speech-act-pragmatic

Classifies corpus texts by speech act type to build a pragmatic signature for voice replication.

13|2|Updated Feb 10, 2026
One-click install
npx skills add https://github.com/aaddrick/written-voice-replication --skill speech-act-pragmatic-aaddrick
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: speech-act-pragmatic
Source: https://github.com/aaddrick/written-voice-replication/tree/main/.claude/skills/speech-act-pragmatic
Command: npx skills add https://github.com/aaddrick/written-voice-replication --skill speech-act-pragmatic-aaddrick

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writers and analysts need to quantify how an author uses language functionally — asserting, advising, explaining, questioning, challenging, or agreeing — rather than just what they say. This Skill measures the distribution of speech acts across a writing corpus to produce a replicable pragmatic signature for voice profiling and style matching. ## Core Features & Use Cases - Speech Act Classification: Categorizes each text unit into six primary acts (asserting, explaining, advising, questioning, challenging, agreeing/supporting) plus secondary acts using signal-phrase dictionaries with contextual override for indirect speech acts. - Pragmatic Signature Generation: Computes proportions, dominance rankings, pragmatic diversity (normalized Shannon entropy), and ratios like challenge-to-agree to produce numeric replication constraints. - Cross-Context Comparison: Compares speech act distributions across topics, time periods, or audiences to detect communicative shifts. - Use Case: Given a Reddit comment export, classify every comment, discover the author explains 35% of the time and challenges 20%, then encode those proportions as constraints for a voice-replication agent. ## Quick Start Analyze the writing samples in this project with the speech-act-pragmatic skill and write the pragmatic signature report to docs/analysis/22-speech-act-pragmatic.md.

Frequently Asked Questions about speech-act-pragmatic

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze speech acts in a writing corpus?▼

Segment the corpus into text units (posts, comments, or paragraphs), then classify each unit by its dominant communicative function using signal-phrase dictionaries combined with contextual reading. Compute proportions across the corpus to build a pragmatic profile.

What is a pragmatic signature in voice profiling?▼

A pragmatic signature is a distributional profile showing what percentage of an author's texts assert, explain, advise, question, challenge, or agree. It captures communicative habits that persist across topics, making it useful as a replication constraint for voice matching.

How much text is needed for speech act analysis?▼

The corpus needs at least 300 words and 10 classifiable text units for reliable proportions. Below these thresholds, only raw counts should be reported, and single-word or very short texts cannot be classified at all.

Can speech act classification detect author intent or sincerity?▼

No. Speech act analysis classifies communicative function, not intent or sincerity. A rhetorical question is classified by its function (often challenging), but the analysis cannot determine whether the author genuinely believes what they wrote.

How are indirect speech acts handled in classification?▼

Indirect speech acts are classified by function rather than surface form. For example, a question like "How can you justify X?" is classified as challenging, and an assertion like "I found X works well" may be classified as advice, with the indirectness noted in the record.

When should speech act analysis not be used?▼

Avoid it for corpora under 300 words, machine-generated or templated text, translated text where pragmatic markers reflect the translator, and when the goal is assessing argument quality or persuasive effectiveness rather than communicative function.