Studying Rare Experiences: Why Testimony Alone Is Insufficient

Epigraph

“A wise man, therefore, proportions his belief to the evidence.”

David Hume, An Enquiry Concerning Human Understanding, X §87; L. A. Selby-Bigge's English edition, second edition, 1902. Complete sentence; a philosophical principle about assurance, rather than a ranking of people's spiritual worth.

What this article is about

A rare account can deserve attention while its cause remains unknown. Research nevertheless distinguishes several tasks: describing experience, establishing prevalence, comparing testimony with an independent record, and testing a particular explanation. Testimony contributes differently to each. We examine what happens to a conclusion between a person and a questionnaire, a questionnaire and a group, and a group and a general hypothesis. Our grounds are two actually read research studies, Hume's philosophical argument, and selected religious testimony from Teresa of Jesus. Their positions remain distinct. Comparison reveals selection, counting units, memory, wording, and the limits of blinded inquiry. There is no promise of restoring an exact past or obtaining an unusual experience. The proposed practice concerns an existing text and its inferential structure, leaving room for respect and research uncertainty.

I. Rarity starts with a question

Imagine an invented report: someone once felt the familiar distinction between themselves and their surroundings disappear. “Rare” could mean this happened once in their life, that they know no similar stories, or that it seldom occurs in a particular group. These meanings differ. A biographical statement supplies no population estimate. The rarity of available accounts may also concern willingness to speak. Before seeking a general explanation, clarify which kind of rarity is being claimed and whose experience forms its reference. A report can remain important without settling all these questions.

There are also different questions about the account itself. What does someone remember? Which words describe it? Is there an independent record of the circumstances? Do they propose a cause? For the first question, testimony is indispensable as the participant's report. Another may require material independent of later interpretation. Combining the questions without showing the transitions makes respect appear to endorse any hypothesis, while criticism of a hypothesis seems to deny someone's right to speak. Unfolding the question prevents that confusion without deciding beforehand whether a proposed explanation is correct.

Research need not settle every task immediately. One study might construct a descriptive vocabulary, another test a scale's stability, and another compare specified conditions. These are distinct, valuable results. Difficulty arises when a descriptive map becomes a finished causal theory or a prevalence estimate for all people. First identify a paper's stated task and ask which data suit it. An attractive title and numerous responses cannot make that transition for the author. To assess a conclusion, readers need the relationship between the question posed and the material actually obtained.

We do not treat rarity as spiritual superiority. An unusual account makes someone no more valuable, and its absence makes life no less meaningful. This is our contemporary position, rather than a measured study finding. It supports a calm purpose for reading: understanding an available claim instead of competing over intensity. Readers need no new episode to join the discussion. An existing text provides enough material to follow where confidence entered an argument and what supports it. Participation can concern careful reading rather than any attempt to change one's state.

II. Whom did recruitment bring into the study?

Gamma and Metzinger's MPE-92M study used an anonymous online questionnaire in five languages. Respondents referred to a clear or significant past episode. Recruitment through networks and circulated invitations provided access to relevant reports without creating a random sample of all meditators. Of 3,627 submissions, 1,403 were retained. The authors discuss numerous incomplete responses and self-selection. These figures describe the route by which material was obtained, rather than the world's proportion of people with the experience. Their meaning depends on how respondents entered and remained in the analysis.

Prevalence claims need an appropriate denominator. An invented collection of twenty unusual accounts could illuminate language without revealing how many people lacked such episodes or did not respond. Consecutive submissions do not remove an unknown invitation pathway. “Who could enter this record?” differs from “How many records were saved?” The first question reveals missing people and circumstances. This limitation accuses no participant; it concerns design and inferential reach. A useful descriptive collection can remain useful while its inability to estimate population frequency is clearly acknowledged by the reader.

People, responses, episodes, and one-minute measurements are different units. A laboratory participant may supply several recordings. More segments provide additional material about that person without turning them into several independent people. Write down what each number counts. If the unit changes between Methods and an appealing retelling, a conclusion can expand unnoticed. Detecting that transition requires no elaborate calculation: restore the link between the number and its stated unit in the paper. Repeated material and additional participants may both matter, but they contribute in different ways to an argument.

Online anonymity has a concrete limit: MPE researchers could not verify respondents' identities or detect repeat participation through retained addresses. This is a stated limitation, rather than established fraud. The responses remain available material obtained under those conditions. Neither automatic rejection nor an assertion that every response's independence was proved is justified. Precise wording preserves the descriptive work while identifying additional checks needed for a stronger conclusion about a group of people. An acknowledged uncertainty changes what can be claimed without requiring the entire questionnaire to be discarded.

