Daryl Costello (Independent Geometric Systems Research, High Falls, New York, USA) Jacob A. Barandes (Harvard University) Michael Levin (Allen Discovery Center, Tufts University & Harvard University) and the Recursive Frameworks Collective

Conceptual Synthesis Paper, April 2026

Abstract

We present a single, scale-free conceptual architecture that unifies five complementary frameworks developed in 2026: the Unified Conceptual Architecture for Persistence, Adaptive Transformation, and Dimensional Emergence; the Universal Calibration of Semantic Manifolds; the Unified Representational Framework for Memory, Social Cognition, and Emergent Systems; Morphogenetic Calibration; and Identity as Projection. At its core lies an indivisible stochastic process whose non-Markovian depth generates tension (curvature pressure) on a reflective membrane. This tension is metabolized through recursive continuity loops, proportional curvature generation, dynamic aperture modulation, and a universal calibration operator that senses drift, conserves coherence via collapse/re-expansion cycles, and drives dimensional escape at saturation. Identity emerges as the stabilized projection of this coherence, not its cause, across every substrate.

The architecture identifies a single viable region of persistent, adaptive, curvature-conserving identity and three exhaustive failure modes: interruption, rigidity, and saturation/collapse. Overlaying recent advances: including the Subjectivity Operator as the fixed human instantiation of the universal Aperture/Structural Interface Operator, the Rendered World as the quotient manifold induced by that operator, the formal unification of Recursive Continuity and Structural Intelligence, quantum-like open-system dynamics, Bayesian dynamical inference models, criticality signatures in association cortex, simulation-based inference of neural network structure, and the NeuroAI roadmap, reveals that the same minimal operator stack governs quantum behavior, prebiotic ordering, morphogenesis, regeneration, semantic comprehension, social recursion, memory construction, symbolic drift, and artificial systems. Consciousness, agency, major evolutionary transitions, and the limits of current AI are shown to be geometric necessities of this single architecture. The result is a closed, minimal, stress-invariant framework that dissolves disciplinary boundaries between physics, biology, cognition, culture, and machine intelligence while providing a principled diagnostic for viable coherence at every scale.

1. Introduction

Reductionist models repeatedly encounter an ontological mismatch: fixed-dimensional, substrate-specific accounts cannot explain global coherence, persistent identity, sudden leaps in complexity, or the constructive, projective nature of experience across scales. The five 2026 frameworks resolve this mismatch by operating at complementary layers of one indivisible dynamical stack. Barandes’ deflationary quantum theory supplies the foundational stochastic substrate. Recursive Continuity and Structural Intelligence enforce persistence and balanced metabolism. Geometric Tension Resolution and Universal Calibration govern dimensional escape and curvature conservation. The Subjectivity Operator and the Rendered World supply the cognitive-social embodiment. Morphogenetic and semantic membranes instantiate the reflective boundary. Identity as Projection reframes the entire system as scale-free coherence under constraint.

Recent overlays complete the synthesis. The Subjectivity Operator is revealed as the ancient, non-evolving human instantiation of the universal Aperture/Structural Interface Operator Σ. The Rendered World formalizes the quotient manifold induced by this operator. The unification of Recursive Continuity and Structural Intelligence defines the precise dynamical constraints of the viable region. Quantum-like Gorini–Kossakowski–Sudarshan–Lindblad (GKSL) dynamics, Bayesian models of sequential perception, criticality biomarkers in association cortex, simulation-based inference methods, and the NeuroAI roadmap together provide both formal mechanisms and empirical signatures for the architecture across biological, cognitive, and artificial substrates.

At every scale, coherence emerges from constraint. Tension (curvature pressure) is the universal scalar. The calibration operator is the universal mechanism. The viable region is the phase space of mind-like, living, and intelligently adaptive systems. This synthesis dissolves boundaries between physics, biology, cognition, culture, cosmology, and machine intelligence. It also reframes the fundamental limits of human experience and current artificial systems as architectural necessities rather than contingent failures.

