Intellectual Lineage

The named thinkers, texts, and research programs that provide the scientific and philosophical scaffolding for UNNATURAL’s three acts — drawn from the author’s own working source graph.

Overview

UNNATURAL is explicitly positioned between science and philosophy: philosophical because it asks what kinds of things there are and how we know them; scientific because it stays answerable to public evidence, bounded cognition, and explicit limits on certainty. The overall.md working source identifies the intellectual clusters that scaffold each act. This page documents those clusters and their relationships to the book’s argument — not as a bibliography but as a map of where the book’s claims are grounded.

The organising principle: each act has a foundation cluster, and the clusters are connected by the book’s overall argument. Act 1’s cluster (wanting, embodiment, homeostasis) is the condition for Act 2’s cluster (distinction, representation, ontology), which is the condition for Act 3’s cluster (measurement, regulation, recursion, economics).

Act 1 foundations: Wanting, embodiment, and regulation

Antonio Damasio. The primary neuroscientific anchor for Act 1. Four books form a coherent sequence: Descartes’ Error (1994) — emotion is structurally necessary for decision-making, not a distraction from it; Self Comes to Mind (2010) — consciousness is grounded in the body’s ongoing self-modelling; Feeling and Knowing (2021) — feeling is the substrate of mind, not a late-arriving cognitive report; The Strange Order of Things (2018) — culture is an outgrowth of life-regulation, a continuation of homeostasis by other means. Together they provide the empirical case that wanting is not optional overlay on a neutral reasoning machine but the regulatory core of any functioning mind. Key search terms: interoception, homeostatic feelings, somatic markers, allostasis.

Embodied and enactivist cognition. Francisco Varela, Evan Thompson, and Eleanor Rosch (The Embodied Mind) establish that cognition is enacted through organism-environment coupling, not computed on detached representations. Action is constitutive of perception; perception is always already motivated and perspectival. This is the philosophical grounding for the claim that the wanting creature is “already a perspectival cut in reality.” Chemero’s Radical Embodied Cognitive Science extends this into contemporary cognitive science, opposing representation-heavy accounts.

Self-determination theory (SDT). Ryan and Deci’s empirical research program identifying autonomy, competence, and relatedness as the central conditions for healthy, self-endorsed action. This provides Act 1’s practical specification: it is not enough that a creature wants; the wanting must be owned (autonomous), effective (competent), and connected (relational) for it to constitute the kind of motivated life the book is describing. SDT distinguishes controlled regulation (acting for external reward or fear) from integrated autonomous regulation, which is the motivational mode the book is pointing toward.

Social baseline theory. James Coan’s research program demonstrating that close others reduce energetic and neural costs through load sharing and risk distribution. Proximity to trusted others is the nervous system’s expected baseline, not a luxury added to individual functioning. This provides the biological grounding for the book’s treatment of love as shared homeostasis rather than sentimental decoration.

Meaning research. A cluster of researchers (Baumeister, Steger, Martela, King, and others) who have operationalised meaning-in-life as three dimensions: coherence (the world makes sense), purpose (there is something to aim for), and significance (one’s existence matters). The book’s Act 1 formulation of meaning maps precisely onto this: wanting with clarity = purpose; understanding constraints and affordances = coherence; felt impact on reality = significance.

Act 2 foundations: Distinction, representation, and ontology

Douglas Hofstadter. The primary guide to recursion, self-reference, and analogy. Gödel, Escher, Bach (1979) establishes that self-reference is the engine of consciousness and creativity, not a paradox to be resolved; I Am a Strange Loop (2007) extends this into a theory of the self as recursive pattern; Surfaces and Essences (2013, with Sander) treats analogy as the core of cognition — every thought is an analogical extension of a prior pattern. Together they provide Act 2’s account of how the Second Blade (world/model) becomes the recursive machinery of mind. Key search terms: object files, core knowledge, psychological essentialism, semantic information, symbol grounding.

Developmental and cognitive research on object representation. Core knowledge research (Spelke and others) establishes that infants track persisting objects, agents, places, and numerosities before language. Object-file research (Kahneman, Treisman, Pylyshyn) treats perception as maintaining indexed representations of persisting objects and updating their features over time. These provide the empirical ground for Act 2’s claim that minds actively carve continuities into followable units — thingification is not a cultural accident but a deep feature of perception.

Psychological essentialism. Medin and Gelman’s research on how people represent categories as having hidden essences — not mere feature lists but causal inner natures. This explains the cognitive stickiness of thingification: once a category has been essentialised, it feels discovered rather than drawn, and the original cut becomes invisible.

Social ontology. John Searle’s analysis of institutional facts (“X counts as Y in context C”) and the distinction between observer-relative and observer-independent features of reality grounds the book’s treatment of institutions as a specific kind of real. Margaret Gilbert’s work on collective intentionality and shared agency extends this into the social layer.

Act 3 foundations: Measurement, regulation, and making

David Marr. Marr’s three levels of analysis — computational, representational/algorithmic, and implementational — provide the theoretical architecture for the Model chapter. A belief is not just matter and not just information; it is a physical state doing semantic work for an organism. Confusing levels produces the same errors as confusing ontological layers.

