience precisely when that region is least equipped to do so — and only in a small, unpredictable subset of the people in whom it should reliably fire.

There is a second issue the TPJ hypothesis cannot avoid: timing. During cardiac arrest, the cortex shuts down within seconds. EEG traces flatten. Cortical metabolism collapses. The TPJ is part of the cortex. If we take the model seriously, then at the very moments NDErs report rich, structured, accurate perception of their environments, the TPJ has no measurable capacity to generate anything at all, much less a coherent virtual reality with verifiable details. The full timing problem is owned by Chapter 17, but the implication for the TPJ hypothesis is straightforward. A brain region cannot manufacture an experience while it is offline.

Finally, there is the issue of what these correlation studies actually show. Set aside the question of veridical perception for a moment. Suppose Blanke's lab is right that TPJ stimulation can produce a partial, fleeting sense of being out of one's body. What does that prove? It proves that the TPJ is involved in our normal sense of bodily self-location. That is not surprising. Of course there is brain activity associated with having a body. There is brain activity associated with everything we experience. From the fact that disrupting the TPJ disrupts the felt location of the self, it does not follow that the TPJ generates the self. To assume that step is to confuse correlation with production.

Here is an old analogy that still works. Cut the cable to a television and the picture goes dark. That does not prove the cable was producing the picture. The cable was transmitting it. The picture was generated elsewhere. A great many philosophers and not a few neuroscientists, going back at least to William James, have argued that the brain may stand in roughly that relation to consciousness — not as producer but as filter or transmitter.²¹ I do not need to settle that debate here, and I will not. Chapter 23 handles it directly. I only need to point out that Blanke's data are equally consistent with both models. They favor neither. The skeptical reading smuggles in a metaphysics it has not earned.

D. Counter-Objections

Two pushbacks deserve a response.

"You are dismissing rigorous neuroscience." I am doing nothing of the sort. Blanke's research is a real contribution to our understanding of how the parietal cortex contributes to body-self awareness. The work is careful and the findings are real. What I am dismissing is the leap from those findings to a sweeping claim about NDE-OBEs. There is no warrant in Blanke's papers for the inference that all OBEs are TPJ-produced illusions. The papers describe a partial, distorted, body-self disturbance in conscious epileptic patients. They do not even attempt to address veridical perception, distant perception, or reports during cardiac arrest. Treating those papers as a refutation of NDE evidence is asking them to carry a weight they were never built to bear.

"Maybe the TPJ produces both fragmentary stimulation experiences and coherent NDE-OBEs." Maybe so. But now the burden has shifted, and it is heavy. The skeptic must explain how a brain region whose normal function is to integrate body-self signals could, when its inputs are catastrophically failing, generate not a more severe distortion but a more accurate picture of the world — including events the patient could never have witnessed by ordinary means. That is the reverse of what brain dysfunction usually does. The hypothesis is not impossible. It is just ad hoc. It is built backward from the data it is meant to explain, with no independent evidence and no testable prediction.

So we set the TPJ argument down where it belongs. It contributes a piece to the neuroscience of normal embodiment. It does not explain veridical NDE-OBEs. The next chapter takes up Marsh's related move — the phantom-limb analogy — and shows why it fares no better.