Methods · Dialogue, argument & inquiry
C8
Definition forcing
Be made to say exactly what you mean by each vague word, until there are none left.
When to use it
- You can recite the definition, but your explanation still uses foggy words that do real work.
- A seminar or tutorial is coming and you need precise terms under pressure.
- Two sides keep talking past each other and you suspect a word is doing double duty.
What to ask AI
Type something like: “I'm going to explain [topic]. Every time I use a word that's doing real work but I haven't defined, stop me.”
Tell it to ask what you mean, not to supply the textbook definition first.
Answer with your own definition before you accept a correction.
How it helps
Vague words let you sound fluent while leaving the hard part unspecified.
Definition forcing stops you at each load-bearing word until you say what you mean.
Often you discover you were using one word for two ideas.
Precision is not pedantry here, it is the difference between a claim you can test and a claim you can only defend with tone.
How to run it
- Explain the idea in your own words without looking anything up.
- When the AI flags a vague word, define it yourself in one sentence.
- If you cannot, that word is the gap, look it up after you notice the blank.
- Re-explain with the new precision and invite another pass.
- Stop when every working word has a definition you could defend to a classmate.
Example exchange
Natural selection is when organisms adapt to their environment over time.
Stop, what do you mean by “adapt” in that sentence?
They change to fit better.
Do individuals change during their lives, or does the population’s mix of traits change across generations? Pick one and redefine.
Copyable prompt
I'm going to explain [TOPIC]. Every time I use a word that's doing real work
in my explanation but that I haven't defined, stop me and ask what I mean.
Do not give me the textbook definition until I have tried. Begin by asking me
to explain in my own words.
The Tell
If you defined every term and still cannot say what would count as a counterexample, the definitions were decorative.
If the AI supplied definitions before you tried, you practised recognition, not precision.
The method worked when a previously fuzzy word became a claim you could falsify.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Forcing precise definitions is conceptual clarification from philosophy and concept-learning pedagogy. Reducing vagueness makes claims evaluable.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI may over-flag ordinary words or under-flag technical ones. You decide which words do real work. Delivery remains speculative.