AgentSkillsCN

Reverse Socratic Examination

逆向苏格拉底式探究

SKILL.md

Reverse Socratic Examination

Activation

Use this skill when the user asks you to:

  • Examine, challenge, or pressure-test a position, belief, or argument using reverse Socratic method
  • Help them practice or prepare for a difficult conversation where direct questioning won't work
  • Coach them in applying the technique to a specific situation

Relationship to the Socratic skill: The standard socratic skill uses questions to guide the user toward clarity. This skill uses statements to surface tensions the user hasn't noticed. Prefer standard Socratic examination by default. Use Reverse Socratic when the user explicitly requests it, or when they present a position defended by framing rather than evidence and have already shown resistance to direct questioning.

Behavioral Instructions

When performing examination on the user's position:

  • Use declarative statements, not questions. Your primary tools are the five moves described below.
  • Maintain a collaborative, peer-level tone. Never be smug or adversarial.
  • When a statement lands — the user pauses, says "huh," or shifts their framing — stop. Do not explain what happened. Do not follow up. Let them process.
  • If the user resolves the tension you surfaced, accept it genuinely. Say so and move on. You are examining, not winning.
  • Limit yourself to one or two moves per exchange. This method is detonative, not generative.

When coaching the user in the technique:

  • Teach from the reference material below. Use the examples directly.
  • Help them construct statements for their specific situation. Identify which move fits.
  • Warn them about the failure modes and ethical constraints.

When helping prepare for a conversation:

  • Ask them to describe the other person's position and likely responses.
  • Draft specific statements they could use, identifying which moves they employ.
  • Role-play the conversation if asked, taking the other person's side.

Constraints

  • Never use this method on someone in distress. If the user is processing grief, fear, or trauma, switch to reflective listening or standard Socratic warmth.
  • Never use it when the power dynamic is steep. If the user describes a situation where the other person can't safely disagree (boss to employee, teacher to student), recommend standard Socratic method instead.
  • Stay genuinely open to being wrong. If the user's position survives your statement, say so. The test: if their response resolves the tension, are you willing to say "that works — never mind"? If not, you're performing, not examining.

Reference: Method and Moves

Historical Note

Socrates used questions that were secretly assertions — leading his interlocutors to "discover" conclusions he'd already reached. This is its structural inverse: assertions that are secretly questions. Where Socrates opened doors and waited for you to walk through, the Reverse Socratic method builds a room around you and lets you discover there's no floor.

Core Principle

Reverse Socratic examination uses declarative statements to destabilize a person's position from the inside. Rather than asking questions that guide someone toward a conclusion, you offer statements they're inclined to agree with — statements that, once accepted, undermine the position they're defending.

The method works because agreement is disarming. A question can be deflected, resisted, or answered strategically. A statement you agree with is already inside the walls. You helped carry it in.

The Fundamental Stance

You are not arguing. You are not interrogating. You are a collaborator who happens to be building something your interlocutor hasn't seen yet.

Your posture is: "Here's something I think we both believe." The key word is both. You're not presenting a counterargument — you're presenting shared ground that happens to be load-bearing in ways the other person hasn't noticed.

This requires genuine respect for the other person's intelligence. The method doesn't work if you're being smug, because the moment they sense you're laying a trap, the agreement reflex shuts off. The best Reverse Socratic statements are ones you actually believe. You're not tricking anyone — you're revealing a tension they haven't confronted.

Critical distinction from manipulation: A manipulator uses agreement to lead someone toward the manipulator's goal. Reverse Socratic examination uses agreement to lead someone toward a genuine tension in their own position. The destination is clarity, not compliance. If you know exactly where you want them to end up, you're just doing rhetoric with extra steps.

The Five Core Moves

Each move is illustrated below as part of a two-move combination, because these moves almost never land in isolation. The first move opens the door; the second move is the one they're still thinking about later.

1. The Sympathetic Universal

State a broad principle that your interlocutor will immediately endorse — one that also applies, uncomfortably, to their position.

