Covolution and its four nearest neighbors

0(0명)
문서 역사

Covolution and its four nearest neighbors

Covolution shares territory with four existing frameworks,
and its claim to independent standing rests on a single move none
of them make: treating the processing of functional information as
the primary variable that sets the rate, direction, and open-endedness
of evolution across scales, with selection as emergent bookkeeping
rather than the fundamental driver.

Versus Niche Construction Theory (Odling-Smee, Laland, Feldman).
NCT adds environment-building as a second evolutionary process that biases natural selection and creates ecological inheritance — but construction modifies selection, while selection remains the engine and the explanatory primitive. Covolution inverts this hierarchy: construction and selection are both surface expressions of an underlying information-processing dynamics, and the symvironment is not a modifier of selection pressure but a co-computed state variable. NCT is selection-first with a feedback term; covolution is information-first with selection emergent.

Versus Active Inference / the Free Energy Principle (Friston).
The FEP explains how a single persisting system behaves as if minimizing surprise relative to an internal model, grounded in the statistics of maintaining a boundary. It is powerful for within-system homeostasis but thin on heredity, lineage, and the mutual construction of shared environments by many interacting systems, and it stays deliberately agnostic on whether the model is real or merely instrumental. Covolution generalizes the internal-model idea from one bounded agent to nested, interacting agents that co-author each other's models and environments, ties it to evolutionary inheritance, and commits to functional information as physically real work rather than an "as if." FEP is one-boundary, one-lifetime, instrumental; covolution is multi-agent, cross-generational, and realist about information.

Versus Autopoiesis (Maturana & Varela).
Autopoiesis defines life as operationally closed self-production and is explicitly anti-representational: the organism does not model its world, it only structurally couples to it, and the theory is nearly silent on evolutionary change. Covolution is the near-opposite on this axis — it makes internal modeling and information processing central, and externalizes that processing into symvironment construction. Autopoiesis is closure without representation; covolution is open, information-bearing, and explicitly a theory of change.

Versus Constructor and Assembly Theory (Deutsch & Marletto; Cronin & Walker).
These supply covolution's physics-level backbone: a substrate-independent account of which transformations are possible (constructor theory) and a measure of how much selection and history are embodied in an object (assembly theory). But both concern possibility and complexity-bookkeeping; neither specifies the generative reciprocal dynamics by which entities and their environments compute each other forward in time. They measure the footprints of construction; covolution proposes the engine that lays them down. This is also covolution's principled bridge from cosmology to biology, because the currency, functional information realized through physical constructors, is the same at both ends.

Covolution's original contribution is therefore synthetic and singular at once: it unifies environment construction (NCT), internal-model-driven action (FEP), self-production (autopoiesis), and the physics of information and complexity (constructor/assembly theory) under one measurable quantity, a system's capacity to process functional information and construct its symvironment, and asserts that this quantity, not external selection gradients, primarily sets evolutionary rate, direction, and open-endedness.

댓글 0