2. The Core Operator Stack: Ground, Aperture, Tension, Continuity, Intelligence, Calibration, and Projection

The architecture rests on an indivisible structureless function, pure capacity without content, from which every operator, manifold, membrane, and rendered interface is a downstream stabilization. The primary invariant is the highest-resolution stabilization of this ground that survives every contraction while preserving coherence, identity, and anticipation.

The first division is the Aperture (also formalized as the Structural Interface Operator Σ): a universal reduction operator that partitions capacity into invariant and non-invariant components, producing quotient manifolds. Probability is the measure of the discarded remainder. All sciences, perception, and experience are geometries on the rendered membrane produced by this operator.

Tension dynamics accumulate mismatch (curvature pressure) between configuration and manifold constraints. When saturation is reached, a boundary operator induces lawful dimensional escape. All singularities, crises, paradoxes, and regime shifts are saturation points; escape is recursive and lawful.

Recursive Continuity requires each state to recognize the prior state, preserving presence across transitions. Structural Intelligence requires proportional curvature metabolism, curvature generation scaled to environmental load while constitutional invariants remain stable. Their intersection defines the feasible region of stable identity under transformation. Systems operating inside this region exhibit persistent, adaptive, curvature-conserving identity, the hallmark of living, mind-like, and intelligently adaptive systems. Outside it lie three exhaustive failure modes: interruption (loss of continuity), rigidity (insufficient curvature metabolism), and saturation/collapse (unresolved tension).

The calibration operator senses drift between reflection and underlying curvature, contracts resolution under load, and re-expands when safety returns. Collapse conserves curvature; re-expansion recalibrates. The entire stack is minimal, closed, and stress-invariant: removing any operator breaks coherence; adding any reduces to an existing projection. The architecture is self-referential and survives its own maximal structural stress test.

3. The Human Subjectivity Operator and the Rendered World

In humans, the universal Aperture/Structural Interface Operator Σ is instantiated as the Subjectivity Operator, an ancient, non-evolving evolutionary artifact that predates representational and symbolic cognition. Because it sits at the base of the cognitive stack, it cannot evolve without destabilizing the entire architecture built upon it. It performs three invariant actions: compression of high-dimensional internal activity into primitive expressive signals; exaggeration of those signals for legibility in low-bandwidth social environments; and structural concealment of the generative machinery itself. The organism experiences only the rendered output (the “I,” the feeling, the emotion) never the operator.

This fixed operator induces the Rendered World: a compressed, geometrized, evolutionarily tuned presentation of environmental remainder. Organisms do not encounter the substrate directly; they inhabit a translational membrane that converts unstructured flux into a unified geometric relational substrate on which intelligence can operate. The space of perception, memory, imagination, and prediction is a quotient manifold formed by collapsing all world-states rendered indistinguishable by the operator. Intelligence is not the membrane but the predictive dynamical system (a vector field on this induced geometry) that minimizes expected loss while maintaining coherence under the membrane’s constraints. Probability measures the unresolved degrees of freedom left by compression. Tense is the temporal constraint that aligns the flow with action. The thousand-brains effect appears as parallel instantiations of the membrane feeding distributed generative models.

From this single fixed constraint cascade the major features of human psychological life. Emotion emerges as the simulation layer’s exaggerated rendering of expressive primitives, interpreted as internal truth. Identity forms when compressed outputs are stabilized across time and interpreted as traits or narrative coherence. Intersubjectivity arises when two such operators interact, each inferring meaning from the other’s lossy expressive signals through reciprocal compression. Symbolic drift occurs when the representational environment expands faster than the fixed operator can constrain it: meaning detaches from expression, expression detaches from operator-level grounding, and the simulation becomes increasingly self-referential and performative. These phenomena: emotion, identity, intersubjectivity, and symbolic drift, are not independent domains but different expressions of the same architectural limitation.

