Most people read research the wrong way.
They collect papers, skim abstracts, highlight a few conclusions, and stop. That creates familiarity—not understanding. You learn what each paper says, but not how the field actually works.
The real value of a literature review is structural: where authors agree, where they conflict, what assumptions they inherit, and what questions are still open.
These nine prompts are designed for that exact job.
They are not summary prompts. They are analysis prompts. Used in sequence, they turn a pile of PDFs into a working map of a field.
1) The Contradiction Finder
Use this to surface direct disagreements and diagnose their causes.
Across all uploaded papers, identify every clear contradiction between two or more authors. For each contradiction, provide:
- the exact positions being contradicted
- the papers and authors involved
- the likely reason for the disagreement, such as methodology, dataset, time period, assumptions, or scope Present the result as a table.
Why it matters: “Paper A says X, Paper B says Y” is not insight. The insight is why they diverge.
2) The Citation Chain
Use this to reconstruct how core concepts evolved.
From these papers, identify the 3 most frequently cited concepts or ideas. For each one, trace the intellectual lineage:
- who introduced it first
- who challenged it
- who refined or extended it
- what the current consensus appears to be, if any Show the lineage as a clear chronology or family-tree style outline.
Why it matters: Good ideas do not appear fully formed. They mutate through criticism and refinement.
3) The Intake Protocol
Run this before any synthesis so you can map the landscape first.
I am uploading [X] papers on [topic]. Before answering any specific question, do the following:
- List every paper with author, year, and a one-sentence core claim.
- Group the papers into clusters based on shared assumptions or themes.
- Flag any papers that directly conflict with one another. Do not summarize the literature yet; map the landscape first.
Why it matters: Literature is organized by camps and assumptions, not by upload order.
4) The Master Synthesis
Use this only after mapping and contradiction analysis.
You now have the full literature set. Write a synthesis that does not summarize each paper individually. Instead, state:
- what the field broadly agrees on
- what remains contested
- what has been established with high confidence
- the single most important unanswered question Keep it under 400 words and remove all filler.
Why it matters: A real synthesis is a decision map, not a sequence of mini-abstracts.
5) The Gap Scanner
Use this to find research questions worth pursuing.
Based on all uploaded papers, identify the 5 most important research questions that remain unresolved. For each gap, explain:
- why it remains open
- which paper came closest to answering it
- what method, dataset, or study design would be needed to close it Focus on gaps that matter, not just gaps that are easy to name.
Why it matters: The best gaps are structurally important, not merely unfilled.
6) The Methodology Audit
Use this to identify evidence concentration and fragility.
Compare the methodologies used across all uploaded papers. Group them by research type, such as surveys, experiments, simulations, meta-analyses, or case studies. Then identify:
- the dominant methodology and why it likely dominates
- the most underused methodology
- the weakest methodology in the set, with a brief explanation of why it is weak
Why it matters: Method shapes claims. A field can look diverse while relying on one evidentiary mode.
7) The Knowledge Map Builder
Use this to generate an orientation model you can reuse.
Create a structured knowledge map of this literature. Format it as an outline with:
- the central claim the field revolves around
- 3–5 supporting pillars
- 2–3 contested zones
- 1–2 frontier questions
- 3 papers a newcomer should read first, with one-line reasons for each Keep it clean, hierarchical, and non-narrative.
Why it matters: This gives you a compact architecture for onboarding, teaching, and strategic reading.
8) The “So What” Test
Use this to pressure-test whether understanding transfers beyond experts.
Pretend I need to explain this entire body of research to a smart non-expert in 5 minutes. Give me:
- the one-sentence version of what this field has shown
- the one honest thing it still does not know
- the single most important real-world implication Use plain language. Avoid jargon and avoid hedging.
Why it matters: If it cannot survive plain-language translation, the understanding is probably shallow.
9) The Assumption Killer
Use this to expose hidden dependencies in the literature.
List the assumptions that most of these papers share but do not explicitly test or justify. For each assumption, provide:
- a clear statement of the assumption
- 1–2 papers that rely on it most heavily
- what would happen to the field if the assumption turned out to be false Focus on assumptions that could materially change the conclusions.
Why it matters: Paradigm shifts often come from breaking assumptions, not incrementally improving results.
Recommended Order (Use Them as a System)
If you run these prompts in random order, results get noisy. Use this sequence instead:
- Intake Protocol (map)
- Contradiction Finder (compare)
- Citation Chain (trace)
- Methodology Audit (stress-test evidence)
- Master Synthesis (compress)
- Gap Scanner (prioritize next questions)
- Knowledge Map Builder (package structure)
- So What Test (translate)
- Assumption Killer (challenge foundations)
Final Thought
These prompts are not productivity tricks. They are a reading system.
Used together, they move you from paper collection to field mastery: first map, then compare, then trace, then synthesize, then challenge.
If you want, I can publish a companion post with copy-ready templates for ChatGPT, Claude, and Gemini so you can run this workflow in one pass.