Type: book
Abstract: This series of eight papers develops a unified formal framework for embedded epistemic systems — systems that must model the world from within it, without access to an external vantage point. It begins from an ontological ground and ends by showing what it is like to inhabit that ground from the inside, and what must be chosen there. The series is preceded by an ontological paper establishing that demarcation and abstraction — the co-arising operations of differentiation-without-division and unification-without-annihilation — are prior to time and space. Their shared primitive, unity-in-difference, is what must be in place for questions of ordering, locating, or relating to arise at all. The formal architecture of the series presupposes this primitive at every level. From that ground the series derives a layered architecture spanning formal epistemology, the philosophy of science, the dynamics of agent–environment interaction, the phenomenology of perception, and moral philosophy. The architecture stabilized through the same recursive process it describes. The founding epistemic constraint is this: an embedded epistemic system can at most classify the ways in which it classifies the world, within the world itself. A central formal result is the derivation that some classifiers belong to the observation space. This is derived as the transcendental condition on any system capable of doubting its own representations. From it follow the fixed-point structure of self-consistent world models and the location of will as the irreducible remainder of the inference–implication loop. The series establishes: that scientific reproducibility is the stability of relational invariants under quotient compression of observational data; that the extension schema — the pattern by which a partially specified relational configuration is completed into a coherent structure that asymmetrically contains its parts and participates in further extension — recurs across holonic composition, simplicial horn filling, and analogical abstraction, and that the identification of this recurrence is itself a further instantiation of it; and that the Koopman eigenfunction structure of relational observables becomes linearly recoverable through prediction error alone in a minimal artificially intelligent agent. The penultimate paper describes what it is like to inhabit the framework from the inside: what it means that perception is always already semiotic, that there is no view from nowhere, and that faith, logic, and experience are not three alternative foundations but three aspects of a single recursively self-correcting loop — the phenomenological form of the inference–implication loop the series formalizes. The final paper closes on the moral consequence the loop cannot derive but can only demand. Will is the irreducible remainder of the inference–implication loop; what is done with it must be chosen. The foremost moral principle is therefore to seek moral principles. Moral admissibility then requires universalizing not an action alone but a tuple of action and motivation set. The series demonstrates its own thesis. The loop closed on a fixed point.
Authors: Tracy, Mark, Nabaala, Salash Tolan
Language: English
Subfield: Philosophy
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7154879452