Jacek Domeredzki

This article proposes an interpretation of evolution as a continuous process of escalation of control architectures. Rather than organizing physical, chemical, biological, cognitive, and technological evolution around ontological levels or the emergence of new types of entities, the paper analyzes evolution in terms of increasing control power and successive reorganizations of regulatory architecture. The framework identifies regulatory thresholds at which existing architectures lose stability and must be restructured in order to maintain coherent system trajectories. The approach is functional, non-normative, and non-teleological, and is intended as a conceptual basis for further formal and empirical work, including the analysis of stability and safety in high-power technological systems.

We propose a unified relational dynamical framework for the emergence of structure across physical, biological, cognitive, and technological systems. The starting point is the notion of minimal difference as the simplest deviation from a trivial regime in which relations are globally reducible and generate no structural effects. The key mechanism is the transition from minimal difference to returning difference, understood as an effect that does not vanish along closed trajectories. We show that a returning difference constitutes the minimal form of feedback, enabling the accumulation of effects and the onset of nontrivial dynamics. A central role in the proposed framework is played by recursion, understood not as mere iteration of a rule, but as a structural property of dynamics in which transformations act on the results of previous transformations. Recursion is one of the fundamental mechanisms underlying the generation of structure: it enables the formation of feedback loops, hierarchical organization, and multi-level regulatory systems. In this sense, returning difference can be viewed as the elementary form of recursion grounded in relational dynamics. We formalize relational dynamics in terms of local relations and their non-closure (curvature), which, in the presence of appropriate operators, generate iterated feedback. The introduction of selection and projection mechanisms leads to invariant sets and concentration of trajectories, interpreted as the formation of stable structures. Extending the model to the statistical level allows for the description of probabilistic dynamics, metastability, and the role of fluctuations. In the continuous limit, the framework leads to effective field descriptions, and when coupled with geometry, to the co-evolution of fields and spatial structure. At a global level, entropy, information, and elements of geometric structure can be understood as emergent properties arising from relational feedback dynamics. The proposed approach is structural and effective: it does not replace existing physical theories, but identifies common mechanisms of organization across domains. In particular, we argue that the transition from minimal difference to returning difference—as the elementary form of recursion and feedback—defines a universal threshold at which the emergence of complex structures becomes possible.

This paper proposes a conceptual framework for analyzing the order of observed dynamical systems in terms of the escalation of feedback architectures. Rather than organizing physical, chemical, biological, and technological systems according to ontological levels or evolutionary narratives, the framework treats them as successive classes of stable regulatory architectures, distinguished by the geometry of their attractors and basins of attraction. Random variations and reorganizations of feedback architectures may enable increases in control power, understood as the capacity to constrain and sustain system trajectories over time. However, only those configurations whose attractors remain dynamically stable are preserved, implying a form of stability-based selection rather than adaptation or optimization. The framework is functional and non-normative: it does not presuppose goals, intentionality, or teleology at any level of analysis. Escalation is shown to proceed through qualitative reorganizations of regulatory architecture, from dynamically closed systems, through energetic and informational regulation, to metaregulatory technological systems capable of simulating and selecting architectures prior to physical realization. While later architectures achieve unprecedented control power, they do so at the cost of increasingly narrow basins of attraction, leading to heightened global fragility and sharp stability cliffs. The paper situates this account in relation to dynamical systems theory and clarifies its relation to the free-energy principle, which is interpreted as a condition of trajectory stability rather than a theory of architectural structure. Overall, the framework offers a unifying, non-teleological perspective on order, stability, and control across domains, emphasizing architecture rather than substrate as the primary axis of analysis.