Near-death experiences: testimony and the limits of explanation

Epigraph

Let not him who chooses first be careless, and let not the last despair.

Plato, The Republic, X, 619b, the myth of Er; Benjamin Jowett translation, based on the third edition of 1888. A complete sentence of counsel about choosing within a philosophical myth, not a promise about medical outcomes.

What this article is about

Accounts given after resuscitation can profoundly affect how people understand their lives. Some describe peace, complex images, encounters, or a reconsideration of their actions; others report no such experience. How can we listen while understanding the limits of research? We distinguish a clinical event, a later recollection, an independent test, and an explanatory hypothesis. Our grounds are the 2001 study by van Lommel and colleagues, AWARE II from 2023, and the selected myth of Er in Plato’s Republic. These sources have different purposes and provide no combined proof or refutation of an afterlife. Observations deserve accurate description; stronger interpretations need further grounds. The comparison helps us keep patient groups separate, avoid attaching any account to any brain recording, and distinguish personal importance from conclusions binding on others. The practice concerns reading and respectful conversation, without danger, altered treatment, or seeking another episode.

I. What happened and what was reported

The expression “near-death experience” brings together an event, an account, and existing ideas about death. Careful reading can separate them. A person may have suffered cardiac arrest, undergone resuscitation, and later reported an unusual experience. A study can describe the circumstances and content of that report. It does not automatically obtain a recording of every moment of inner life. “During,” “after,” and “in connection with” therefore make different claims. When an experience’s timing is available only through a later account, that boundary belongs beside its most striking description, rather than disappearing as the account becomes more vivid.

Cardiac arrest followed by successful resuscitation in these studies is also different from established irreversible brain death. The researchers studied patients, some of whom survived and could participate in interviews. We cannot rename the entire group as people returned after final death and then present that description as a finding. For our question, keeping the actual study conditions is sufficient; we offer no independent diagnosis of death. The conditions establish a serious clinical crisis and the possibility of a subsequent account. They provide no ready answer to every question about consciousness and existence, however meaningful those questions may be.

An account can convince its narrator regardless of how precisely external circumstances have been established. Recognising this differs from agreeing with a causal explanation. A listener can acknowledge an event’s importance without confirming every interpretation. A researcher can preserve a report without claiming its origin is settled. These positions require no humiliation. They clarify which questions the evidence addresses and why personal certainty cannot replace an independent check of timing, content, or cause. Respect and precision can remain part of the same conversation.

Nobody must have such an experience in order to live a meaningful life. The absence of an account may involve unavailable memory; in a particular case, its reasons can remain unknown. We do not call people without these recollections less open or less spiritually developed. This is our contemporary position, rather than the result of testing patients’ spiritual qualities. It helps readers avoid competition over unusual experiences and allows a person to leave silence unexplained. Published sources are sufficient for this comparison. Ordinary life need not become the material for a dangerous independent experiment designed to obtain a more impressive account.

II. What the 2001 study collected

Van Lommel and colleagues included 344 successfully resuscitated patients at ten Dutch hospitals. Sixty-two reported experiences classified by the researchers as near-death experiences; 41 described experiences meeting a stricter category. These figures represent 18 and 12 per cent of that group. The article also reports 509 resuscitations: some people experienced more than one event. A person and an episode are different counting units here. Changing the denominator without explanation changes the percentage and creates a misleading impression of either contradiction or a general rate among all people. The number needs its population and definition to remain useful.

Most initial interviews took place within the following days, although the interval before interview had a wider range. Prospective inclusion describes a specified recruitment pathway after a clinical event. It does not turn a subsequent account into a direct recording of consciousness at an exact second. The design allows comparisons of groups and the conditions under which reports were obtained, while preserving limits involving memory, interview availability, and observation. We neither call it ideal nor dismiss it because it cannot settle every question. Its actual task should remain visible in a summary, especially when a striking narrative attracts more attention than recruitment.

Follow-up at two and eight years allowed the authors to discuss reported changes in life. However, the available group became smaller through deaths, refusals, and revisions to some original categories. Data from remaining participants can be described without automatically becoming the history of everyone initially included. This was no random allocation to different ways of transforming personality, and no trial of a therapeutic procedure. An important life account retains research value while the causal conclusion about what produced each change remains narrower than an appealing claim about the transformative power of experience. Attrition belongs beside the reported changes, not outside their interpretation.

In the Discussion, the authors offer a strong philosophical interpretation and challenge the sufficiency of ordinary physiological explanations. These statements should be attributed as their argument, separately from the observations collected. The absence of an established association with some investigated conditions does not exclude every possible mechanism. This study obtained no continuous EEG recording of every moment to which an account might refer. We therefore do not turn it into direct confirmation of consciousness during established absence of all brain activity. The mechanism remains an important question precisely where the data still provide no conclusive answer to it.

