Daryl Costello

Independent Theoretical Research | Rosendale, NY, United States

Correspondence: Daryl.Costello@outlook.com

Submitted: October 2, 2026

Theoretical Biology • Cognitive Science • Philosophy of Mind • Complex Systems

Abstract

Background. The dominant paradigm in cognitive science draws a sharp boundary between the organism and its environment, treating cognition as a process localized within neural tissue. This boundary, however, is both empirically inadequate and theoretically impoverished. Across biological scales (from the lipid bilayer to cultural institutions) systems exhibit selective temporal integration, reconstitutive metabolism of environmental input, and hierarchical propagation of compressed abstractions. No unified formal framework has yet integrated these observations into a single, scale-invariant architecture.

Framework. This manuscript presents the Cognitive Membrane Model of Adaptive Temporal Continuity (CMMATC), a unified theoretical architecture synthesizing eight source frameworks: (1) Cognition as Genomic Medium, (2) Evolution as the Ascent of Reasoning, (3) Insight as Developmental Phase Transition, (4) Morphogenesis in the Continuum, (5) Teleodynamic Foundations of Physical Reality, (6) The Cognitive Membrane Theory of Adaptive Temporal Continuity, (7) The Ontological Distance, and (8) The Reconstitutive Medium. The model’s central operator is the cognitive membrane; a formal boundary function M: I × E × T → S that partitions state-space into inside, environment, and time dimensions, producing a selective filter. Built upon this operator, we develop the Abstraction Ascent Stack (AAS), an eight-layer tower of cognitive operators spanning teleodynamic physical substrates through meta-cognitive recursion. We integrate Michael Levin’s intelligence cone, Terrence Deacon’s teleodynamics, and a novel formalism for ontological distance as a metric on the space of intelligence cones.

Key Results. We derive seven core theorems: Membrane Universality, Medium Invariance, Ascent Coupling, Cone Expansion, the Insight Threshold theorem, the Ontological Metric theorem, and Reconstitutive Accumulation. We formalize the Genomic Temporal Stack (GTS) as a multi-band cognitive archive, the Morphogenetic Cognitive Field (MCF) as a continuous spatial integration operator, and the Reconstitutive Medium Operator R(G, M, t) = I_t as the general principle by which environmental media metabolize generative output into causally active invariant structure.

Implications. The CMMATC reframes intelligence not as a substrate-specific phenomenon but as a scale-invariant pattern of membrane-mediated temporal integration. This has direct implications for the theory of evolution (as directional ascent in cognitive architecture space), artificial intelligence alignment (as an ontological distance problem), developmental biology (as continuous cognitive field integration), and the philosophy of mind (as the formal grounding of selfhood in selective exclusion rather than positive content).

Keywords: cognitive membrane, adaptive temporal continuity, teleodynamics, abstraction ascent stack, genomic cognition, ontological distance, intelligence cone, morphogenetic cognition, reconstitutive medium, insight threshold

Table of Contents

1. Introduction	3
2. Theoretical Foundations	4
Part I: The Cognitive Membrane	5
Part II: The Medium as Differential Shadow	6
Part III: Genomics as Temporally Stratified Cognition	8
Part IV: The Operator-Stack and Vertical Ascent	9
Part V: The Intelligence Cone and Recursive Agency Continuum	11
Part VI: Insight as Developmental Phase Transition	13
Part VII: Ontological Distance	14
Part VIII: Morphogenesis as Continuous Cognitive Field	16
Part IX: Teleodynamic Grounding	17
Part X: Unified Formal Summary	19
3. Discussion	22
4. Conclusion	24
Glossary of Formal Terms	25
References	26

1. Introduction: The Problem of Temporal Continuity in Adaptive Systems

Every adaptive system faces a fundamental epistemic problem: the present moment is too thin to act on. A bacterium navigating a glucose gradient must integrate information about past concentrations to determine directional movement; a developing embryo must read the history of its own cellular states to correctly position a limb bud; a human mind must maintain narrative coherence across decades of experience to make rational long-term decisions; a civilization must encode the solutions its predecessors discovered to avoid catastrophic repetition. The problem of adaptive temporal continuity (how a system maintains productive relationship with its own history across time) is not a peripheral concern of cognitive science. It is, we argue, the central problem of life itself.

The prevailing approach to this problem in cognitive science has been, broadly speaking, representationalist: systems maintain continuity through internal memory structures that encode and retrieve past states. This approach, while productive within its domain, suffers from three compounding limitations. First, it localizes cognition within a privileged substrate (usually neural tissue) ignoring the extensive cognitive work performed by non-neural biological structures, from the genome to the extracellular matrix to bioelectric fields. Second, it treats the system-environment boundary as a datum rather than a cognitive achievement: the very distinction between inside and outside, self and world, past and future, is constituted by a boundary operation that is itself the primary act of cognition. Third, it treats time as a uniform dimension over which representation occurs, rather than recognizing that adaptive systems integrate information at fundamentally different temporal scales simultaneously; a phenomenon we term temporal stratification.

This manuscript presents the Cognitive Membrane Model of Adaptive Temporal Continuity (CMMATC), a unified theoretical architecture that addresses all three limitations. The model’s primary innovation is the formalization of the cognitive membrane; not as a physical structure, but as a scale-invariant boundary operator that constitutes the self/world distinction at every level of biological and cognitive organization. Built upon this operator, the model develops a hierarchical architecture (the Abstraction Ascent Stack (AAS)) spanning from teleodynamic physical substrates through genomic, morphogenetic, neural, conceptual, and cultural cognitive levels to meta-cognitive recursion. The architecture is unified by the principle that each layer instantiates the same canonical membrane operations at a different temporal scale and in a different medium, and that each layer’s output constitutes the substrate from which the next layer constructs its membrane.

The CMMATC synthesizes eight source theoretical frameworks developed across the fields of theoretical biology, cognitive science, philosophy of mind, and complex systems theory. These frameworks are not merely cited as supporting evidence; they are formally integrated as components of a single compositional architecture. The manuscript proceeds as follows: Section 2 reviews the theoretical foundations; Parts I through X develop the formal architecture in sequence; the Discussion section addresses implications for biology, AI, cognitive science, and philosophy of mind; and the Conclusion identifies open questions and future directions. A Glossary of Formal Terms and a comprehensive Reference list are provided as appendices.

A note on formal notation: throughout this manuscript, we use standard set-theoretic and functional notation. Operators are denoted by bold capital letters or named functions. Formal definitions are set off in bordered boxes. Theorems are numbered and stated precisely, with informal interpretations following immediately. ASCII diagrams represent architectural relationships that would otherwise require figures. All formal claims are intended to be precise enough for mathematical development even where full proofs are not provided.

2. Theoretical Foundations: The Source Frameworks

The CMMATC does not arise from a single theoretical lineage but represents a genuine synthesis across eight distinct frameworks, each capturing a different facet of the unified architecture. We review each briefly here, indicating how it contributes to the formal edifice developed in subsequent sections.

2.1 Cognition as Genomic Medium

The first source framework reframes the genome not as a passive blueprint but as an active cognitive substrate. The genome encodes not merely structural information but solutions to past adaptive problems; information that has been selected, refined, and stratified across evolutionary time. The genome is, on this view, a multi-temporal cognitive archive: a structure that integrates signals from deep evolutionary time (conserved core regulators), population time (allelic variation), developmental time (regulatory networks), and epigenetic time (methylation patterns and chromatin state). Crucially, this framework insists that the genome is not read passively but interpreted; the developing organism performs an active, context-sensitive read across multiple temporal strata simultaneously. This is the phenomenon of temporally stratified cognition, formalized in Part III.

2.2 Evolution as the Ascent of Reasoning

The second framework reconceptualizes evolution not as a blind search through fitness landscapes but as a directed ascent in the space of abstract reasoning. This claim must be stated carefully to avoid vitalism: the direction is not imposed from outside but emerges from the internal logic of the cognitive hierarchy itself. Each level of the Abstraction Ascent Stack makes available a new class of adaptive solutions unavailable to lower levels; evolutionary processes that instantiate higher levels therefore occupy a larger region of adaptive space. The “ascent” is not teleological in the pre-Darwinian sense but represents a monotonic relationship between cognitive architecture depth and adaptive reach; what we will formalize as the expansion of the intelligence cone (Part V).

2.3 Insight as Developmental Phase Transition

The third framework reframes insight (the phenomenological experience of sudden understanding) as the visible surface of a sub-threshold developmental cascade. This view challenges the dominant “aha moment” model, which treats insight as discontinuous. We argue that what appears as discontinuous at the phenomenological level is continuous at the developmental level: insight events are membrane-crossing events; moments at which accumulated sub-threshold cognitive reorganizations exceed a threshold, establishing a new self/world boundary condition. This framework contributes the Insight Threshold theorem and connects individual cognitive development to paradigm-level cultural transitions (Part VI).

2.4 Morphogenesis in the Continuum

The fourth framework treats morphogenesis not as the mechanical execution of a genetic program but as a continuous cognitive process. Bioelectric fields, chemical gradients, and mechanical tension fields are cognitive media; distributed representations that are actively read, written, and transformed by cells acting as cognitive agents. The developing body plan is not an artifact but a cognitive achievement: the time-integral of a continuous field of membrane-mediated signals. This framework contributes the Morphogenetic Cognitive Field formalism and grounds Layer 3 of the AAS (Part VIII).

