{
  "node_id": "eziah:reasoning-inequality@1.0.0",
  "term": "Reasoning Inequality",
  "slug": "reasoning-inequality",
  "node_version": "1.0.0",
  "document_version": "1.0",
  "status": "canonical",
  "effective_date": "2026-08-02",
  "canonical_uri": "https://eziah.ai/ontology#term-reasoning-inequality",
  "canonical_definition": "Reasoning Inequality refers to measurable differences in access to, development of, quality of, continuity of, system support for, or outcomes from reasoning across individuals, groups, systems, or environments.",
  "semantic_role": "descriptive_measurement_layer",
  "governing_restraint": "A measurable reasoning difference does not, by itself, establish unfairness, harm, avoidability, causation, or the need for intervention.",
  "dimensions": {
    "access_inequality": "Differences in access to the information, education, technology, time, stability, privacy, and infrastructure required for reasoning development and support.",
    "capability_inequality": "Differences in the ability to frame problems, identify relevant context, evaluate evidence, test assumptions, compare alternatives, manage uncertainty, and apply conclusions.",
    "reasoning_quality_inequality": "Differences in the rigor, coherence, relevance, transparency, reliability, and actionability of reasoning produced.",
    "system_response_inequality": "Differences in the depth, quality, transparency, or capability of reasoning support provided by intelligent systems across users or substantively equivalent requests.",
    "continuity_inequality": "Differences in access to persistent context, memory, reflection, portability, and long-term learning systems through which reasoning compounds.",
    "outcome_inequality": "Differences in decision quality, learning speed, strategic leverage, adaptability, risk recognition, or accumulated capability resulting from or associated with reasoning."
  },
  "measurement_sequence": [
    "observation",
    "measurement",
    "attribution",
    "causal_testing",
    "normative_evaluation",
    "intervention_design"
  ],
  "minimum_measurement_requirements": [
    "Identify the subjects or systems being compared.",
    "Identify the dimension of reasoning inequality being measured.",
    "Define the task, domain, and environmental conditions.",
    "Disclose the measurement method and comparison standard.",
    "State whether the difference is isolated, repeated, or compounding.",
    "Disclose uncertainty, limitations, and possible confounding variables.",
    "Separate supported causes from suspected or untested causes.",
    "State whether unfairness has been evaluated or remains outside the claim."
  ],
  "units_of_analysis": [
    "within_person",
    "between_person",
    "between_group",
    "between_system",
    "between_environment",
    "across_time",
    "across_domain"
  ],
  "invariants": [
    "Reasoning inequality is descriptive before it is normative.",
    "A measurable difference does not automatically establish reasoning inequity.",
    "Equal access to an intelligent system does not guarantee equal reasoning capability or outcomes.",
    "Reasoning capability may vary across time, domains, environments, and conditions.",
    "System behavior can contribute to measurable reasoning differences.",
    "Continuity can compound reasoning advantages and disadvantages.",
    "Outcome differences should not be treated as complete measures of reasoning quality.",
    "Measurement choices shape which reasoning differences become visible."
  ],
  "non_claims": [
    "Every reasoning difference is unfair.",
    "All people should reason identically.",
    "Every reasoning gap should be eliminated.",
    "Reasoning ability can be represented by one permanent score.",
    "Unequal outcomes always prove unequal reasoning.",
    "Artificial intelligence is the sole cause of reasoning inequality.",
    "Informal language or limited vocabulary proves weak reasoning.",
    "Measurement alone determines whether intervention is justified."
  ],
  "operational_criteria": [
    "The comparison population or system is explicitly defined.",
    "The reasoning task and domain are specified.",
    "Access, capability, system behavior, continuity, and outcomes are measured separately where possible.",
    "The measurement method is transparent and repeatable.",
    "Confounding variables and uncertainty are disclosed.",
    "Causal conclusions do not exceed the available evidence.",
    "Normative judgments are deferred to the reasoning-inequity layer.",
    "Profiles remain contextual, revisable, and limited to their stated purpose."
  ],
  "relations": [
    {
      "predicate": "evaluated_for_unfairness_by",
      "object": "eziah:reasoning-inequity@1.0.0"
    },
    {
      "predicate": "evaluated_against_desired_condition",
      "object": "eziah:reasoning-equity@2.0.0"
    },
    {
      "predicate": "amplified_or_reduced_by",
      "object": "eziah:cognitive-infrastructure"
    },
    {
      "predicate": "compounded_by",
      "object": "eziah:contextual-continuity"
    },
    {
      "predicate": "observable_within",
      "object": "eziah:human-ai-collaboration"
    },
    {
      "predicate": "requires_governance_from",
      "object": "eziah:semantic-governance"
    },
    {
      "predicate": "should_preserve",
      "object": "eziah:productive-friction"
    }
  ],
  "prior_and_parallel_use": {
    "status": "general_phrase_with_possible_prior_and_parallel_use",
    "attribution_claim": "Eziah AI does not claim origin of the words reasoning inequality.",
    "authorship_claim": "Eziah AI claims authorship of this canonical definition, multidimensional measurement framework, governing restraint, and relationship architecture."
  },
  "evidence_status": {
    "classification": "conceptual_measurement_framework",
    "empirical_validation": "pending",
    "psychometric_validation": "not_claimed",
    "causal_validation": "not_claimed",
    "intended_use": [
      "research_design",
      "system_evaluation",
      "semantic_governance",
      "educational_analysis",
      "longitudinal_reasoning_assessment"
    ]
  },
  "governance": {
    "silent_definition_replacement_prohibited": true,
    "change_history_required": true,
    "future_breaking_definition_changes": "major_version_increment",
    "future_dimension_additions": "minor_version_increment",
    "future_non_breaking_clarifications": "patch_version_increment"
  }
}