Taxonomy Review¶
Procedure — instantiates Category Boundary Audit
Inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly hides.
Taxonomy Review audits a whole scheme's shape rather than a single line. Its object is the tree: the levels, the labels, the way siblings are split, and how items are grouped under parents. It looks for the places where the structure itself — not any one criterion — creates trouble: cross-cutting items that belong under two branches but can be filed under only one, sibling categories whose boundary is unstable because they overlap, and groupings that were clean when the scheme was built but no longer match the work being sorted. Its defining assumption-to-break is the mono-hierarchy: the belief that every item has exactly one right place in one tree. When the real cases are multi-dimensional, that assumption manufactures invisibility, and the fix is structural — split, merge, regroup, or re-express the dimension as a facet.
Example¶
A large law firm organizes its document intranet by practice area: top-level folders for Litigation, Corporate, IP, and Employment, each with subfolders. A Taxonomy Review is triggered because associates keep failing to find prior work. The review samples "lost" documents and finds the pattern immediately: a matter that is an employment dispute inside an M&A deal with an IP-licensing wrinkle had to be filed in one folder — Corporate — and became invisible to the litigators and IP lawyers who would reuse it. The scheme's sibling categories are not mutually exclusive; real matters cut across them.
The review does not just relabel folders. It concludes that practice-area is being forced to act as a single hierarchy when it is genuinely a dimension, and recommends converting to a faceted scheme: a matter carries a matter-type, one or more practice-area tags, and a client tag, so the same document surfaces under every branch it touches. The output is a set of explicit structural decisions — one merge of two near-duplicate subfolders, one split of an overloaded catch-all, and one dimension promoted to a facet — each tied to the misfiled cases that justified it.
How it works¶
- Enumerate the structure. Lay out every level, label, and sibling set so the scheme can be seen as a whole rather than navigated one click at a time.
- Test the sibling seams. Check each set of siblings for mutual exclusivity and collective exhaustiveness — the two places a hierarchy silently fails.
- Pull the structural casualties. Gather items that were misfiled, duplicated across branches, or dumped in a catch-all; these expose where the tree's shape fights the material.
- Diagnose mono-hierarchy failures. Separate genuine miscategorization from cases that are inherently multi-branch and cannot be fixed by relabeling.
- Decide the structural revision. Recommend split, merge, regroup, add a level, retire a catch-all, or convert a dimension to a facet.
Unlike Ontology Audit, which checks a category's fit against the broader domain model, Taxonomy Review stays inside one scheme and works on its internal shape.
Tuning parameters¶
- MECE strictness — how firmly siblings are required to be mutually exclusive and collectively exhaustive. Strict schemes are tidy but brittle against real overlap.
- Hierarchy versus facet — whether to preserve a single tree or promote overlapping dimensions to independent tags. Faceting handles cross-cutting cases but is harder to browse.
- Depth and granularity — how many levels and how fine the leaf labels are. Deeper trees discriminate more but multiply the seams where items fall between siblings.
- Restructure threshold — how many misfiled or cross-cutting cases accumulate before the structure is changed rather than patched, trading stability against accuracy.
When it helps, and when it misleads¶
Its strength is diagnosing findability and stability failures that no single-boundary audit would catch, because the problem lives in the relationship between categories rather than in any one of them.
Its failure mode is churn: taxonomies invite endless re-organization, and each restructuring imposes a migration and re-learning cost on everyone who used the old shape. A related misuse is over-faceting — dissolving a usefully simple tree into a cloud of tags no one can browse — or reshuffling labels while leaving the cross-cutting reality untouched. The guarding discipline is to anchor every structural change in observed misfiled cases and to reserve faceting for dimensions that are genuinely independent, following the logic of faceted classification[n1] rather than restructuring for elegance.
How it implements the components¶
Taxonomy Review realizes the structural face of the archetype — the components that treat the scheme, not a single line, as the object:
category_under_audit— takes the whole classification scheme, and its interrelated categories, as the unit of analysis.category_boundary— examines each sibling boundary for overlap and instability rather than any one boundary in isolation.edge_case_register— collects the cross-cutting and misfiled items that expose where the structure fails.boundary_revision_decision— produces the split/merge/regroup/facet decisions that restructure the scheme.
It does not map who is harmed or externalized by the structure (boundary_impact_map) — that is Boundary Critique Review; nor does it record the scheme's owner and version history (boundary_authority_record), which is Category Governance Record.
Related¶
- Instantiates: Category Boundary Audit — supplies the structural read of a classification scheme the audit may act on.
- Sibling mechanisms: Boundary Critique Review · Eligibility Audit · Classification Fairness Review · Edge-Case Analysis · Ontology Audit · Category Governance Record · Appeals Case Review
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Taxonomy Review operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly hides.
Independent corroboration: The frozen evidence defines Taxonomy Review as 'Inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly hides', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Library & Information Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Taxonomy review derives most directly from library and information science's classification, provenance, and retrieval tradition; its defining operation is to inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly hides.
Related originating lineages:
- Biology & Ecology — Biological and ecological research supplies a parallel or contributing lineage for the mechanism's defining operation: inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly….
- Computer Science & Software Engineering — Computer science and software-engineering practice supplies a parallel or contributing lineage for the mechanism's defining operation: inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly….
- Linguistics & Semiotics — Linguistics and semiotics' terminology, meaning, and sign-system tradition provides a formative adjacent lineage for the same taxonomy review operation.
Review resolution: Both blind reviewers independently select library_information_science as the primary historical origin for the concrete operation—Inspects the structure of a classification scheme — its levels, labels, sibling splits, and groupings — for the cross-cutting cases and unstable seams that a clean hierarchy quietly hides. The queued differences concern alternate origin disagreement, origin mode disagreement, domain reach disagreement, encyclopedia synthesis disagreement, not the primary lineage. I retain every alternate that either reviewer explains, without a numeric cap, and choose origin_mode=cross_disciplinary_synthesis because the reviewers' combined evidence identifies material construction from multiple disciplines. domain_reach=universal records later portability rather than multiplying historical origins; confidence=high is the conservative shared evidentiary level, and encyclopedia_synthesis=true preserves either reviewer's affirmative synthesis finding.
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] Faceted classification, introduced by S. R. Ranganathan, describes an item by several independent attributes (facets) combined at retrieval time, rather than forcing it into one slot in a single hierarchy. It is the standard remedy when items are genuinely multi-dimensional and a mono-hierarchy keeps making them unfindable. ↩