Skip to content

Vocabulary Drift Audit

Checklist — instantiates Geometric Primitives Vocabulary Constraint

A recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.

A Vocabulary Drift Audit points backward, at what has actually shipped. It samples the artifacts a system has produced and checks each against the established grammar: are these built only from approved primitives, using only permitted transformations, obeying the composition rules — or has the language quietly eroded, one unapproved corner radius and one "just this once" custom shape at a time? Its defining move is that it measures conformance of outputs over time, not the capacity of the vocabulary: it assumes the language is settled and asks whether practice is still faithful to it. The whole point is that drift is invisible per-instance and obvious in aggregate — no single deviation looks like a problem, but a quarterly sweep reveals the trend line, and catching it early is the difference between a correction and a relapse into the pre-constraint chaos the vocabulary was built to end.

Example

A city transit authority rolled out a unified wayfinding system three years ago: a fixed sign vocabulary of a few pictograms, one arrow primitive, a single type family, and strict composition rules for how directions stack. Today the network looks consistent from the master files — but the signs on the street are made by a dozen contractors across many stations. The audit is the reality check. An auditor samples 40 recently installed signs and runs a fixed checklist against each: only approved pictograms? arrow at the sanctioned angles and weights only? panel spacing and stacking per the composition rules? The sweep surfaces the drift the master files can't show — one contractor has been using a slightly bolder arrow because it "read better," a station added a bespoke accessibility icon nobody approved, and three renovated platforms crammed the panels tighter than the grammar allows. Individually forgivable; collectively, the system is fraying at the edges. The audit's output is a scored conformance report and a targeted fix list, plus one flag pushed upstream: the recurring bespoke accessibility icon may indicate a real gap that belongs in a sufficiency review, not just a compliance ticket.

How it works

The audit is a periodic conformance sweep with a fixed rubric:

  • Sample the population. Pull a representative set of shipped artifacts — biased toward the newest and the most peripheral, where drift lands first.
  • Check against the grammar. For each artifact, run the checklist: approved primitives only (no smuggled-in forms past the exclusion boundary), permitted transformations only, composition rules obeyed. Each check is a concrete pass/fail against the published standard.
  • Aggregate the pattern. Read the results in bulk, not per-item: a conformance rate, the specific rules most often broken, and the sources (teams, contractors, contexts) where drift concentrates.
  • Route, don't just log. Correctable violations become a fix list; a recurring violation is escalated as a signal — either the standard needs enforcement teeth, or it is fighting a real need that belongs upstream in a sufficiency review.

The value is the trend and its distribution, which tells you not only that the vocabulary is eroding but where and why.

Tuning parameters

  • Audit cadence — how often the sweep runs. Frequent audits catch drift early but cost standing effort; rare ones let deviation compound before it is seen.
  • Sample size and bias — how many artifacts, weighted toward which sources. Larger, edge-weighted samples detect drift sooner but raise cost; small central samples miss the periphery where drift begins.
  • Severity weighting — whether all violations count equally or a smuggled-in new primitive counts more than a minor spacing slip. Weighting focuses fixes on what erodes coherence most, but encodes a judgment that can be gamed.
  • Pass threshold — the conformance rate that counts as "healthy." A high bar keeps the language crisp but generates constant fix churn; a low bar tolerates erosion.
  • Escalation trigger — how many repeats turn a violation from a fix ticket into an upstream signal. A tight trigger surfaces vocabulary problems quickly but risks crying wolf over noise.

When it helps, and when it misleads

Its strength is that it makes an invisible, gradual failure measurable and locatable: it converts "the system feels like it's slipping" into a conformance rate, a broken-rule ranking, and a map of where drift originates — so maintenance targets the real sources instead of repainting everything. Catching small deviations before they normalize is the whole game, the design equivalent of the broken-windows intuition that unrepaired small disorder invites larger disorder.[1]

Its failure mode is mistaking conformance for health. An artifact can pass every checklist item and still be bad, and an audit that only measures rule-compliance can enforce a vocabulary all the way into irrelevance — policing adherence to a language that has stopped serving the work. The classic misuse is treating every recurring violation as a discipline problem to be stamped out, when a repeated deviation in the same place is often the vocabulary telling you it is too small — a sufficiency question, not a compliance one. The guarding discipline is to route recurring violations upstream rather than merely correcting them, and to remember the audit checks fidelity to the language, never whether the language is still the right one.

How it implements the components

Vocabulary Drift Audit fills the are-artifacts-still-compliant slice of the archetype — the ongoing enforcement of the settled vocabulary:

  • exclusion_boundary — the audit's core check is whether artifacts have let disallowed forms across the boundary the vocabulary drew.
  • composition_rule_grammar — it verifies that shipped assemblies still obey the sanctioned composition rules rather than improvising new joins.
  • allowed_transformation_set — it checks that only permitted transformations (weights, radii, angles, scales) appear, catching the slow slide of "slightly off" variants.

It does not test whether the vocabulary is expressively big enough (primitive_sufficiency_test) or maintain the governed register of approved additions (variant_expansion_register) — deciding the alphabet's adequacy is the job of its nearest twin, Primitive Sufficiency Review; this audit polices whether artifacts obey the language, not whether the language is sufficient.

Editorial Notes

Form Classification

Form family: Assessment, Review & Assurance

Rationale: Vocabulary Drift Audit operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it a recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.

Independent corroboration: The frozen evidence defines Vocabulary Drift Audit as 'A recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms', so its operative form is Assessment, Review & Assurance.

Review outcome: Independent reviewer agreement; high confidence.

Origin Attribution

Primary origin: Linguistics & Semiotics

Origin pattern: Single lineage

Present-day reach: Universal

Rationale: ISO 704: Terminology work — Principles and methods documents that terminology practice aligns concepts, definitions, and labels and tracks semantic drift across communities. This is direct, mechanism-specific evidence for linguistics semiotics as the best-evidenced historical home of the operation—A recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.—rather than evidence merely that the operation is useful there. The retained alternates record genuine adjacent lineages; later portability is represented separately by domain_reach=universal.

Related originating lineages:

  • Accounting & Auditing — Accounting Auditing supplies a historically relevant adjacent lineage or formative practice for the operation—A recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.—but the adjudicated evidence more directly locates the defining lineage in linguistics semiotics.
  • Communication & Media Studies — Communication and media research supplies a parallel or contributing lineage for the mechanism's defining operation: a recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.
  • Library & Information Science — Library and information-science stewardship supplies a parallel or contributing lineage for the mechanism's defining operation: a recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms.
  • Statistics & Experimental Design — Statistics' variance, sampling, inference, and experimental-design tradition contributes a separate formative lineage to the mechanism's vocabulary drift audit logic.

Review resolution: The blind reviewers disagree on primary lineage (accounting_auditing versus linguistics_semiotics). The defining operation is: A recurring review that checks whether shipped artifacts still obey the approved primitive grammar, or have quietly drifted back toward one-off forms. The researched ISO 704: Terminology work — Principles and methods establishes that terminology practice aligns concepts, definitions, and labels and tracks semantic drift across communities. That source therefore supports linguistics semiotics as the historical origin. accounting auditing remains in the uncapped alternates where it contributes a formative practice, but application or governance is not itself proof of origin. origin_mode=single_lineage records lineage construction; domain_reach=universal separately records later applicability.

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] The broken windows hypothesis (James Q. Wilson and George L. Kelling, 1982) argues that visible unrepaired disorder signals that norms are unenforced and invites further disorder. A vocabulary drift audit applies the same logic to a form language: uncorrected small deviations advertise that the grammar is optional, so early, visible correction is what keeps the language governing behavior. withdrawn registry