Love OS — Theory

Theory

Love is a developmental system.

Love OS = Attraction × Viability

Formation ≠ Maintenance ≠ Dissolution

The Core Insight

The Similarity Paradox

Similarity predicts attraction reliably.

But regulation predicts durability.

That is the similarity paradox.

Different variables govern different phases. The same factors that bring people together do not determine whether they stay together.

FORMATION MAINTENANCE DISSOLUTION Similarity Attraction Formation drives Regulation Viability Durability Hazard h(t) dissolution risk drives
Developmental Trajectory Logic — Formation ≠ Maintenance ≠ Dissolution. The same variables do not govern every phase.

The Framework

The Two Layers

Compatibility is not one dimension. It is a phase space.

Layer 1

Attraction

Reward-based mechanisms.

  • Similarity
  • Domain-specific complementarity
  • Passion activation
Read Attraction theory →

Layer 2

Viability

Regulation-based mechanisms.

  • Emotional stability
  • Attachment organization
  • Autonomy–closeness alignment
  • Commitment stabilization
Read Viability theory →

Compatibility States

The Four Compatibility States

Compatibility is not one dimension. It is a phase space.

Viability Attraction Low High Low High HIGH A / HIGH V Durable Romantic Compatibility High chemistry + stability Low dissolution hazard Optimal state HIGH A / LOW V Intense but Structurally Fragile High chemistry Low regulation Rising hazard LOW A / LOW V Unstable and Weakly Bonded Low chemistry Low stability Very high hazard LOW A / HIGH V Companionate but Low-Intensity Low chemistry High stability Low hazard C(t) = A × V(t)
Compatibility Phase Space — C(t) = A × V(t). Each quadrant represents a distinct relational state.
Academic model (formal) →

Formal coupling

C(t) = A × V(t)

Hazard of dissolution

h(t) = h₀(t) · exp(−V(t))

Notes

Attraction governs formation probability. Viability governs durability and dissolution hazard. Formation, maintenance, and dissolution are treated as distinct phases with distinct governing variables.