Branch-Pruning Protocol¶
Procedure — instantiates Contingency-Visibility Across Scales
Runs proposed counterfactual branches through ordered exclusion gates, cutting the ones that need too many unsupported changes, breach a documented constraint, or rewrite a commitment the path had already locked.
A Branch-Pruning Protocol is a stepwise elimination procedure applied to a set of proposed counterfactual branches. It does not generate alternatives and does not score them on a grid; it removes the ones that cannot stand — branches demanding too many unsupported changes to the actual record, branches that violate a documented constraint, and branches that quietly rewrite commitments the path had already locked in. Its defining discipline is subtractive and rule-bound: every surviving branch has earned its place by passing explicit exclusion tests, so the counterfactual space that reaches the final account is small, cheap to defend, and free of alternate-history sprawl.
Example¶
An incident review board examines a runway excursion — a jet that overran the end of a wet runway on landing. The room fills with counterfactuals: "if the crew had gone around," "if the tower had reported the standing water," "if autobrake had been set to MAX," "if the airline flew a different tire compound," "if the runway had an arrestor bed." The protocol runs each through three gates in order.
Unsupported-change gate: "different tire compound" requires inventing a fleet-wide procurement decision with no record — pruned. Documented-constraint gate: "arrestor bed" violates published airport infrastructure the crew could not change in flight — pruned as a crew counterfactual, retained separately as an infrastructure recommendation, which is a different question. Path-commitment gate: "go around" is checked against the point of no return — past touchdown and derotation a go-around was no longer available, so the branch is cut after that moment and kept before it. What survives — "autobrake MAX" and "tower water report" — are the few branches worth the reconciliation step, each defensible on the record.
How it works¶
The protocol is an ordered filter, not a score. Each candidate branch passes through gates in sequence; failing any one gate ends it. Gate 1 — minimal rewrite: count the unsupported historical changes the antecedent requires; over threshold, prune. Gate 2 — constraint: does the branch require breaching a mapped structural constraint the actor could not lift? If so, prune it, or reclassify it to the level that could have changed it. Gate 3 — path commitment: at the branch's moment, had the path already foreclosed this option? If so, cut it or time-bound it. Every surviving branch carries a one-line survival rationale, and every pruned branch stays in a visible log with the gate that killed it.
Tuning parameters¶
- Rewrite threshold — how many unsupported changes before a branch is cut. Loose keeps rich alternatives but invites sprawl; tight may cut real possibilities.
- Constraint strictness — treat mapped constraints as hard walls or soft costs. Hard prunes aggressively; soft keeps "expensive but possible" branches alive.
- Reclassify versus discard — whether a branch that fails at one actor level is thrown out or moved to the level that could have acted. Reclassifying preserves systemic insight.
- Commitment resolution — how finely the path's point of no return is dated. Finer dating time-bounds branches more precisely.
When it helps, and when it misleads¶
Its strength is that it cures counterfactual sprawl and the debate that multiplies because no one will discard a branch; it leaves a short, defensible set that the reconciliation step can actually use.
Its failure mode is over-aggression: a protocol that prunes toward inevitability, cutting every alternative until the outcome looks forced, is macro determinism smuggled in through a filter. The classic misuse is setting the rewrite threshold so tight that only the actual path survives, then declaring nothing could have been otherwise. The guard is to apply the minimal-rewrite rule[n1] symmetrically — the antecedent should require few changes, but the pruning must not require rewriting the record either — and to keep the visible log of what was cut and why, so a reader can contest any single prune.
How it implements the components¶
counterfactual_branch_filter— the protocol is this filter rendered as an executable, gate-by-gate procedure.structural_constraint_map— Gate 2 reads directly off the constraint map, pruning branches that would breach an unliftable constraint.path_dependence_checkpoint— Gate 3 tests each branch against the commitments the path had already locked, cutting or time-bounding options the path had foreclosed.
It does NOT rate the surviving branches on plausibility, evidence, and scale — that is contingency_marker_set and alternate_scale_lens_review, owned by Counterfactual Sensitivity Matrix: the Protocol eliminates branches, the Matrix scores the ones that remain.
Related¶
- Instantiates: Contingency-Visibility Across Scales — the protocol keeps the archetype's counterfactual work disciplined and bounded.
- Consumes: Hinge-Point Timeline supplies the raw set of forks and near-misses the protocol prunes down.
- Sibling mechanisms: Counterfactual Sensitivity Matrix · Hinge-Point Timeline · Actor-Constraint Map · Aggregate Trend Overlay · Nested Scale Timeline · Paired Micro/Macro Case Study · Interpretive Weighting Memo · Scale-Shift Review Prompt · Source Resolution Table
Editorial Notes¶
Form Classification¶
Form family: Protocol, Workflow & Routine
Rationale: The mechanism sends each counterfactual branch through ordered minimal-rewrite, constraint, and path-commitment exclusion gates and stops on the first failure, so its operative form is a pruning protocol.
Nearest alternative: Decision, Gate & Allocation — Each branch receives a keep-or-prune disposition, but ordered reusable gate enactment rather than free case selection defines the mechanism.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: History & Historiography
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Testing counterfactual branches against documentary constraints and path commitments belongs to disciplined historical counterfactual analysis.
Related originating lineages:
- Futurism & Strategic Foresight — Scenario practice supplies explicit branching futures and feasibility filters.
- Philosophy — Counterfactual semantics and minimal-change reasoning ground rejection of branches requiring unsupported world changes.
Review resolution: History and historiography is the agreed primary lineage because the protocol tests causal branches against evidence and removes unsupported counterfactual continuations. Philosophy and foresight materially shape causal warrant and scenario pruning; the generalized protocol is an Encyclopedia synthesis.
Attribution caveat: Scenario planning uses similar pruning prospectively, but the mechanism's evidence discipline and primary examples are historiographic.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The minimal-rewrite rule, from the methodology of disciplined counterfactual history (developed in Tetlock and Belkin's framework for evaluating counterfactuals), holds that a credible counterfactual antecedent should require the smallest possible departure from the documented record — change one thing and let the rest of the world run on. Branches needing many unsupported changes are, by this rule, too speculative to carry explanatory weight. ↩