Requirements Engineering Analysis

Foundational model

Requirements engineering treats communication as iterative elicitation, negotiation, specification, validation, traceability, rationale, and controlled evolution. Ambiguity is partly inevitable; the target is managed uncertainty, not merely unambiguous prose.

Recent and industrial evidence

Only 28 of 174 mapped ambiguity papers had empirical evaluation and most focused on detection rather than reduction [EV-REQ-2026-48A0]. A recent automotive dataset exposes trace links, decisions, deviation rationales, and domain references across requirement transformation [EV-REQ-2026-D781].

Competing interpretations and counterevidence

Formal precision can exclude users; traceability can become compliance theater; ambiguity can enable negotiation; requirements can become wrong as context changes.

Reusable methods and outcomes

Ambiguity taxonomies; stakeholder validation; trace coverage; rationale capture; change-impact analysis; defect leakage; volatility; decision latency.

Coverage and gap

RE covers action-bearing artifacts, rationale, traceability, validation, and evolution. Its common unit is intended system behavior, not transient work state, and empirical evidence for many language tools remains weak.

Assumptions challenged

Detection is not resolution; trace completeness is not decision quality; natural language flexibility is not pure defect.

Discriminating experiment

Compare a lightweight requirements package with the proposed profile on continuation correctness, negotiation quality, update cost, and error discovery.

Confidence and limits

High overlap confidence; Medium confidence transient state needs more.

Ten future questions

1. Which handoff statements are requirements versus observations?
2. When is ambiguity productive?
3. Does rationale traceability reduce incorrect continuation?
4. What minimum trace graph has decision value?
5. Can task state use validation economically?
6. Which stakeholders must validate a handoff?
7. Does formal notation justify its exclusion cost?
8. How should expired assumptions propagate?
9. Does ambiguity detection change outcomes?
10. What incremental outcome remains after mature RE practice?

Completion assessment

Bounded rapid-map gate satisfied; traceability intervention review remains debt.