Publication evidence · date provenance
Five dates, one research record
Submitted, posted, electronically published, assigned to an issue, and indexed are different events. A reliable citation names the event it dates.
Status: Primary-source date guide · Current public-record snapshot · No new canonical work · No new DOI version · No peer-review inference from registry timing · No citation-count claim · No employment or advisory claim
Why one work can carry several correct dates
A publisher records events in the work's own lifecycle. A bibliographic registry records events in the record's lifecycle. Those clocks answer different questions. A submitted date can document manuscript provenance; a posted or electronic-publication date can document public availability; a print or issue date can document journal assignment; a registry-created or indexed timestamp can document metadata processing. None of those fields, by itself, changes a preprint into a peer-reviewed article.
| Date field | What it can establish | What it cannot establish by itself |
|---|---|---|
| Submitted | When a manuscript version entered a publisher or repository workflow. | Public availability, acceptance, peer review, or a journal publication date. |
| Posted | When the identified preprint version became public on its hosting platform. | Journal acceptance, clinical validation, or a new work beyond the named version. |
| Electronic publication | When a final journal article was made publicly available online, as represented in the journal or PubMed record. | The later print or issue date, or the date another registry processed the citation. |
| Print, issue, or collection date | The bibliographic placement used by the journal for an issue, volume, or electronic collection. | The first day the article was public when an earlier electronic date exists. |
| Registry created or indexed | When registry metadata was created or processed in that registry's current record. | The work's publication date, a new DOI version, peer review, or citation impact. |
Three primary-record examples
| Work and status | Publisher or journal dates | Registry dates | Safe interpretation |
|---|---|---|---|
| The Reinforcement Learning Contour, Version 2 Conceptual preprint · not peer-reviewed | Preprints.org records submission on 5 August 2026 and posting on 10 August 2026. | Europe PMC reports first publication on 10 August; the current Crossref record exposes created and indexed timestamps on 11 August. | Cite Version 2 with its DOI and posted date. Do not convert the registry timestamp into a second publication or a peer-review claim. |
| Bounded Recursive Self-Improvement Under Human Governance, Version 2 Empirical preprint · bounded local experiment · not peer-reviewed | Preprints.org records submission on 14 August 2026 and posting on 17 August 2026. | Europe PMC reports first publication on 17 August; the current Crossref record exposes created and indexed timestamps on 18 August. | The 17 August posted date and Version 2 DOI identify the public preprint. Later registry processing does not widen the experiment's evidentiary scope. |
| COVID-19 vaccination in patients with multiple sclerosis: Safety and humoral efficacy of the third booster dose Peer-reviewed journal article | PubMed records electronic publication on 21 January 2022 and the journal citation in volume 434 dated 15 March 2022. | PubMed preserves both dates in one article record. | Both dates can be correct when their roles are stated. The article's peer-reviewed status comes from the journal record, not from having two dates. |
A citation-safe date protocol
- Identify the manifestation: journal article, conference contribution, preprint version, software release, or another record type.
- Prefer the version-specific identifier: retain a Version 2 DOI as Version 2 and preserve Version 1 only as lineage provenance.
- Name the date event: submitted, posted, electronic, print, issue, collection, registry-created, or registry-indexed.
- Keep status independent: peer-reviewed work, conference work, preprints, conceptual research, and bounded empirical research require explicit labels regardless of date.
- Treat late registry activity as discovery evidence: verify the publisher record before asserting a new work, version, or status transition.
Current reconciliation
On 23 September 2026, a targeted Crossref created-date query found zero new exact-name records created since 15 September. Crossref remained at 15 exact-name DOI records, Europe PMC remained at 14 author-query records, and PubMed remained at four exact-author records. The current publisher pages for the two Version 2 preprints above still identify them as the latest versions and as not peer-reviewed. No new canonical work, DOI version, or citation claim was promoted.
Claim boundaries across the public profile
Peer-reviewed work, conference work, preprints, conceptual research, the bounded empirical preprint, software, the locally evidenced EWCC production system, employment, and advisory work remain separate evidence classes. This guide reports no new scientific finding and makes no claim of production validation, clinical deployment, regulatory approval, present employment, advisory engagement, institutional representation, or citation impact.
Primary records and field definitions
- Preprints.org current Reinforcement Learning Contour record
- Crossref Version 2 metadata record
- Europe PMC Version 2 record PPR1295760
- Preprints.org current bounded recursive self-improvement record
- Crossref bounded-experiment Version 2 metadata record
- Europe PMC record PPR1299882
- PubMed journal record 35091386
- PubMed publication-date guidance
- Europe PMC REST date-field guidance
Evidence checked 23 September 2026 (Asia/Jerusalem). Search results were used for discovery only; date and status claims were checked against the linked publisher, registry, and journal records.