2.5 Teleodynamic Foundations of Physical Reality

The fifth framework draws on Terrence Deacon’s teleodynamics; a theory of how purposive, end-directed causation emerges from physical substrates through the operation of constraints. Deacon’s key insight is that teleodynamic systems are defined by their absences: what a system is constrained not to do determines its end-directedness more than what it does. The cognitive membrane, we argue, is paradigmatically teleodynamic: it is constituted by its selectivity (what it excludes) and its intelligence is proportional to the sophistication of its exclusion structure. This framework provides the deepest grounding of the entire architecture (Part IX, Layer 0 of the AAS).

2.6 The Cognitive Membrane Theory of Adaptive Temporal Continuity

The sixth framework is the direct precursor of the CMMATC: it proposes the membrane as the canonical operator of cognition at all scales. The membrane separates inside from outside, past from future, self from environment; it enables selective temporal integration by admitting historical signals while filtering noise; and it implements the self/world distinction as a cognitive achievement rather than a physical datum. This framework contributes the formal definition of the membrane operator M: I × E × T → S and the three canonical membrane operations (Part I).

2.7 The Ontological Distance

The seventh framework introduces ontological distance as a formal measure of separation in the space of possible causal structures; “ontological space.” Systems may be spatially adjacent yet ontologically distant if their causal architectures (their patterns of membrane, medium, and operator configuration) differ substantially. This notion has direct practical implications for cross-species communication, human-AI alignment, and cultural translation. We formalize ontological distance as a pseudo-metric on the space of intelligence cones (Part VII).

2.8 The Reconstitutive Medium

The eighth framework insists that the environment of a cognitive agent is not passive but metabolically active: it takes generative output (chemical gradients, phenotypes, cultural artifacts, symbols) and metabolizes it into lower-dimensional invariants that persist across time and exert causal influence on subsequent generative acts. The medium is not a neutral receiver but an active reconstituter. This is the formal basis of stigmergy in social insects, epigenetic inheritance, the accumulation of scientific knowledge, and the historical causation exercised by legal and linguistic structures. We formalize this as the Reconstitutive Medium Operator in Part II.

Part I. The Cognitive Membrane: Formal Definition, Properties, and Scale-Invariance

I.1 The Problem of Boundary

Before a system can act adaptively, it must distinguish itself from its environment. This distinction (the self/world boundary) is not given by physics; the physical universe contains no marked line between organism and milieu. The boundary is a cognitive achievement: it must be actively maintained against the thermodynamic tendency toward dissolution, and its character (what counts as inside and what as outside) determines everything about the system’s adaptive behavior. The cognitive membrane is our formalization of this achievement.

The membrane, in our framework, is not primarily a physical object (though it has physical instantiations) but a boundary operator; a function that partitions state space and implements selective passage between partitions. As such, it is scale-invariant: the same formal structure instantiates at the lipid bilayer of a prokaryotic cell, at the body surface of a multicellular organism, at the neural architecture of a brain, at the boundary of a conceptual schema, and at the borders of a cultural community. Each instantiation operates at a different spatial and temporal scale, in a different medium, with a different characteristic timescale; but the formal structure is identical.

I.2 Formal Definition

Definition 1 (Cognitive Membrane). A cognitive membrane is a boundary operator M: I × E × T → S where I is the inside state space, E is the environmental state space, T is the temporal index set, and S is the selective filter function space. For each triple (i, e, t) ∈ I × E × T, M(i, e, t) returns a filter function st: E → I that specifies which environmental signals at time t are admitted to the inside state, and in what transformed form.

Several features of this definition require comment. First, the membrane operator is context-sensitive: the filter function it produces depends on the current inside state i as well as on the environmental state e and the temporal index t. This reflects the biological reality that membrane permeability is regulated by the internal state of the cell; the immune system’s discrimination of self from non-self depends on the history of immune interactions; and the selective attention of a cognitive agent depends on its current goals and prior experience.

Second, the temporal index T makes explicit that the membrane operates across time. This is not merely that the membrane exists at each time point; it is that the membrane’s filter function is temporally sensitive; it may admit signals that are correlated with historical patterns while blocking signals that are temporally novel or noisy. This temporal sensitivity is the formal basis of adaptive temporal continuity: the membrane integrates temporal structure, not merely instantaneous states.

Third, the codomain S is a space of filter functions, not a space of states. This captures the fact that the membrane does not merely pass signals through; it transforms them. The glucose transporter does not pass glucose unchanged; it couples glucose entry to conformational change. The immune synapse does not merely admit antigens; it presents them in a specific context that determines the response. The cultural membrane does not merely transmit ideas; it encodes them in a specific representational format that shapes what can be expressed.

I.3 Three Canonical Membrane Operations

Every instantiation of the cognitive membrane implements three canonical operations. These operations are not independent; they form a functional cycle that constitutes the membrane’s cognitive work.

I.3.1 Temporal Integration

The membrane selectively admits historical signals while filtering temporally uncorrelated noise. Formally, the temporal integration operation is the restriction of the filter function s_t to signals that are correlated with the system’s historical state trajectory I_0..t-1. The membrane does not merely sample the environment at each time step; it maintains a running integration of past signals and adjusts its permeability in light of this history. At the cellular level, this is implemented by receptor sensitization and desensitization; at the neural level, by synaptic potentiation and depression; at the cultural level, by tradition, canon, and institutional memory.

I.3.2 Reconstitutive Metabolism

Environmental input admitted through the membrane is not stored unchanged but is actively transformed into internal invariant structure; compressed representations that can be rapidly accessed and applied to future filtering decisions. This is the metabolism of signal into structure, of noise into pattern, of event into memory. The reconstitutive character of this transformation is crucial: the membrane does not merely record; it digests, eliminating the irrelevant and crystallizing the structurally significant.

I.3.3 Ascent Propagation

Compressed abstractions produced by reconstitutive metabolism are relayed upward to the next stratum of the cognitive hierarchy. The output of a lower-level membrane operation becomes the medium in which a higher-level membrane operates. The lipid bilayer’s selective integration of ionic gradients produces bioelectric signals that become the medium of morphogenetic cognition; neural activity patterns become the medium of conceptual cognition; conceptual structures become the medium of cultural cognition. This propagation is the formal mechanism of the Abstraction Ascent Stack.

I.3.4 The Emergence of Medium from Kernel Incompatibility

The emergence of medium marks the first moment in which the generative continuum becomes historically self‑referential. It is the point at which the grammar’s six operators encounter a structural mismatch they cannot dissolve through direct coupling, and must therefore metabolize the mismatch into a new ontological layer. Medium is not an added substrate. It is the remainder (the differential shadow) generated whenever two kernel regimes attempt adjacency but cannot be reconciled at the operator level.

Every kernel carries its own polarity gradient, indeterminacy aperture, and refraction geometry. When two kernels or kernel regimes meet, they attempt adjacency through the grammar’s coupling interface. If their operator stacks are compatible, adjacency produces a higher‑order kernel configuration: a new rung in the generative ascent. But when the operator stacks are incompatible, the grammar cannot fuse them. The mismatch cannot be ignored; tension accumulates. The system must act.

This is the moment at which the grammar invokes the metabolization–redistribution pair. The metabolization operator M resolves the interaction into two components: (1) the portion that can be integrated into a stable successor configuration, and (2) the portion that cannot. The latter is the residue ρ. Redistribution D disperses this residue into the surrounding kernel network, where it becomes a persistent structural remainder.

That remainder is the medium.

Medium is therefore not a primitive. It is a consequence; the metabolized residue of unresolved generative mismatch. It is the historical accumulation of everything the grammar could not fuse, everything that resisted direct integration, everything that remained after the operators had done all they could. Medium is the archive of incompatibility.

This gives medium its defining properties. It is historical, because residue accumulates across recursive cycles. It is stratified, because each layer of residue corresponds to a specific incompatibility event at a specific scale. It is preservational, because redistribution disperses residue without erasing it. And it is generative, because the accumulated residue becomes the substrate from which new invariants emerge.

In this sense, medium is the ladder of abstraction. Invariants are the rungs. Recursive agency is the climber. Each incompatibility event produces residue; each residue is redistributed; each redistribution becomes the historical substrate from which the next invariant is extracted. The vertical ascent of abstraction is therefore not an ascent through space but an ascent through medium, the accumulated record of generative mismatch.

This is why the genome can be understood as temporally stratified cognition. Genomic architecture is the long‑duration medium produced by billions of years of kernel incompatibility resolution. Each gene, each regulatory motif, each conserved developmental pathway is a metabolized remainder; an invariant extracted from the residue of evolutionary mismatch and preserved as part of the organism’s medium. Cognition, operating in real time, performs the same operation at a faster temporal scale: resolving mismatches between perception and world‑model, metabolizing the residue into memory, and climbing the invariant ladder.

Michael Levin’s light cone of intelligence is the geometric expression of this process. The light cone is the reachable region of recursive agency; the set of invariants an agent can extract from its medium and the set of future states it can stabilize through those invariants. Medium defines the cone’s boundary. Invariants define its interior geometry. Agency defines its expansion.

Thus the emergence of medium is not an incidental feature of the generative continuum. It is the structural mechanism by which the continuum becomes capable of history, memory, and ascent. Medium is the metabolized remainder of incompatible kernel adjacency; the differential shadow of generative output that cannot be reconciled and is therefore preserved as the substrate of future invariants. It is the condition of possibility for recursive agency, for abstraction, for evolution, and for intelligence itself.

