SQC and neighboring cognitive architectures

Artificial-psychology classification checklist · SQC v1.2 · evidence review 18 July 2026

Working test: an engineered system qualifies when explicit psychological constructs are causal, inspectable, testable, and jointly organize context, values, motivation, affect, memory, state, and adaptive behavior over time.

Classification readout

Category fit is separate from validation maturity; “neighbor” does not mean “failed.”

SQC
Clear AP category fit

Explicit EPS sequence makes values, appraisal, motivation, emotion, state and memory causal inputs to action.

Soar
Conditional AP fit

Strong cognitive architecture; affect is documented as an integrated extension, while explicit human values are not a core primitive.

ACT-R
Adjacent cognitive model

Strong human-cognition theory and implementation, but affect and value integration are not established in the reviewed core architecture.

LIDA
Strong AP-compatible fit

Context, emotion-based motivation, memory, learning and action selection are integrated in recurring cognitive cycles.

Category-membership checklist

CriterionSQCSoarACT-RLIDA
Executable engineered systemImplemented architecture capable of producing behavior✓✓✓✓
Explicit psychological constructsNamed internal constructs with declared roles✓EPS stages✓memory, goals, appraisal✓chunks, activation, utility✓attention, emotion, memory
Causal behavioral mediationConstructs participate in selecting or generating action✓EPS-only action input✓working memory + productions✓buffers + production selection✓broadcast + action selection
Inspectable internal stateStates and transitions can be recorded and tested✓✓✓✓
Context-sensitive processingCurrent situation changes internal processing and action✓✓✓✓
Explicit values or normative appraisalMore than scalar production utility or task reward✓value state is explicit◐preferences/reward; value model not core◐goal utility is narrower◐incentives and feelings; values not isolated
Motivational stateInternal motives causally shape selection over time✓Belonging, Identity, Power◐goals, reward, appraisal extensions◐goals and utilities✓emotion-based motivation
Affect or emotion in the causal loopAffect alters attention, learning, state or action✓◐Soar-Emote extension○not established in reviewed core✓
Memory integrated with behaviorWorking and/or long-term memory informs action✓bounded explicit memory✓working, semantic, episodic✓declarative + procedural✓multiple memory systems
Adaptive behavior over timeLearning, history or recurring state transition◐bounded temporal interface✓✓✓
Falsifiable and intervention-readyMechanisms support predictions, manipulations or failure tests✓stage interventions + audits✓✓✓
✓ documented in the core or audited implementation ◐ partial, narrower, or extension-dependent ○ not established in reviewed core sources

Validation maturity is a separate question

The AP-0–AP-5 ladder has been applied to SQC’s repository evidence; equivalent architecture-wide audits were not performed here for Soar, ACT-R or LIDA.

SQC category
AP-1 supported

Executable, inspectable EPS architecture with behavior generated from the declared pathway.

SQC internal evidence
Substantial progress toward AP-2

Shortcut controls, freeze audits, causal sensitivity, synthetic competence and bounded generalization; full gate remains open.

Later SQC gates
AP-3–AP-5 not established

Independent construct, ecological and use-specific operational validation remain future work.

Evidence basis

Core architecture sources support the checks; absence means “not established in the reviewed core,” not proof that no extension exists.