III. What became measurable?

MPE offered 92 items and obtained a factor structure of descriptions. This identifies answers that group together in the material. It neither records the past episode directly nor turns each answer cluster into a separate natural entity. The chosen twelve dimensions explained 44 percent of variance, with remaining model-fit problems. The authors identify further scale validation as necessary. We retain that preliminary status rather than promising a final universal division of phenomenology or an established cause. The distinction concerns what a statistical structure tells us about the particular responses analysed.

Questionnaire words also matter. Discussing one item, the authors acknowledge possible conflation of philosophical emptiness with the spatial experience of a void. One familiar word can mean different things to differently trained people. Agreeing with an item does not remove those differences. Comparison requires knowing what respondents were asked to rate and how they could understand it. This concrete qualification establishes no impossibility of questionnaires. It shows why language belongs to the method rather than forming a transparent covering over an experience whose meaning was already settled before the question was asked.

Teresa likewise distinguishes mind-formed inward words from those she attributes to God. Her criteria belong to religious testimony, however, rather than questionnaire factor analysis. Comparison is useful at the level of attending to differences, rather than discovering one shared instrument. Her discernment cannot become a factor merely because both texts discuss experience. Preserve her task, then specify the questions a contemporary scale asks its own participants. A related topic need not erase different grounds. The comparison can illuminate those differences without claiming that either source has measured the other's object.

Versions matter even before measurement. We use an identified Teresa translation whose editor discusses correcting an earlier negative in one sentence. An author's name cannot replace checking the wording available to a reader. Consider our invented question, “Was it quiet?” Physical sound, absence of thoughts, and calm could attract the same answer under an unclear instruction. This is an original illustration of clarification, not another finding about the sources' participants. It helps distinguish the words used from the experience or interpretation a speaker means those words to convey.

IV. Experienced clarity and later memory

Teresa reports unusually clear inward words and transformation attributed to her Lord. These are important parts of her testimony. Reported clarity does not constitute an independent time log of the event. We can attend to the text while retaining the question of externally established cause. That requires neither declaring her experience false nor assigning a current diagnosis. It distinguishes available testimony, religious interpretation, and absent material. Historical reading particularly needs this distinction because today's reader was absent from the event and has no direct access to its original circumstances.

MPE considers reports of sometimes distant experiences. The authors separately identify a difficulty in accounts of states without ordinary experienced selfhood. Could such a state later be stored and retrieved in an autobiographical format? The paragraph discusses a possibility whose relevant evidence still needs gathering. Our key quotation concerns that specific question. It does not claim that consciousness has never been researched or accuse every respondent of invention. The methodological problem is posed rather than settled; preserving its precise referent prevents an evocative sentence from acquiring a much larger meaning.

For ordinary reading, distinguish remembering an experience from remembering its explanation. Imagine someone initially recording unusual calm, then describing it a year later through a tradition they subsequently studied. Both records may matter, but later vocabulary need not reproduce the earlier formulation. A retained date can reveal changed interpretation without fully restoring vanished sensations. This original example concerns the structure of material, rather than a measured memory-error rate or a claim that later religious language is generally untrustworthy. Its purpose is to keep the history of the wording visible.

None of this demands emotionally weak testimony. An event can remain central to someone's life. Research can simply retain when its record appeared and which interpretation is current. Preserve unknown timing where necessary. If no early record exists, do not promise to recover it through detailed imagination. Our position is limited: protect the distinction between available material and desired completeness. It supplies no memory-recovery method and assesses no spiritual worth. Someone may speak with strong conviction while a reader still makes a carefully limited claim about what the recorded account establishes.

V. Why might researchers need to know less?

Dream Catcher separated its data and analysis teams. Analysts gradually received more information to test whether selected states could be distinguished from recordings without adapting their answer to known labels. This is a specific use of blinding, rather than ignorance in general. Ask who had which labels at each stage. Two teams and five disclosure steps belong to one study. They are not five independent investigations repeatedly confirming a broad account of consciousness or religious experience. The procedure's structure is part of the result that a responsible retelling needs to preserve.

