2561 people saw the same thing on DMT. Did anyone measure what happened the next day?

Johns Hopkins surveyed 2561 DMT users and found that each one, independently, described the same visual: kanji-like glyphs scrolling vertically. Imperial College recorded the EEG correlate while the substance was active. Both studies are the most rigorous record available on the phenomenon. Both stop at the same point: when the session ends, the instrument stops. Nobody measured a single night afterward.

The Finding

In 2020, Alan Davis and his team at Johns Hopkins published results from the largest retrospective survey of DMT experiences ever conducted. The sample was 2561 people who had used the substance at some point in their lives. The method: detailed questionnaires about what participants remembered having perceived.

One pattern stood out. Subjects with no contact with each other, from different cultures and languages, described the same type of image. Geometric glyphs. Vertical movement. A structure that repeated across contexts where it had no reason to repeat.

The year before, in 2019, Christopher Timmermann and the team at Imperial College London published something different: the real-time EEG correlate of the experience. With electrodes on the scalps of volunteers, they captured the signal from the central nervous system while the substance ran its course. Not a memory. A direct reading.

Study	Institution	Sample	Method	Window	Limit
Davis et al. 2020	Johns Hopkins	2561	Retrospective survey	Post-experience recall	Measures memory, not physiology
Timmermann et al. 2019	Imperial College London	—	Real-time EEG	During the experience	No follow-up measurement

The Gap

Here is where H4 opens: the proxy is not the thing.

A retrospective survey measures what a person remembers. Not what their body went through. This does not invalidate Davis et al. It bounds what the study can claim. The real datum is: 2561 people shared a structurally similar memory, described in the language available to them, from a body that was no longer in the state it was during the session.

Timmermann et al. is more direct: real-time signal, electrodes, data unfiltered by memory. But the equipment was switched off when the experience ended. The team did not continue measuring while participants slept that night. That window does not exist in any published dataset.

Why does that night matter? Two reasons.

The first is physiological. If the body went through something of intensity, the hours of sleep that follow are part of the process. Resting heart rate, heart rate variability, respiratory rate — all of these shift during sleep and reflect what the autonomic nervous system is processing. If something happened, it may appear there.

The second is methodological. Memory consolidates during sleep. If the physiological state of that night was different from baseline, then the memory formed in an altered context. The survey data cannot be separated from the state in which it was built.

Neither study measured that night. Nobody did.

The Transfer

This is not a remote laboratory problem. It is the same gap that appears every time someone reports what they felt after an intense practice.

Long meditation sessions. Controlled hyperventilation protocols. Extended fasting. Any kind of intake. In each case, the standard method is identical: ask how you felt, collect the account, file it as data.

What that method does not capture is what happened in the body during the night that followed. Not because it is unimportant — because no sensor was placed there.

The gap that turned 2561 accounts from Davis et al. into a document of memories rather than a physiological record is the same one that exists with any practice for which no nighttime measurement was kept. You do not need to believe anything about what those 2561 people saw. The gap applies regardless.

The Instrument

What closes H4 in this context is not another survey. It is a device on the wrist before sleep.

A wearable that records through the night captures what happens in the hours following any intervention: resting heart rate, heart rate variability, respiratory rate. Each variable has a measured noise floor specific to the subject. A change that clears that floor is detectable. Not as a confirmation of a subjective experience, but as a record of what the body did.

If the change exists, it can be measured. If it does not, the null is also information. What cannot be done is assert that something happened without having measured it — or assert that nothing happened on the same grounds.

The instrument's limit is attached to its advantage: it records the body, not the experience. A shift in heart rate variability says nothing about glyphs or geometry. It indicates only that the autonomic nervous system behaved differently that night compared to its own baseline pattern. Interpreting that requires more nights, more data, and the willingness to publish when the result is zero.

The Limit

Davis et al. 2020 measures memories. Not physiological states. 2561 people described their experiences in the language available when they answered the questionnaire. That is a genuine finding about the phenomenology of recall. It is not a physiological record.

Timmermann et al. 2019 measures the EEG during the experience. Not before. Not after. The signal ends when the session ends.

Neither study answers the question this piece is asking. The first night of sleep after the experience does not exist in any published dataset. This piece does not contain it either. What it does is name the gap with enough precision for a protocol to be designed around it.

The Inherited Absence

That 2561 people share a pattern of recall does not answer how long any physiological change persists, if one exists. Is it one night? Four? Twenty-eight? Duration does not appear in either study because the follow-up was never done. That is what the next measurement opens.

---

Mental Work — all protocols · related file: dmt-laser-matrix-code · open science — the dataset

Sources: Davis et al. 2020 — Johns Hopkins University · Timmermann et al. 2019 — Imperial College London