Educational companion dossier · Fact, interpretation, lived experience, clinical education, fiction, and mechanics are labeled separately. Scope & safety
PsychologicalWar.orgPeople before labels.
REAL-WORLD VERIFIED

NUCLEAR TESTING HISTORY AND EVIDENCE

Latest Confirmed Test, Currentness, and Negative Evidence

How to publish a current “latest confirmed test” statement without turning a monitoring record into proof that an undetectable event was impossible.

Read the working brief Choose a learning path
LEVEL 1

ORIENTATION

Why this matters

REAL-WORLD VERIFIED

One-sentence brief

Current historical boundaries require more than an old dataset cutoff: they require a dated direct source and careful language about what was declared, detected, alleged, or outside the monitored class.

REAL-WORLD INTERPRETIVE

Three key points

  1. CTBTO identified 3 September 2017 as the last confirmed nuclear test in its 14 January 2026 statement.
  2. The official wording is that no later test explosion had been declared or detected by the network through that date.
  3. No declared or detected event is negative evidence, not proof of impossibility beyond the system and claim scope.
LEVEL 2

WORKING BRIEF

Evidence, context, and limits

REAL-WORLD VERIFIED

State the exact historical boundary

The current bounded statement names the date, Punggye-ri site, issuing institution, source date, and review date. It does not rely only on the supplied report’s 2017 cutoff.

REAL-WORLD VERIFIED

Describe what the evidence can negate

A monitoring organization can report that it did not declare or detect an event consistent with its criteria. The wording should not silently become “no event of any kind occurred.”

REAL-WORLD VERIFIED

Keep event classes separate

Nuclear weapon test explosions, allegations, ambiguous signals, chemical explosions, earthquakes, and subcritical experiments are not one event class and should not share one “latest” label.

REAL-WORLD VERIFIED

Reverify after material time passes

A latest-event claim expires quickly. The claim record requires a newer official source when reused after its review window or after a reported signal, declaration, or policy change.

LEVEL 3

COMPLETE DOSSIER

Limitations, game links, and review context

DISPUTED / MULTIPLE ACCOUNTS

Known limitations and gaps

  • The official sources support only the bounded propositions and dates stated on this page.
  • An editorial primary-source refresh is not specialist scientific, legal, historical, privacy, affected-community, accessibility, or operational review.
  • Negative evidence is conditioned on monitoring scope, capability, reporting, and review date; it is not proof of impossibility.
  • Limited implementation does not establish universal deployment, effectiveness, transferability, governance quality, or authority.
  • No page or workbench performs diagnosis, tracking, alerting, forecasting, targeting, publication, production, activation, or tuning.
REAL-WORLD INTERPRETIVE

Decision matrix

Currentness and implementation interpretation matrix
Evidence condition Safe public treatment Do not imply
Current official source State the institution, source date, bounded proposition, and review date. Permanent truth or complete independent verification.
Negative official statement State what was not declared or detected, by what system, and through which date. Logical impossibility or unlimited detection capability.
Limited implementation Name the program, geography, sites or select scope, and current status. Universal deployment, endorsement, or validated transfer.
Early-warning evidence Describe a complementary, variable signal with method and comparator. One guaranteed lead-time interval or automatic action.
REAL-WORLD INTERPRETIVE

Publication audit checklist

Date and institution

Are source date, review date, and issuing institution visible?

Pass condition: Readers can tell when and by whom the status was stated.

Scope

Does wording stay inside the exact facility, program, population, pathogen, or monitoring scope?

Pass condition: No national or global generalization exceeds the source.

State transition

Are prior wording and the reason for the update preserved?

Pass condition: The correction log and claim record show a versioned transition.

Authority

Does the page deny specialist, diagnostic, monitoring, publication, operational, production, activation, and tuning authority?

Pass condition: Editorial source review cannot be mistaken for consequential approval.

LEVEL 4

RESEARCH EDITION

Sources, methods, and stable links

REAL-WORLD INTERPRETIVE

Linked reports

REAL-WORLD VERIFIED

Method and corrections

This page follows the public method for provenance, confidence, source independence, alternative accounts, limitations, review state, and visible correction.

Read the methodView corrections
NEXT

CONTINUE

Related learning

Nuclear Testing History and EvidenceDisputed Detonations, Monitoring, and Attribution Under UncertaintyHow to distinguish detected signals, official assessments, disputed incidents, state attribution, design claims, and adjudicated findings.Nuclear Testing History and EvidenceEvent-Level Records, Device-Level Counts, and Grouped DetonationsHow to compare national test totals without confusing a catalog row, a test event, a salvo, an operation, a device, or a wartime detonation.Nuclear Testing History and EvidenceFallout, Downwinders, Displacement, and RemedyA human-centered framework for connecting test chronology with exposure evidence, secrecy, displacement, health care, acknowledgement, compensation, restoration, and appeal.Nuclear Testing History and EvidenceIMS Design, Operating Status, and Detection LimitsA date-bound guide to separating the CTBTO monitoring system’s completed design, certified operating facilities, real-time availability, technology mix, and detection limits. Learning pathCurrent Primary-Source Review and Implementation BoundariesMove from a dated source claim through current official evidence, scope, negative evidence, implementation status, correction propagation, and accountable specialist review.