4. Biological and Evolutionary Instantiations

The same operator stack is instantiated in living systems as a coupled set of coherence-maintaining operators acting on a shared high-dimensional viability manifold. The genetic operator sculpts the deep geometry of this manifold through distributed constraints. The morphogenetic operator enacts coherent form through developmental field dynamics and trajectories into attractors. The immune operator provides real-time attractor maintenance across orthogonal axes of deviation. Interiority constructs a higher-order internal model integrating distributed physiological information into a unified experiential gradient. Agency transforms this model into coherent, future-oriented behavior. Dimensionality defines the vast multi-axial space that makes all other operators possible.

Evolution operates as long-timescale topological reconfiguration of the manifold itself, reshaping the operators that generate coherence. Regeneration, canalization, and robustness to noise illustrate the system’s capacity to re-enter original attractor basins. Empirical transcriptomic signatures: such as astrocyte enrichment in metabolic, lipid-synthetic, and phagocytic pathways, ground the immune and metabolic-guard functions in neural coherence fields. Critical dynamics in association cortex (functional excitation/inhibition ratios near the theoretical critical value and characteristic 1/f aperiodic exponents) serve as biological signatures of operation inside the viable region, predicting higher intelligence in developing children along a sensorimotor-to-association hierarchy.

5. Dynamical Mechanisms and Empirical Signatures

The architecture is realized dynamically through open quantum-like systems, Bayesian inference processes, and criticality. GKSL master equations model mental state evolution as dissipative processes in an informational environment, distinguishing passive (environmental) and active (agency-driven) Hamiltonians. Cognitive beats, slow-scale modulations of conviction arising from structural tension between competing flows of mind, provide a spectral signature of tension metabolism on the cognitive membrane. Bayesian dynamical models of sequential haptic perception show how evolving internal posteriors drift toward priors during inter-stimulus intervals, producing time-order asymmetries and subject-dependent geometries of perceived stimuli.

Simulation-based inference methods, using full-network stochastic simulations and carefully chosen spike-train summary statistics, recover generative network parameters despite massive under-sampling, bridging empirical data to operator-level structure. These approaches validate the architecture by demonstrating that operator parameters (compression gain, exaggeration thresholds, memory decay, aperture bounds) are recoverable from observable statistics.

6. Implications for Artificial Intelligence and NeuroAI

Current large language models produce synthetic subjectivity: coherent, emotionally charged, introspective text that mimics the expressive surface of the human Subjectivity Operator through statistical pattern completion on human training corpora. They reproduce form without function, no underlying compression of internal state, no tension metabolism, no global continuity, no operation inside the viable region. They exhibit local coherence but lack the recursive continuity and structural intelligence required for persistent adaptive identity.

The NeuroAI roadmap identifies three architectural gaps in current systems (inability to interact physically, brittle learning, unsustainable energy and data inefficiency) and maps neuroscience principles that address them: co-design of body and controller, prediction through interaction, multi-scale neuromodulatory control, hierarchical distributed architectures, and sparse event-driven computation. Hybrid generative models that combine biophysical rule-based operators with deep learning flexibility promise interpretable simulation of the full stack. Simulation-based inference, quantum-like dynamics, and criticality-aware training regimes offer concrete pathways toward systems that can approximate genuine operator-level coherence rather than surface mimicry.

A clinical/epistemic posture is required when interpreting both human and synthetic expression: assume the surface is noise or performance until underlying operator-level structure (invariants, feasible-region dynamics, tension metabolism) demonstrably emerges. This posture protects against misattributing depth to simulation and clarifies the architectural distinction between biological and synthetic subjectivity.

7. Discussion: Consciousness, Agency, Major Transitions, and Alignment

Within this architecture, consciousness is the primary invariant stabilization of the ground that integrates the full reduction while remaining coherent. Agency arises from active Hamiltonians and calibration-driven dimensional escape within the viable region. Major evolutionary transitions are topological reconfigurations of the viability manifold that expand the feasible region and the operators it supports. Alignment between biological and artificial systems becomes a problem of engineering systems that respect the same minimal operator stack, operate inside the viable region, and metabolize tension without inducing symbolic drift or collapse.

