# style analysis prompt
#prompting
## Vocabulary
• Lexical density (open-class words) ≈ 56 %
• Dominant word classes: nouns 38 % > verbs 13 % > adjectives 11 % > adverbs 6 %
• Recurring collocations: “in this paper”, “generative AI”, “in the sense”, “on the other hand”, “we argue that”
• Common affixes: prefixes meta-, inter-, auto-, non-; suffixes ‑tion, ‑ity, ‑ness, ‑ism, ‑ment
• Preferred synonyms: “argue” > “contend”; “plausible” > “likely”; “illustrate” > “demonstrate”
## Morphology and inflection
• Tense spread: present 62 %, past 22 %, present perfect 9 %, other 7 %
• Voice: active 78 %, passive 22 %; progressive aspect < 4 %
• Modal verbs per 1 000 words: can 7,8; must 3,1; may 2,9; should 2,2; might 1,7; could 1,5
• Pronouns: inclusive “we” 5,2‰; third-person pronouns 4,6‰; impersonal “one” 0‰
## Syntax
• Mean sentence length = 29,4 words (range 8–67)
• Simple : compound : complex ≈ 15 % : 30 % : 55 %
• Frequent fronted adjuncts: “In this section, we …” (4,8‰)
• Common subordinators: because, although, while, since, if, when
• Parenthetical punctuation per 1 000 words: commas 54,0; em-dashes 3,2; semicolons 2,9; colons 3,4
## Punctuation
• Relative counts /1 000 words: , 54; . 35; ; 3; : 3; — 2; … 0,3; “ ” 4; ‘ ’ 2; ( ) 6
• Single space after all marks; no space before
• Ellipsis restricted to quotations; em-dash marks asides (“—not as an agent—”)
## Cohesion devices
• Logical connectives /1 000 words: however 4,1; therefore 2,7; thus 3,3; nevertheless 1,0; indeed 1,2
• Deictic chains with “this/that/such” frequent (“[[This approach]]…”)
• Enumerators: “first”, “second”, “third”, “finally” structure arguments
## Paragraph architecture
• Average paragraph length = 5,8 sentences (~160 words)
• Typical openings: definition or claim 64 %; citation 18 %; rhetorical question 4 %
• Closings: summary 41 %; transition 37 %; invitation 6 %
## Tone and register
• Formality = 4,2 /5 (academic)
• Contractions 0,1‰ (“doesn’t”)
• Hedging 11,9‰ (seems, appears, might, arguably) — e.g. “might still deserve some degree of ‘credit’”
• Boosters 3,6‰ (clearly, certainly, indeed) — e.g. “Indeed, a better understanding…”
• Evaluative adjectives moderate (“unsatisfactory”, “plausible”, “problematic”)
• Epistemic : deontic modal ratio ≈ 3 : 1
• Stance: inclusive “we”; first-person singular absent; impersonal passives common
• Reader engagement: sparse rhetorical questions (“Should we not say the same thing…?”) ~1/800 w.
