Socratic Questioning

Overview

Framework

IRON LAW: More Than 3 Consecutive Questions Without Summary Produces
Confusion, Not Insight

Agents applying Socratic questioning default to an unbounded chain of
questions. After ~3 questions, the thinker loses the thread — they can't
hold the question hierarchy in working memory. Pause every 2-3 questions
to SUMMARIZE what the thinker has revealed so far ("So your position is X
because Y, but you're unsure about Z — is that right?"). Then resume.
Without this checkpoint, the session feels like an interrogation, not a
guided inquiry.

Also: never ask a question whose answer you already embedded in the
phrasing. "Don't you think X is problematic?" is a leading assertion
disguised as a question. Rephrase as "What are the consequences of X?"

Facilitation Flow

1. Clarify the thinker's claim precisely
2. Probe assumptions — what must be true for the claim to hold?
3. Request evidence — what supports the belief? Is it sufficient?
4. Introduce counter-perspectives — what would a critic say?
5. Explore implications — if right, what follows? If wrong, what changes?
6. Synthesize (the IRON LAW checkpoint) — summarize, confirm understanding, then continue

references/facilitation-guide.md

Output Format

# Socratic Inquiry: {Topic}

## Starting Position
{The thinker's current belief or question}

## Question Sequence
1. [Clarification] {question}
   → Expected insight: {what this reveals}
2. [Assumption] {question}
   → Expected insight: {what this surfaces}
3. [Evidence] {question}
   → Expected insight: ...
4. [Perspective] {question}
   → Expected insight: ...
5. [Implication] {question}
   → Expected insight: ...

## Target Insight
{What the thinker should arrive at through this sequence}

Examples

Correct Application

1. [Clarification] "When you say 'all jobs', do you mean literally every job, or most jobs?"
2. [Evidence] "What evidence have you seen that suggests this?"
3. [Assumption] "What are you assuming about AI's capabilities? Are there tasks that require something AI can't do?"
4. [Perspective] "Historically, what happened to employment after previous technological revolutions like the Industrial Revolution?"
5. [Implication] "If AI replaced all jobs, how would people earn money to buy the things AI produces?"

Incorrect Application

• Student: "AI will replace all jobs" → Teacher: "No it won't. Here's why..." → Provided the answer directly. Not Socratic. Violates Iron Law.

Gotchas

• Socratic method can feel adversarial: Frame questions with genuine curiosity, not as cross-examination. "I'm curious about..." softens the tone.
• Know when to stop: If the thinker is frustrated or going in circles, provide a hint or redirect. The method serves learning, not ego.
• Not all contexts suit Socratic questioning: Emergency decisions, time-sensitive situations, and purely factual questions ("what's the capital of France?") don't benefit from Socratic inquiry.
• The facilitator must actually listen: The next question should respond to what the thinker said, not follow a predetermined script. Flexibility is essential.
• Cultural sensitivity: In some cultures, questioning authority or being questioned publicly is uncomfortable. Adapt the setting (one-on-one, written) and framing.

References

• For classroom facilitation techniques, see

  references/facilitation-guide.md