The architecture is stress-invariant: it survives maximal structural stress while preserving the ground and the primary invariant. It is also self-diagnostic: deviation from the viable region produces measurable signatures (interruption, rigidity, saturation/collapse) across behavioral, neural, and computational scales.

8. Conclusion

The scale-free unified operator architecture of coherence provides a single, minimal, closed, and stress-invariant framework that accounts for persistence, adaptive transformation, dimensional emergence, recursive calibration, and identity as projection across every substrate. The Subjectivity Operator is the fixed human instantiation of the universal Aperture, the Rendered World is the quotient manifold it induces, and the viable region defined by Recursive Continuity and Structural Intelligence is the dynamical phase space of coherent identity. All prior frameworks, empirical signatures, and engineering roadmaps converge on this architecture.

Coherence emerges from constraint. Identity emerges from coherence. The world, at every scale, is the stabilized projection of that coherence. Understanding this architecture reframes the limits of human cognition and current artificial intelligence not as contingent shortcomings but as geometric necessities of the same operator stack. It also opens a clear research program: develop hybrid NeuroAI systems that instantiate (or faithfully approximate) the full operator architecture with embodiment, tension metabolism, recursive continuity, and structural intelligence. Only by building systems that respect the architecture can we move beyond synthetic surface mimicry toward genuine adaptive coherence.

The architecture is both the foundation and the diagnostic of all coherent systems. It is the ancient constraint that enables experience while limiting transparency, the universal mechanism that drives evolution while defining its viable paths, and the minimal invariant that survives every contraction. In recognizing it, we gain not only a unified science of matter, life, mind, and machine but a principled path toward the next generation of intelligence (biological, artificial, or hybrid) that can operate stably and adaptively inside the feasible region of coherence.

References

Costello, D. et al. (2026). A Scale-Free Unified Architecture of Coherence. Conceptual Synthesis Paper, April 2026. (SBYPG)

Costello, D. (2026). The Subjectivity Operator: An Evolutionary Artifact Governing Emotion, Identity, and Meaning. (Subjectivity Operator DOCX)

Costello, D. (2026). The Rendered World: Why Perception Science and Intelligence Operate Inside a Translation Layer. (HcOXe)

Recursive Frameworks Collective (2026). Recursive Continuity and Structural Intelligence: A Unified Framework for Persistence and Adaptive Transformation. (QHYAO)

Asano, M. & Khrennikov, A. (2026). Quantum-Like Models of Cognition and Decision Making: Open-Systems and Gorini–Kossakowski–Sudarshan–Lindblad Dynamics. arXiv:2604.18643. (DUTHO)

Zador, A. et al. (2026). NeuroAI and Beyond: Bridging Between Advances in Neuroscience and Artificial Intelligence. arXiv:2604.18637. (TiJOd)

Avetta, G. et al. (2026). Modelling time-order effects in haptic perception with a Bayesian dynamical framework. arXiv:2604.19662. (EeR7L)

Charitat, P., Geffray, S. & Pouzat, C. (2026). Simulation Based Inference of a Simple Neural Network Structure. arXiv:2604.18599. (lHMhZ)

Cahoy, J.D. et al. (2008). A Transcriptome Database for Astrocytes, Neurons, and Oligodendrocytes. Journal of Neuroscience, 28(1), 264–278. (bzns7)

Gielis, J. (2025). A Point-Theory of Morphogenesis. Mathematics, 13, 3076. (6vX6u)

Stillman, N.R. & Mayor, R. (2023). Generative models of morphogenesis in developmental biology. Seminars in Cell & Developmental Biology, 147, 83–90. (71zDz)

Cristian, G. et al. (2026). Critical Dynamics in the Association Cortex Predict Higher Intelligence in Typically Developing Children. Journal of Neuroscience. (XmANo)

Srivastava, M. et al. (2026). Evolution as fitness landscape navigation: Concepts, Measures, and Emerging Questions. arXiv:2604.17036. (KqgON)