Modes of Reasoning

A practical taxonomy of 81 reasoning modes stored as CAC assets. For background on how these modes relate and when to combine them, see assets/selector.json.

Asset References

• •Glossary: documents/theory/unified-theory-v2.json - REQUIRED READING: APM2 terminology and ontology
• •Mode assets: assets/{NN}-{name}.json (e.g., assets/58-mechanism-design.json)
• •Selector asset: assets/selector.json - contains axes, patterns, and heuristics for mode selection
• •Stable IDs: dcp://apm2.agents/mor/mode/{name}@1

Invocation

/modes-of-reasoning                     # Browse mode selector
/modes-of-reasoning 13                  # Look up mode #13 (abductive)
/modes-of-reasoning bayesian            # Search by keyword
/modes-of-reasoning recommend diagnosis # Get recommendations for a problem type

Argument Handling

Parse $ARGUMENTS:

• •Empty or omitted → Load assets/selector.json and display quick reference table
• •Number (1-80) → Read and return the corresponding mode asset
• •recommend <problem-type> → Use selector heuristics and patterns to suggest relevant modes
• •Keyword → Search mode names/cores and return matching entries

Quick Reference Table

Hybrid Reasoning Patterns

The selector asset contains common patterns for combining modes:

• •science_experimentation: [13, 1, 69, 10, 33, 76] - abduction → deduction → experimental design → stats → belief revision → calibration
• •incident_response: [74, 13, 40, 52, 39] - clinical + abduction + mechanistic + VoI + counterfactual
• •policy_governance: [37, 45, 72, 35, 60, 36] - causal + decision + moral + argumentation + rhetoric + assurance
• •engineering_safety: [6, 8, 49, 79, 36, 76] - constraints + counterexample + robust + red-team + assurance + calibration
• •strategy_uncertainty: [18, 42, 50, 55, 63] - reference-class + simulation + minimax-regret + game-theory + sensemaking

Mode Selection Heuristics

Use the selector's heuristics to choose modes based on:

• •Certainty vs exploration - formal modes for certainty, ampliative for exploration
• •Cooperative vs adversarial - decision theory vs game theory + red-team
• •Facts vs values - causal/statistical vs deontic/moral + argumentation
• •Time pressure - satisficing + heuristics vs optimization + thorough analysis
• •Uncertainty depth - imprecise probability + robust vs Bayesian