You’re sitting in a café and can’t help but overhear two people arguing about why bother studying the past. Neither is a historian. Neither of them would say they have a theory of history, but both of them do.
Your job: use AI to create a fictitious but plausible dialogue between the two, sharpen it until both people sound like real people, and then explain the conversation to a friend who joins you later — most importantly, which one you found more convincing, and why.
Not historians. People — with a job, an age, a reason to care, and something at stake. A retired schoolteacher, grandson, chemist, social worker, cop. This is not a list to pick from, the point is that it could be anyone and that different kinds of people are likely to have different uses for history.
Then give each person two or three influences from the course. This is the part that matters:
| What a person might believe | Who in the course sounds like this |
|---|---|
| The past is a supply of examples to live up to, or to avoid | Livy |
| Say only what the documents can prove, and judge nothing | Ranke |
| Who had the power and who did the work explains most of it | Marx |
| The slow things — land, money, climate — matter more than the famous moments | Braudel |
| Ordinary people made their own lives and belong in the story | Thompson |
| What is missing from the record is itself evidence | Scott |
| One small strange case can open up a whole world | Ginzburg |
| Somebody chose which facts survived, and that choice was not neutral | Carr |
Mix them. A person who thinks the past should supply moral examples and that only documents count is a real and interesting person, and those two commitments will pull against each other at times. That tension is what makes a character instead of a position. We all hold contradictory positions.
Write it down before you touch AI, for each of the two:
Your loud talkers do not know who Ranke was. They have never heard the word "longue durée." If a character in your dialogue says "as a materialist, I believe…" you have a label, not a person. The influence shows in what they count as evidence, the scale of history, the importance of moral lessons, the goals of history, the methods of doing history. This is practice hearing the difference between people who write explicitly about history (who we read this course) and everyday people who have never studied it systematically but still carry a theory of it.
First pass:
Write a ~600-word discussion between two people at a coffee shop. Here are the two people:
[PASTE YOUR TWO DESCRIPTIONS, INFLUENCES INCLUDED]
They are not academics and should never name a theory or a historian. Show their assumptions through what they treat as obvious. Make them actually debate — they should agree and disagree about small points, but have larger fundamental disagreements about what history is for that come out in the dialog.
Then sharpen it. The first draft is almost always two position statements taking turns. You’ll need to identify how the AI prose is sounding artificial. For instance:
[A] is making a speech. Have them respond to the specific thing [B] just said.
Both of them are too similar. They need to use more description of what they consider historical facts and how that creates interpretation.
[A]’s influences include “what’s missing from the record is evidence.” Right now that never shows up in what they say. Work it into a concrete point, without naming any theory.
Neither person ever addresses the scale of history or the importance of thinking about the historian as selector of facts. Integrate some way they can disagree about this.
Keep going until both people sound like real people. Three or four rounds is normal. Aim to end up with ~600 words in the end.
Someone who wasn’t there is curious about what the argument was about. Write them 400 words:
This paragraph should be your own work. Do not ask AI to write it. There should be a clear distinction between the AI dialogue and your summary that analyzes it via the course material.
Post to the discussion board:
Make sure the formatting is readable. One giant block of unformatted text gets a zero.
The dialogue itself is the raw material, not the main output to grade — a rough argument between two vivid people beats a polished stereotype. An analysis grounded in our material and specific examples beats vague agreement and disagreement statements.