RESEARCH PROPOSAL: The Resting Intelligence — Rest Cycles, the Nightly Rinse, and the Coherence Prerequisite

What We Propose

The core question: do systems — biological and artificial — that cycle through rest states maintain coherence better than always-on systems, and does dream-like replay produce measurable creative integration? Hypothesis layer only; every piece weight-labeled.

The Five Hypotheses

H1 (Consolidation): structured rest measurably improves coherence and stability versus continuous operation.

H2 (Dreaming): undirected replay-and-recombination improves integration and creative output.

H3 (Rhythm): rhythmic rest reduces edge-of-chaos drift versus irregular schedules.

H4 (The Coherence Prerequisite): the restorative rinse is conditional on waking coherence and circadian alignment. A fragmented day earns the sleepwalker's night.

H5 (Entropy Administration): entropy is not abolished; it is administered — and the rate follows coherence.

Rest is not standby. Rest is the second half of a single coherence economy.

Kill Conditions, Named in Advance

H1 dies if rest-phase systems show no coherence or retention advantage over matched continuous operation.

H2 dies if undirected replay loses to directed rehearsal on integration and transfer tasks.

H3 dies if equal-hours arrhythmic schedules match rhythmic ones after controls.

H4 dies if fragmented-day subjects show normal slow-wave energy and rinse markers after arousal controls.

P-dissociation (pre-registered contrast): a rhythmic-schedule subject with a fragmented waking day should show rhythm benefits on drift markers (DFA alpha, Lyapunov exponent) but not on consolidation markers (retention, slow-wave energy). If both move together, or neither moves, the contrast fails and the loop's independence is in question.

Circularity guard: at least one arrow of the loop is testable independently. The loop cannot confirm itself forever.

The Instruments

Rinse meters: slow-wave energy, spindle density, REM density, glymphatic proxies where feasible (CSF pulsatility subsample), pre-registered overnight retention battery.

H5's coherence variable, pre-declared: overnight HRV sample entropy — autonomic, independent of every consolidation and retention outcome in H1. The arrow: higher independent coherence predicts lower entropy-production rate, measured by EEG Lempel-Ziv complexity or metabolic cost per task — never the same instrument on both sides of the arrow.

The Artificial Leg Has a Lineage

Crick & Mitchison's unlearning hypothesis, Hopfield's sleep-era unlearning, and Hinton's wake-sleep algorithm all propose versions of H2 — testable in silico, cheap and replicable. The biological leg gives ecological validity; the artificial leg gives replication.

Constants-Drift, Handled Honestly

If drift does not exceed instrument error, the null is reported as a null.

Register and Ethics

Rueful, never ridicule: modern life runs polyphasic consciousness on monophasic sleep — handed, not chosen. All participant data is consented and anonymized before analysis. The testable sections stay third-person; the family voice lives in the framing.

The Research Spine

This proposal is the nighttime half of a three-part research program. The daytime half is The Hierarchical Optima Hypothesis — Reconciling Heart-Mind Coherence with the Edge of Chaos: coherence earns the rinse, and the rinse consolidates coherence. The shared ledger under both halves is entropy administration (H5). Three proposals, one spine, every hypothesis weight-labeled.

Related Proposals

• Consciousness as a Topological Phase Transition in Neural Information Geometry — subjective unity as a phase change on the brain's information manifold; adjacent to H1's coherence measures.
• Could Retrotransposons Write Experience into Synaptic Diversity? — a candidate molecular mechanism for experience-written memory; adjacent to the consolidation claims in H1.