III. AWARE II: different parts of one project

AWARE II combined a prospective hospital study, an additional investigation of survivors’ accounts outside that group, and a physiological monitoring substudy. These parts complement one another but answer different questions. The hospital study included 567 cases; 53 people survived to discharge and 28 could complete interviews. Recruitment occurred during working hours under specified inclusion conditions. The interviewed group cannot therefore be presented, without qualification, as the whole population of cardiac-arrest survivors. The pathway from a clinical event to an available account already introduces a substantial limit, before any interpretation of the experiences themselves begins.

Eleven of the 28 interviewees reported memories or perceptions in a broad sense; six fell within the category the authors call recalled experience of death. Categories in the thematic analysis could overlap. They should not be added as though each person necessarily belonged to only one. The project distinguishes reports concerning medical procedures, later emergence from coma, dreamlike images, and other experiences. This distinction preserves the material’s variety. It does not establish that every account of one kind has the same ultimate cause. Readers need both the category and the counting rule when interpreting the numbers attached to these reports.

The additional part included 126 people already reporting recollections. It helped describe themes more fully, but supplied no further independent denominator for overall prevalence. We cannot simply add these accounts to the hospital group and calculate the frequency of a single phenomenon among all patients. For each part of the project, readers can place its name beside the route into it and its analytical purpose. Rich descriptions remain valuable without performing the work of random selection that this part did not provide. The difference in purpose becomes part of the finding, rather than an inconvenience to remove from the summary.

The study also shows why interview conditions matter. People could be too unwell to participate, and interviews depended on recovery. Researchers used open questions and a specified scale, but received subsequent reports. The ability to tell an account becomes a condition for inclusion in the analysis of accounts. It differs from the presence or absence of any experience during the clinical crisis. We infer neither a final theory of unconsciousness from silence nor demonstrated continuity of consciousness. Data availability constrains both broad conclusions, not merely the one a reader happens to dislike. Keeping this symmetry helps prevent uncertainty from becoming selective evidence.

IV. Testing with an independent stimulus

A separate AWARE II procedure used predetermined visual and auditory stimuli. Its purpose differed from collecting meaningful accounts: a response could be compared with independent material. Reading the full article requires preserving the distinction between explicit recollection and selection in an implicit-learning test. No participant explicitly recalled the presented image or auditory stimulus. Nobody selected the required image, while one selected the correct fruit combination. This last response cannot be renamed as confirmed out-of-body visual perception. The result concerns the task actually administered, with its available participants and limitations, rather than a different test imagined by a later summary.

A brief retelling can quietly remove the procedure’s name. “The correct answer was selected” then begins to sound like “the person certainly saw a hidden picture.” Reading Methods and Results together helps prevent this shift. We need the question, presentation method, response format, and limitation of the small available group. The authors themselves note how low survival constrained testing implicit learning. We retain that boundary instead of making one response an overall success for the protocol. It also provides no proof that all accounts are false or that no afterlife exists. A limited test answers its specified question.

The researchers separately compared some reports with clinical records. A record confirming that a shock was delivered clarifies a particular external circumstance. It differs from successful recognition of an independent image and does not automatically establish how the person acquired the information. These forms of corroboration can be useful when their identities remain clear. In the earlier study, the nurse’s account concerning dentures likewise has the status of a separate testimony, rather than a blinded test of a predetermined target. A story’s vividness cannot replace the described procedure with whatever stronger evidence we would prefer to have obtained from it.

An ordinary invented example makes an independent ground clearer. Someone remembers a bell ringing during a meeting. A recording confirms the bell, without explaining when or how the recollection formed. A different question—distinguishing several sounds saved beforehand—needs a corresponding test. Calling both tasks “confirmation” does not combine them. This is our original example, not a patient’s story. It shows how a specific agreement retains value without becoming a universal ground for further conclusions about memory or consciousness. An ordinary comparison clarifies the question without supplying an explanation of a clinical experience.

V. What can be linked to a brain recording

The physiological part of AWARE II obtained EEG and cerebral oxygenation data under difficult resuscitation conditions. Usable material was not available for everyone. The article identifies 85 participants with EEG data and 53 with interpretable recordings. Movement artefacts and technical difficulties constrained the material. We have not reanalysed the original signals and do not diagnose individual states from a published image. For our question, the available subgroup, number of segments, and people providing accounts need to remain distinct. These sets cannot be declared identical without checking which measurements and reports actually belong together within the project.