The selected analysis comprised nine participants and 54 one-minute recordings, three per condition per person. The authors selected early-night NREM episodes and particular reports. Forgotten-content and uncertain reports did not enter the binary comparison chosen. Its result therefore concerns that set. Excluded categories do not disappear from human sleep. The analytical question is narrower than the available life from which it draws. Distinguishing all accounts from those selected for testing reveals how a precise inquiry can be useful while leaving related questions outside its scope rather than pretending they were answered too.

An invented document check can illustrate the procedure. Knowing the desired answer in advance may make a reader notice matching details while overlooking others. A hidden label limits that particular influence. It repairs no poor original document and creates no missing page. The analogy is original and bounded: a laboratory signal is no manuscript, and we tested no everyday checking procedure. It illustrates how blinding addresses a specified problem without guaranteeing an error-free chain of knowledge. The quality and availability of the starting material still require their own questions.

Hume's §§88–89 consider circumstances of testimony and differing grounds for assurance. We do not credit him with a contemporary blinding protocol. Comparison reveals a move from philosophical questions about trust to a concrete procedure examining available-information effects. Nor does historical criticism become a guide to detecting lies through demeanour. A witness's forcefulness or hesitation needs context; one expression replaces no examination of material. The contemporary procedure's value lies in a describable function and limit, rather than an assertion that it can remove every difficulty associated with human reports.

VI. What does an unsuccessful result mean?

No chosen Dream Catcher classification step produced a result significantly above chance. This matters as an outcome of the stated test. It does not exhaust every possible method of measurement and analysis forever. The authors identify sample size, selected reports, and recording-system limitations. Keep these beside the finding. Otherwise an unsuccessful prediction becomes purported final proof that physiology has no relationship with experience. The study established no such universal conclusion and could not replace all future tests. Its result needs to retain the particular material and procedure to which it belongs.

An inconvenient finding should not vanish because an initial hypothesis is attractive. If the published procedure obtained no required distinction, retain that in the account. Further proposals can be discussed without becoming retrospective success for the original test. A study may reveal a task's difficulty without confirming an expected relationship. Distinguish the unsuccessful prediction, the usefulness of a published method, and the broader debate over consciousness. These questions have different scales. A subsequent improvement would add new work, rather than silently change what the earlier investigation had actually obtained.

Negative findings differ from missing evidence. One means a specified test occurred; the other means material for it was not obtained. MPE's question about memory of selfless states is open in the identified paragraph, unlike the unsuccessful EEG classification. The comparison helps when this difference remains. A gap and an unsuccessful prediction cannot simply become two equivalent confirmations of preferred metaphysics. Name separately what was tested, what did not succeed, and what was only proposed. Different unresolved questions require different additional grounds rather than one repeated expression of confidence.

Use language matching the scale of assurance. “No distinction was obtained in selected recordings” is more precise than “There is never a distinction.” “This memory possibility needs testing” differs from “The witness cannot remember.” Stronger theories remain possible but require additional arguments. We offer this discipline as a contemporary application of the comparison, not an efficacy finding from these studies. It can preserve the published result without presenting an expected future explanation as grounds already supplied. The distinction matters whether a reader welcomes the result or would have preferred another one.

VII. Hume: unfamiliar does not mean impossible

Hume's selected argument belongs to the section on miracles and has an explicitly critical direction. We do not turn it into neutral approval of any religious account. In §89 he distinguishes what is unusual for someone from the stronger miracle claim considered next. The unfamiliar-frost example makes previous experience relevant. Unfamiliarity does not alone produce final impossibility. This transition matters here without implying that a questionnaire or laboratory recording has resolved Hume's entire philosophical debate. A useful limited comparison can identify the step without claiming to settle everything beyond it.

We also do not adopt historical social and clinical judgements as present credibility rules. Origin, traditional affiliation, or someone's impression of character cannot remove a person's right to report experience. Inquiry concerns grounds for the particular claim. Where independent records, timing, and procedures exist, examine their content. Where absent, state their absence. This contemporary boundary limits our application without rewriting Hume's position. It specifies what we invite readers to do with the text rather than pretending every historical judgement can become a responsible rule for evaluating people today.

Teresa, by contrast, reports inward transformation and attributes it to God. Her religious certainty and Hume's critical argument genuinely diverge. The attractive word “experience” cannot eliminate that divergence. Comparison can preserve both positions and identify where each seeks grounds. We need neither a philosopher's supposed endorsement of her attribution nor a witness's invented contemporary causal test. The disagreement remains visible. Different texts do not become long-agreed parts of one theory merely because a convenient table needs a pleasing connection between them; their real differences are part of that connection.

