v0.1 · 2026-04-30
Temperature
Temperature is the affective load every Vaquum surface carries, fixed system-wide at one value. Neutral places the firm between warm — evaluative, reassuring — and cool — forensic, distancing.
What temperature is
Temperature is the affective load of every surface the firm produces. The single value the parameter takes is neutral — between warm language that evaluates or reassures and cool language that is forensic or distancing. The firm reports observation; it does not react. The same neutral applies to a memo, a working paper, a research note, and a documentation page.
One value, neutral
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: neutral applies to every surface the firm produces, and the validator runs the same closed rule set against each. Subject, audience, and channel do not relax the value, because none of the three are parameters of the voice.
Sub-properties
Temperature declares 18 sub-properties in its typed source. Each names a distinct facet of affective load — emotional vocabulary, evaluative adjectives, reassurance language, punctuation affect, and reporting verb bias. Of the 18, 13 are backed by typed rules that run during validation. Two — intensifiers and exclamations — share their checks with register. Three remain editorial: scare quotes, figurative language, and judgment without data.
Typed rules
The first exhibit below renders the 13 typed rules read directly from @vaquum/voice/parameters/temperature. Each row carries the sub-property the rule guards, the rule’s prose description, and the mechanism the validator uses to detect a violation. The rule set is closed and the validator reads it against every voice-bearing surface in the repository.
Mechanisms
Temperature carries two 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; they enumerate forbidden affect-laden tokens, evaluative diction, ingratiating concessions, and reporting verbs that lean. Four rules are regex patterns; they match emoji code points, doubled punctuation, quantified feeling, and direct emotional attribution.
How a violation reads
A temperature violation in source prose returns a typed report from the validator. The report names the rule that fired with an identifier such as temperature.affect-laden, points to the source section in Voice-Addendum.md, and carries the offending text span. The second exhibit below reads neutral against warm and cool failures across reporting verbs and punctuation. The reader compares the Vaquum row against the warm and cool rows; each property reads differently when temperature drifts from neutral.
Source: Voice-Addendum.md § Temperature
Source: Voice-Addendum.md § Temperature