The Dyadic Universe Model: a process theory of centralization, compression, and autonomy in romantic relationships

Contemporary relationship science provides rich accounts of attachment, intimacy, and self-expansion, yet remains comparatively underspecified regarding a class of high-intensity romantic relationships in which partners progressively become each other's primary source of identity and meaning. We introduce the Dyadic Universe Model (DUM), a process-theoretic framework developed to address this explanatory gap, since existing theories such as attachment, self-expansion, differentiation of self, and the suffocation model explain important aspects of closeness, growth, or strain but do not fully capture the distinctive pattern in which expansion later turns into perceived constraint. This framework is formalized using three core variables: dyadic centralization (C), external-world investment (E), and dispositional pressure (ψ). Dyadic centralization captures the extent to which the relationship displaces external domains of identity and engagement; external world investment reflects ongoing participation in those domains; and dispositional pressure represents individual differences in the drive for novelty, growth, and identity diversification. The model proposes a recursive dynamic in which increasing centralization is associated with reduced external engagement, limiting access to novel inputs and contributing to the accumulation of subjective compression. For individuals higher in dispositional pressure, this compression is predicted to intensify over time, eventually activating autonomy-restoration processes that reduce centralization or destabilize the relationship. From this minimal structure, the model derives a characteristic four-phase lifecycle, which are recognition, centralization, saturation, and autonomy restoration, linking the emergence of exceptional relational intensity to its potential transformation into psychological constraint. The DUM integrates and extends existing theories of self-expansion, differentiation of self, and relational suffocation by specifying the temporal dynamics through which immersive relationships evolve. It advances a set of falsifiable propositions concerning lifecycle trajectories, individual differences, and dyadic asymmetries, and outlines a coordinated empirical research program across longitudinal, dyadic, and experience-sampling designs. While elements of these dynamics are suggested across existing empirical findings, the model formalizes them into a unified process-level framework and generates explicit predictions requiring systematic empirical testing. By articulating the interplay between centralization, external engagement, and dispositional pressure, the DUM provides a parsimonious account of how the same processes that generate profound intimacy may also create conditions for instability.