Decision Record

Context

The program prompt asks for a “DF record” that selects writer approaches by purpose, audience, medium, and risk. Canonical governance defines DF- as a Decision Record, not a decision-framework artifact. The pilot also finds that whole-writer rankings conceal genre, authority, and proof-regime differences.

Decision

Use research/frameworks/WRITER-PURPOSE-FIT-FRAMEWORK-v1.0.md as the human- readable selection framework. This DF- records and authorizes that choice. Recommendations must declare the communication contract, select bounded functional patterns, apply ethical/accessibility gates, and name an outcome test. Do not issue a universal writer ranking or a pseudo-precise score.

Alternatives

1. Assign each writer a numeric style score. Rejected: no validated scale or outcome model exists.
2. Create a DF- typed as “decision framework.” Rejected: conflicts with canonical REP v2.0 semantics.
3. Recommend one writer per document type. Rejected: audience, risk, and use mode materially alter fit within a document type.

Evidence

Five primary-text records and two contextual scholarship records support function-level differences [EV-COMM-2026-D6B2 through EV-COMM-2026-7D8D; HY-COMM-2026-D246]. Existing TH-COMM-2026-0001 predicts conditional fit.

Consequences

The framework is auditable and avoids style impersonation, but it requires more up-front context and cannot automate selection until outcome evidence is available. Concept and glossary records remain embedded provisionally in the REP because no canonical registry/schema exists.

Reversibility

High. The framework is versioned and can be replaced without changing primary evidence.

Validation

Revisit after two comparative audience experiments or when a larger corpus produces a counterexample that changes a selection rule.