넓은 리포트 주제를 답할 수 있는 질문으로 좁히는 프롬프트

Narrow this topic into an answerable research question.

Work through four axes, showing each narrowing step:
1. Scope — which part of the topic
2. Population or case — whose, where
3. Period — when
4. Point of contention — what is actually disputed, since a question with no dispute produces a summary, not a paper

Output:
- Three candidate questions at different levels of ambition.
- For each: what you would need to answer it, roughly how many sources, and whether it fits my length.
- The one I should pick, and why.
- Questions that look narrow but are not — a common trap where a specific-sounding question still requires the whole field.

Rules:
- *Do not propose a question that cannot be answered with library and web sources* unless I said I can collect data.
- A good question has a plausible answer that could turn out either way. Reject candidates that can only be answered one way.