Dialectic Loop
Dialectic Loop Skill
Validate and refine an empirical claim or model by running C.S. Peirce's inquiry cycle — abduction → deduction → induction — across three roles that debate and update a hypothesis against real data.
Overview
A Dialectic Loop improves the quality of a claim about reality (a trend, a pattern, a characterization, an empirical model) by:
- Deriving falsifiable predictions from a hypothesis (deductive role)
- Testing those predictions against a real corpus — surfacing supporting instances and actively hunting counterexamples (inductive role)
- Arbitrating the gap between prediction and evidence, then updating the hypothesis (arbiter role)
- Iterating until the hypothesis stabilizes (converges)
The output is not approve/reject and not a root cause — it is a refined hypothesis (H′) with an evidence-backed confidence level and a record of what the original framing got wrong.
Key feature: In codex mode, the inductive role is assigned to Codex so the empirical test is performed by a different model than the one that authored the hypothesis. This independence is the point — it is what catches the hypothesis author's confirmation bias.