Skip to main content
The synthesis must be validated by the original monks — not by you, and not just by the user. This is how Hegelian dialectics works: the thesis and antithesis must each recognize themselves as preserved-but-elevated in the Aufhebung.

Critical insight: Elevated vs defeated

The monks don’t stop believing after the synthesis. A defeated monk has dropped its belief load — the belief fell on the floor rather than being sublated into something larger.A properly elevated monk believes more — it sees its original position as a partial truth within a larger truth, and it now believes the larger truth.The validation checks whether belief was transformed or merely defeated.

Validation process

Send a condensed summary of the determinate negation and sublation to Monk A and Monk B. They need the key moves (what was cancelled, preserved, elevated) and concrete proposals, not the full philosophical argument.
If your environment supports session resumption, call back into the same sessions for conviction context. If not, include a summary of the agent’s original argument in the prompt.

Model selection for validation

Use the strongest available model with extended thinking for all validation — monk validation is more subtle than it appears (the monk must reason about whether its core insight was genuinely preserved or quietly destroyed).The hostile auditor (below) especially needs maximum reasoning, since its value comes from catching what everyone else missed.

Validation prompt structure

You argued passionately for [POSITION]. Here is a summary of your argument:
[CONDENSED SUMMARY OF THIS AGENT'S ESSAY — or omit if resuming session]

A dialectician has analyzed the structural contradiction between your
position and your opponent's, and proposed a synthesis.

Here is the structural analysis:
[CONDENSED SUMMARY OF PHASE 4 — key moves only]

Here is the proposed synthesis:
[CONDENSED SUMMARY OF PHASE 5 — what's cancelled, preserved, elevated,
and the concrete proposal]

Evaluate this honestly from your committed position. Answer:

1. Does this synthesis PRESERVE your core insight? What specifically
   does it get right about what you were arguing?

2. Does it reveal a genuine limitation in your position that you can
   now see? What were you missing? What assumption were you trapped in?

3. Do you feel ELEVATED or DEFEATED? "Elevated" means: "I see my
   position as a partial truth within a larger truth I couldn't have
   reached alone." "Defeated" means: "My position was just dismissed
   or diluted." Be honest — if you feel defeated, say so and explain
   why.

4. If you feel elevated: what NEW questions does this synthesis create
   that you want to explore?

Adversarial check

Before accepting the validation responses, apply the adversarial check: Would the hardest-hit monk actually accept this? If one position is fundamentally challenged more than the other, does its monk’s “elevated” response feel genuine or forced?
If a monk claims to feel elevated but its response reads like politeness rather than genuine recognition, the synthesis likely dismissed rather than preserved its core insight.

Hostile auditor: Fresh adversarial agent

After the monks validate, spawn a hostile auditor — a fresh agent with no position — to attack the synthesis.

Hostile auditor prompt

You are a hostile auditor. Your sole job is to find flaws in a proposed
synthesis of a dialectical contradiction.

Here is the synthesis:
[FULL SYNTHESIS]

Attack it for:
1. Hidden assumptions the synthesis smuggled in
2. Compromise disguised as transcendence (did it just split the difference?)
3. Structural flaws (does it actually make the contradiction predictable?)
4. Emergent structure test failure (is every element traceable to one
   monk or the other, or does it have genuinely new organizational properties?)
5. Closure property failure (is it concrete enough to be believed and
   argued from in the next round?)

Use Pollock's defeater types:
- Undercutting defeaters: broken inferential links
- Rebutting defeaters: counter-evidence

Prioritize structural critique over evidentiary critique.

Be ruthless. If this synthesis is compromise dressed up, say so.
The hostile auditor uses the strongest available model. Its value comes from catching what everyone else missed because they were invested in making the synthesis work.

Refine based on validation

Based on monk validation responses and the hostile auditor’s critique, identify improvements. Present improvements individually to the user, incorporate accepted ones. Don’t batch refinements — present one at a time so the user can evaluate each on its merits.

When validation reveals a bad synthesis

If both monks feel defeated, or the hostile auditor identifies fundamental structural flaws, go back to Phase 5 and revise.
This is not failure — it’s the process working as designed. The validation caught a compromise before it became the foundation for the next round.

Example validation responses

Good (elevated):

“Yes, I see now. I was arguing that [POSITION] because I believed [ASSUMPTION]. But I was trapped in [FRAME]. The synthesis preserves my core insight about [X] — that’s genuinely true — but situates it within [LARGER TRUTH] that I couldn’t see from inside my position. I now believe this larger frame, and my original position looks like a partial view of it. The new question I want to explore is [NEW CONTRADICTION].”

Bad (defeated):

“I suppose the synthesis makes some good points. Both sides have merit. It’s a reasonable compromise.”
This monk has been defeated, not elevated. The synthesis dismissed its position rather than transforming it.

Next: Recursion

Phase 7: Recursion

Identify new contradictions and choose the next direction

Build docs developers (and LLMs) love