Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› soctheory-conceptual-exhibits
soctheory-conceptual-exhibits
GitHub用于构建社会学理论手稿中的概念图表，包括机制图、过程模型和类型学。旨在通过可视化展示理论机制与分类，确保图表具有可读性且每个元素均有明确定义，绝不包含数据回归表或实证结果图。
Trigger Scenarios
需要为已构建的理论制作机制图或类型学以增强可读性
图表仅为无实质机制的方框箭头
类型学的单元格未清晰定义为独立概念
试图在纯理论文章中插入数据图或回归表
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill soctheory-conceptual-exhibits -g -y
SKILL.md
Frontmatter
{
"name": "soctheory-conceptual-exhibits",
"description": "Use when building the figures and typologies of a Sociological Theory (ST) manuscript — mechanism diagrams, process models, 2x2 typologies, and concept maps that carry theoretical work (never regression tables or data plots). Designs conceptual exhibits; it does NOT build the underlying theory (soctheory-theory-construction) or write the prose (soctheory-writing-style)."
}
Conceptual Exhibits: Typologies, Mechanism Diagrams, Process Models (soctheory-conceptual-exhibits)
When to trigger
- The theory is built and you need a figure or typology to make it legible
- Your figure is "boxes and arrows" with no readable mechanism
- A typology's cells are not all distinct, defined concepts
- You are tempted to include a data plot or a results table (ST has neither)
Every exhibit at ST is conceptual. There are no regression tables, no scatterplots, no coefficient figures — ST does not test, so it has nothing to plot. The exhibit must carry theoretical work: it shows a mechanism, a process, or a classification that the prose alone conveys less clearly.
The exhibit types ST uses
| Exhibit | Carries | Design rule |
|---|---|---|
| Mechanism diagram | the social process linking concepts | every arrow is a stated mechanism, not a vague "influences" |
| Process model | sequence/feedback over time | each stage is a defined concept; loops are theorized, not decorative |
| Typology (2x2 / nxm) | a classification along dimensions | dimensions are clearly named; every cell is a distinct, defined type |
| Concept map | the relations among a concept family | hierarchy/overlap shown deliberately; no orphan nodes |
| Comparison table | this theory vs. rivals, conceptually | rows are theoretical properties, never empirical results |
Design discipline (the "earn its keep" test)
A conceptual exhibit must do work the prose cannot do as efficiently. Before including it, pass all three tests:
- Readability test. Can a reader reconstruct the core argument from the figure plus its caption alone? If not, it is decoration.
- Concept test. Is every box/cell/node a concept that is defined in the text and labeled identically? A box with an undefined label is a hole in the theory.
- Mechanism test. Does every arrow / transition name the mechanism it represents? "A → B"
with no mechanism is exactly the failure
soctheory-theory-constructionwarns against, reproduced graphically.
Typology construction
A good ST typology is a theoretical claim, not a filing system:
- The dimensions must be conceptually motivated (why these two axes?), not convenient.
- Every cell must be a real, distinct type — name it and define it; an empty or forced cell signals the dimensions are wrong.
- The typology should generate propositions (e.g., types in cell 3 behave differently
because of mechanism M), tying back to
soctheory-theory-construction. - Watch scope: a typology that implies coverage beyond the stated domain over-claims (see
soctheory-boundary-conditions).
Propositions table (optional, for proposition-heavy papers)
If the paper has many propositions, a compact table listing P1…Pn with their one-line mechanisms helps reviewers track the theory. It is a theoretical table — propositions and mechanisms only, never coefficients or significance.
Format notes
- Figures should meet the publisher's resolution/format requirements; confirm exact specs on the SAGE/ST author page (检索于 2026-06；以官网为准).
- Figures and tables count toward ST's overall length limit (all material is inclusive of the
word cap; see
soctheory-writing-style/soctheory-submission), so an exhibit must earn its space against the cap, not just against the page.
Checklist
- Every exhibit is conceptual (mechanism / process / typology / concept map / theory-comparison)
- No data plot, scatterplot, coefficient figure, or results table anywhere
- Each box/cell/node is a concept defined in the text with an identical label
- Each arrow/transition names its mechanism
- The figure passes the readability test (argument recoverable from figure + caption)
- Typology dimensions are motivated and every cell is a distinct, defined type
- Exhibits respect the stated boundary conditions and the word/length cap
Anti-patterns
- Boxes-and-arrows with undefined labels or unlabeled arrows
- A data plot or regression table imported from an empirical template (impossible at ST)
- A 2x2 with a forced or empty cell, revealing mis-chosen dimensions
- A figure that merely restates a sentence, adding nothing
- Labels in the figure that differ from the concept names in the text
- A typology that implies coverage outside the theory's domain
Output format
【Exhibit type】mechanism diagram / process model / typology / concept map / theory-comparison
【Concepts in exhibit】all defined in text + identically labeled: yes / fix [...]
【Arrows】each names a mechanism: yes / fix [...]
【Readability test】argument recoverable from figure+caption: yes / no
【Typology】dimensions motivated; every cell distinct: yes / n/a
【Next step】soctheory-contribution-framing
Version History
- 1839142 Current 2026-07-05 14:27