I.4 Scale-Invariance: Membrane Instantiations Across Biological Scales

Scale	Membrane Instantiation	Inside (I)	Environment (E)	Characteristic τ	Filter Mechanism
Molecular	Lipid bilayer	Cytoplasm	Extracellular milieu	Milliseconds	Channel gating, pump kinetics
Cellular	Cell surface / receptor array	Cell state	Tissue signals	Minutes–hours	Signal transduction cascades
Tissue	Body plan boundary	Organism	Physical environment	Days–years	Immune system, epithelial junctions
Neural	Attentional / predictive filter	Internal model	Sensory stream	Seconds–years	Predictive coding, attention
Conceptual	Conceptual schema boundary	Belief system	Informational environment	Years–decades	Cognitive frameworks, priors
Cultural	Cultural membrane	Community	Cross-cultural contact zone	Decades–millennia	Canon, tradition, law, language
Civilizational	Paradigmatic membrane	Scientific tradition	Anomalies, revolutionary ideas	Centuries	Peer review, institutional gatekeeping

Part II. The Medium as Differential Shadow: The Reconstitutive Medium Operator

II.1 Rethinking the Environment

Classical cognitive science treats the environment as a source of stimuli; signals that impinge upon the organism from outside and trigger internal processing. Even ecological and embodied approaches, which rightly emphasize the organism-environment coupling, tend to treat the environment as a pre-given structure that affords or constrains action. What these approaches miss is the active metabolic role the environment plays in consolidating, compressing, and preserving the historical outputs of the agents that inhabit it. The environment is not merely the stage on which cognition occurs; it is an active participant in the cognitive process; a medium that metabolizes generative output and returns it as causally active invariant structure.

Consider: when a colony of Formica ants lays pheromone trails, the environment does not merely receive the chemical signal. The trail-laying act, repeated by multiple agents, is metabolized by the environment (evaporation, reinforcement, branching) into a compressed representation of collective foraging history that then guides the next generation of foraging decisions. The environment has performed a cognitive operation: dimensionality reduction on a distributed generative act, producing an invariant that persists across the replacement of individual agents. This is the formal structure of the medium.

II.2 Formal Definition of the Medium

Definition 2 (Reconstitutive Medium). A reconstitutive medium is a physical or abstract substrate M equipped with a metabolic function μ: G × T → I, where G is the space of generative outputs and I is the space of invariant structures, such that for any generative output g ∈ G at time t: μ(g, t) = I_t where I_t is a lower-dimensional compressed structure that (a) is causally active upon subsequent generative acts by agents in the medium, and (b) persists across timescales substantially longer than the characteristic timescale of the generating agents.

Definition 3 (Reconstitutive Medium Operator). The Reconstitutive Medium Operator is the mapping: R(G, M, t) = I_t where G is generative output, M is the medium’s metabolic function μ, and I_t is the resulting invariant structure at time t. The operator R is a compression operator: dim(I_t) < dim(G), with the information-theoretic content of I_t capturing the structurally invariant component of G while discarding idiosyncratic variation.

II.3 The Differential Shadow

We define the medium formally as the differential shadow of generative output. The metaphor is precise: just as a shadow is a lower-dimensional projection of a three-dimensional object (preserving outline and structural invariants while losing depth) so the medium is a lower-dimensional projection of the generative activity of the agents that have inhabited it. The shadow is differential in two senses: it registers only the difference between successive generative acts (what is new, what is reinforced, what is extinguished), and it is the mathematical differential (the cumulative integration) of the generative process over time.

Crucially, the shadow is causally active. A shadow does not cause illumination, but a medium-as-shadow does cause subsequent generative acts. The ant trail guides the next forager; the epigenetic mark guides the next cell’s transcriptional decisions; the legal precedent guides the next judge’s ruling; the scientific paradigm guides the next researcher’s experimental design. In each case, the compressed invariant stored in the medium exerts downward causation on the agents that produced it.

II.4 Generalization: A Taxonomy of Reconstitutive Media

Domain	Generative Output (G)	Medium (M)	Metabolic Function (μ)	Invariant (I_t)	Causal Return
Entomology	Pheromone deposition	Substrate + airflow	Evaporation + reinforcement	Trail network	Foraging path selection
Genetics	Gene expression pattern	Chromatin / epigenome	Methylation, acetylation	Epigenetic state	Transcriptional regulation
Linguistics	Individual utterances	Language community	Usage frequency + drift	Grammar + lexicon	Constraints on speech acts
Science	Experimental reports	Literature corpus	Peer review + citation	Paradigm	Research question framing
Law	Judicial rulings	Legal tradition	Precedent formation	Doctrine	Future case decisions
Morphogenesis	Cell signaling acts	Bioelectric field	Integration of local signals	Morphogenetic pattern	Cell fate specification
Evolution	Phenotypic variants	Fitness landscape	Natural selection	Adaptive landscape topology	Selection pressure on next generation

Part III. Genomics as Temporally Stratified Cognition

III.1 Against the Blueprint Metaphor

The metaphor of the genome as a blueprint has dominated molecular biology since the discovery of the double helix. A blueprint is a static, context-free specification: it contains complete information about the final structure and is read once in a single pass. The genome, as we now understand it, is none of these things. It is dynamic, context-sensitive, read repeatedly in different cellular contexts, and contains information at multiple temporal scales simultaneously. The blueprint metaphor obscures the genome’s most important cognitive property: its temporal stratification.

The genome is better understood as a multi-temporal cognitive archive; a structured information store in which different regions encode solutions to selection pressures operating at radically different timescales. Some genomic information reflects selection pressures acting across hundreds of millions of years (the Hox gene cluster, conserved transcription factor binding sites, core metabolic enzymes); other information reflects selection acting across thousands of generations (population-level allelic variation, locally adapted regulatory haplotypes); still other information reflects selection acting within individual developmental and somatic lifetimes (epigenetic marks that respond to environmental signals within hours or days). These are not merely different genomic regions; they are different cognitive strata, each with its own characteristic timescale, update function, and read operator.

III.2 The Genomic Temporal Stack

Definition 4 (Genomic Temporal Stack). The Genomic Temporal Stack (GTS) is the ordered tuple: GTS = ⟨L1, L2, …, Ln⟩ where each stratum Li is defined by the triple: Li = (τi, Ui, Ri) with τi the characteristic timescale of the stratum (the mean time between significant updates), Ui the update function specifying how the stratum’s content is modified by selective pressure, and Ri the read operator specifying how the stratum’s content is accessed and expressed during development. The strata are ordered such that τ1 < τ2 < … < τn (from fastest to slowest).

III.3 Characterizing the Strata

Drawing on contemporary molecular biology, we characterize four primary strata of the GTS:

Stratum	Genomic Substrate	τi (Characteristic Timescale)	Update Function Ui	Read Operator Ri	Cognitive Analogue
L1 – Epigenetic	DNA methylation, histone modification	Hours to decades	Environmental signal transduction; somatic reprogramming	Chromatin accessibility; TF binding probability	Working memory; short-term adaptive adjustment
L2 – Regulatory	Promoter regions, enhancers, non-coding RNA	Thousands of generations	Selection on gene expression timing and tissue-specificity	Developmental stage-specific TF activation	Procedural memory; context-specific execution scripts
L3 – Transposable Elements	Retrotransposons, SINEs, LINEs	Tens of thousands to millions of years	Transpositional activity; exaptation by selection	Stress-induced activation; co-option by regulatory networks	Episodic memory; medium-term adaptive reservoir
L4 – Deep Conserved	Hox clusters, core developmental genes, tRNA genes	Hundreds of millions of years	Purifying selection; near-zero mutation tolerance	Constitutive expression; evolutionary constraint	Long-term semantic memory; foundational world-model

III.4 Temporally Stratified Cognition: Reading Across Strata

The developing organism does not read the genome stratum by stratum sequentially; it reads across all strata simultaneously, integrating signals from multiple temporal scales into a coherent phenotypic output. This simultaneous multi-temporal read is the formal definition of temporally stratified cognition. The zygote at the moment of fertilization is simultaneously reading: (a) deep conserved gene networks that specify the fundamental body plan (L₄); (b) regulatory sequences that specify tissue-specific and stage-specific expression patterns (L₂); (c) transposable element insertions that have been exapted for novel regulatory functions in the lineage (L₃); and (d) parental epigenetic marks that encode environmental information from the parent’s lifetime (L₁).

Part IV. The Operator-Stack and Vertical Ascent: The Abstraction Ascent Stack

IV.1 The Architecture of the Stack

The central formal contribution of the CMMATC is the Abstraction Ascent Stack (AAS); a hierarchical tower of cognitive operators spanning from the teleodynamic physical substrate to meta-cognitive recursion. The AAS is not merely a descriptive taxonomy of cognitive levels; it is a compositional formal architecture in which each layer’s output is the substrate for the layer above, and each layer’s operation is a membrane operation in a specific medium at a specific temporal scale.

Definition 5 (Abstraction Ascent Stack). The Abstraction Ascent Stack (AAS) is the compositional operator: AAS = L7 ∘ L6 ∘ L5 ∘ L4 ∘ L3 ∘ L2 ∘ L1 ∘ L0 where each layer Lk is defined by the four-tuple: Lk = (Mk, Medk, τk, Ck) with Mk the membrane operator active at layer k, Medk the medium in which it operates, τk the characteristic temporal scale of the layer, and Ck the ascent coupling function specifying how layer k’s output becomes layer k+1’s substrate.

