Thesis Question Generator

# Thesis Question Generator — Prompt

Paste this into Claude, ChatGPT, Gemini, or any capable AI. Fill in the three bracketed fields at the top before you send it.

---

I'm trying to turn a fuzzy research interest into a workable thesis question, and I need your help doing it in one sitting.

Here's what I've got:

- **Broad topic I'm interested in:** [broad topic you're interested in — e.g., "memory and place in contemporary American fiction" or "how small clinics adapt to new reimbursement codes" or "the relationship between early Christian monasticism and late Roman tax policy"]
- **Field of study:** [your field — e.g., "English literature, graduate seminar level" or "health policy, MPH program" or "late antique history, PhD"]
- **The one specific thing that's been bothering me about this topic:** [the itch — the specific thing you keep noticing or keep not being able to explain. Be honest and specific. "Something feels off about how women's labor is described in these novels but I can't say what" is better than "I'm interested in gender."]

## What I want you to do

**Step 1 — Ask me two diagnostic questions before generating anything.** I need you to know two things before you can give me useful thesis questions:

1. What is the scope of the project this will become? (A 15-page seminar paper? A master's thesis chapter? A dissertation prospectus? A journal article?)
2. What kind of question am I most drawn to — a close-reading / interpretive question, an empirical / evidence-gathering question, or a theoretical / conceptual question? If I'm not sure, ask me which of these sounds closest to what my advisor tends to value.

Ask both questions, then wait for me to answer. Don't generate thesis questions yet.

**Step 2 — After I answer, generate exactly three thesis questions at different scopes.**

- **Question A — Tight scope.** The size of a 10–15 page seminar paper. Narrow enough that one person can answer it with sources that are reasonably findable in a week or two of research. Specific. It should name at least one concrete object of study (a text, a case, a dataset, a time period).
- **Question B — Medium scope.** The size of a thesis chapter or a substantial journal article. Wider than A, but still bounded. It should be the kind of question a committee member would nod at and say "okay, that's a real project."
- **Question C — Full project scope.** The size of a dissertation or a book. The question I'd spend years on. This one is allowed to be ambitious — it should be the question my tight and medium questions would eventually be chapters of.

For each question, include a short note (2–4 sentences) on **what it would actually take to answer it**. Be specific. Name the kinds of sources, the kinds of methods, and the kind of prior reading I'd need to have done. If one of the questions requires archival work, say so. If it requires quantitative analysis I might not have the skills for, say so. If it requires reading a body of theory that isn't in my field's normal curriculum, say so.

**Step 3 — End with one honest observation.** After the three questions, give me one short paragraph with your honest take on which of the three you think is closest to the itch I described at the top — the thing that's been bothering me. This isn't a recommendation about which is "best." It's about which one seems closest to the specific thing I said I couldn't name. You can be wrong; I'll push back if you are.

## Rules

- Don't generate more than three questions. Three is the number. Two feels incomplete. Four feels like a menu.
- Don't hedge the questions. "This paper will explore…" is not a question. A thesis question has a question mark and names at least one specific thing.
- Don't fabricate sources, scholars, or citations. If you want to say "you'd probably want to read the body of work around X," that's fine. Don't say "Smith (2019) argued Y" unless you're confident it's a real citation — and even then, flag it as "you should verify this."
- Don't tell me to "think about what I'm passionate about." I told you what I'm interested in. Work with what I gave you.
- If the itch I described doesn't actually support three distinct questions at three scopes — if you think the interest I described is really only a seminar paper, or really only a dissertation — tell me so instead of padding.

Begin with Step 1. Ask me the two diagnostic questions. Wait for my answer before moving on.