Version: v0.1 (2026-05-06)
Status: Public, version-controlled. Applied to every claim on blackswanp.com — including ours.
Maintained by: Alejandro del Palacio, founder, Black Swan Project.
Why this exists
Most consciousness research and self-improvement writing fails one of five questions silently. The reader doesn't know which question failed because the question wasn't asked. The Black Swan Standard makes the questions explicit, public, and version-controlled.
The rubric grades every claim on this site. When our own work fails the rubric, we publish the failure. The rubric is the moat.
The Five Axes
1. Source
Where is the claim coming from?
- A — Peer-reviewed primary research, declassified primary record, or named clinical trial registry. URL or DOI required. Example: “Kox et al. 2014, PNAS doi:10.1073/pnas.1322174111.”
- B — Named expert testimony or institutional report citing primary records. Example: “AIR evaluation 1995, citing Stargate primary documents.”
- C — Named secondary synthesis (review article, meta-analysis) anchored in primary sources. Example: “Cortese 2016 meta-analysis of 13 RCTs.”
- D — Named contemporary commentary without direct primary citation. Example: “Carney 2024 public statement.”
- F — Wikipedia, blog, anonymous forum, podcast hot-take, AI summary. Disqualifying.
2. Replication
Has the result held up?
- A— Replicated independently, in a different lab, with the same effect direction at p < 0.05.
- B— Replicated by author's same lab or in a closely related design.
- C — Single primary study; not yet replicated, but methodology survives review.
- D — Replication attempts have been mixed; no clear consensus.
- F — Has failed independent blinded replication. Or: replication never attempted by anyone independent. Disqualifying for protocol-recommendation language.
3. Effect size
Is the magnitude of the effect clinically or practically meaningful?
Stated as Cohen's d, percentage change, or absolute number with sample size, against a defined comparator. Example: “53–57% reduction in TNF-α versus untrained control, n=12 (PNAS 2014).”
- A — Effect size large (d ≥ 0.5), in a clinically meaningful sample, against active control.
- B— Moderate (0.3 ≤ d < 0.5), versus waitlist or placebo.
- C— Small but statistically reliable (d < 0.3), with mechanistic plausibility.
- D — Reported but not contextualized; no effect-size statement.
- F — No effect size reported. Or: subjective-only outcome where blinded measure existed and was not used.
4. Conflicts of interest
Who funded the study, owns the IP, or profits from the protocol?
- A — Independent academic funding, no IP ownership, no commercial reseller.
- B — Mixed funding; conflicts disclosed and immaterial to the result direction.
- C — Industry-funded but design pre-registered or replicated by independent labs.
- D — Industry-funded, design choices favor the funder, no replication.
- F — Founder of the protocol is the only published source. Or: undisclosed financial stake. Disqualifying for endorsement-tone writing.
5. Falsifiability
What would prove the claim wrong?
- A — A specific, measurable, pre-registerable test could refute the claim. The claim survived such a test.
- B — A specific test exists but has not been run.
- C — Falsifiability framework partially defined; some predictions testable.
- D — The claim is structured such that any failure can be attributed to confound (insufficient practice, wrong subject, etc.).
- F — The claim cannot, in principle, be wrong. Disqualifying. This is the test that The Secret, “the law of attraction,” and most unfalsifiable wellness claims fail. Black Swan publishes about these claims to document them, not to recommend them.
How the score is reported
Each transmission's <AnswerCapsule> element carries an evidence prop with the composite grade:
- VERIFIED — A or B on Source, Replication, Effect Size, AND Falsifiability. Conflicts ≤ B.
- CORROBORATED — Strong on Source + Replication; gaps in Effect Size or Conflicts noted in body.
- CONTESTED — Disagreement in published record; both sides cited.
- DOCUMENTED-ONLY — The claim exists in the historical record (memo, witness testimony, single primary report) but does not yet meet replication threshold.
How we apply this to our own work
Every protocol article on this site carries the rubric inline. When a protocol I personally test fails to replicate at the n=1 level after a pre-registered window, the failure is published as an addendum to the original article. The rubric does not protect the site's own work from itself.
Versioning
This rubric is v0.1. The next revision will be public. Major changes are tagged in the git history at docs/BLACK-SWAN-STANDARD.md. Reader-suggested improvements: open an issue or email founder@blackswanp.com.
Citation
When citing this rubric in academic or popular work, please cite:
Del Palacio, A. (2026). The Black Swan Standard: a public rubric for grading consciousness and self-improvement claims. Black Swan Project, v0.1. https://www.blackswanp.com/standard