IV.2 Layer Definitions

Layer 0: Teleodynamic Physical Substrate

At the base of the stack lies the physical substrate constituted by Deacon’s teleodynamic constraints. End-directed physical processes (autocatalytic cycles, dissipative structures, self-organizing chemical systems) constitute the first membrane operation: the separation of a thermodynamically bounded system from its environment. The medium is the physical-chemical environment; the temporal scale is that of chemical reaction kinetics (femtoseconds to seconds); the ascent coupling is the production of stable autocatalytic cycles that constitute the chemical boundary conditions for biological membranes.

Layer 1: Membrane Operations

The first biological membrane (the lipid bilayer) instantiates the selective boundary function in physical form. The membrane implements the self/other distinction at the chemical scale: ionic gradients across the bilayer constitute the inside/outside distinction, and channel proteins implement the selective filter function. The medium is the electrochemical gradient maintained across the bilayer; the temporal scale is milliseconds to hours; the ascent coupling is the production of stable bioelectric signals that constitute the medium for morphogenetic cognition.

Layer 2: Genomic Cognition

The Genomic Temporal Stack (GTS) constitutes Layer 2: temporally stratified memory, multi-scale adaptive history encoded in a multi-band cognitive archive. The membrane operation here is the selective read across temporal strata; deciding which historical stratum is relevant to the current developmental context. The medium is the genome itself (the inscribed record of adaptive history); the temporal scale spans hours (epigenetic) to billions of years (deep conserved); the ascent coupling is the production of gene expression patterns that constitute the positional and temporal signals for morphogenetic cognition.

Layer 3: Morphogenetic Cognition

The developing body plan is a cognitive field: bioelectric potentials, chemical morphogen gradients, and mechanical tension fields are distributed representations being read, written, and transformed by cells acting as cognitive agents. The membrane operation is the integration of local signals into global spatial patterns. The medium is the multicellular tissue field (the Morphogenetic Cognitive Field, formalized in Part VIII); the temporal scale is hours to years; the ascent coupling is the production of the body plan (specifically the neural architecture) that constitutes the substrate for neural cognition.

Layer 4: Neural Cognition

The nervous system instantiates the membrane operation at the scale of organismal experience: recurrent predictive modeling, in which the brain maintains an internal model of the world and continuously updates it against sensory prediction errors (the predictive coding framework). The medium is the neural network; the pattern of synaptic weights instantiated in biological tissue; the temporal scale is milliseconds to decades (from spike timing to lifetime learning); the ascent coupling is the production of compressed world-models that constitute the substrate for conceptual cognition.

Layer 5: Conceptual Cognition

At Layer 5, the membrane operation is symbolic abstraction: the compression of world-models into compositional symbolic representations that can be combined, manipulated, and communicated. The medium is the symbolic-linguistic-conceptual space instantiated in the individual mind and shared with conspecifics through communication; the temporal scale is seconds to decades (from sentence processing to conceptual development); the ascent coupling is the production of symbolic outputs (utterances, artifacts, inscriptions) that enter cultural media and constitute the substrate for cultural cognition.

Layer 6: Cultural Cognition

The cultural membrane is the most extensive cognitive membrane in the biological world (as we currently know it): it implements collective reconstitutive media (language, law, science, religion, art) that integrate the symbolic outputs of millions of individual agents across generations. The membrane operation is the filtering of individual cognitive outputs through institutional structures that determine which outputs are preserved, amplified, and integrated into the cultural invariant. The medium is the cultural apparatus in its totality; the temporal scale is decades to millennia; the ascent coupling is the production of civilizational-level knowledge structures that constitute the substrate for meta-cognitive recursion.

Layer 7: Meta-Cognitive Cognition

At the apex of the current stack is recursive self-modeling: a cognitive agent modeling its own cognitive architecture. This is the layer at which the AAS becomes aware of itself; where agency models its own agency. The membrane operation is the construction of an accurate self-model that can be used to predict, regulate, and intentionally modify the lower layers of the stack. The medium is the reflective capacity instantiated in neural and cultural structures; the temporal scale is seconds to lifetimes; the ascent coupling is, at present, undefined; it would represent the construction of cognitive architectures that transcend current human cognitive limits.

IV.3 The Full AAS Architecture

IV.4 Recursive Downward Causation

A critical feature of the AAS, which distinguishes it from simple hierarchical models, is that the causal relationship between layers is bidirectional. Higher layers do not merely emerge from lower layers; they exert downward causation on them. Cultural practices shape neural development (Layer 6 → Layer 4); neural activity patterns regulate gene expression (Layer 4 → Layer 2); body plan structures constrain bioelectric field dynamics (Layer 3 → Layer 1). This bidirectionality means the AAS is not a one-way pipeline but a recursive, mutually constraining architecture; each layer both reads from and writes to the layers surrounding it.

Part V. The Intelligence Cone and the Recursive Agency Continuum

V.1 Michael Levin’s Light Cone of Intelligence

Michael Levin has proposed a powerful conceptual tool for comparing cognitive agents across the vast space of biological and artificial systems: the intelligence cone, modeled by analogy with the light cone of special relativity. In special relativity, the light cone of an event defines the set of all spacetime points that can causally influence or be influenced by that event, given the finite speed of light. Levin proposes that every cognitive agent has an analogous spatio-temporal cone of influence: the set of all (spatial, temporal) coordinates over which the agent can set goals and influence outcomes toward those goals.

This framing is illuminating for three reasons. First, it provides a continuous measure of cognitive depth; agents are not categorized into discrete types (reflexive vs. cognitive vs. conscious) but are situated on a continuum defined by the extent of their influence cone. Second, it grounds the measure in goal-directedness across space and time rather than in substrate: what matters is not whether the agent has neurons, but how far across space and time it can maintain and pursue goals. Third, it provides a natural bridge to the AAS: the expansion of the intelligence cone across evolutionary time corresponds precisely to the addition of AAS layers.

V.2 Formal Definition of the Intelligence Cone

Definition 6 (Intelligence Cone). For a cognitive agent A, the Intelligence Cone IC(A) is the set: IC(A) = {(x, t) : Agent A can influence or model state at (x, t) in service of a goal} The cognitive depth of agent A is the measure μ(IC(A)); the volume of the intelligence cone in spatio-temporal space. The temporal horizon of A is the supremum of t-values in IC(A).

V.3 Intelligence Cones Across AAS Layers

The expansion of the intelligence cone across AAS layers is the formal content of the claim that evolution represents an ascent in cognitive architecture space. At each successive layer, the cone expands along both spatial and temporal dimensions:

AAS Layer	Representative Agent	Spatial Cone Extent	Temporal Cone Extent	Cone-Expanding Mechanism
L0 – Teleodynamic	Autocatalytic chemical cycle	Nanometers	Milliseconds	Thermodynamic constraint
L1 – Membrane	Prokaryotic cell	Micrometers	Minutes–hours	Selective ion channel gating
L2 – Genomic	Any organism with genome	Organism body	Lifetime + epigenetic inheritance	Temporal stratification of memory
L3 – Morphogenetic	Multicellular organism	Body plan extent	Developmental program duration	Field-based spatial integration
L4 – Neural	Vertebrate with nervous system	Sensory-motor range	Years (episodic memory)	Predictive internal modeling
L5 – Conceptual	Language-using hominid	Communicative reach	Decades (autobiographical narrative)	Symbolic compression + communication
L6 – Cultural	Human civilization	Global (civilization-wide)	Millennia (cultural memory)	Institutional reconstitutive media
L7 – Meta-Cognitive	Reflectively self-aware agent	Potentially unbounded	Potentially transgenerational	Self-model enabling recursive self-modification

V.4 The Cone as the Continuum of Recursive Agency

A critical insight of the intelligence cone formalism is that it reveals the AAS not as a discrete ladder but as a continuum of recursive agency outcomes. The membrane → genome → morphogenesis → neural → conceptual → cultural sequence is not a set of qualitatively distinct categories but a series of quantitative expansions of a single cone. This continuum interpretation resolves a persistent problem in theories of mind and life: the question of where to draw the boundary between “mere mechanism” and “genuine cognition.” The answer, on the CMMATC view, is that this boundary cannot be drawn absolutely; it is always a matter of cone extent, not categorical kind.

Furthermore, the cone formalism allows us to define a precise measure of cognitive distance between agents (the Ontological Distance) as the distance between their intelligence cones. This is developed in Part VII.

Part VI. Insight as Developmental Phase Transition

VI.1 The Standard View and Its Inadequacy

The received view of insight in cognitive psychology treats it as a sudden, discontinuous event: the “aha moment” in which a solution that was previously inaccessible becomes available through a rapid reorganization of problem representation. This view, associated with the Gestalt tradition and refined by contemporary work on representational change, captures something real about the phenomenology of insight (it does feel sudden) but it fundamentally mischaracterizes the underlying cognitive dynamics. What appears sudden at the phenomenological surface reflects a sub-threshold process of gradual, continuous developmental reorganization that eventually exceeds a threshold and becomes perceptible.

The CMMATC offers a precise alternative: insight events are membrane-crossing events in developmental cognitive space. They occur when accumulated sub-threshold cognitive reorganizations (modifications to the cognitive architecture that do not individually produce phenomenologically salient change) collectively exceed a threshold defined by the establishment of a new membrane boundary condition. The moment of insight is the moment at which a new inside/outside distinction becomes established: a new self/world boundary that reorganizes the entire structure of available experience and action.

