Axiodynamics v5.11 — Triadic Dynamics: Coalition Formation, Mediation Mechanisms and Triangulation Effects

Dyadic models cannot adequately represent coalition formation, mediation, or scapegoating when the presence of a third agent transforms the relation between two existing agents. This preprint extends the axiodynamic framework to the minimal triadic configuration and treats the triad as the first relational level that may be irreducible to an aggregation of independent dyads. The model introduces directed conative and inhibitory affective loads for each relational channel, while preserving agent-global dissipation and affective capacity. A triadic coherence tensor distinguishes intra-channel ambivalence from adaptive differentiation across relational targets. A fully endogenous triangulation operator formalises the distortion of the perceived axiological distance between two agents in the presence of a salient third agent. Its intensity is derived from the third agent’s affective salience and the dyad’s inertial permeability, without introducing an additional free parameter. Six canonical triadic configurations are formally characterised through static and dynamic signatures: coalition, mediation, tertius gaudens, tertius separans, scapegoat, and balanced triad. Structural balance theory is recovered as a discrete and static limiting case of the proposed continuous dynamics. The model additionally represents ambivalence, path dependence, relational lock-in, and hysteresis. The computational implementation instantiates four of the six candidate regimes under a single parameter set: balanced, coalition, scapegoat, and mediation-like configurations. Factorial ablation compares the full triadic model with perception-only, structure-only, and dyadic-aggregate architectures. The executed simulations also include basin mapping and an exclusion–reintegration protocol producing plateau-dependent hysteresis areas. The preprint formulates six falsifiable empirical predictions concerning coalition thresholds, mediator position and salience, scapegoat irreversibility, target-specific ambivalence, third-party distortion, and the dynamic instability of ambivalent ties. Status: This work presents a theoretical and computational candidate model. The simulations establish computational adequacy and reproducibility but do not constitute empirical validation. Regime thresholds remain uncalibrated, and the proposed validation protocols have not yet been executed. Code and reproducibility materials are available in the public GitHub release v5.11-sim-1.0: https://github.com/Alan-Kleden/Axiodynamic-Framework/releases/tag/v5.11-sim-1.0