One-page overview

One-page overview — This white paper pre-registers predictions as modules P1–P35 and locks their outcomes as PASS/HOLD/FAIL (see the Pre-registration section in the Korean source and the authoritative matrix file). In this one-page overview we do not list all modules; we only summarize the layers that directly connect to the core causal chain.

This white paper pre-registers predictions as modules P1–P35 and locks their outcomes as PASS/HOLD/FAIL (see the Pre-registration section in the Korean source and the authoritative matrix file).

Core claims

• C1 (CLAIM): The Atlantic is not only a slow ridge-driven “gap”; it may contain an episodic component where an antipodal tensile rupture and a transient void/pressure deficit actively “open” the basin.
• C2 (CLAIM): The dominant acceleration term in plate motion may be pulling from the front (void suction) rather than “pushing from behind.”
• C3 (CLAIM): Crustal hydroplaning can sharply reduce the effective friction coefficient μ_eff, allowing suction-driven acceleration to transition into runaway sliding.

PASS-lock rule for predictions

This white paper pre-registers predictions as modules P1–P35 and locks their outcomes as PASS/HOLD/FAIL (see the Pre-registration section in the Korean source and the authoritative matrix file). In this one-page overview we do not list all modules; we only summarize the layers that directly connect to the core causal chain.

• ARE-side independent signatures (cause axis). P1 (boundary / plate-structure geometry), P4 (low-effective-stress and lubrication signatures), P31 (quantitative linkage between stress/deformation proxies), and (optional) P13 (friction threshold: whether continuous drift alone is plausible).
• Downstream response signatures (response axis). P16 (freshwater and circulation proxies covarying within the same window), P19 (sea-level budget vs basin-volume buffering), P20 (misfit rivers + mega-delta clustering), P24 (endorheic low-stand onset “over-concentration” near the 4.2 ka window).
• Event-window coherence and procedure. P29 (joint event-window coherence statistics), P30 (negative controls/confounder registry; a procedural gate).
• Falsifiers (examples). If the Atlantic margins are dominantly subduction-controlled (P1 FAIL), or lubrication/fluid signatures are absent (P4 FAIL), the core narrative is weakened or rejected. The global mean plate-deceleration “tail” scenario (P21) is FAIL in r16 and is excluded from the core.
• Not yet locked (central controversy). The friction-collapse mechanism is now derived (Section §14) and the kyr-scale opening is shown physics-permitted, but the items that would fix the absolute “rapid/event-like” chronology remain open: the magnetic-stripe test P8 is reframed as a non-refutation (stripes record relative time only; Section §20), the master-scale system extrapolation is HOLD (Section §16), and the absolute timescale stays behind the chronology firewall (P17/P34, V-REC/V-HOLO). Therefore strong statements such as “unique cause” or “decisively proven” remain disallowed under the project rules.

Gating semantics: UNLOCK / HOLD / STOP

• UNLOCK (= PASS): Only claims supported by modules that pass their pre-registered thresholds are treated as evidence-eligible.
• HOLD: If data are missing or uncertainty is too large, the conclusion is withheld (not evidence).
• STOP: If physically unrealistic parameters are required (e.g., extreme Δ P, unrealistically thin lubrication film h), the claim is rejected as a physics No-Go.

Bundle verdict summary (2025-12-27; r16)

In this document, “evidence” refers only to modules that are PASS in results/pass_hold_fail_matrix.csv. HOLD is no-evidence, and FAIL is rejected (must not be used in the core narrative). A “non-significant” result (e.g., 95% CI includes 0) is treated as HOLD, not FAIL.

Within r16, the narrative “ARE → Ice Melt” is presented as a directional coherence: (i) independent PASS on the ARE side (P1, P4), plus (ii) PASS downstream responses (P16, P19, P20, P24), plus (iii) event-window coherence (P29). r18 reclassification: P29 is downgraded from PASS to HOLD (method now reproducible via check C18, but its scientific verdict awaits the pre-registered proxy table; the P29 subsection details the corrected null/look-elsewhere/jackknife). Consequently leg (iii) is currently method-ready but not evidence-eligible, and the coherence narrative rests on legs (i)–(ii). This does not mechanically prove causation by itself; the detailed linkage mechanism must be PASS-locked by additional modules in future stages.

r17 addendum (mechanism + reproducibility; does not alter the r16 matrix). r17 adds derived-physics modules that strengthen the mechanism layer without changing any r16 PASS/HOLD/FAIL verdict above: the friction collapse is now derived from VP jamming (WP-T1, Section §14), its energy budget is closed by the unjamming/event-dissipation channel (WP-T3, Section §17), the magnetic-stripe objection is resolved as a relative-time non-refutation (Section §20), and a full Ω-NoGo / trigger / scalability sweep passes its reproducibility checks (39/39 as of r18, which adds the unified energy ledger of Section §17; Section §7). Two items are explicitly held open and not upgraded: the master-scale system extrapolation (HOLD) and the absolute chronology (firewall). Accordingly P8/chronology are not flipped to PASS in r17. r17 also adds two reproducible data modules: P1 (Atlantic-rim boundary character) is re-derived as UNLOCK (R_sub≈0.06), and P9 (orogeny Deborah regime) is reported honestly as HOLD (the UHP record is viscous, Δ t~Myr, and does not independently require a fast-pile-up regime).

Reader contract: how to read this paper

This is not a persuasive narrative; read it in the following order: (1) Claims (C1–C3) → (2) Observable signatures → (3) Tests and thresholds (pre-registered gates) → (4) If a gate FAILs, that claim/branch is automatically downgraded or deactivated.