Theory of Framework Engineering v0.1: research frontier

Scope and source artifacts

• research/theory/theory-of-framework-engineering-v0.1.md
• research/theory/confidence-assessment.md
• research/theory/falsification-criteria.md

These files are treated as one analysis unit because they share an evidence base or form one accepted/current package. Generated derivatives, templates, prompts, private working packets, and superseded runs are not counted as independent conclusions.

Knowledge extraction

• Primary objective: Explain frameworks through identity/capability, primitive reasoning, composition, redesign, and possible grammar separation.
• Primary claims: Identity and capability separation may improve characterization. A finite primitive vocabulary may represent procedural reasoning. Reasoning and execution grammars may differ.
• Methodology: Theory synthesis from FE-008, FE-011A, FE-012A, and FE-012B exploratory evidence.
• Accepted hypotheses/findings: Framework Engineering currently functions most defensibly as a characterization/redesign methodology.
• Rejected or unsupported hypotheses: The theory is validated by internal same-model evidence.
• Assumptions: Exploratory experiments generalize. Primitive labels correspond to stable constructs.
• Limitations: Independent human tests and strong controls are absent. Vocabulary and grammar remain provisional.
• Known uncertainties: Vocabulary saturation; Grammar separability; Independent convergence; Incremental utility.
• Confidence: bounded by the source package; no confidence is promoted by this frontier analysis.

Adversarial challenge

The theory retains discipline-scale hypotheses; the boundary REP finds no unique mechanism and very-low support for discipline status. Evidence that would materially change the assessment includes independent replication, preregistered negative controls, matched causal comparison, or a demonstrated failure of the current measurement instruments. The analysis deliberately treats planned validation as a dependency, not as completed evidence.

Top five opportunities

Semantic duplicates were merged into the linked repository-wide records. The five selections maximize uncertainty reduction and repository reuse, rather than creating five unique labels for each source package.