# Communication Theory

> Audit whether a communication theory explains a proposed mechanism, select defensible theoretical anchors, and surface rival explanations. Use for theory selection, mechanism claims, theory-driven RQ development, 理论选择/机制解释/传播理论审计. Do not use for generic statistical or writing work without a theoretical-claims question.

- Skill: `haochengw372-hash/communication-theory` (Agent Skill, multi-file: 4 files)
- Install (CLI): `npx skillmds@latest add haochengw372-hash/communication-theory`
- Raw SKILL.md: https://api.skillmd.com/api/skills/haochengw372-hash/communication-theory/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- Author: haochengw372-hash (https://skillmd.com/u/haochengw372-hash)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/haochengw372-hash/communication-theory

---


# Communication Theory

## Overview

This Skill runs the Theory Gate: it decides whether a theory is genuinely explaining a mechanism rather than sharing keywords with the topic, and it returns defensible anchors, boundaries, and rival explanations.

## Required start

1. State the phenomenon and the mechanism claim to be explained.
2. List the candidate theories and what each would predict differently.
3. Read [references/theory-anchors.md](references/theory-anchors.md) only when choosing among the common communication anchors, and [references/css-theory.md](references/css-theory.md) when the phenomenon is a general social-science process (contagion, segregation, collective action, attention, cultural evolution) rather than a communication-specific mechanism.

## Theory Gate checklist

For each candidate theory, answer all of:

- **Mechanism**: what mechanism does the theory name, and is that mechanism present in this phenomenon?
- **Keywords**: is the theory cited because of concepts, or because its terms co-occur with the topic?
- **Boundary conditions**: under what conditions does the theory apply, and is the study inside or outside them?
- **Rival explanations**: which alternative mechanisms predict the same observation?
- **Construct mapping**: which constructs does the theory require, and are they distinguishable in the design?
- **Falsification**: what observable pattern would count against the theory, and can the design produce it?
- **Evidence fit**: does the chosen design actually test the mechanism, or only an outcome the theory also predicts?

A theory that fails the mechanism or falsification item is not an anchor; it is a label.

## Anchor selection rules

- Prefer the narrowest theory whose mechanism matches the phenomenon; add a second theory only when it explains something the first cannot.
- Never combine theories just to look rich. If two theories are needed, state the boundary where each applies.
- Treat China or any non-Western setting as a boundary test, not an automatic theory failure: name the boundary conditions and what the setting revises.

## Output shape

End with a theory memo:

1. phenomenon and mechanism claim;
2. selected anchors with one-line mechanism and boundary;
3. rejected theories and the reason (keyword-only, wrong mechanism, untestable);
4. rival explanations the design must address;
5. constructs the theory implies (ready for the construct gate);
6. the falsifiable prediction that anchors the RQ/H.

## Common anti-patterns

- Citing "media equation" or "CASA" because an AI is involved, without the mechanism (mindless social attribution).
- Citing "trust" or "credibility" as if they were theories.
- Using "perceived X" as the mechanism with no theory behind the perception.
- Importing a theory from another discipline without arguing that its boundary conditions transfer.
- Claiming a theory is supported by data that a rival theory predicts equally well.

## Reference

- [references/theory-anchors.md](references/theory-anchors.md) contains a compact anchor table for common communication theories with mechanism, boundaries, suitable evidence, and common misuses.
- [references/css-theory.md](references/css-theory.md) contains the companion table for computational social science theory anchors (threshold contagion, homophily, structural holes, segregation/tipping, social identity, polarization, collective action, cultural evolution, attention economy, computational grounded theory).