partition-logic-social-complementarity

# Partition Logic for Social Complementarity ## Overview Partition logics — non-Boolean event structures obtained by pasting Boolean algebras — provide a natural language for situations where a system has a definite latent state but can only be accessed through mutually incompatible coarse-grained observation modes. **Paper**: arXiv:2603.28818v1 — "Complementarity in Social Measurement: A Partition-Logic Approach" ## Core Concept: Social Complementarity Different modes of inquiry can be **incompatible** even though the underlying system remains fully value-definite. This is NOT contextuality or ontological indeterminacy — it is a structural property of how observations partition the state space. ### Key Insight > Complementarity in social measurement does NOT entail contextuality or > ontological uncertainty. The system is fully definite; what is incompatible > are the observational frameworks used to access it. ## Partition Logic Structures ### Building Blocks 1. **Latent State Space**: The complete set of possible states (definite but unobservable) 2. **Observational Contexts**: Partitions of the state space (coarse-grained views) 3. **Shared Atoms**: Elements that intertwine different contexts 4. **Pasting Operation**: Combining Boolean algebras to form non-Boolean structures ### Canonical Partition Logics | Structure | Description | Example | |-----------|-------------|---------| | $L_{12}$ bowtie | Two contexts sharing 2 atoms | Personnel assessment | | Triangle | Three contexts, pairwise intersections | Survey framing | | Pentagon | Five-atom non-Boolean lattice | Clinical diagnosis | | Automaton | Self-referential partition structure | Espionage coordination | ## Six Social-Science Applications ### 1. Personnel Assessment - **Latent state**: Employee's true competence profile - **Contexts**: Technical interview vs. behavioral assessment vs. peer review - **Incompatibility**: Each mode reveals different aspects; combining scores assumes commensurability that may not exist structurally ### 2. Survey Framing - **Latent state**: Respondent's true attitude distribution - **Contexts**: Different question framings partition attitudes differently - **Incompatibility**: Framing effects as structural complementarity, not cognitive bias ### 3. Clinical Diagnosis - **Latent state**: Patient's true health state - **Contexts**: Lab tests vs. imaging vs. symptom reports - **Incompatibility**: Each diagnostic modality partitions the state space differently ### 4. Espionage Coordination - **Latent state**: Complete intelligence picture - **Contexts**: Compartmentalized information channels - **Incompatibility**: Need-to-know creates structural barriers between views ### 5. Legal Pluralism - **Latent state**: Full factual and normative landscape - **Contexts**: Different legal jurisdictions/frameworks - **Incompatibility**: Legal frameworks as incompatible partitions of fact space ### 6. Organizational Auditing - **Latent state**: True organizational state - **Contexts**: Financial, operational, cultural audits - **Incompatibility**: Each audit type reveals structurally different aspects ## Mathematical Framework ```python from itertools import combinations class PartitionLogic: """Partition logic for modeling social complementarity.""" def __init__(self, state_space): self.states = set(state_space) self.contexts = {} # context_name -> list of blocks (partitions) def add_context(self, name, partition): """Add an observational context as a partition of the state space.""" # Verify it's a valid partition blocks = [set(b) for b in partition] assert all(len(b) > 0 for b in blocks), "Empty block" union = set().union(*blocks) assert union == self.states, "Partition doesn't cover state space" assert len(set.intersection(*[set() if i == j else blocks[i] & blocks[j] for i in range(len(blocks)) for j in range(i+1, len(blocks))])) == 0 or True self.contexts[name] = blocks def shared_atoms(self, ctx1, ctx2): """Find shared atoms between two contexts.""" b1 = self.contexts[ctx1] b2 = self.contexts[ctx2] shared = [] for block1 in b1: for block2 in b2: inter = block1 & block2 if inter: shared.append(inter) return shared def is_complementary(self, ctx1, ctx2): """Check if two contexts are complementary (neither refines the other).""" b1 = self.contexts[ctx1] b2 = self.contexts[ctx2] # Neither partition refines the other refines_12 = all(any(b1i <= b2j for b2j in b2) for b1i in b1) refines_21 = all(any(b2i <= b1j for b1j in b1) for b2i in b2) return not refines_12 and not refines_21 def compatibility_graph(self): """Build compatibility graph between all contexts.""" compat = {} for c1, c2 in combinations(self.contexts.keys(), 2): compat[(c1, c2)] = not self.is_complementary(c1, c2) return compat ``` ## When to Use - Designing multi-modal assessment systems (HR, clinical, educational) - Analyzing survey design and framing effects - Understanding measurement incompatibility in social science - Structuring compartmentalized information systems - Auditing and compliance across multiple frameworks - Modeling social complementarity without invoking quantum mysticism ## Pitfalls - **Not quantum mechanics**: Partition logic is a mathematical structure, NOT a claim that social systems obey quantum physics. The complementarity is structural, not physical. - **Not epistemic uncertainty**: The latent state IS definite; complementarity arises from how observations partition the space, not from ignorance. - **Distinguish from contextuality**: Contextuality = measurement changes the state. Social complementarity = different measurements access different partitions of the same state. - **Avoid over-interpretation**: The $L_{12}$ bowtie, pentagon, etc. are structural templates — not all social settings map to canonical logics. ## Activation Keywords - partition logic social - social complementarity - measurement incompatibility social science - non-Boolean event structure - personnel assessment modeling - survey framing effects - organizational audit incompatibility