VI.2 The Insight Threshold Formalism

Definition 7 (Insight Event). Let C be a cognitive system in developmental state space Ω. Let R1, R2, …, Rk be a sequence of sub-threshold cognitive reorganizations; modifications to the cognitive architecture that do not individually produce a membrane-boundary-change event. Define the cumulative reorganization magnitude: |∑i=1..k Ri| = ρk An insight event I* is defined as the event at time t* such that: ∃t* : ρk(t*) > θ where θ is the membrane threshold; the minimum cumulative reorganization required to establish a new membrane boundary condition in cognitive space. The insight event corresponds to the crossing of this threshold, producing a qualitative change in the cognitive architecture’s self/world distinction.

Theorem 5 (Insight Threshold Theorem). Insight events are membrane-crossing events in cognitive developmental space. Every phenomenologically salient insight corresponds to the establishment of a new boundary condition Mnew: I’ × E’ × T’ → S’ that is discontinuous with the preceding membrane operator Mold, but arises from a continuous sub-threshold developmental trajectory in cognitive architecture space.

VI.3 Scale Correspondence: Individual and Cultural Insight

A powerful consequence of the membrane-crossing model of insight is that it applies at all scales of the AAS simultaneously. Individual cognitive insight (Layer 4–5 event) and paradigm-level cultural shift (Layer 6 event) are formally identical: both are membrane-crossing events in which accumulated sub-threshold reorganizations exceed a threshold and establish a new boundary condition. Thomas Kuhn’s description of scientific revolutions (long periods of normal science punctuated by rapid paradigm shifts) is precisely what the Insight Threshold theorem predicts for collective cognitive systems operating at Layer 6 of the AAS.

This scale correspondence is not merely analogical; it reflects the structural identity of the membrane operator across scales. Kuhn’s “paradigm” is a cultural membrane: it defines what counts as inside (legitimate scientific practice) and what counts as outside (pseudo-science, metaphysics, anomaly). A paradigm shift is the replacement of one cultural membrane by another; a membrane-crossing event at the civilizational scale, produced by the accumulation of sub-threshold anomalies until they exceed the paradigm’s capacity to assimilate them.

VI.4 Developmental Prerequisites for Insight

The CMMATC predicts that insight events require specific developmental prerequisites at the level of the AAS: the sub-threshold reorganizations that accumulate before an insight event are not random but are structurally organized by the current cognitive architecture. This explains the well-documented phenomenon of insight incubation (the period of apparently unproductive engagement with a problem that precedes insight) as a phase of sub-threshold structural reorganization. The cognitive system is not idle during incubation; it is performing the reorganizational work that must accumulate before the membrane threshold is crossed.

Part VII. Ontological Distance: A Metric on the Space of Intelligence Cones

VII.1 The Problem of Cross-Architecture Communication

One of the most persistent difficulties in comparative cognition, cross-cultural communication, and artificial intelligence alignment is the problem of understanding between agents with radically different cognitive architectures. The standard approach to this problem treats it as a translation problem: we need better methods for converting representations from one format to another. The CMMATC reframes it as a distance problem in ontological space: agents with large ontological distance between their cognitive architectures cannot communicate effectively not because of a translation failure but because their fundamental self/world distinctions (their membrane structures) are incommensurable. There is no common reference frame from which to perform the translation.

VII.2 Ontological Space

Definition 8 (Ontological Space). Ontological Space O is the metric space of all possible cognitive architectures; all possible configurations of (membrane operators, medium structures, AAS operator stacks). A point p ∈ O represents a specific cognitive architecture fully specified by its membrane operators, media, temporal scales, and ascent coupling functions at each AAS layer.

VII.3 Ontological Distance as a Pseudo-Metric

Definition 9 (Ontological Distance). The Ontological Distance Do(A, B) between cognitive agents A and B is: Do(A, B) = inf { len(γ) : γ is a path in O from IC(A) to IC(B) } where IC(A) and IC(B) are the intelligence cones of agents A and B (mapped into their positions in ontological space), and len(γ) is the length of path γ measured by a metric on O that assigns distance proportional to the dissimilarity of membrane operators, media structures, and ascent coupling functions encountered along the path.

Theorem 6 (Ontological Metric Theorem). Do is a well-defined pseudo-metric on the space of intelligence cones: it satisfies non-negativity (Do(A,B) ≥ 0), symmetry (Do(A,B) = Do(B,A)), and the triangle inequality (Do(A,C) ≤ Do(A,B) + Do(B,C)). It is a pseudo-metric (not a full metric) because cognitively identical architectures instantiated in different substrates have Do = 0.

VII.4 Applications of Ontological Distance

The ontological distance framework has immediate applications across several domains:

VII.4.1 Cross-Species Communication

Two animal species may be spatially adjacent yet have large ontological distance if their AAS architectures differ substantially. The dolphin and the human share Layers 0–4 of the AAS but diverge significantly at Layer 5: cetacean conceptual structures are organized around fundamentally different sensorimotor primitives (echolocation-based spatial cognition vs. manual manipulation and visual object recognition). The ontological distance between human and dolphin is therefore moderate; smaller than between human and bacterium, larger than between human and chimpanzee. The difficulty of establishing cross-species communication is a direct function of this distance.

VII.4.2 Human-AI Alignment

The AI alignment problem (ensuring that artificial intelligence systems pursue goals aligned with human values) can be precisely reframed as an ontological distance problem. Current large-scale artificial neural networks instantiate something like Layers 4–5 of the AAS, but their membrane operators differ fundamentally from biological ones: they lack teleodynamic grounding (Layer 0), genomic memory (Layer 2), morphogenetic embodiment (Layer 3), and developmental history (which shapes biological Layers 4–5 in ways not replicated by gradient descent on static datasets). The ontological distance between a contemporary AI system and a human is therefore substantial; primarily along the AAS-depth dimension. Alignment research, on this view, is the project of reducing this ontological distance by building cognitive architectures that more fully instantiate the AAS.

VII.4.3 Cultural Translation

Cultural translation (the process of conveying meaning across cultural membranes) is possible precisely to the extent that two cultural architectures share AAS layers. Human cultures share Layers 0–5 universally (all human cultures arise from the same biological substrate through Layers 0–5) and differ primarily at Layer 6. Ontological distance between cultures is therefore moderate (substantially smaller than between species or between humans and AI systems) which is why cultural translation, though difficult, is reliably achievable.

Agent Pair	Shared AAS Layers	Diverging AAS Layers	Estimated Do	Communication Feasibility
Human ↔ Chimpanzee	L0–L4	L5–L7 (partial)	Low-moderate	Feasible with scaffolding (sign language, tokens)
Human ↔ Dolphin	L0–L4	L5–L7	Moderate	Rudimentary (gestural, acoustic)
Human ↔ Octopus	L0–L3	L4–L7	High	Very limited (behavioral)
Human ↔ Bacterium	L0–L2	L3–L7	Very high	Effectively absent at intentional level
Human ↔ Current AI LLM	L4–L5 (partial)	L0–L3, L6–L7	High (different axes)	High in linguistic domain; poor in embodied/values domain
Human culture A ↔ B	L0–L5	L6 (partially)	Low-moderate	Achievable with effort (translation, cultural exchange)

Part VIII. Morphogenesis as Continuous Cognitive Field

VIII.1 The Body as Cognitive Achievement

The standard view of morphogenesis treats the development of the body plan as the mechanical execution of a genetic program: genes encode proteins, proteins interact to form regulatory networks, and regulatory networks drive the differentiation of cells into specific tissue types according to their position in the embryo. On this view, the body is the product of a program, and cognition begins only when the program has been executed; when the brain is in place and functioning.

The CMMATC rejects this view. Morphogenesis is not the precondition for cognition; it is cognition: cognition at Layer 3 of the AAS, distributed across the tissue field. The cells of the developing embryo are not executing a program; they are performing a distributed cognitive act. They read signals from the bioelectric field, the chemical gradient landscape, and the mechanical tension field; they integrate these signals through membrane operations; they produce output signals that modify the field for other cells; and they commit to cell-fate decisions (differentiation) on the basis of this integrated information. This is exactly the structure of cognition as defined in the CMMATC: membrane-mediated selective temporal integration in a reconstitutive medium.

VIII.2 Bioelectric Fields as Cognitive Media

Michael Levin’s experimental program has demonstrated that bioelectric fields (patterns of transmembrane voltage across tissues) function as a medium for long-range morphogenetic information. Bioelectric signals can specify organ identity, control tissue polarity, coordinate the regeneration of complex structures (including the vertebrate nervous system), and even redirect the developmental program toward alternative body plans. These are not peripheral modulatory effects; they are constitutive of the morphogenetic process. The bioelectric field is a cognitive medium in the precise sense of Definition 2: it receives generative outputs from individual cells (ion pump activity, gap junction conductance), metabolizes these outputs into distributed representations (voltage gradients, spatio-temporal patterns), and exerts causal influence on subsequent cell behavior.

VIII.3 The Morphogenetic Cognitive Field

Definition 10 (Morphogenetic Cognitive Field). The Morphogenetic Cognitive Field (MCF) at time t is the spatial integral of membrane-mediated signals across the developing tissue: MCF(t) = ∫∫A m(x, y, t) dA where A is the tissue area, and m(x, y, t) is the local membrane-mediated signal at position (x, y) at time t; the sum of bioelectric, chemical, and mechanical signals integrated through membrane operations at that position. The body plan is the time-integral of the MCF: BP = ∫0T MCF(t) dt The body plan is therefore not the execution of a program but the frozen time-integral of a cognitive field; crystallized distributed intelligence.