• Intensifiers: very 1,5‰; quite 0,9‰; rather 0,6‰; somewhat 0,5‰
• Emotional valence neutral; occasional “promising” / “problematic”
• Jargon density ≈ 9 % (ontology, recalcitrance, [[machina naturans]])
• Humour/irony: negligible
• Idioms/clichés rare (“on the other hand”, “prima facie”) 2,3‰
• Taboo/colloquial language absent
• Rhythm: hypotaxis dominant; parataxis for emphasis in introductions
• Objectivity : subjectivity ≈ 70 % factual / 30 % evaluative
• Politeness: soft directives (“let us consider”); gratitude confined to acknowledgements
## Capitalisation and numbering
• Sentence-initial capitals observed; proper nouns standard
• Heading style: sentence-case within articles; Title-case in [[main title]]
• Numbers: words for ≤ ten, digits for 11 +; years unspaced; no thousands separator
• Citation style: author–year parenthetical (“Helliwell 2023”)
## Other recurrent features
• Parentheses for clarification and citations 6,9‰
• Footnotes rare; endnotes in proofs only
• Imperatives occasional in thought experiments (“Consider the following…”)
• Idiosyncrasies: em-dash insertions, and the formula “in relation to” as subsection links
# style analysis prompt
#prompting
## Vocabulary
• Lexical density (open-class words) ≈ 56 %
• Dominant word classes: nouns 38 % > verbs 13 % > adjectives 11 % > adverbs 6 %
• Recurring collocations: “in this paper”, “generative AI”, “in the sense”, “on the other hand”, “we argue that”
• Common affixes: prefixes meta-, inter-, auto-, non-; suffixes ‑tion, ‑ity, ‑ness, ‑ism, ‑ment
• Preferred synonyms: “argue” > “contend”; “plausible” > “likely”; “illustrate” > “demonstrate”
## Morphology and inflection
• Tense spread: present 62 %, past 22 %, present perfect 9 %, other 7 %
• Voice: active 78 %, passive 22 %; progressive aspect < 4 %
• Modal verbs per 1 000 words: can 7,8; must 3,1; may 2,9; should 2,2; might 1,7; could 1,5
• Pronouns: inclusive “we” 5,2‰; third-person pronouns 4,6‰; impersonal “one” 0‰
## Syntax
• Mean sentence length = 29,4 words (range 8–67)
• Simple : compound : complex ≈ 15 % : 30 % : 55 %
• Frequent fronted adjuncts: “In this section, we …” (4,8‰)
• Common subordinators: because, although, while, since, if, when
• Parenthetical punctuation per 1 000 words: commas 54,0; em-dashes 3,2; semicolons 2,9; colons 3,4
## Punctuation
• Relative counts /1 000 words: , 54; . 35; ; 3; : 3; — 2; … 0,3; “ ” 4; ‘ ’ 2; ( ) 6
• Single space after all marks; no space before
• Ellipsis restricted to quotations; em-dash marks asides (“—not as an agent—”)
## Cohesion devices
• Logical connectives /1 000 words: however 4,1; therefore 2,7; thus 3,3; nevertheless 1,0; indeed 1,2
• Deictic chains with “this/that/such” frequent (“[[This approach]]…”)
• Enumerators: “first”, “second”, “third”, “finally” structure arguments
## Paragraph architecture
• Average paragraph length = 5,8 sentences (~160 words)
• Typical openings: definition or claim 64 %; citation 18 %; rhetorical question 4 %
• Closings: summary 41 %; transition 37 %; invitation 6 %
## Tone and register
• Formality = 4,2 /5 (academic)
• Contractions 0,1‰ (“doesn’t”)
• Hedging 11,9‰ (seems, appears, might, arguably) — e.g. “might still deserve some degree of ‘credit’”
• Boosters 3,6‰ (clearly, certainly, indeed) — e.g. “Indeed, a better understanding…”
• Evaluative adjectives moderate (“unsatisfactory”, “plausible”, “problematic”)
• Epistemic : deontic modal ratio ≈ 3 : 1
• Stance: inclusive “we”; first-person singular absent; impersonal passives common
• Reader engagement: sparse rhetorical questions (“Should we not say the same thing…?”) ~1/800 w.