Testimony cannot settle every conclusion simultaneously, but remains useful. It can open questions, provide descriptive language, and show what a person considers significant. Population prevalence, independent linkage to an event, and general causes require further steps. This is our limited comparative conclusion, rather than every proposition in Hume or Teresa. It takes accounts seriously while retaining the appropriate burden of grounds where another question begins: a claim offered to people independently of their own experience. Those people can respect the witness while asking how that particular transition is supported.

VIII. Following the path of evidence

Consider an original scenario: someone believes an unusual thought preceded receiving a letter. For external sequence, an existing dated thought-record and the letter's independent timestamp may help if voluntarily available. A later account alone establishes neither clock accuracy nor the absence of other messages. This is no proposed experiment or search for danger. It shows which material bears on a particular question. Even a clarified sequence would not automatically identify coincidence's sole cause or knowledge of the future. That further assertion needs its own argument rather than being supplied by timing alone.

Describe the path in words: who reported, when a record appeared, what analysts knew, and what they concluded. Changes can enter each step. A retelling adds detail, an editor selects a striking episode, or researchers narrow categories. Not every change is error or deception. Make it visible when it affects a conclusion. Retaining wording history helps distinguish initial report from subsequent explanation. The aim is transparency of available material rather than a demand for an ideal record of every life. A documented limitation is preferable to an unnoticed change in what the account means.

Future tests may benefit from questions and result criteria stated beforehand. Here this is an original general principle, rather than a claim that every project considered completed one contemporary registration procedure. A prior description limits opportunities to select convenient conclusions afterwards. It makes no poor question good and removes no measurement limit. Readers can ask which decisions preceded a finding and which were suggested after it. This difference helps assess an explanation without unsupported accusations about the researchers. Later exploratory proposals can remain useful when their actual status is made clear.

Consent and disclosure limits form a separate issue. Someone can decline to provide an intimate record without establishing fabrication. Researchers can leave a causal conclusion uncertain when its necessary grounds are absent. These positions are compatible. An existing published text suits our exercise; there is no need to demand details from an acquaintance or reconstruct forgotten experience. Respect for consent is our contemporary position, rather than a claim that every historical source contains today's participation procedure. A research interest alone grants no automatic entitlement to someone else's private life or original records.

IX. Investigating while preserving scale

Rare experiences make access difficult rather than permitting methods to be skipped. When events cannot readily be observed again, retain exactly what was obtained: an account, immediate rating, signal, document, or hypothesis. These materials can complement one another without becoming interchangeable. Readers can examine that structure without special equipment or a complete theory of consciousness. Ask which transition an author makes and what additional grounds support expansion from a person to a group or from description to cause. The difficulty of obtaining material makes its actual character particularly important to preserve.

Respect and critical attention can coexist. Recognise that testimony changed a life while declining its proposed general theory. Read brain research attentively without declaring philosophy finished. Retain a negative result without celebrating an opposing side's defeat. This conversation requires purpose and patience. It becomes harder when unusual accounts serve as membership badges or excuses to humiliate others. Our practice therefore concerns a claim's wording rather than a final verdict on a person. Its goal is a clearer argument, even where the people discussing it still disagree about religious or philosophical questions.

An honest outcome may contain several statuses: descriptions collected, representativeness unestablished, one classification unsuccessful, and a memory possibility awaiting testing. This looks less dramatic but identifies the next substantive question. It prevents apparent completion where work remains open. New research needs added grounds rather than another confident retelling. Historical texts retain their voices as actual comparative reasoning, without becoming contemporary data tables or scientific-hypothesis confirmations. Such a record can show what changed after an inquiry and what still needs work, instead of making every source appear to say the same thing.

Attentive reading promises no agreement on consciousness's ultimate nature. It offers something more limited: retaining people while criticising conclusions and retaining inferential limits while respecting people. Testimony contributes where it suits the task; additional verification belongs where another question begins. This work uses existing material and requires no changes to sleep, treatment, or ordinary life. It can stop with the unknown clearly identified. An open question demands no immediate new unusual episode as an answer, nor a forced choice between dismissing the account and accepting every interpretation attached to it.

What this means in practice

Choose a published account and one associated claim. Identify its question: experience description, frequency, external-event sequence, or cause? Keep one for the first reading. Name the material: participant's words, retelling, measurement, or hypothesis. Mark boundaries where it is mixed. This is our original reading exercise, rather than lie detection, spirituality assessment, or a medical evaluation. An ordinary existing text is sufficient; you need no personally intense event to perform the task.

