The Right Field at the Right Boundary: How Field Differentiation Shapes Agentic and Multi-Agent Solution Quality
Agentic and multi-agent systems are frequently compared through agent count, model capability, orchestration topology, memory and tool access. These characteristics do not by themselves explain which interpretation, evidence, authority and consequence obligations must survive a functional boundary. This paper studies a more selective problem: when does a distinction among Field categories preserve a target-relevant obligation that a simpler, equally informed architecture does not already preserve? It proposes a conceptual framework of Field Functional Signatures for seven candidate categories—Subfields, Large Language Fields (LLFs), Large Signal Fields (LSFs), Overfields, Metaoverfields, Metafields and Manifields—and a boundary-relative alignment schema. The signatures are structured as purpose, assumptions, obligations, evidence, authority, failure conditions and qualified Return. A comparative selection approach admits justified differentiation, conventional equivalence, redundancy, incompatibility and unresolved qualification. Four adversarial thought experiments explore premature outcome claims, conventional sufficiency, coordination-regime conflicts and uncertain signals requiring suspension. The approach is compared with earlier Field studies and established methods in interfaces, contracts, provenance, security, distributed systems and agent evaluation. It neither identifies a universal Field hierarchy nor asserts that every functional responsibility needs a new software component or agent. The proposed symbolic conditions are conceptual requirement schemata, not mathematical theorems. No new empirical results are reported. The residual contribution is the candidate mapping from required boundary obligations to appropriate Field responsibilities, with explicit counterfactual comparisons and falsifiers. Its independent scientific distinctness remains a question for further verification.
Authors
- Rogério Figurelli (ORCID: https://orcid.org/0009-0002-5726-1765)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-17
- DOI
- https://doi.org/10.5281/zenodo.24308517
- Primary Topic
- Multi-Agent Systems and Negotiation
- Type
- preprint