What has actually been measured
The theory board at /theories is a library of proposals. This page is the other half of it: the measurements, the published work, and the methods that would settle the question. A theory sitting on this site is not evidence, and nothing here is a finding until it has survived a blinded test.
Every record named below comes from the research library or the typed trials table, both of which anyone can download under CC-BY-4.0. Where a section holds no records yet, it says so rather than filling the space.
Direct tests: the laser observation itself
Work that tested the observation rather than the compound. Every record is typed on https://dmtcode.com/trials, so a pilot report is never mistaken for a registered clinical trial. The published pilot report is Goler's 2025 account in IPI Letters, DOI 10.59973/ipil.158.
Stage one of this project is screening, not the experiment: open, self selected, unblinded, with priming not ruled out. Stage two captures the memory before the observer sees the catalogue. Stage three is a randomized blinded arm, designed and not run. Nothing on this site settles the question.
Every typed record, including the community experiments and the reported replication. The case against.
Mechanistic science
What optics and vision science already know that could produce a repeatable form without anything exotic: laser speckle, the geometry of the visual cortex, predictive processing, and the classic hallucinatory form constants. If one of these accounts for the reports, that is the answer, and it would be a real one.
Additions to this strand are the ones most worth having, since a conventional optical or cortical account would settle the question. Browse the library and send additions to info@dmtcode.com.
DMT science
What has been measured about the compound itself: EEG and fMRI during the experience, pharmacology and receptor work, and the phenomenological literature that tries to describe what people report. The library is stance scored, so skeptical, neutral and supportive sources are indexed side by side.
Methodology
How a claim like this would be tested properly: blinded matching, similarity scoring that survives rotation and stroke width, preregistration of the outcome before recruitment begins, and honest handling of null results. The draft study design is published in full, including the sample size correction that followed an error in an earlier version.
- Methods: blinding design, control conditions, sample size calculation, and the limits of similarity scoring.
- Null reports: reports of seeing nothing, published alongside the positive ones.
- Critiques: the case against, kept where it can be read first.
- Protocol guide: the reported equipment, including the laser class the pilot actually used.
Open projects
Collaborations are listed on this page as they are agreed, with the collaborator named, and not before.
- Proposed studies: the blinded arm is designed and unfunded. The design is public on Methods and the questions it would settle are tracked on open questions.
- Research recruitment: analysts, recorders and translators can volunteer through join. Observers who want to record before browsing start at capture.
- Datasets needing analysis: the corpus is downloadable at /data.json under CC-BY-4.0, with field definitions on dataset. The two symbol counts are not synonyms; the object model explains what each one counts.
- Active collaborations: listed here as they are agreed. Researchers who want to use the corpus or run a study with it can write to info@dmtcode.com. The licence is CC-BY-4.0, so nothing here needs our permission.