The vocabulary of writing measurement

Plain-English definitions of the terms behind writing measurement, each with real numbers, cited sources, and a worked example.

• Voice Match
  Voice Match is ScriptGrain's measure of how closely a piece of writing aligns with a defined writing-voice profile, across the style features that profile measures. It is computed by comparing measured features of the draft against the profile, not by asking a model whether the text sounds right.
• Burstiness
  Burstiness measures how much writing varies across a document, most visibly in sentence length and structure. Human writing tends to alternate short and long sentences; AI-generated text is typically more uniform, producing low burstiness.
• Perplexity
  Perplexity measures how predictable a text is to a language model: the average surprise per word. Low perplexity means the model finds each word easy to predict; AI-generated text typically scores lower than human writing.
• Sentence Length Variance
  Sentence length variance measures how widely sentence lengths spread around their average in a passage, reported as variance or standard deviation in words. High variance signals varied rhythm; low variance signals uniform, metronomic prose.
• Flesch Reading Ease
  The Flesch reading ease score rates text readability from 0 to 100 using average sentence length and average syllables per word. Higher scores mean easier reading; 60-69 is standard, readable by high school students.
• Type-token ratio (TTR)
  The type-token ratio (TTR) measures vocabulary diversity by dividing the number of distinct words (types) by the total number of words (tokens) in a text. Higher values mean less repetition; scores are only comparable between equal-length texts.
• Hapax legomenon
  A hapax legomenon is a word that occurs exactly once within a given context: a single text, an author's complete works, or a language's entire written record. From Greek, 'said once'; plural: hapax legomena.
• Function words
  Function words (the, of, and, to, was) carry grammatical structure rather than content meaning. They form a small closed class, used at high frequency and largely unconsciously, which makes their usage rates powerful evidence in authorship attribution.
• Lexical density vs lexical diversity
  Lexical density is the share of a text's words that are content words (nouns, verbs, adjectives, adverbs). Lexical diversity is the range of different words used. A text can score high on one and low on the other.
• Stylometric fingerprint
  A stylometric fingerprint, also called a writeprint, is the measurable pattern of a writer's unconscious habits, from word choice to punctuation, that stays stable enough across texts to identify authorship, much as a physical fingerprint identifies a person.
• Idiolect
  An idiolect is one person's individual version of a language: the vocabulary, grammar, spelling and phrasing habits unique to them. In writing, your idiolect is effectively your fingerprint; no two people use language identically.
• Voice vs tone vs style
  Voice is a writer's stable personality on the page; tone is how that personality flexes for context and audience; style is the concrete, countable choices, such as sentence length and punctuation, that carry both.
• Passive voice
  The passive voice is a grammatical construction in which the subject receives the action rather than performing it, formed with a form of 'be' plus a past participle: 'the report was submitted'. It is a choice, not an error.
• Humanised AI content
  Humanised AI content is machine-drafted text revised so that its measurable habits (sentence rhythm, contractions, phrasing, punctuation) fall inside the range of human writing, as distinct from text rewritten only to fool a detector. The revision can be measured: before-and-after counts of the tells, and a voice-match score against a real writer.
• Tone matching
  Tone matching is making a piece of writing carry the same register as a reference: the same formality, warmth, humour and confidence. It is one layer of voice matching, not the whole of it; a draft can match a writer's tone and still miss their sentence rhythm, punctuation and vocabulary, which is why tone alone is a weak test.
• Brand voice
  Brand voice is the consistent set of writing habits a brand's published copy shares: its register, sentence rhythm, vocabulary, punctuation and structure. Described, it is a page of adjectives; measured, it is a set of attributes extracted from the brand's own writing that any new draft can be scored against.
• Voice drift
  Voice drift is the gradual movement of a writer's, brand's or client's published writing away from its own established habits: sentences lengthen or flatten, contractions disappear, the register formalises, new fillers creep in. It is measurable as the change in voice-match score between a baseline and later pieces, and it is what multi-writer teams, ghostwriters and AI assistants most reliably introduce.
• Authorship attribution
  Authorship attribution is the task of working out who wrote a text by comparing its measurable style habits, above all its function words, with writing by candidate authors. It is the oldest application of stylometry, from the Federalist Papers to forensic cases, and it works best when the texts compared are the same kind of writing.
• AI-isms
  AI-isms are the words, phrases and sentence shapes that language models overuse, so often that readers learn to spot them: "delve", "it's worth noting", "not X but Y", a list of three in identical grammar. Each one appears in human writing too; what gives AI text away is the density, several in a short piece.