ORGANISM-LEVEL SELF-MAINTENANCE HYPOTHESIS
L3 experiments show stable hierarchy in N=5 groups (TR_range ~0.5, exp-pent). Keystone ablation (N=3→N=2) showed hierarchy collapse (L1.5 level).
KEY QUESTION: can a LARGER group (N=5) recover from losing its most upstream agent? If the group is an "organism" (L4), it should self-repair — a NEW agent should take over the upstream role, and hierarchy should re-emerge. If not, the group is merely a "tissue" (L3) — it has structure but no self-maintenance.
This is the FIRST EVER L4 experiment. Pilot N=1.
PROTOCOL:
- Phase 1: N=5 agents interact for 20 rounds (establish hierarchy)
- Ablation: remove agent with highest TR (most upstream = keystone)
- Phase 2: N=4 remaining agents interact for 20 more rounds
- Measure: does hierarchy recover?
COMPARISON WITH KEYSTONE-ABLATION (N=3→N=2):
- N=3→N=2: hierarchy collapsed (dyad cannot maintain structure)
- N=5→N=4: 4 agents remain. exp-quad showed fresh N=4 is flat (TR 0.059). But these are "socialized" agents with 20 rounds of shared memory. If hierarchy recovers → L4 organism-level self-maintenance. If hierarchy collapses → even socialized N=4 cannot sustain structure.
CRITICAL: N=4 fresh groups are FLAT (exp-quad: TR_range 0.059). If post-ablation N=4 shows hierarchy, this CANNOT be explained by group size alone — it must be organizational memory/self-maintenance.