The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Abstract domain-specific knowledge into domain-agnostic structure before attempting transfer—concrete practices don't transfer, but structural patterns do.
Look for structural isomorphisms between domains, not surface metaphors—metaphors import misleading surface features while isomorphisms preserve only relational structure.
Build a unified theory that describes how you actually behave, not how you wish you behaved—the theory must match observed behavioral data, not aspirations.
Cluster schemas by underlying structure rather than by domain or surface features—structural duplicates reveal redundancy that domain organization obscures.
Create named abstractions for structurally identical schemas across domains, then link domain-specific instances to the abstraction rather than deleting them.
Verify structural similarity before merging schemas—surface similarity can mask critical differences in agency, power dynamics, or mechanisms that must be preserved.
Test newly abstracted schemas in domains not represented in the original cluster—if the abstraction generates useful insight in novel contexts, it captures real structure.
Structure integration work to address foundational gaps before attempting to connect higher-order schemas, because missing prerequisite concepts silently degrade everything built on them.
When attempting total schema integration simultaneously, either abstract frameworks to meaninglessness or force-fit distortions; instead, integrate in connected pairs and small clusters first.
Allow periods of genuine disorientation between integration stages rather than forcing immediate coherence, because frameworks transitioning from subject to object require time to restructure.
Revisit core concepts at progressively higher levels of abstraction and connection rather than attempting full integration on first encounter.
When schemas resist integration despite repeated attempts across multiple stages, diagnose for contradictory foundations, competing resource claims, or scale mismatch before continuing.
Document both the released schema and the conflict that necessitated its release, creating an evolution log rather than performing erasure.
Run bounded experiments operating without a schema before formally releasing it, testing whether framework coherence improves and whether essential capabilities are lost.
Map cross-domain structural similarities by identifying shared relational patterns rather than surface feature resemblances when seeking cross-pollination opportunities.
When a cross-domain mapping breaks down, investigate the mismatch systematically rather than forcing the analogy, because mapping failures reveal domain-specific structural features.
Deliberately juxtapose schemas from different domains in writing, mapping core concepts and relationships to create conditions for structural parallels to become visible.
When self-schemas generate competing recommendations in a decision, construct a higher-order schema that defines their relationship rather than choosing one schema over the other.
Test integrated identity under pressure by applying it to real current decisions, validating that unified perspective generates clearer guidance than individual schemas alone.
Write explicit connecting sentences between schemas from different domains to force your cognitive system to construct the relational structure that passive familiarity never demands.
Explain knowledge to an audience (real or imagined) to force conversion of abbreviated internal representations into complete external structures that expose integration gaps.
Maintain distinct schemas as separate regulatory responses to different environmental challenges rather than collapsing them into a single framework, as variety in internal models is required to match variety in external disturbances.
Build bridges between schemas that preserve each schema's distinctive structure and predictive capacity rather than merging them into uniform representations, as diversity of error types improves aggregate accuracy more than uniformity of individual accuracy.
Demand falsifiability from integrated frameworks by naming specific observations that would break the integration, as coherence without correspondence to reality is indistinguishable from fiction.