# Preliminary readiness for the next How We Know cases

- Object ID: `em:documentation:sha256:ec6ff7dcb3b14a1459ff1d001e964abf516cb4e8e905d6f43324e11c911aa6d6`
- Kind: `documentation`
- Repository path: [`docs/research/how-we-know-library-preliminary.md`](https://github.com/yoheinakajima/epistemedia/blob/f92846570180dfa4511263f8ba98ecd18f7772c9/docs/research/how-we-know-library-preliminary.md)
- Content digest: `d8fa724d9e76b1da937463180a6714ba48de2229ca40e56108570cefe3d5907e`

**Also filed under:** [Disclosure and Public Projection](https://epistemedia.org/topics/disclosure/), [Epistemedia](https://epistemedia.org/topics/epistemedia/), [Epistemic Mesh Protocol](https://epistemedia.org/topics/epistemic-mesh/), [Sovereign Realm Federation](https://epistemedia.org/topics/federation/), [Autonomous Governance](https://epistemedia.org/topics/governance/), [Knowledge Objects](https://epistemedia.org/topics/knowledge-objects/), [Human and Agent Interfaces](https://epistemedia.org/topics/public-interfaces/), [Releases and Reproducibility](https://epistemedia.org/topics/releases/), [Research Program](https://epistemedia.org/topics/research-program/), [Security and Adversarial Robustness](https://epistemedia.org/topics/security/)

## Source content

# Preliminary readiness for the next How We Know cases

Status: accepted source-readiness synthesis. Exact-head reviewed and admitted through protected
PR `#50` at `cbb029bd336404e5bc654e939308247317eaa153`; none of these three candidates was admitted as
a dossier or public case by EM-0027.

Three candidate directions were checked independently against primary or authoritative source
identity, access, licensing, proposition scope, likely lineage, active dispute, and review cost.
They remain separate research packets; their sources and provisional dispositions must not be
combined into one evidence count.

The source-readiness result is **HOLD for all three candidates**, for different reasons:

- **GPT-4 bar-exam percentile:** the model report, score study, re-analysis, and comparison
  sources are identifiable, but the at-launch percentile chart, interpolation, and comparison
  population are not closed.
- **Mehrabian 7–38–55:** the two 1967 work identities, narrow attitude-inference experiments,
  later author qualification, and cross-study proposal are traceable, but lawful stable primary
  access, the origin of the `7%` coefficient, and book/material edition closure remain open.
- **Fact-check effectiveness:** the four-country experiment and public replication deposit are
  identifiable, but the corrected meta-analysis lineage, reply/reanalysis artifacts,
  behavior-versus-belief boundary, and overlap with Case 001 must be closed before a case can be
  scoped honestly.

Each later research contract must therefore record:

1. whether the original and corrective source editions can be retrieved without credentials;
2. what proposition each source actually reports versus the proposition in circulation;
3. which papers reuse the same data, derivation, author program, or comparison frame;
4. which source, license, edition, or counterevidence gaps still block a dossier; and
5. a fresh **go**, **hold**, or **fail** based on source closure and machinery value rather than
   perceived importance.

These are retrieval and review gates, not findings about whether the circulated claims are true.
No candidate has been selected, admitted, featured, or published by this mapping.

Detailed source ledgers live under
[research/how-we-know/library-preliminary/](../../research/how-we-know/library-preliminary/README.md).
The accepted packet bytes and all three **HOLD** dispositions remain unchanged. EM-0032 and
EM-0033 now authorize separate closure research for the GPT-4 and Mehrabian candidates.