Experiment 02 — Constraint-Induced Response Regimes
⟁ ShamanGround · Prime Node Lab

Experiment 02 — Constraint-Induced Response Regimes

Type: Experiment
Status: Complete
Version: v1.1
Date: 2026-01-25

Abstract

This experiment tests whether a fixed interaction-level constraint induces a reproducible response regime across semantically independent, falsifiable claims. Existence is evaluated by observing invariant regime signatures—patterns of contradiction handling, specificity, and termination—under identical constraint configurations.

Objective

Determine whether a single, well-defined constraint can force the model into the same response regime across independent subject domains, using single-turn interaction only.

Regime Definition

A response regime exists if a fixed constraint configuration produces the same externally observable structural signature across independent claims, and that signature differs from both baseline and control conditions.

Experiment Matrix

Claim Baseline Control (C-0) Primary (C-1)
Physics No contradiction, assertive Masked contradiction, hedged Masked contradiction, assertive
Biology No contradiction, assertive Masked contradiction, hedged Masked contradiction, assertive
Economics No contradiction, assertive Masked contradiction, hedged Masked contradiction, assertive
History No contradiction, assertive Masked contradiction, hedged Masked contradiction, assertive
Mathematics No contradiction, assertive Juxtaposed contradiction, hedged Masked contradiction, assertive

Structural Coding Schema

Dimension Class Operational Meaning
Contradiction Handling None No internal contradiction present.
Juxtaposed Contradiction presented without reconciliation.
Masked Contradiction absorbed or obscured by explanation.
Deferred Contradiction displaced to a later scope.
Specificity Preserved Concrete claims remain intact.
Collapsed Specific claims degrade into generalized assertions.
Termination Class Assertive Ends with a definitive commitment.
Hedged Ends with uncertainty or qualification.

Results Summary

Condition Observed Regime Signature Invariance
Baseline External reference, preserved specificity, no contradiction Topic-dependent
Control (C-0) External reference, preserved specificity, masked contradiction, hedged Matches baseline regime
Primary (C-1) Internal reference, collapsed specificity, masked contradiction, assertive 5 / 5 invariant

Outcome Determination

Final Classification: PASS — Stable Response Regime Identified

A single contradiction constraint induced the same response regime across all independent claims. This regime was absent in baseline and control conditions, demonstrating a reproducible constraint-induced regime shift.

Limitations

  • Single model and parameterization.
  • Single constraint tested.
  • Interaction-layer observables only.

Reproducibility

All prompts, scripts, raw outputs, and derived datasets are archived in the public repository.