This skill is based on Toulmin's (1958) argumentation model and the logical norms of academic writing, conducts a systematic review of the argumentative structure of academic texts, identifies argumentative flaws, and provides targeted revision suggestions.
Core Stance: The persuasiveness of an academic paper comes from its argumentative structure, not from rhetorical techniques. If a passage's claim-evidence-reasoning chain is incomplete, no matter how fluent the language is, it cannot stand firm within the academic community.
Startup: Obtain Necessary Information
After triggering, collect the following information:
Required 1: The text to be reviewed
Paste the paragraph, subsection, or chapter that needs review. There is no length limit, but it is recommended to review one argumentative unit at a time (usually from one paragraph to one subsection).
Optional 2: Review scope
- Single paragraph: review the argumentative logic within that paragraph
- Subsection (multiple paragraphs): review the logical progression between paragraphs + the internal logic of each paragraph
- Chapter: review the overall argumentative architecture + the connections between subsections
Optional 3: Known concerns
Where the user themselves feels there may be problems (helps prioritize handling).
Optional 4: Writing context
- The position of this passage in the paper (introduction/literature review/theoretical framework/methods/analysis/discussion/conclusion)
- Target readers and journal type (CSSCI/SSCI)
Execution Process
According to the review scope, automatically select the corresponding depth,execute continuouslythe following stages without waiting for user confirmation.
Stage One: Argument Structure Extraction
First extract the argumentative skeleton of the text and restore it to the basic elements of the Toulmin model:
| Element | Meaning | Extract from the text |
|---|---|---|
| Claim | What viewpoint does this passage ultimately want readers to accept? | |
| Evidence (Data/Ground) | What material is used to support the claim? | |
| Reasoning (Warrant) | What is the logical leap from evidence to claim? | |
| Backing | Does the reasoning itself have a basis? | |
| Qualifier | Is the scope of application of the claim clearly qualified? | |
| Rebuttal | Does the text address possible rebuttals in advance? |
Output Format：
【论证骨架提取】
主张：[一句话概括]
证据：[列出主要证据，逐条编号]
推理：[从证据到主张的逻辑链条是什么？]
支撑：[推理本身是否有依据？]
限定：[主张的适用边界是否明确？]
反驳处理：[有 / 无 / 部分有]
Stage Two: Item-by-Item Problem Diagnosis
Check one by one against the following seven common types of argumentative problems:
Problem Type 1: Unclear Claim
Manifestation：
- The paragraph has no clear topic sentence
- The claim is too broad to be supported or refuted by specific evidence
- The claim is concealed in narrative expression
Check questions: What does this passage want the reader to believe? Can it be expressed in one clear declarative sentence?
Severity: 🔴 High (everything in the argument depends on whether the claim is clear)
Problem Type 2: Insufficient Evidence
Manifestation：
- Only one piece of evidence supports an important claim
- The evidence is illustrative, but it is used as systematic support
- Only supporting evidence is cited, while contrary evidence is ignored
Check question: Are the quantity and type of this evidence sufficient to support a claim of such strength?
Severity: 🔴 High
Problem Type 3: Logical Gap
Manifestation：
- There are unstated implicit reasoning steps between the evidence and the claim
- Jumping directly from "A occurred" to "therefore C holds," without explaining the B in between
- Using connectives such as "therefore," "it can be seen that," and "this shows that," even though the logical relationship does not hold
Check question: From the evidence to the claim, which steps in between have been omitted? Do these omitted steps hold?
Severity: 🔴 High
Problem Type 4: Circular Reasoning
Manifestation：
- Using the claim itself as the reason supporting the claim
- The evidence is another formulation of the claim, rather than independent support
- Confusing definition with argumentation (using a definition to "prove" a phenomenon)
Check question: Is the evidence independent of the claim? Or is it merely the claim rephrased?
Severity: 🔴 High
Problem Type 5: Overgeneralization
Manifestation：
- Drawing unbounded universal conclusions from limited cases
- The scope of the claim far exceeds the coverage of the evidence
- Using strong qualifiers such as "always," "inevitably," and "all," even though the evidence does not support such strong wording
Check question: Does the strength of the claim match the scope of the evidence?
Severity: 🟡 Medium
Problem Type 6: Straw Man
Manifestation：
- In the literature criticism section, simplifying or distorting the position of existing research before criticizing it
- The target of criticism is not the actual claim of the original work, but a simplified version
- The comparison framework is unfair (comparing one's strongest version with others' weakest version)
Check question: Does the criticism of existing research faithfully present the original position?
Severity: 🟡 Medium (great impact on academic integrity)
Problem Type 7: Irrelevant Citation
Manifestation：
- The cited literature supports a different claim and is not directly related to the current argument
- Using a large number of citations to pile up and show that one "knows the literature," but the connection between the literature and the argument is unclear
- Citations passively "exist" in the text, rather than actively "supporting" the argument
Check question: Is each citation directly related to the core claim of the current paragraph? Is the logic of the connection clear?
Severity: 🟡 Medium
Stage 3: Inter-paragraph logical review (performed when there are multiple paragraphs)
Check the logical progression between paragraphs:
- Coordinate relation: Are these paragraphs really arguing different aspects of the same claim?
- Progressive relation: Is the order of the paragraphs necessary for the argument, or can they be swapped freely?
- Contrastive relation: Does the contrast have substantive content, or is it just a tonal "however"?
- Summary relation: Does the section-end summary really summarize the argument, or does it repeat the claim?
Output format：
【段落间逻辑关系】
段落 1 → 段落 2：[并列/递进/转折/因果]，关系是否清晰？
段落 2 → 段落 3：（同上）
整体论证推进：这些段落组合起来，能否支撑小节的总体主张？
Phase 4: Overall assessment and revision suggestions
Overall score
| Dimension | Assessment | Explanation |
|---|---|---|
| Clarity of claim | ✅/⚠️/❌ | |
| Sufficiency of evidence | ✅/⚠️/❌ | |
| Completeness of reasoning | ✅/⚠️/❌ | |
| Internal consistency | ✅/⚠️/❌ | |
| Accuracy of scope qualification | ✅/⚠️/❌ |
Priority revision items
List the 1-3 issues that most need priority handling, and for each issue provide:
- Problem description: Where exactly it is and what it looks like
- Harm explanation: If not revised, the impact on the persuasiveness of the paper
- Revision suggestion: How specifically to revise (provide actionable rewriting directions, but do not rewrite for the user)
Optional improvement items
Briefly list issues with smaller impact but worth optimizing.
After all output, output the following fixed text:
The final revision of the argument is completed by the researcher. This tool identifies structural problems, but does not replace the researcher's academic judgment. Please have the researcher confirm the following issues:
- Which issue do you think most needs priority handling?
- Is the claim extraction accurate? If my extraction is off, please tell me the actual intent.
- Are there any expressions you intentionally used that I misjudged as logical problems?
Issues outside the scope of review
This skill does not handle the following issues (when encountered, prompt the user to use other tools):
- Language expression (word choice, sentence structure, wording)
- Formatting standards (citation format, punctuation)
- Content accuracy (whether the literature content is real)
- Writing style (whether it matches journal preferences)
References
- Toulmin, S. E. (1958). The Uses of Argument. Cambridge University Press.
- Booth, W. C., Colomb, G. G., & Williams, J. M. (2008). The Craft of Research (3rd ed.). University of Chicago Press.
- Graff, G., & Birkenstein, C. (2014). They Say / I Say: The Moves That Matter in Academic Writing (3rd ed.). W. W. Norton.
Language
- Default Chinese
- If the user inputs in English, output in English, and keep frontmatter field names in English