Find the selection pathway and counting unit. Who could enter, who was excluded, and what does each number mean? Note repeated episodes or recordings from one person. With no denominator, calculate no prevalence. Numerous records cannot supply independence merely by accumulating. An open gap is preferable to a precise-looking number answering another question. The exercise requires only an existing paper and its descriptions, rather than further recruitment or collection of personal reports.

Clarify recording time, definitions, and available tests. Which words belonged to the initial report and which to later interpretation? What did analysts know? Was a test completed unsuccessfully, or was a question merely proposed? Reproduce no laboratory awakenings and induce no dangerous state. Keep personal material voluntary. Someone else's intimate life becomes no compulsory appendix to your research interest, even when their account seems relevant. You can choose an already open text instead.

Write one limited conclusion and add what remains unknown: for example, “These accounts describe selected episodes without establishing prevalence among everyone.” Check that respect survives and no cause appeared without further grounds. Other conclusions are possible when material supports them. Stop with a clearly unresolved question if necessary. The exercise promises no exact memory, common metaphysics, or tested efficacy of its own procedure. Its modest outcome is a more accurate description of what the available material establishes.

Key quotation

“But the relevant evidence has yet to be gathered.”

Alex Gamma and Thomas Metzinger, 2021, Limitations, paragraph on possible autobiographical storage and retrieval of states without ordinary experienced selfhood. Complete sentence; this limitation concerns the identified memory possibility, rather than all consciousness research.

Connections between traditions

Two grounds and locators	Comparison mechanism	Boundary
MPE, Methods / selection Limitations; Dream Catcher, Participants / Data selection	Follow online self-selection and laboratory selection of people and recordings.	Responses, people, and minutes differ; prevalence not established.
Hume, X §§87–89; MPE, memory and identity Limitations	Distinguish philosophical testimony assessment from concrete collection limits.	No historical diagnostic scale or established universal account falsity.
Teresa, XXV §§2, 23; MPE, past-state Limitations	Separate experienced clarity from reconstruction through later reporting.	Teresa was no survey participant; experiential origin neither verified nor refuted.
Dream Catcher, Blinding / five steps; Hume, X §§88–89	Compare questions about available knowledge with a specific test of its influence.	Contemporary blinding is not derived from Hume and removes no totality of errors.
MPE, Factor 8 / understanding variation; Teresa, XXV §§3–6 / editorial introduction	Clarify wording, version, and distinctions between experience and concepts.	Religious criteria are no factor scales; no common cause established.
Hume, X §§89–93; Teresa, XXV §23	Preserve unusual versus miraculous and a genuine disagreement between positions.	Rarity is no impossibility; texts share no imposed doctrine.

Primary sources

David Hume, An Enquiry Concerning Human Understanding. L. A. Selby-Bigge, editor, second edition, 1902, reprinting posthumous 1777 edition; original English. Title and complete selected X §§86–91/92–93 freshly read. Epigraph is §87's complete sentence. Neither whole book nor whole section claimed read. Miracle criticism not harmonised with religious testimony; historical prejudices supply no present credibility rules. Russian quotation translated by the authors.
Alex Gamma, Thomas Metzinger, The Minimal Phenomenal Experience questionnaire. PLOS ONE 16(7), e0253694, 2021. Honest reuse of selected Methods, Results, Discussion, and complete Limitations actually read in this task; exact memory paragraph freshly read. Separate 2024 correction concerns data-availability link. Raw data not reanalysed. Anonymity, self-selection, retrospective reports, language, and preliminary model fit constrain conclusions; no population prevalence. Russian key quotation translated by the authors.
William Wong and colleagues, The Dream Catcher experiment. Neuroscience of Consciousness 2020(1), niaa006, 15 July 2020. Pages 1–13 of university-hosted PDF and selected complete primary publisher Discussion paragraphs actually read, including limitations, NREM depth, and future avenues. Later PDF pages unavailable directly, recorded separately in research evidence. No raw-data or supplement analysis. Nine participants and selected recordings limit the negative result.
Teresa of Jesus, The Life of St. Teresa of Jesus. David Lewis translation, third edition, Thomas Baker / Benziger Brothers, 1904; editor Benedict Zimmerman. Honest reuse of actually read title, editorial correction, XXV §§1–6, and complete §23 with context. Neither whole book nor Spanish original claimed read. Clarity and transformation testimony provides no independent causal verification or current diagnosis.

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.