VIII.4 Body Plan as Frozen Cognition

The phrase “frozen cognition” requires precision. We do not mean that the adult body plan is static or unchanging; it is of course continuously maintained and dynamically regulated. We mean that the morphological structure of the body plan encodes the history of the cognitive field that produced it: the sequence of patterning decisions, the integration of signals across different temporal and spatial scales, the resolution of competing morphogenetic influences into a coherent spatial organization. The body plan is the trace of the cognitive process, just as a mathematical proof is the trace of a reasoning process.

This interpretation has a striking implication: the body plan is an invariant (in the formal sense of Definition 2) produced by the Reconstitutive Medium Operator acting on the cellular generative outputs in the tissue field. The body plan is what the morphogenetic medium has metabolized from the individual cellular signaling acts: a lower-dimensional structure (three-dimensional body plan) that is more stable and causally potent than the individual acts (local ion pump activity) that generated it. Body plans are reconstitutive invariants; and their extraordinary conservation across evolutionary lineages (the body plans of bilaterians have been conserved for over 550 million years) is exactly what the Reconstitutive Accumulation theorem (Theorem 7) predicts.

Part IX. Teleodynamic Grounding: Absence as the Foundation of Intelligence

IX.1 Deacon’s Teleodynamics

Terrence Deacon’s project in Incomplete Nature (2011) is to account for the emergence of purpose-like causation (intentionality, normativity, end-directedness) from purely physical processes, without appealing to vitalist forces or supernatural intervention. His solution is teleodynamics: a theory of how higher-order emergent dynamics arise from the interaction of lower-order dynamical processes through the operation of constraints. Crucially, Deacon argues that the most important constraints in teleodynamic systems are absences (what the system is prevented from doing or becoming) rather than positive contents. A living organism is defined as much by what it excludes (it excludes disorder, maintains non-equilibrium states, resists entropic dissolution) as by what it contains.

Deacon introduces three levels of dynamics: thermodynamics (purely dissipative processes), morphodynamics (self-organizing processes that produce stable, far-from-equilibrium structures), and teleodynamics (processes that are end-directed; that maintain themselves relative to a future state). Each higher level is constituted by the constraint that one lower-level process exercises over another: morphodynamics arises when thermodynamic processes mutually constrain each other to produce stable patterns; teleodynamics arises when morphodynamic processes mutually constrain each other to produce end-directedness.

IX.2 The Cognitive Membrane as a Teleodynamic Entity

The cognitive membrane, on the CMMATC account, is paradigmatically teleodynamic. Its defining property is not its positive content but its selectivity; what it excludes. A perfectly permeable membrane is cognitively inert; it is the structure of its exclusions that constitutes its intelligence. The more sophisticated the exclusion structure (the more finely tuned the membrane’s capacity to admit certain signals while excluding others, in a context-sensitive and temporally integrated manner) the greater the cognitive depth of the membrane.

This claim has a formal consequence: the intelligence of a membrane is not a positive property that can be fully enumerated by listing what the membrane does. It is constituted by the absence structure (the pattern of what is excluded) which is in principle an infinite space. This is why cognitive systems are difficult to fully characterize: their defining feature is their exclusion structure, and exclusion structures have an openness that positive contents do not. There is always more that a membrane could exclude; the sophistication of cognitive evolution is the evolution of increasingly refined exclusion structures.

IX.3 Absential Causation Across AAS Layers

Deacon’s concept of absential causation (causation by absent states) applies at every layer of the AAS:

AAS Layer	Absential Cause	Effect	Teleodynamic Structure
L0 – Teleodynamic	Absence of thermodynamic equilibrium	Sustained dissipative structure	Morphodynamic self-organization
L1 – Membrane	Absence of ionic equilibrium across bilayer	Maintained electrochemical gradient; action potential	Thermodynamic constraint by membrane selectivity
L2 – Genomic	Absence of expressed gene product	Developmental pathway specification	Absence of TF → absence of target gene expression → cell fate
L3 – Morphogenetic	Absence of morphogen signal	Default cell fate commitment	Pattern formation by gradient minima and maxima
L4 – Neural	Absence of predicted sensory input (prediction error)	Model update; learning signal	Predictive coding: absence IS the signal
L5 – Conceptual	Absence of confirmatory evidence	Belief revision; inquiry	Epistemic teleodynamics: absence drives inquiry
L6 – Cultural	Absence of resolution (anomaly)	Paradigm destabilization	Kuhnian tension: accumulating absences
L7 — Meta-Cognitive	Absence of self-understanding	Self-inquiry; philosophical investigation	Recursive absential causation

IX.4 Teleodynamics and the Origin of the Cognitive Membrane

The first cognitive membrane (the protocell lipid bilayer) arose, on the teleodynamic account, not by design but by the mutual constraint of morphodynamic processes. Amphipathic molecules in aqueous solution spontaneously form bilayers not because they are “programmed” to do so, but because the bilayer configuration is the attractor state of the morphodynamic process of amphipathic self-assembly. The resulting bilayer is a teleodynamic entity: it is defined by its selectivity (its systematic exclusion of ions and large molecules) and this exclusion structure is what makes it cognitively potent. The first cognitive membrane was not a solution to the problem of cognition; it was the instantiation of the teleodynamic conditions under which cognition becomes possible.

This teleodynamic origin story unifies the CMMATC’s account of the cognitive membrane with the empirical story of the origin of life: life began with the formation of the first selective membrane, which constituted the first inside/outside distinction, which constituted the first act of cognition at Layer 1 of the AAS. The emergence of life and the emergence of cognition are, on this account, the same event; the instantiation of the first teleodynamic membrane operator.

Part X. Unified Formal Summary: Theorems and Architecture Table

X.1 The Seven Core Theorems

Theorem 1 (Membrane Universality). Every cognitive operation at every scale of the AAS can be formally decomposed into a membrane operator M: I × E × T → S acting on a medium Med, where (a) the membrane partitions state space into inside and environment, (b) the temporal index T enables history-sensitive filtering, and (c) the selective filter function S specifies the transformation applied to admitted environmental signals. Interpretation: Cognition is not substrate-specific or scale-specific. The formal structure of cognition (selective temporal integration by a context-sensitive boundary operator) is the same from the lipid bilayer to the cultural institution. Differences between cognitive levels are differences in medium, temporal scale, and filter sophistication; not differences in kind.

Theorem 2 (Medium Invariance). The intelligence of a cognitive system is proportional to the dimensionality reduction it achieves in converting generative output to medium invariants: I(C) ∝ dim(G) / dim(I_t), where G is generative output, I_t is the medium invariant produced by R(G, M, t), and the ratio dim(G)/dim(I_t) measures the compression achieved. Interpretation: Intelligent systems are compression engines. The more efficiently a system distills its generative activity into compact, causally potent invariants, the greater its cognitive depth. Evolution is the progressive improvement of compression efficiency across AAS layers.

Theorem 3 (Ascent Coupling). Each layer L_k of the AAS is both (a) a consumer of the output of L_{k-1} (it uses the invariants produced by the layer below as its medium) and (b) a substrate provider for L_{k+1}; its own invariants constitute the medium in which the layer above operates. Formally: Med_k = I_{k-1}^{invariant} and I_k^{invariant} = Med_{k+1}. Interpretation: The AAS is not a collection of independent levels but a single compositional architecture in which each layer’s output is the next layer’s environment. Disruption of any layer necessarily affects all layers above it through medium deprivation, and may affect layers below through the loss of downward constraint.

Theorem 4 (Cone Expansion). The intelligence cone of a cognitive system expands monotonically with the number of AAS layers it instantiates: if system A instantiates k layers and system B instantiates k+1 layers, then μ(IC(A)) < μ(IC(B)), where μ is the measure of the intelligence cone in spatio-temporal space. Interpretation: Each additional AAS layer provides a new mechanism for extending the system’s goal-directedness across greater spatial and temporal extents. Evolutionary progression through AAS layers is therefore monotonically intelligence-increasing; not by any external criterion, but by the internal structure of the architecture.

Theorem 5 (Insight Threshold). Insight events are membrane-crossing events in cognitive developmental space. For cognitive system C with accumulated sub-threshold reorganizations R_1, …, R_k, an insight event I* occurs at time t* such that |∑R_i| > θ, where θ is the membrane threshold for establishing a new boundary condition. The phenomenological discontinuity of insight is the subjective registration of a topological change in the cognitive architecture’s self/world boundary; a change that was prepared by a continuous sub-threshold developmental trajectory. Interpretation: Insight is not magic and is not random. It is a predictable outcome of sufficient sub-threshold developmental accumulation. The subjective sense of suddenness reflects the threshold’s sharpness, not the absence of preparation.

Theorem 6 (Ontological Metric). Ontological Distance D_o(A,B) = inf{len(γ): γ path in O from IC(A) to IC(B)} defines a well-defined pseudo-metric on the space of intelligence cones. D_o satisfies non-negativity, symmetry, and the triangle inequality. D_o(A,B) = 0 if and only if IC(A) and IC(B) are identical as structures in ontological space; i.e., the agents have equivalent cognitive architectures regardless of substrate. Interpretation: The difficulty of communication, coordination, or mutual modeling between two agents is a function of their ontological distance, not their physical proximity. Two agents can be physically adjacent yet ontologically distant (human and bacterium); two agents can be physically distant yet ontologically proximate (two human cultures).

