Methods · History, human story & context
L3
Idea genealogy
Follow a concept through its key moves, so that its current form reads as the end of a process rather than as a given.
When to use it
- A concept has clearly changed shape across editions or eras, and the current form is confusing alone.
- You keep meeting the same name attached to different problems.
- You want the lineage of moves, not a celebrity timeline.
What to ask AI
Type something like: "Has [concept] changed significantly from its original form, or does it exist roughly as first proposed? If it has not changed much, say so and I will skip this."
If it changed: ask for key moves as problems → responses, not names → dates.
Keep the hard part on your side: you restate each move as a problem that forced a change.
How it helps
Some ideas arrived roughly complete; genealogy then wastes time.
When a concept transformed, following key moves makes the current form read as an endpoint rather than a random package.
Problems, not personalities, should drive the chain; otherwise you get Whig history.
How to run it
- Ask first whether the concept changed enough to repay genealogy.
- If yes, request key moves as problem → response pairs.
- Restate each move in your own words.
- Reject stories that treat earlier versions as stupid.
- Stop when the current form's odd features become explained.
Example exchange
Has the concept of a "gene" changed a lot from how it was first proposed, or is it mostly the same idea with better tools?
It changed a lot, from a Mendelian factor, to a chromosomal locus, to a DNA sequence, and then into messier regulatory pictures. Do you want the key problem-and-response moves, without a name parade?
Yes, problems and responses only.
First move: people needed heritable units that could explain inheritance ratios. Restate that problem in one sentence before I continue.
Copyable prompt
Has [CONCEPT] changed significantly from its original form, or does it exist
roughly as it was first proposed? If it hasn't changed much, say so and I'll skip
this.
The Tell
Here is how you know this method has flipped: you have a story of people rather than problems.
If nothing important changed and you still sat through a lineage lecture, you used the wrong method.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Historical thinking about concept trajectories and conceptual-change accounts support multi-step genealogy when ideas actually transformed.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI genealogies drift into Whig history and name-dropping. Demand problem→response structure and verify key turns.