• Intensifiers: very 1,5‰; quite 0,9‰; rather 0,6‰; somewhat 0,5‰
• Emotional valence neutral; occasional “promising” / “problematic”
• Jargon density ≈ 9 % (ontology, recalcitrance, [[machina naturans]])
• Humour/irony: negligible
• Idioms/clichés rare (“on the other hand”, “prima facie”) 2,3‰
• Taboo/colloquial language absent
• Rhythm: hypotaxis dominant; parataxis for emphasis in introductions
• Objectivity : subjectivity ≈ 70 % factual / 30 % evaluative
• Politeness: soft directives (“let us consider”); gratitude confined to acknowledgements
## Capitalisation and numbering
• Sentence-initial capitals observed; proper nouns standard
• Heading style: sentence-case within articles; Title-case in [[main title]]
• Numbers: words for ≤ ten, digits for 11 +; years unspaced; no thousands separator
• Citation style: author–year parenthetical (“Helliwell 2023”)
## Other recurrent features
• Parentheses for clarification and citations 6,9‰
• Footnotes rare; endnotes in proofs only
• Imperatives occasional in thought experiments (“Consider the following…”)
• Idiosyncrasies: em-dash insertions, and the formula “in relation to” as subsection links
# style analysis prompt
#prompting
## Vocabulary
• Lexical density (open-class words) ≈ 56 %
• Dominant word classes: nouns 38 % > verbs 13 % > adjectives 11 % > adverbs 6 %
• Recurring collocations: “in this paper”, “generative AI”, “in the sense”, “on the other hand”, “we argue that”
• Common affixes: prefixes meta-, inter-, auto-, non-; suffixes ‑tion, ‑ity, ‑ness, ‑ism, ‑ment
• Preferred synonyms: “argue” > “contend”; “plausible” > “likely”; “illustrate” > “demonstrate”
## Morphology and inflection
• Tense spread: present 62 %, past 22 %, present perfect 9 %, other 7 %
• Voice: active 78 %, passive 22 %; progressive aspect < 4 %
• Modal verbs per 1 000 words: can 7,8; must 3,1; may 2,9; should 2,2; might 1,7; could 1,5
• Pronouns: inclusive “we” 5,2‰; third-person pronouns 4,6‰; impersonal “one” 0‰
## Syntax
• Mean sentence length = 29,4 words (range 8–67)
• Simple : compound : complex ≈ 15 % : 30 % : 55 %
• Frequent fronted adjuncts: “In this section, we …” (4,8‰)
• Common subordinators: because, although, while, since, if, when
• Parenthetical punctuation per 1 000 words: commas 54,0; em-dashes 3,2; semicolons 2,9; colons 3,4
## Punctuation
• Relative counts /1 000 words: , 54; . 35; ; 3; : 3; — 2; … 0,3; “ ” 4; ‘ ’ 2; ( ) 6
• Single space after all marks; no space before
• Ellipsis restricted to quotations; em-dash marks asides (“—not as an agent—”)
## Cohesion devices
• Logical connectives /1 000 words: however 4,1; therefore 2,7; thus 3,3; nevertheless 1,0; indeed 1,2
• Deictic chains with “this/that/such” frequent (“[[This approach]]…”)
• Enumerators: “first”, “second”, “third”, “finally” structure arguments
## Paragraph architecture
• Average paragraph length = 5,8 sentences (~160 words)
• Typical openings: definition or claim 64 %; citation 18 %; rhetorical question 4 %
• Closings: summary 41 %; transition 37 %; invitation 6 %
## Tone and register
• Formality = 4,2 /5 (academic)
• Contractions 0,1‰ (“doesn’t”)
• Hedging 11,9‰ (seems, appears, might, arguably) — e.g. “might still deserve some degree of ‘credit’”
• Boosters 3,6‰ (clearly, certainly, indeed) — e.g. “Indeed, a better understanding…”
• Evaluative adjectives moderate (“unsatisfactory”, “plausible”, “problematic”)
• Epistemic : deontic modal ratio ≈ 3 : 1
• Stance: inclusive “we”; first-person singular absent; impersonal passives common
• Reader engagement: sparse rhetorical questions (“Should we not say the same thing…?”) ~1/800 w.
• Intensifiers: very 1,5‰; quite 0,9‰; rather 0,6‰; somewhat 0,5‰
• Emotional valence neutral; occasional “promising” / “problematic”
• Jargon density ≈ 9 % (ontology, recalcitrance, [[machina naturans]])
• Humour/irony: negligible
• Idioms/clichés rare (“on the other hand”, “prima facie”) 2,3‰
• Taboo/colloquial language absent
• Rhythm: hypotaxis dominant; parataxis for emphasis in introductions
• Objectivity : subjectivity ≈ 70 % factual / 30 % evaluative
• Politeness: soft directives (“let us consider”); gratitude confined to acknowledgements
## Capitalisation and numbering
• Sentence-initial capitals observed; proper nouns standard
• Heading style: sentence-case within articles; Title-case in [[main title]]
• Numbers: words for ≤ ten, digits for 11 +; years unspaced; no thousands separator
• Citation style: author–year parenthetical (“Helliwell 2023”)
## Other recurrent features
• Parentheses for clarification and citations 6,9‰
• Footnotes rare; endnotes in proofs only
• Imperatives occasional in thought experiments (“Consider the following…”)
• Idiosyncrasies: em-dash insertions, and the formula “in relation to” as subsection links