The ARCH model: a neuroevolutionary framework for behavioral execution
Tahir Rahman, Charles F. Zorumski, J. Reid Meloy
Abstract
Behavior arises from the convergence of multiple constraints rather than single causes. The ARCH × Φ model formalizes this process as a computational grammar of behavior, in which Archetype (A), Drive (D), and Culture (C) interact multiplicatively, and expression occurs only when a context-sensitive threshold (Φ) is crossed. This scalar-vector framework specifies behavior as probabilistic and testable, supporting hypotheses that can be evaluated across neurobiological, behavioral, and symbolic domains. We define a provisional taxonomy of ten archetypal systems (Systema Behavorum), such as Agonix (competition), Theromata (caregiving), and Sacrifex (self-sacrifice), which serve as structured inputs to the grammar. ARCH × Φ integrates ethology, affective neuroscience, psychiatry, and cultural psychology, reframing archetypes not as metaphors but as conserved neural scripts subject to scalar amplification and symbolic modulation. The framework supports falsifiable predictions, operational definitions, and clinical applications in decoding motivation, threshold dysregulation, and symbolic distortion. ARCH × Φ thus reframes behavior as an emergent property of convergent constraints across biology, affect, culture, and context.