phg-argument-synthesis
GitHub用于将映射文献转化为Progress in Human Geography综述论文的核心论点。通过概念整合多源证据,构建超越单一来源的累积性主张,避免描述性总结,确保论点具有原创性和逻辑连贯性。
Trigger Scenarios
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill phg-argument-synthesis -g -y
SKILL.md
Frontmatter
{
"name": "phg-argument-synthesis",
"description": "Use when turning a mapped literature into one cumulative argument for a Progress in Human Geography (PiHG) review essay — synthesizing many sources into a claim the field could not see from any one of them. Builds the synthetic spine; it does not gather sources (see phg-literature-positioning) or make the critical turn (see phg-critical-intervention)."
}
Argument Synthesis (phg-argument-synthesis)
The single most important craft in a PiHG essay: synthesis. A review is not an ordered summary of what
others found; it is an argument the sources add up to but none of them made alone. Reviewers and
editors reject "annotated bibliography" pieces on sight. This skill turns a mapped literature
(phg-literature-positioning) into a cumulative claim organized by your concept
(phg-theory-building).
When to trigger
- You have the corpus and the concept, and must now build the spine
- The draft reads as "A found X, B found Y, C found Z" — a list, not an argument
- A reviewer said "this is descriptive," "where is the argument?," or "just a summary"
- You cannot state the essay's thesis in one sentence
Summary vs. synthesis (the core distinction)
| Summary (reject) | Synthesis (PiHG) |
|---|---|
| Organized by source (paper by paper) | Organized by your concept / claim |
| "X studied A; Y studied B" | "Across X and Y, the field has assumed Z — which fails because…" |
| Coverage is the goal | An argument is the goal; coverage serves it |
| Reader learns what exists | Reader learns how to think about the field |
If a paragraph's topic sentence names an author rather than an idea, you are summarizing. Rewrite so ideas are the subjects and sources are the evidence.
Build the cumulative spine
- State the thesis in one sentence. The whole essay is the defense of this claim. If you cannot, go
back to
phg-critical-intervention. - Decompose the thesis into 3–5 sub-claims. These become the sections — organized by argument, not by the literature's own buckets.
- Under each sub-claim, marshal the evidence from across the corpus: which works support it, which complicate it, and how they combine into the point.
- Make the sources talk to each other. Put works that never cite each other into the same paragraph when your concept shows they are addressing the same relation.
- Handle disconfirming work honestly — a synthesis that ignores contrary evidence is advocacy, not a review. Fold it in and show why the thesis survives (or scope it).
The "one sentence" and "adds up" tests
- One-sentence test: can you say what the essay argues (not covers) in a sentence? If not, there is no spine yet.
- Adds-up test: does the conclusion state something no single cited work says? A review whose conclusion is already in one of its sources has synthesized nothing.
Reconcile, periodize, or adjudicate
Three reliable synthetic structures (pick the one your literature needs):
- Reconcile — show two camps that talk past each other are compatible under your concept.
- Periodize — narrate how the field's thinking moved through phases, to intervene in the next.
- Adjudicate — weigh a live disagreement and argue which side the evidence, read together, favors.
Checklist
- Thesis stated in one sentence; essay is its defense
- Sections organized by sub-claims, not by the literature's buckets or by source
- Paragraph topic sentences name ideas, not authors
- Works that never cite each other are brought into contact where the concept links them
- Disconfirming evidence engaged, not ignored
- Conclusion states something no single source states (adds-up test)
Anti-patterns
- Annotated-bibliography structure (one paragraph per paper)
- Topic sentences that start with an author's name
- "Coverage" mistaken for contribution — comprehensive but argument-free
- Cherry-picking supportive work and dropping the inconvenient
- A conclusion that merely restates the most-cited paper's point
Output format
【Thesis】the one-sentence claim the essay defends
【Sub-claims】the 3–5 argument-sections
【Synthetic structure】reconcile / periodize / adjudicate
【Sources in contact】which non-citing works your concept joins
【Adds-up?】does the conclusion exceed any single source? [Y/N]
【Next】phg-critical-intervention
Supplementary resources
../../resources/worked-examples/01-introduction.md— before→after: annotated list vs. cumulative argument../../resources/external_tools.md— CAQDAS and synthesis methods (thematic, meta-narrative, critical interpretive)
Version History
- 9f86f09 Current 2026-07-19 17:22


