What a Lattice Lacks
The chemistry a bare cellular automaton never had — authored in, visibly.
Real molecular systems supply things a bare lattice does not, and for the Bio-Kernel sims to test the framework honestly the substrate must author an explicit stand-in for each one. Because that stand-in is authored, its classical metrics are placed up front — so an authored gate is never mistaken for an emergent one.
The five roles chemistry plays
Each maps to a construct the framework can adjudicate but not generate:
| Role | What chemistry supplies | What RAPT provides (judge, not engine) |
|---|---|---|
| Action | the actual transitions happening | transaction accounting; Σ/T/I/Λ |
| Long-term structure | heritable, copyable substrate | ALPHA (α-trace — residue, not copyable) |
| Candidates | recombination as a variation generator | none — RAPT is a filter, not a source |
| Energy gating | feasibility + maintenance gates | ADM/IST; T0-Φ/MBC |
| Substitution | interchangeable parts, so a distinction can stop mattering | reserved — the Compression Descriptor slot, undefined |
The fifth role is different from the other four
For action, long-term structure, and energy gating, the substrate supplies and RAPT judges with a construct already in force. For candidates, there is no RAPT counterpart at all, by design. Substitution is different from both: its RAPT counterpart is reserved and undefined. A bare lattice has no substitutability — every cell is positionally load-bearing, and no two subconfigurations are interchangeable at a site without downstream consequence. Real chemistry has this constantly: swap a functional group and the molecule's reactivity is unchanged. That interchangeability is what makes a distinction stop mattering — and a distinction that does not matter is slack. Without substitutability there is nothing for compression to remove. So this is the first role where the substrate could hand the framework a definition rather than merely be judged by it.