FE-MISSION-001 pre-registration

Status: Frozen before external-source interpretation
Date: 2026-07-28
Provider: openai-codex
Parent: FE-EVAL-REP-2026-07-23
Hypothesis under test: FEH-001

Research question

Does Framework Engineering have a useful non-redundant boundary relative to adjacent disciplines, or is it better classified as an integrated method profile or a repository-specific research program?

Frozen candidate explanations

1. Distinct discipline — FE has a coherent object of practice and at least one non-subsumed mechanism or outcome whose independent utility can be tested.
2. Integrated method profile — FE composes established mechanisms around frameworks as the unit of analysis; the integration may be useful without being a new discipline.
3. Repository-specific research program — FE is currently an internally coherent local program whose transfer and generality are unproven.
4. Full subsumption — adjacent disciplines collectively or individually cover FE without meaningful loss; discipline-building should stop.

Frozen comparison dimensions

1. Primary object and unit of analysis.
2. Intended users, decisions, and contexts.
3. Lifecycle coverage: discovery, design, validation, operation, evolution, retirement.
4. Core operations and claimed causal mechanisms.
5. Representation, semantics, and implementation independence.
6. Evidence, measurement, verification, validation, and falsification.
7. Configuration, versioning, provenance, traceability, and change control.
8. Context adaptation, tailoring, and boundary conditions.
9. Human roles, organizational adoption, governance, and ethics.
10. Automation, tooling, interoperability, and machine/agent use.
11. Outcomes, failure modes, burden, and cost-effectiveness.
12. Institutional maturity: body of knowledge, education, standards, community, credentialing, and independent practice.

Adjacent fields fixed before review

• Systems engineering.
• Method engineering and situational method engineering.
• Requirements engineering.
• Knowledge engineering and ontology engineering.
• Decision analysis/science.
• Cybernetics and systems/control thinking.
• Quality engineering and quality management.
• Organizational learning and knowledge management.
• Agent orchestration and multi-agent systems engineering.
• Design science research as a cross-cutting comparator.

Inclusion and evidence rules

• Prefer standards bodies, official handbooks, original peer-reviewed research, and direct technical records.
• Use secondary sources only for orientation or explicit critique.
• Record inaccessible standards as abstract-only rather than inferring their contents.
• Separate direct source claims, repository observations, synthesis inferences, and missing evidence.
• Do not count terminology, integration, documentation volume, or the existence of tools as evidence of distinctiveness.
• Frequency of a mechanism is descriptive, not proof of value.

Decision rules

• Distinct requires a coherent boundary plus at least one plausible non-subsumed mechanism or outcome and a feasible discriminating experiment.
• Integrated profile is preferred when FE's object-level integration is coherent but mechanisms are inherited.
• Repository program is preferred when coherence exists but transfer, independent use, and comparative utility are missing.
• Full subsumption is supported when an adjacent field covers the same object, lifecycle, mechanisms, and outcomes without meaningful loss.
• Literature can establish boundary overlap and research hypotheses; it cannot establish FE's causal utility.

Planned outputs

Source registry, cited comparison matrix, competing-definition analysis, FEH-001 confidence update, profile defect log, successor executive definition, architecture implications, exact matched-utility experiment, and full REP.