Among the 28 interviewees, two had EEG data, and neither reported explicit recollections. This is a crucial limit on individual comparison. We cannot take a vivid account from one part of the study and a physiological segment from another participant, then claim that the specific content was recorded at a specific moment. A common project makes such questions investigable without supplying an absent data pair. Readers can first ask whether the account and measurement concern the same person and event, and what establishes their timing, before judging the causal explanation proposed for the apparent relationship between them.

The EEG procedure used brief periods that reduced interference; it was no uninterrupted recording of every inner thought. The authors report some physiological patterns and propose associations with possible conscious activity. Recognising a pattern, however, does not read an account’s content or determine an experience’s meaning. The limitations retain questions about different interpretations of alpha activity and the need for fuller testing. In our account, this uncertainty must survive an attractive phrase about a brain that has finally explained or refuted every report. Physiological observation gains clarity when the remaining interpretive step is visible, rather than presented as already completed.

The resuscitation conditions must also remain explicit. Measurements obtained while care was being provided do not permit every record to be described as a brain after the definitive ending of all function. These studies offer no procedure for readers to repeat independently. We consider clinical methods only insofar as they clarify the grounds of a conclusion. Curiosity about experience never justifies creating danger or delaying assistance. The action boundary is clear: published material can be studied without another hazardous state or an attempt to obtain personal proof where preserving a life is the immediate concern. Reading needs no reenactment.

VI. Where observation becomes explanation

In AWARE II, the authors discuss possible restoration of coordinated activity and different approaches to explaining consciousness. They also explicitly acknowledge that its mechanism and relationship with brain function remain unresolved. Our key quotation belongs in this context. Strong expressions about clarity or new dimensions of reality remain interpretations by the researchers, rather than independent proof of an entire ontology. We can attend to their hypotheses while asking which specific part of the data distinguishes one proposed explanation from another. The question preserves the observations; it asks whether the explanatory transition has obtained the further grounds it requires.

Establishing a cause requires work beyond discovering an association. Even when two types of data are related, questions remain about event order, collection conditions, and alternatives. If no individual pair exists, the limit is narrower still. This applies both to a preferred brain-based explanation and to a claim that consciousness exists independently. We select no final cause on the basis of which expression sounds more reassuring. The distinction preserves a hypothesis as a hypothesis and prevents confident presentation from being treated as completed testing of all competing explanations. Confidence in language cannot supply an observation absent from the design.

An unknown cause does not automatically confirm any particular cause. Physiological investigation also does not require declaring personal meaning empty. A person may understand relationships and life’s finitude differently while the research question remains open. These conversations complement one another when their boundaries are named. An account of meaning does not replace a timed recording; a timed recording cannot decide how someone will live afterwards. Our comparison retains differences between empirical description, religious faith, and philosophical explanation. They can meet without becoming interchangeable forms of evidence.

We therefore announce neither proof nor refutation of an afterlife from the works considered. This is a limit of our conclusion, not a prohibition on philosophical positions. Source authors may take stronger positions; we name them and examine the transition from data to interpretation. Readers may likewise hold faith or doubt. What matters is not presenting someone with a final research finding where only one possible understanding of the material has been offered. Agreement and disagreement can then concern the actual grounds. They need not depend on an article title silently expanded into a claim its methods never tested.

VII. The myth of Er: another genre and purpose

In Book X of the Republic, Socrates tells of Er returning with a message about another world. Judgment, consequences of actions, choices of future lives, and forgetting shape the narrative. We read the complete selected tale while retaining its ethical setting. It is a philosophical myth within a dialogue, rather than a hospital record with confirmed resuscitation conditions. We use the identified Jowett translation and acknowledge the note about the composite electronic transcription. An ancient author’s name does not make a literary narrative an independently verified clinical case or allow us to diagnose its protagonist in contemporary medical terms.

Comparison with research accounts can begin with reconsidering life and relationships with others. A shared theme, however, establishes no common cause and confirms no complete description of the world portrayed in the myth. Plato’s story directs attention towards choice and justice; a contemporary project systematises subsequent accounts obtained under specified conditions. These purposes deserve to be named before images are compared. The historical text then retains substance, while clinical research avoids becoming merely an appendix to a previously chosen picture of an afterlife. That picture was not independently tested simply because a later account contains a related moral theme.

The epigraph belongs to counsel about choosing lives. Within the tale, a person’s position in the order of choice removes neither the need for attention nor the possibility of hope. It promises no recovery and explains no difference between one patient’s survival and another’s death. We do not use the myth to blame someone for a clinical crisis or assess their moral worth. Applying its language to responsibility for ordinary decisions is our own contemporary step. It must remain identified so that ethical language does not acquire the force of a medical causal proof that the comparison has never supplied.

