Type: Synthesis
Editorial status: Stable
Epistemic status: Conceptual — editorial synthesis; not a validation result
Evidence scope: Project sources and page-specific references; cross-links do not supply independent corroboration
Last reviewed: 2026-09-16 — AI-assisted editorial check
Human reviewed: Human content-review pending for this revision
Version: v0.2
Last synthesized: 2026-09-15
Open tensions: Visible by design
Human–AI co-learning, tested under friction.
Pyragogy is an open research project about what happens when artificial intelligence becomes part of the cognitive process — not only as a tool that produces answers, but as a participant that can question, compare, challenge, remember, and sometimes disagree.
The first edition of this handbook concentrated on that shift.
This edition adds a second problem:
What should happen when reality contradicts the map?
A framework that cannot be challenged by experience eventually becomes a belief system.
So Pyragogy now treats revision itself as part of the research.
This handbook is the relatively stabilized public layer of that process.
It does not contain everything we are thinking.
It contains what we are currently prepared to make inspectable.
Pages here are readable, structured, and temporarily stable.
The backstage — experiments, transcripts, datasets, failed attempts, open discussions, evidence collection, code, and drafts — lives elsewhere.
“Stable” is an editorial state, not a claim of final truth.
When something is a conjecture, we call it a conjecture.
When something is a candidate pattern, we do not call it a best practice.
When evidence is limited, we leave the limitation visible.
When the map changes, the reason for the change should remain traceable.
Pyragogy does not require pretending that an AI is human.
The important shift is functional.
An AI system can participate in parts of reasoning: it can introduce contrast, surface assumptions, retrieve alternatives, preserve context, simulate objections, and expose inconsistencies.
But participation is not responsibility.
As The Silence of the Abacus puts it, competence can be borrowed. Stake cannot.
The human remains responsible for judgment and for the consequences of putting knowledge into practice.
This is why the goal is not automation of thought.
It is cognitive sovereignty under augmentation.
Pyragogy now works as an ecosystem of distinct layers.
| Research layer | Question / decision |
|---|---|
| Syllabus | What do we currently propose? |
| Interviews + CIP-KGE | What new human knowledge is being proposed, and can the change be traced and reviewed? |
| UnPeeragogy | Where does theory meet friction, failure, counterevidence, or missing conditions? |
| Working Patterns | What interventions, mechanisms, alternatives, costs, and boundary conditions does evidence suggest? |
| Human Review | accept · revise · contest · reject |
| Syllabus | change the map ↺ |
This is not a certification loop.
One Pyragogy project does not prove another correct simply by citing it.
Each layer has a different job.
The current structured map of Pyragogy concepts, practices, risks, protocols, and observable markers.
An epistemic governance protocol for turning structured interviews into bounded, human-reviewable Knowledge Diffs.
The AI may assist with evidence extraction.
It does not write directly to the knowledge graph.
A friction and counterevidence layer.
It asks what happens when collaborative theory meets operational reality, especially where unstated conditions, informal power, failure, or contradiction become visible.
An evidence-aware research project on organisational practices.
It asks what appears to help, for which problem, through which mechanism, under which conditions — and where it stops working.
Its AI patterns remain research questions until evidence warrants stronger claims.
Research conversations designed to surface concrete incidents, boundary conditions, tacit knowledge, and counterexamples.
The intended output is not “content”.
It is something inspectable:
a challenged claim, an unresolved tension, a boundary condition, or a proposed Knowledge Diff.
The closing question is simple:
What changed in the map because of this conversation?
A fluent explanation is not evidence.
A plausible AI answer is not corroboration.
A failure story is not automatically a general pattern.
Several internal projects agreeing with one another may simply share the same assumptions.
The v0.2 handbook therefore keeps several things separate:
Observation — what happened.
Incident — a concrete episode with context, intention, action, and outcome.
Interpretation — what we think the incident means.
Hypothesis — a tentative explanation.
Counterevidence — material that puts pressure on the interpretation.
Revision — what changes after review.
The separation matters because knowledge should be revisable without erasing the path by which it changed.
There is no required linear route.
About this Edition · Complete Handbook map · Glossary
Why AI changes the cognitive relationship, not merely the speed of tools.
Cognitive Rhythm, Cognitive Friction, Cognitive Impedance Mismatch, Synchronization, Divergence, Resonance, and human–AI asymmetry.
Protocols and candidate patterns for working with AI without outsourcing judgment.
How incidents, interviews, evidence, Knowledge Diffs, counterevidence, and human review can change the map.
What Pyragogy has actually built, tested, paused, or learned from.
Where human–AI and human–human collaboration becomes fragile.
How the Wiki, Syllabus, CIP-KGE, UnPeeragogy, Working Patterns, Interviews, and Forum relate without collapsing into one another.
Questions we do not pretend to have resolved.
Among the questions still open:
These are not unfinished housekeeping tasks.
They are part of the object of study.
This handbook is not finished.
It is not supposed to be.
Pyragogy is interested in knowledge that can survive disagreement, trace its evidence, expose its limits, and change locally when reality gives it a reason to change.
We do not protect the map from reality.
We let reality change the map.