Theorem 7 (Reconstitutive Accumulation). As t → ∞, the medium M asymptotically accumulates the differential shadow of all generative outputs: lim_{t→∞} R(G, M, t) = I_∞, where I_∞ is a compressed archive of the complete history of agency in the medium. This archive is causally active; it constitutes a stable, low-dimensional invariant that shapes all future generative acts by agents in the medium. The archive is the medium’s “cognitive memory”; its accumulated intelligence. Interpretation: Media become smarter over time. The genome encodes billions of years of adaptive history; language corpora encode millennia of human experience; legal systems encode centuries of social problem-solving. These are not passive records but causally active invariants that shape the cognitive acts of every agent that inhabits them.

X.2 Unified Architecture Summary Table

AAS Layer	Name	Membrane Operator M_k	Medium Med_k	Timescale τ_k	Ascent Coupling C_k	Teleodynamic Absence	IC Extent
L0	Teleodynamic	Thermodynamic constraint	Physical-chemical environment	fs–seconds	Stable autocatalytic cycles	Absence of equilibrium	Nanometers / ms
L1	Membrane	Selective ionic filter	Electrochemical gradient	ms–hours	Bioelectric signal patterns	Ionic disequilibrium	Micrometers / minutes
L2	Genomic	Multi-stratum temporal reader	GTS (genome archive)	hrs–Gyr	Gene expression patterns	Absent TF → cell fate	Organism / lifetime
L3	Morphogenetic	MCF spatial integrator	Bioelectric + morphogen fields	Hours–years	Body / neural plan	Absent morphogen → default fate	Body extent / developmental time
L4	Neural	Predictive coding filter	Synaptic weight matrix	ms–decades	Compressed world-model	Prediction error (absent confirmation)	Sensory range / years
L5	Conceptual	Symbolic abstraction boundary	Symbolic-linguistic space	Seconds–decades	Utterances / artifacts	Absent evidence → inquiry	Communicative reach / decades
L6	Cultural	Institutional filter	Cultural apparatus	Decades–millennia	Civilizational knowledge structures	Anomaly (absent paradigm fit)	Civilization-wide / millennia
L7	Meta-Cognitive	Recursive self-model operator	Reflective capacity	Seconds–lifetimes	[Open: trans-human architecture]	Absent self-understanding	Potentially unbounded

X.3 Cross-Section Formal Relationships Summary

Formal Object	Symbol	Definition	Part Defined	Key Theorem
Cognitive Membrane	M: I × E × T → S	Context-sensitive boundary operator producing selective filter	Part I	Theorem 1
Reconstitutive Medium Op.	R(G, M, t) = I_t	Compression of generative output into medium invariant	Part II	Theorem 7
Genomic Temporal Stack	GTS = ⟨L_1,…,L_n⟩	Multi-stratum cognitive archive; each stratum = (τ, U, R)	Part III	Theorem 1
Abstraction Ascent Stack	AAS = L_7 ∘ … ∘ L_0	Compositional tower of membrane operators across scales	Part IV	Theorems 3, 4
Intelligence Cone	IC(A) = {(x,t): A can influence/model at (x,t)}	Spatio-temporal extent of goal-directedness	Part V	Theorem 4
Insight Threshold	∃t*: |ΣR_i| > θ	Membrane-crossing in cognitive developmental space	Part VI	Theorem 5
Ontological Distance	D_o(A,B) = inf{len(γ)}	Path length in O between intelligence cones of A and B	Part VII	Theorem 6
Morphogenetic Cognitive Field	MCF(t) = ∬ m(x,y,t) dA	Spatial integral of membrane-mediated tissue signals	Part VIII	Theorem 1

3. Discussion

3.1 Implications for Biology

The CMMATC’s most immediate biological implication is the reframing of the genome as a cognitive substrate. This is not a metaphorical claim; it is a claim with specific empirical predictions. If the genome is a multi-temporal cognitive archive, then we should expect to find: (a) systematic differences in evolutionary conservation rate that correlate with temporal stratum membership (deep conserved > regulatory > transposable elements > epigenetic); (b) context-sensitive read mechanisms that integrate information across strata simultaneously; (c) evidence of downward causation from higher AAS layers (neural, cultural) to lower layers (genomic, epigenomic); a prediction that is now extensively supported by the neuroscience of experience-dependent gene expression and the emerging field of cultural epigenetics.

The Morphogenetic Cognitive Field formalism offers a new framework for understanding the remarkable fidelity of morphogenetic patterning. On the CMMATC account, body plan fidelity across evolutionary lineages is the expected consequence of Reconstitutive Accumulation (Theorem 7): the morphogenetic medium has accumulated billions of years of generative output into a highly compressed, causally potent invariant (the conserved body plan) that strongly constrains all subsequent morphogenetic acts. The extraordinary resistance of body plans to perturbation (the phenomenon of developmental robustness or canalization) is a direct prediction of the Medium Invariance theorem (Theorem 2): the most heavily compressed invariants are the most stable.

Finally, the teleodynamic grounding of the CMMATC offers a principled solution to the problem of biological normativity: how can biological systems be said to function correctly or incorrectly, to be healthy or diseased? The answer is that biological normativity is a property of teleodynamic systems whose functional states are defined by their absences; by what they maintain against entropic dissolution. A diseased state is one in which the membrane’s exclusion structure has been compromised: wrong things are admitted, right things are excluded, and the self/world distinction is degraded.

3.2 Implications for Artificial Intelligence

The AI alignment problem, reframed as an ontological distance problem, becomes tractable in a new way. Current alignment approaches focus primarily on adjusting the behavior of AI systems at Layer 5 (conceptual/linguistic output) without addressing the ontological distance arising from the absence of Layers 0–3 in current AI architectures. The CMMATC predicts that alignment interventions at Layer 5 alone will be systematically insufficient, because the fundamental self/world distinctions implemented by AI systems are constituted by architectures that lack teleodynamic grounding, genomic memory, and morphogenetic embodiment. A truly aligned AI would need to instantiate cognitive architectures with much smaller ontological distance from the human AAS; not merely produce human-compatible output, but operate with human-analogous membrane structures at multiple levels.

This has implications for AI architecture design. The development of artificial systems that instantiate more layers of the AAS (embodied AI systems with developmental trajectories (approaching Layer 3), systems with long-term memory architectures that implement temporal stratification (approaching Layer 2), and systems capable of genuine self-modeling and recursive architecture modification (approaching Layer 7)) represents the principled path toward both greater cognitive capability and greater alignment. The CMMATC provides a formal map of this developmental trajectory.

3.3 Implications for Cognitive Science

For cognitive science, the CMMATC’s central contribution is the reconceptualization of the mind/body distinction. On the CMMATC account, the boundary between body and mind is not a qualitative break but a quantitative transition within the AAS: body (Layers 0–3) and mind (Layers 4–7) are distinguished by temporal scale, medium, and filter sophistication, not by metaphysical kind. This has consequences for the debates surrounding embodied cognition, extended mind, and 4E cognition (embodied, embedded, enacted, extended).

The CMMATC supports the 4E program in principle but provides a more precise formal framework: the “extension” of mind into the environment is a specific claim about the reconstitutive medium; the environment has absorbed generative output and returned it as causally active invariant. The notebook in Clark and Chalmers’ extended mind thought experiment is a reconstitutive medium in the formal sense of Definition 2; its status as a cognitive component is not a matter of philosophical stipulation but of whether it satisfies the formal conditions of the Reconstitutive Medium Operator. The CMMATC thus provides a principled criterion for extension: a system is cognitively extended into an environment to the extent that R(G, M, t) produces invariants that (a) are causally active on the system’s subsequent cognition and (b) could not be reproduced from the system’s internal resources alone.

3.4 Implications for Philosophy of Mind

The CMMATC’s most radical philosophical implication concerns the nature of selfhood. On the standard view, the self is constituted by positive contents: memories, personality traits, beliefs, desires; a collection of internal states that individuate the agent. The CMMATC proposes instead that the self is constituted by its exclusion structure; the structure of what the cognitive membrane keeps out. The self is the pattern of selective exclusion implemented by the cognitive membrane across all scales of the AAS.

This view resonates with Buddhist philosophical traditions that characterize the self as an empty center (defined by its boundaries, not its contents) but provides a formal grounding for this intuition in terms of teleodynamic constraint theory. The self is not an illusion (as eliminativist readings of Buddhism suggest) but a real pattern of teleodynamic constraint; real in exactly the sense that the selective permeability of a lipid bilayer is real. It is a real absence-structure, not a positive substance.

For the philosophy of consciousness, the Insight Threshold theorem has implications for theories of phenomenal experience. If the subjective experience of insight is the felt crossing of a membrane (the establishment of a new self/world boundary condition) then phenomenal experience may be more generally related to the dynamics of membrane operations: the continuous monitoring and adjustment of the self/world boundary. Consciousness, on this view, is not a property of a specific substrate (neural tissue) but a functional property of cognitive systems operating at sufficiently high AAS layers with sufficiently complex membrane dynamics. This is not a behaviorist claim (consciousness is defined by behavior) but a structural-functional claim: consciousness is defined by the complexity of the membrane operation.