At the tale’s close, Er cannot describe the means by which he returned. This literary boundary invites comparison between a narrative and a mechanism, without turning the text into an experimental admission of error. An author can construct a philosophical story that conveys significance without explaining every transition. A research article must state what was measured and how. Neither genre should be judged merely as an unsuccessful copy of the other. The comparison helps when it clarifies their purposes and preserves the text’s ethical importance without presenting it as a contemporary collection of clinical data. Difference is part of the connection.

VIII. Life afterwards and ownership of an account

An experience may be important, difficult, ambiguous, or accompanied by no available account. We require no single response. Research materials contain different categories and themes; readers should not extend the brightest example to everyone. An attentive conversation begins with whether a person wants to discuss the event at all and what help or form of listening they choose. This contemporary position concerns respect, rather than claimed therapeutic effectiveness. The studies did not test our conversational exercise and supply no ready instruction for transforming personality. Listening can acknowledge variety without requiring another person to fit the example that most impressed us.

Imagine someone wanting more time with loved ones after a serious event. A listener could help discuss a concrete step without requiring unusual details or a general conclusion about death. Personal importance becomes a ground for the person’s own decision, without deciding for others how they should live. An invitation retains the possibility of refusal or another time. We thus distinguish felt significance, a chosen action, and mutual agreement. This is an original example, not a research participant’s story. It shows the ordinary scale of responsibility for a later action, rather than pretending to establish the source of the original experience.

Changes in relationships can be discussed independently of a final theory of the event. People can report what changed, what remained difficult, and which expectations were disappointed. These observations broaden material about consequences without automatically answering the cause of the initial experience. The earlier study’s follow-up was likewise no randomised trial of a method for obtaining a good life. Both scales matter: consequences deserve attention, but they do not turn every interpretation of an experience into a confirmed causal model. Importance and evidential strength need not coincide completely for an account to remain worth hearing and a question worth investigating.

Ownership of an account includes leaving some experience private. An unpublished record proves no fabrication; missing material also supplies no confirmation of external events. When a question concerns a shared decision, the available action can be discussed without requiring intimate disclosure. When it concerns a published conclusion, researchers can identify the grounds’ limitations. These forms of respect are compatible. They let us attend to a person while remaining precise about a formulation, using neither scientific uncertainty to humiliate nor a person’s vulnerability to impose an untestable general demand. Privacy leaves some questions unanswered, rather than answering them through speculation.

IX. What precise reading can contribute

The comparison yields several separate points. Accounts were collected after a clinical crisis; categories were described; independent tests provided no general confirmation; physiological data have their own subgroup and limitations; mechanisms remain under investigation. None should be replaced by a single victorious phrase. Precision does not drain the subject of interest. It shows what is already available and which next question needs further grounds. The historical myth retains its ethical content in this conversation, without being added to the patient count or the number of successful independent tests. Its contribution has another form, and deserves to be described accordingly.

Interest in consciousness’s mystery can coexist with attention to ordinary conditions of knowledge. Expressive language matters alongside recruitment, interview timing, comparisons, and links between a project’s different parts. Readers need not independently recalculate every datum to notice that two groups differ. They can check a claimed transition against the article’s description. This makes discussion more substantial because disagreement concerns the argument’s actual structure, rather than an overall impression that science has confirmed or rejected everything connected with near-death experiences. A smaller, accurately stated question may advance discussion further than a sweeping answer whose grounds have disappeared during retelling.

An open question obliges nobody to abandon personal faith. Faith, however, cannot be presented to others as an independent research test already completed. A preferred physiological hypothesis must likewise keep the status warranted by its data. These boundaries permit discussion by clarifying its level. Meaning, method, and the ultimate nature of consciousness can be discussed separately. Disagreement in one question need not end the possibility of respect or useful conversation in another. Our proposal is limited to this discipline of reading and responsibility for words. It offers no guarantee that every participant will adopt the same philosophical position afterwards.

Reading can end with a clear statement of what remains unknown. There is no need to seek another crisis or change sleep, treatment, or ordinary conduct for the sake of proof. Consciousness deserves investigation without making life an instrument of hazardous testing. Respect for a survivor also requires no announcement of a final metaphysical truth we do not possess. Preserving the account, naming its conditions, and limiting the conclusion leaves room for personal meaning and further work. This is a modest outcome of comparison, without promising universal agreement or completion of inquiry into the nature of consciousness.

