Precision

What precision is

Precision is the discipline applied to every word the firm chooses for its referent. The single value the parameter takes is exacting — the precision of a well-edited research publication. Exacting precision requires the noun and verb that fit the observation, every numerical claim carrying its unit, every reference naming its antecedent, and no vague quantifiers. The same precision applies to a memo, a working paper, an alert, and a documentation page without softening.

One value, exacting

The parameter takes exactly one value across the system, and the typed module rejects any other string at parse time. The fixed value carries an invariance: exacting precision applies to every surface, and the validator runs the same closed rule set against each source file. Tone and audience do not relax the value, because neither is a parameter of the voice.

Sub-properties

Precision declares 18 sub-properties in its typed source, grouped along the lexical, numeric, and referential axes the parameter governs. Lexical sub-properties cover verb specificity, noun specificity, and vague quantifiers. Numeric sub-properties cover quantity-and-unit pairing, significant-figure discipline, ranges, and dimensional consistency. Referential sub-properties cover time references, place references, named benchmarks, and actor specificity. The remaining sub-properties carry technical terms, abbreviations on first use, and placeholder language.

Of the 18 sub-properties, 11 are backed by 12 typed rules in the validator. Eleven of those rules run during validation; the parse rule that pairs every numerical claim with its unit remains pending implementation. The other seven sub-properties are specified in Voice-Addendum.md and are not yet backed by typed rules.

Typed rules

The first exhibit below renders the 12 typed rules read directly from @vaquum/voice/parameters/precision. Each row carries the sub-property the rule guards, the rule’s prose description, the mechanism the validator uses, and the declared severity. The 11 implemented rules run against every voice-bearing surface in the repository. The remaining rule — parse mechanism, editorial severity — remains pending implementation.

Mechanisms

Precision carries three mechanisms in its typed rule set, drawn from the closed set of four — regex, dictionary, parse, and editorial — declared at the validator level. Nine rules are dictionaries; the second exhibit below renders the tokens each rule catches, grouped by sub-property and declared severity. Two rules are regex patterns: one rejects tilde approximation in front of a numeral, and the other rejects a hyphen-minus joining two integers in a numeric range. The remaining rule is a parse-mechanism check that pairs every numerical claim with its unit.

How a violation reads

A precision violation in source prose returns a typed report from the validator. The report names the rule that fired with an identifier such as precision.vague-quantifiers, points to the source section in Voice-Addendum.md, and carries the offending text span. Every report attaches one of three severity values drawn from a closed set: fail, flag, or editorial. A fail blocks the build; a flag surfaces in the validator output without blocking; an editorial result waits for review.