Cairn Documentation

Why Dead Paths Matter

Most engineering documentation shows you what was built. It doesn't show you what wasn't built — the alternatives that were considered, evaluated, and rejected along the way.

This is a problem.

The Pruned Alternatives Thesis

Every system is the result of a tree of decisions. At each branch point, engineers evaluated options: fuel cell vs. battery, tracks vs. wheels, centralized vs. distributed processing. One path was selected. The others were pruned.

Those pruned paths aren't failures — they're engineering knowledge. Each one carries:

• Why it was rejected: specific technical, economic, or schedule reasons
• What would have to change: the conditions under which it becomes viable again
• What it depended on: cross-branch coupling that influenced the decision

This information is valuable. It prevents re-litigation of settled decisions. It accelerates onboarding ("why didn't we use fuel cells?" is answered in the model). It enables what-if analysis ("if battery energy density improves 40%, which alternatives reopen?").

What Gets Lost

In traditional documentation, this knowledge is scattered:

• Trade study reports (if they were written)
• Meeting notes (if they were taken)
• Email threads (if you can find them)
• Engineers' memories (if they're still on the project)

Six months after a decision, the rationale is effectively gone. New team members don't know why things are the way they are. Stakeholders ask "did you consider X?" and the answer requires archaeology.

Worse, rejected alternatives get re-proposed. Someone who wasn't in the original discussion thinks fuel cells are worth evaluating — not realizing the team already did that evaluation and rejected it for specific reasons. The meeting happens again. The analysis happens again. Time is wasted.

Dead Paths in Cairn

The Dendritic Lens makes pruned alternatives first-class citizens:

• They appear in the system tree alongside active nodes
• They're visually distinguished (muted, marked as "pruned")
• They carry structured evaluation metadata:
  • Prune reason (engineering prose)
  • First principle (the physics or constraint that drove rejection)
  • Evaluation phase (when in the design process it was pruned)
  • Decision type (goal, physics, engineering, mission, autonomy)
  • Cross-dependencies (what other nodes it was coupled to)

This means the decision history is always available. Anyone can see what alternatives existed, why they were rejected, and what conditions would change the calculus.

The dead paths are the point. They're half the engineering knowledge — and without them, you're only seeing half the picture.