What this means in practice

Choose a published study or an account made voluntarily available. Write down one claim and identify its task: experiential content, an external event, frequency, a consequence, or a cause? Keep one question for the first reading. Mark participants’ words, measurement results, and authors’ interpretations separately. This practice diagnoses nobody and tests no spirituality; it works with an existing text rather than seeking new intimate information.

Find who was interviewed and who supplied an independent recording. Do the groups match? How are interview and measurement times specified? Does a number count patients, events, recordings, or selected accounts? Keep different denominators separate unless grounds justify combining them. If a particular report lacks a link to the same person’s signal, preserve that gap. A large common project does not create the pair by itself. No personal data need to be collected for this exercise.

Check the corroboration procedure’s name. Was it explicit stimulus recollection, selection among options, a clinical record, or another person’s later testimony? Retain the result of that particular check, including failure and limitations. Repeat no clinical intervention independently and create no danger to obtain another account. If the conversation concerns someone you know, ask their agreement and purpose for discussing it. Declining to disclose an experience proves no falsehood.

State the outcome without an extra transition to a general cause. Separately name what remains unknown and the personal significance someone reported. That significance can be respected without requiring shared faith or happiness. The exercise ends when the claim’s structure becomes clearer; it promises no accurate recovery of memory, therapeutic effectiveness, or answer about an afterlife. Uncertainty is acceptable as a named limit of available material, rather than an invitation to take risks in order to eliminate it.

Key quotation

The mechanism of consciousness and its relationship with brain resuscitation and function remain undiscovered.

Sam Parnia and colleagues, AWARE II, 2023, Discussion, page 9. A complete sentence expressing the explanatory uncertainty discussed by the authors. Hypotheses and causal testing retain different statuses.

Connections between traditions

Two grounds and locators	Substantive connection	Boundary
AWARE II, Methods / Results, pages 2, 6–7; van Lommel, Methods / Results	Distinguish inclusion, survival, interviews, selected accounts, and events.	Different denominators supply no common prevalence; the complementary group was selected for recollections.
AWARE II, Table 2 / thematic Results; Plato, X 614b–615c, 617d–620d	Reported moral reconsideration meets the ethical task of choice and justice.	A shared theme establishes no common verified ontology; myth is no clinical case.
AWARE II, independent stimuli and record corroboration, pages 3, 6–7; van Lommel, nurse account	Distinguish later corroboration from a predetermined independent test.	Agreement proves no entire class of accounts; an unsuccessful test refutes no afterlife.
AWARE II, EEG subgroup / limitations, pages 7, 9; van Lommel, timing / Discussion	Separate a signal, event timing, and a subsequent report.	No individual RED–EEG pair established; the older design provided no continuous EEG of every moment.
Van Lommel, follow-up; Plato, X 619b–621d	Distinguish reported changes from a narrative’s ethical significance.	Neither a therapeutic procedure nor an assessment of every survivor’s morality.
AWARE II, Discussion, page 9; van Lommel, Discussion	Trace the transition from observation to stronger philosophical interpretation.	No final cause, proof, or refutation of an afterlife announced.

Primary sources

Sam Parnia and colleagues, AWAreness during REsuscitation — II. Resuscitation 191, 109903, 2023; original English. Pages 1–11 of the published PDF actually read: complete Methods, Results including table text and captions, Discussion and limitations, funding and authorship information. Supplements and original datasets were not obtained; EEG images were not independently rescored. Low survival, interview availability, and the physiological subgroup constrain conclusions. The exact key sentence was read within its full discussion; Russian translation is our own.
Pim van Lommel, Ruud van Wees, Vincent Meyers, Ingrid Elfferich, Near-death experience in survivors of cardiac arrest. The Lancet 358(9298), 2039–2045, 15 December 2001. Complete Methods, Results with tables and the selected testimony, and the entire Discussion actually read and honestly reused in this article. No original-data reanalysis. A registered 2002 correction is acknowledged; its contents could not be obtained, and pharmacological claims are not used here. Survivor selection, interview timing, and follow-up losses constrain conclusions; the authors’ stronger philosophical interpretation remains distinct from data.
Plato, The Republic of Plato, Book X. Benjamin Jowett translation, third edition of 1888; the electronic transcription is based largely on it, with Stephanus numbering from the 1908 edition, as stated in the actual transcriber note. Complete selected tale 614b–621d, title information, and the nearby ethical transition read. Epigraph: complete sentence at 619b. Neither the Greek original nor the whole book is claimed read. Philosophical myth and its attributed immortal-soul setting supply no independently verified clinical case. Russian quotation translation is our own.

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of traditions, intercultural dialogue, and contemporary spiritual discipline. Instagram: @andrey.zarinsh.

Maria Evilkut is a co-author of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.