Between philosophy and engineering lies a missing layer — the governance syntax of complex systems.
What is Dynamic Sustenance Theory (DST)?
DST is a structural grammar for governing complex adaptive systems. It asks not "what a system is" but "how a system sustains itself amid flux" — and turns that perspective into a diagnosable, intervenable grammar.
Its ontological ground is the domain of sustenance: a system persists not by staying unchanged, but by maintaining structural continuity under flux.
Core mechanism — force-rule symbiosis: the dynamic layer (Three Waves) and the rule layer (Three Phases) co-adapt to keep a system alive. Key concepts include Ti-Xiang Correlation, Layered Asynchrony, and Syntax Lock-in — three structural invariants that together define where and how intervention can take hold.
Existing complex-systems theories explain how systems evolve. DST works on the other question: how a system sustains itself.
Selected Publications (in English)
The treatise is written in Chinese; nine papers of the series are published so far. The English papers below are the entry point for international readers.
Zenodo, 2026 · DST English Extension Series No. 4 · DOI: 10.5281/zenodo.23328739
Complexity science has produced powerful descriptive tools, yet it still lacks a shared structural language for intervention—a gap that lies not in method but in the problem domain. This paper argues that the proper question for the study of complex systems is not "what a system is" but "how a system sustains itself amid flux," and that governance is where this question becomes operational. AI governance, climate governance, financial stability, and urban emergency response repeatedly encounter the same kind of predicament—multi-agent self-adaptation, cross-level temporal asynchrony, and constraint structures that lose their adjustability under pressure—yet there is no common structural language for comparing and transferring lessons among them. This paper argues that the root of this predicament lies not in specific methods but in a triple misalignment among problem domain, epistemology, and methodology. The paper offers an operationalization of this alignment—a general structural grammar—comprising five basic constituents, a logical walkthrough from diagnosis to action (using platform governance as a simplified scenario), differential criteria against existing frameworks, and a four-rung validation ladder.
Zenodo, 2026 · DST English Extension Series No. 3 · DOI: 10.5281/zenodo.22884353
Drawing on Dynamic Sustenance Theory (DST v15.0), this paper develops the concept of Layered Asynchrony to diagnose why law always lags AI. The dynamic layer advances cumulatively on a month-to-quarter scale and cannot be paused; the rules layer proceeds through procedural cycles on a year-to-decade scale.
Zenodo, 2026 · DST English Extension Series No. 2 · DOI: 10.5281/zenodo.22698811
This paper presents structural disclosure (Xianying) as an operative methodology for human-AI division of labor. It replaces the vague human/AI boundary with a burden-bearing criterion: a judgment is non-delegable when a human occupies a pressure-bearing position and must bear the consequences of being wrong.
SSRN Working Paper, 2026 · DST English Extension Series No. 1 · DOI: 10.2139/ssrn.7453298
Current AI governance frameworks rely on a static-compliance paradigm, treating AI systems as discrete, bounded objects amenable to fixed classification and rule-based control. This paper argues that this paradigm rests on an ontological category mistake. Drawing upon Dynamic Sustenance Theory (DST), it proposes that AI systems are not static entities but pattern-continuous processes of sustenance: their identity persists not through unchanging components, but through the continuous, topologically recognizable coordination of forces and rules.
Contact