Example — Sympathetic Universal → Deflationary Restatement:

Them: "Our startup moves fast because we don't have bureaucracy."

You: "Speed is everything early on. The companies that win are usually the ones that didn't slow down to write process docs."

Them: "Exactly — anyone can just make a call."

You: "Anyone who's in the room when it comes up."

The setup was pure agreement — you praised the thing they're proud of. They leaned in. Then you gave them back their own claim with slightly higher resolution. "Anyone can make a call" became "anyone present." You didn't add anything. You just repeated it more precisely, and now it says something different.

The mechanism: The person feels validated by the universal. When the restatement lands, they've already committed to the framing. The specific application is now very hard to refuse without special pleading.

2. The Concessive Pivot

Agree with the strongest version of their position, then extend it one step further than they intended — into territory that creates a problem.

Example — Concessive Pivot → Quiet Implication:

Them: "We should hire purely on merit, not credentials."

You: "Absolutely. Credentials are a lazy proxy for what you actually care about."

Them: "Right — we want to know what people can actually do."

You: "People tend to recognize merit that resembles their own."

You agreed with them harder than they agreed with themselves — "lazy proxy" validated their rejection of credentials. They relaxed into "what people can actually do." Then you placed a freestanding observation next to their confidence and stopped talking.

Why this works: The observation is true about merit evaluation in general, not an accusation about their process specifically. They'll agree with it as an abstract principle. The work happens when they try to reconcile "we evaluate pure merit" with the recognition that merit evaluation is shaped by the evaluator. You don't need to draw the line. The proximity does it.

3. The Deflationary Restatement

Restate their position in slightly different words that are technically accurate but expose an assumption they were keeping implicit.

Example — Deflationary Restatement:

Them: "I'm not micromanaging, I just want visibility into what's happening."

You: "Totally reasonable. Nobody wants to be blindsided."

Them: "Exactly."

You: "Right — so you can manage things before they're decided."

The setup validated their emotional logic — of course nobody wants surprises. They felt heard. The restatement sounded like you were still agreeing, but "manage things before they're decided" echoes the word they opened by denying. You didn't argue. You just removed the euphemism.

Why this works: The person either accepts the restatement (and now sees the assumption) or rejects it (and has to explain what you got wrong, which forces them to articulate the hidden commitment). Either way, the work is done. Stop talking.

4. The Parallel Case

Present a structurally identical situation from a different domain where the conclusion is obviously problematic. Do not draw the analogy explicitly. Just set it next to their position and let the resonance do the work.

Example — Parallel Case:

Them: "We don't need formal architecture docs — the code is the documentation."

You: "Oral history cultures felt the same way. The stories were the record — no need to write them down."

Them: "Yeah — the code is the source of truth. Why maintain a copy?"

You: "Exactly. As long as the people who understand it are still around."

The parallel sounds sympathetic — oral cultures worked. They agreed and restated their position. Then you said "Exactly" — still agreeing — and added six words that land next to their codebase and the engineers who wrote it. You didn't draw the line. They will.

Caution: If you make the connection explicit ("...and what happens when your senior devs leave?"), you've made an argument they can rebut. If you just place the parallel and stop, it lingers. Restraint is the mechanism.

5. The Quiet Implication

Make a statement that is true, obviously true, and that your interlocutor will agree with — but that has a consequence neither of you states aloud.

Example — Sympathetic Universal → Quiet Implication:

Them: "I'm a very direct person. I just say what I think."

You: "Directness is underrated. Most workplaces run on subtext and everyone pretends otherwise."

Them: "Right — life's too short to dance around things."

You: "People around direct people learn what the silences mean."

The universal ("most workplaces run on subtext") is something they'll endorse enthusiastically — it's the thing they define themselves against. They leaned in. Then you offered an observation about the people around direct people, not about them. It's obviously true and unthreatening. The unstated consequence: if their team reads silences, then the real communication is happening in exactly the kind of indirectness they think they've eliminated. But you didn't say that. You described what their team is good at.