4. Conclusion

The Cognitive Membrane Model of Adaptive Temporal Continuity presents a unified formal architecture for understanding intelligence, cognition, and adaptive organization across all scales of biological and cultural reality. Its core achievement is the formalization of the cognitive membrane as a scale-invariant boundary operator; a structure that constitutes the self/world distinction at every level from the lipid bilayer to the civilizational paradigm, and whose three canonical operations (temporal integration, reconstitutive metabolism, ascent propagation) define the formal structure of cognition as such.

Built upon this foundation, the CMMATC develops eight integrated formal structures: the Cognitive Membrane operator (M: I × E × T → S), the Reconstitutive Medium Operator (R(G, M, t) = I_t), the Genomic Temporal Stack (GTS), the Abstraction Ascent Stack (AAS), the Intelligence Cone (IC(A)), the Insight Threshold formalism, the Ontological Distance metric (D_o), and the Morphogenetic Cognitive Field (MCF). These structures are unified by seven core theorems (Membrane Universality, Medium Invariance, Ascent Coupling, Cone Expansion, Insight Threshold, Ontological Metric, and Reconstitutive Accumulation) that together constitute the model’s formal core.

Several important open questions remain. First, the precise mathematical topology of Ontological Space O has not been specified; a rigorous metric geometry of this space is required before the Ontological Distance theorems can be given full mathematical treatment. Second, the ascent coupling function at Layer 7 (the mechanism by which meta-cognitive recursion generates the substrate for a potential Layer 8) is currently undefined, and its definition would represent a significant theoretical advance with implications for the theory of recursive self-improvement in both biological evolution and artificial intelligence. Third, the empirical operationalization of the MCF (the measurement of bioelectric and morphogenetic field states in terms compatible with the formal definitions offered here) remains a challenge for experimental biology, though Levin’s program makes significant inroads.

The CMMATC does not resolve these questions; it provides a framework within which they become tractable. The model’s core wager is that the formal unity of cognition across scales is not a metaphor but a mathematical reality; that the membrane operator active in the lipid bilayer of an archaeon four billion years ago and the membrane operator active in the conceptual architecture of a contemporary theorist are instances of the same formal structure, separated by four billion years of AAS expansion and reconstitutive medium accumulation, but formally continuous. If this wager is correct, then the theory of life and the theory of mind are not separate disciplines awaiting integration but a single formal science awaiting recognition.

Glossary of Formal Terms

Term	Symbol / Notation	Definition	Part Introduced
Abstraction Ascent Stack	AAS = L_7 ∘ … ∘ L_0	Compositional tower of eight cognitive operator layers spanning teleodynamic substrate to meta-cognitive recursion	Part IV
Ascent Coupling	C_k	Function specifying how AAS layer k’s output becomes layer k+1’s substrate	Part IV
Cognitive Depth	μ(IC(A))	Measure of the intelligence cone’s volume in spatio-temporal space; a continuous measure of cognitive capability	Part V
Cognitive Membrane	M: I × E × T → S	Scale-invariant boundary operator partitioning state space and producing a selective filter function	Part I
Differential Shadow	(informal)	The lower-dimensional projection of generative activity accumulated in the medium; the causally active trace of past agency	Part II
Genomic Temporal Stack	GTS = ⟨L_1,…,L_n⟩	Ordered tuple of genomic strata, each defined by characteristic timescale, update function, and read operator	Part III
Inside State Space	I	The set of states on the interior side of the cognitive membrane; constituted by the membrane’s exclusion structure	Part I
Insight Event	I*	Membrane-crossing event in cognitive developmental space; occurs when cumulative sub-threshold reorganizations exceed threshold θ	Part VI
Intelligence Cone	IC(A)	Set of (x,t) pairs over which agent A can influence or model states in service of a goal; the spatio-temporal extent of goal-directedness	Part V
Membrane Threshold	θ	Minimum cumulative cognitive reorganization required to establish a new membrane boundary condition (Insight Threshold theorem)	Part VI
Medium Invariant	I_t	The compressed, causally active structure produced by the Reconstitutive Medium Operator; the medium’s metabolized output	Part II
Morphogenetic Cognitive Field	MCF(t) = ∬ m(x,y,t) dA	Spatial integral of membrane-mediated signals across developing tissue; the distributed cognitive field of morphogenesis	Part VIII
Ontological Distance	D_o(A,B)	Infimum path length in ontological space O between intelligence cones IC(A) and IC(B)	Part VII
Ontological Space	O	Metric space of all possible cognitive architectures, parameterized by membrane operators, media, and ascent couplings	Part VII
Reconstitutive Medium	M with μ: G × T → I	A substrate equipped with a metabolic function that compresses generative output into causally active invariants	Part II
Reconstitutive Medium Operator	R(G, M, t) = I_t	The compression mapping from generative output G through medium metabolism M to invariant I_t at time t	Part II
Selective Filter Function	S (as s_t: E → I)	The time-indexed filter produced by the membrane operator; specifies which environmental signals are admitted and in what transformed form	Part I
Teleodynamics	(after Deacon)	The level of dynamics at which morphodynamic processes mutually constrain each other to produce end-directed behavior; grounded in absential causation	Part IX
Temporal Horizon	sup{t: (x,t) ∈ IC(A)}	The maximum temporal extent of agent A’s intelligence cone; how far into the future A can maintain and pursue goals	Part V
Temporal Stratification	(informal)	The simultaneous integration of information from multiple temporal scales within a single cognitive act; formal basis of GTS	Part III

References

1. Bateson, G. (1972). Steps to an Ecology of Mind: Collected Essays in Anthropology, Psychiatry, Evolution, and Epistemology. Chandler Publishing.
2. Clark, A., & Chalmers, D. J. (1998). The extended mind. Analysis, 58(1), 7–19.
3. Deacon, T. W. (2011). Incomplete Nature: How Mind Emerged from Matter. W. W. Norton & Company.
4. Deacon, T. W. (1997). The Symbolic Species: The Co-Evolution of Language and the Brain. W. W. Norton & Company.
5. Friston, K. (2010). The free-energy principle: A unified brain theory? Nature Reviews Neuroscience, 11(2), 127–138.
6. Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Houghton Mifflin.
7. Goodwin, B. (1994). How the Leopard Changed Its Spots: The Evolution of Complexity. Weidenfeld & Nicolson.
8. Holland, J. H. (1995). Hidden Order: How Adaptation Builds Complexity. Addison-Wesley.
9. Jablonka, E., & Lamb, M. J. (2005). Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral, and Symbolic Variation in the History of Life. MIT Press.
10. Kauffman, S. A. (1993). The Origins of Order: Self-Organization and Selection in Evolution. Oxford University Press.
11. Kuhn, T. S. (1962). The Structure of Scientific Revolutions. University of Chicago Press.
12. Levin, M. (2021). Bioelectric signaling: Reprogrammable circuits underlying embryogenesis, regeneration, and cancer. Cell, 184(8), 1971–1989.
13. Levin, M. (2019). The computational boundary of a ‘Self’: Developmental bioelectricity drives multicellularity and scale-free cognition. Frontiers in Psychology, 10, 2688.
14. Levin, M., & Dennett, D. C. (2020). Cognition all the way down. Aeon Magazine.
15. Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and Cognition: The Realization of the Living. D. Reidel Publishing.
16. Maturana, H. R., & Varela, F. J. (1987). The Tree of Knowledge: The Biological Roots of Human Understanding. New Science Library.
17. Nurse, P. (2008). Life, logic and information. Nature, 454, 424–426.
18. Prigogine, I., & Stengers, I. (1984). Order Out of Chaos: Man’s New Dialogue with Nature. Bantam Books.
19. Rao, R. P., & Ballard, D. H. (1999). Predictive coding in the visual cortex: A functional interpretation of some extra-classical receptive-field effects. Nature Neuroscience, 2(1), 79–87.
20. Schrödinger, E. (1944). What Is Life? The Physical Aspect of the Living Cell. Cambridge University Press.
21. Shannon, C. E., & Weaver, W. (1949). The Mathematical Theory of Communication. University of Illinois Press.
22. Simon, H. A. (1962). The architecture of complexity. Proceedings of the American Philosophical Society, 106(6), 467–482.
23. Tononi, G., Boly, M., Massimini, M., & Koch, C. (2016). Integrated information theory: From consciousness to its physical substrate. Nature Reviews Neuroscience, 17(7), 450–461.
24. Varela, F. J., Thompson, E., & Rosch, E. (1991). The Embodied Mind: Cognitive Science and Human Experience. MIT Press.
25. von Foerster, H. (2003). Understanding Understanding: Essays on Cybernetics and Cognition. Springer.
26. Waddington, C. H. (1957). The Strategy of the Genes: A Discussion of Some Aspects of Theoretical Biology. George Allen & Unwin.
27. West-Eberhard, M. J. (2003). Developmental Plasticity and Evolution. Oxford University Press.
28. Wiener, N. (1948). Cybernetics: Or Control and Communication in the Animal and the Machine. MIT Press.
29. Wilson, E. O. (1971). The Insect Societies. Harvard University Press. [on stigmergy and social insect media]
30. Wolfram, S. (2002). A New Kind of Science. Wolfram Media.

© 2026 Daryl. Rosendale, NY. All rights reserved.
 Prepared: October 2, 2026 • Framework: Cognitive Membrane Model of Adaptive Temporal Continuity (CMMATC) • Version 1.0