Split/Merge Decision Tree¶
Template — instantiates Entity Individuation Criteria Design
A branching template that walks a transformation to a verdict: continuity, succession, fork, merger, replacement, or several new entities.
Steady-state identity is the easy part; transformations are where entities quietly multiply, vanish, or swap places without anyone deciding they should. Split/Merge Decision Tree is the reusable branching template that meets a specific transformation event — a division, combination, restructuring, or replacement — and walks it through ordered questions to a named verdict: does the original entity continue, spawn a successor, fork into branches, merge into another, get replaced, or become several countable individuals? Its defining move is that it operates on events, not on pairs of records: it is triggered by something happening to an entity and answers "how many entities exist after, and which of them continue which before." It does not decide whether two static presentations are the same (that is resolution) — it decides what a change does to entity count and continuity.
Example¶
A corporate structuring team faces a reorganization: a division of a parent company is being spun into a subsidiary, which simultaneously absorbs a small acquired firm. Is the resulting subsidiary a new entity, the continuation of the division, or the successor of the acquired firm? Getting this wrong scrambles contract assignment, tax history, and liability. The team runs the event through the decision tree. Did operations continue uninterrupted? Yes. Was legal personhood newly created? Yes — a new incorporation. Does any predecessor's liability follow the assets? Yes for the acquired firm's obligations, no for the division's shared-service history.
The tree resolves it as a succession with partial inheritance: the subsidiary is a new legal entity that succeeds the acquired firm (inheriting its contracts and liabilities) but does not continue the division as the same entity (its prior obligations stay with the parent). That verdict, reached by walking named branches rather than by improvisation, is what tells downstream systems to open a new entity record, link succession lineage to the acquired firm, and leave the parent's history untouched — consistently, the same way the next reorganization will be handled.
How it works¶
- Trigger on an event, not a comparison. The tree is entered when a transformation occurs, and its input is what happened, not a set of look-alike records.
- Ask ordered gating questions. Continuity of operation, creation or dissolution of the identity-providing kind, and inheritance of obligations are asked in a fixed order, each branch narrowing toward one transition type.
- Emit a typed verdict. The leaf is a named outcome — continuity, succession, fork, merger, replacement, or multiplication — with the continuity links it implies.
- Hand off lineage. The verdict specifies which before-entities map to which after-entities, ready for the registry to record as lineage.
Tuning parameters¶
- Branch granularity — how many transition types the leaves distinguish. Finer trees capture exotic cases (partial succession, phased merger) but are harder to apply consistently.
- Question ordering — which gate is asked first. Front-loading the most decisive question resolves common cases fast but can mis-route rare ones.
- Determinism — whether every path ends in a verdict or some route to human judgment. Fully deterministic trees are fast and repeatable but brittle on genuine novelty.
- Kind-specificity — one universal tree versus a tree per entity kind. Per-kind trees fit their domain's transitions better but multiply maintenance.
When it helps, and when it misleads¶
Its strength is consistency at the exact point the archetype names as most dangerous — transitions, where hidden identity assumptions do the most damage. A template makes the reorganization handled today and the one handled next quarter reach the same kind of verdict by the same reasoning.
Its failure mode is the persistence question a tree cannot escape: the Ship of Theseus problem has no purely mechanical answer[1], so a tree that pretends every gradual replacement is decidable by yes/no branches will over-commit on cases that are genuinely matters of convention. A tree can also encode one domain's continuity intuition and misfire when reused where the convention differs. The guarding discipline is to route genuinely conventional cases to explicit adjudication rather than forcing a leaf, to keep the tree's branches traceable so a verdict can be re-examined, and to treat the template as an application of the persistence and succession rules — not as a substitute for having decided them.
How it implements the components¶
split_merge_and_succession_rule— it is the applied form of that rule: the branching logic that classifies a transition as fork, merger, succession, replacement, or multiplication.persistence_through_change_rule— its continuity gates apply the standing rule for which transformations preserve identity versus create a successor.
It does not perform the identity_criterion same-as decision between static records — that belongs to the Entity Resolution Policy; the tree acts on transformation events and their continuity, not on whether two look-alike presentations already denote one entity.
Related¶
- Instantiates: Entity Individuation Criteria Design — the reusable template for resolving transition events into typed verdicts.
- Consumes: Individuation Criteria Charter — supplies the persistence and succession rules the branches apply.
- Sibling mechanisms: Entity Resolution Policy · Versioned Identity Rulebook · Master Entity Registry · Individuation Criteria Charter · Identity and Unity Test Checklist · Entity Definition Workshop · Count Impact Assessment · Edge-Case Adjudication Panel
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: Split Merge Decision Tree operates by asks ordered continuity and identity questions and makes the split-or-merge disposition. That concrete deployed or enacted form is Decision, Gate & Allocation under the frozen taxonomy.
Nearest alternative: Interface, Display & Cue — Although Interface, Display & Cue can support this mechanism, the frozen evidence makes its operative form the act that asks ordered continuity and identity questions and makes the split-or-merge disposition; the alternative is therefore secondary rather than defining.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Law & Governance
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Classifying continuity, succession, fork, merger, replacement, or new entities is legal-identity governance. GLEIF's regulator-backed entity-event rules explicitly distinguish mergers, acquisitions, breakups, transformations, retirement, and successor status; computing supplies the branching template.
Related originating lineages:
- Computer Science & Software Engineering — computer_science contributes computer science and software-engineering practice to this mechanism's defining operation—A branching template that walks a transformation to a verdict: continuity, succession, fork, merger, replacement, or several new entities—without displacing the selected primary historical lineage.
- Engineering & Design — Engineering design, reliability, and systems-safety practice supplies a parallel or contributing lineage for the mechanism's defining operation: a branching template that walks a transformation to a verdict: continuity, succession, fork, merger, replacement, or several new entities.
- Library & Information Science — Authority records distinguish changed and successor entities.
- Organizational & Management Science — Restructuring requires explicit identity disposition.
- Philosophy — Persistence conditions determine when an entity remains the same.
Review resolution: The blind reviewers disagree on primary lineage (law_governance versus computer_science). Authoritative or primary research supports law_governance as the best historical origin: Classifying continuity, succession, fork, merger, replacement, or new entities is legal-identity governance. GLEIF's regulator-backed entity-event rules explicitly distinguish mergers, acquisitions, breakups, transformations, retirement, and successor status; computing supplies the branching template. The cited GLEIF, Legal Entity Event and Data Quality Rules directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=cross_disciplinary_synthesis records lineage, while domain_reach=multi_domain records later applicability separately from provenance.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
References¶
[1] Gallois, A., & Kurtsal, I. "Identity Over Time". The Stanford Encyclopedia of Philosophy, Spring 2026 Edition (2026). Presents the Ship of Theseus as a diachronic-identity puzzle with multiple competing accounts rather than a mechanically settled answer. registry ↩