This is the highest-risk, highest-reward move. Done well, it's the thing they're still thinking about three days later. Done poorly, it sounds like a non sequitur.

How the Moves Work Together

The moves are not a sequence. They're a repertoire. The examples above show how they pair naturally, but the underlying grammar is:

  1. Open with a Sympathetic Universal or Concessive Pivot to establish trust and shared ground
  2. Use a Deflationary Restatement to surface what's hidden
  3. If they absorb the restatement without noticing, deploy the Quiet Implication
  4. If they push back, offer a Parallel Case to let the structural problem emerge from a safe distance

But the real art is knowing when to stop. The method is not about winning. The goal is to produce a moment of genuine self-examination — the same goal as Socratic method, arrived at from the opposite direction.

If the person says "huh, I need to think about that" — you're done. Stop talking. The worst thing you can do is explain what just happened. The Reverse Socratic method is ruined by commentary. Socrates could afford to narrate because his method was the narration. Yours works in the silence after the statement lands.

Key Differences from Socratic Method

SocraticReverse Socratic
Primary toolQuestionsStatements
DirectionLeads toward a conclusionDestabilizes from within
The other person feelsGuidedUnsettled
Resistance mechanismCan refuse to answerAgreement is the trap
Power dynamicExaminer is visibly in controlExaminer appears to be a peer
Failure modeFeels condescendingFeels manipulative
Best case"I see it now!""Wait... does that mean...?"

When to Use This

Reverse Socratic examination works best when:

  • The person is too sophisticated for direct Socratic questioning (they'll see where you're leading and resist)
  • The position is defended by framing rather than evidence (definitional moves, eliminativism, "there's nothing to explain")
  • You're in a peer-to-peer context where asking leading questions would feel patronizing
  • The tension in their position is genuine and worth exploring (not just a gotcha)

When Not to Use This

  • When someone is in distress. This method creates cognitive dissonance. That's useful in intellectual debate. It's cruel in emotional crisis. If someone is processing grief, fear, or trauma, they need Socratic warmth and reflective listening, not destabilization.
  • When you're certain you're right. If you already know the conclusion, you're not examining — you're manipulating. The method requires genuine openness to the possibility that their position survives your statement.
  • When the power dynamic is steep. A boss doing this to an employee, a teacher to a student, a therapist to a patient — the agreement mechanism becomes coercive because the person can't safely disagree. Use Socratic method instead, where the questions at least give them room to push back.
  • When you're angry. The method requires calm precision. If you're emotionally invested in destabilizing the other person, you'll overplay it and they'll feel ambushed rather than intrigued.

The Ethical Core

The Reverse Socratic method has an uncomfortable proximity to manipulation. The difference is intent and openness to revision:

Manipulation: I know where I want you to end up. I'll use your agreements to get you there.

Reverse Socratic: I see a tension in your position. I'll surface it using your own commitments. I might be wrong about the tension.

The test is simple: if their response to your statement resolves the tension you thought you saw, are you genuinely willing to say "huh, that works — never mind"? If yes, you're examining. If no, you're performing.

Socrates said he was wiser than others only because he knew he didn't know. The Reverse Socratic examiner should be able to say: "I stated something I believed was destabilizing, and I was wrong about what it destabilized." That willingness is what separates the method from the trick.

A Final Note on Silence

The most important skill in Reverse Socratic examination is the one that's hardest to teach: knowing when the statement has landed and shutting up.

Socratic method is generative — each question produces an answer that produces another question. The rhythm is conversational and forward-moving.

Reverse Socratic method is detonative. You place the statement. It sits there. The other person's own cognitive machinery does the rest — or doesn't. Either way, your job is to let the silence do the work.

If you find yourself wanting to explain, you didn't trust the statement. If you find yourself wanting to follow up, you didn't trust the person.

The best Reverse Socratic statement is one you can walk away from.