Active inference (Karl Friston). The Active Inference framework treats perception, prediction, planning, and action as a single process of minimising prediction error (or more precisely, variational free energy). This contemporary bridge between neuroscience and philosophy provides Act 3 with a rigorous account of what it means to model and to act from a model. The organism is not a passive observer that acts occasionally; it is constantly predicting, constantly updating, and constantly acting to confirm or revise its predictions.

W. Ross Ashby and cybernetics. Ashby’s Introduction to Cybernetics (1956) establishes the law of requisite variety: a controller must have at least as much internal variety as the system it controls, or it will be overwhelmed. This is the formal statement of why the MMM loop must keep running and why a fixed model facing a complex environment will fail. Cybernetics more broadly (Wiener, Ashby) treats living systems as feedback regulators — a framing the book uses throughout.

Chaos theory and Lorenz. Lorenz’s work on weather prediction demonstrated that deterministic systems with sensitive dependence on initial conditions hit intrinsic limits on long-range prediction. The implication for the book: life cannot be managed by one clean extrapolation from a single snapshot. Not because the world is random but because small differences in initial conditions produce large divergences over time. This is why the loop must iterate rather than solve once. Key search terms: chaos predictability limit, Lorenz attractor.

Gary Becker. Becker’s work on the allocation of time and human capital formalises what the MMM loop looks like in economic life. Time is the ultimate constrained resource; skill is embodied capital that future loops can draw on; human capital investment is self-directed MMM across a lifetime. The household production model treats people not as passive consumers but as producers of valued outputs using time, goods, and skills.

Oliver Williamson and transaction cost economics. Williamson’s work explains why some manifestations of the MMM loop produce institutions rather than open-market exchanges. When performance is hard to measure, when commitments require asset-specific investment, and when opportunism is possible, governance structures emerge to reduce the loop’s friction. This provides Act 3’s economic layer for the chapter on making and the civilisational scale of MMM.

Dupoux, LeCun, and Malik (2026). “Why AI systems don’t learn and what to do about it: Lessons on autonomous learning from cognitive science” (arXiv:2603.15381) proposes a three-system architecture for autonomous intelligence: System A (observation learning — learning by watching), System B (active-behavior learning — learning by doing), and System M (meta-control — choosing between them). This is a structural echo of Measure / Model / Manifest arriving from contemporary AI research. The book treats it as independent corroboration, not influence: when the three-layer shape recurs across independent attempts to describe agency, it is probably tracking something real. Chapter 16 names this the evidence that MMM is the most meta process we can observe.

The book’s position in its intellectual neighbourhood

The overall.md source identifies the book’s closest academic neighbours: embodied cognition, philosophy of action, social ontology, meaning-in-life research, cybernetics, and behavioral economics. It explicitly positions the project in a productive middle ground: philosophical (asks what kinds of things there are and how we know them) but scientific (stays answerable to public evidence and bounded cognition).

The deepest intellectual move is to let these otherwise separate traditions hang together as one argument. Damasio explains why wanting is biological. SDT specifies what healthy wanting looks like. Hofstadter explains how self-referential models of wanting produce selves. Ashby explains why those models must keep updating. Lorenz explains why they can never solve once. Becker and Williamson explain how the resulting loops produce economic life. That cascade is the book.

Key claims

The three acts have distinct intellectual foundations that connect into one argument: wanting → distinction → recursive regulation. — Source: /dump/overall.md
Damasio’s four books provide a coherent line from emotion to consciousness to culture as homeostatic continuation. — Source: /dump/overall.md
Active inference unifies perception, prediction, planning, and action as a single loop — the rigorous version of MMM at the neural level. — Source: /dump/overall.md
Ashby’s law of requisite variety is the formal statement of why the MMM loop must keep running in a complex environment. — Source: /dump/overall.md
Lorenz’s chaos work explains why re-measurement is structurally necessary, not an admission of failure. — Source: /dump/overall.md
Dupoux, LeCun, and Malik’s 2026 A/B/M architecture is structural convergence with MMM from AI research — independent corroboration that the three-layer shape tracks something real about agency. — Source: /src/content/chapters/16-mmm-as-lived-experience.md

Connections

[[Volo Ergo Sum]] — Damasio, enactivism, and SDT are its primary scientific scaffolds
[[The MMM Loop]] — Marr, Ashby, active inference, Lorenz, Becker, and Williamson are Act 3’s intellectual backbone
[[Want and Will]] — social baseline theory and SDT ground the claims about love and aligned wanting
[[Thingification]] — Hofstadter, psychological essentialism, and social ontology ground Act 2
[[Layered Realism]] — Searle, Marr, and philosophy of artifact ground the ontological taxonomy
[[Wisdom Traditions]] — the Stoic, Buddhist, Gita, and Daoist connections in overall.md are also part of the intellectual lineage

Open questions

The book draws on these traditions selectively; where they conflict with each other (e.g., active inference’s continuous loop vs. MMM’s three-step articulation), the book does not always adjudicate.
Some of the cited traditions (SDT, social baseline theory, meaning research) are active empirical programs whose findings continue to evolve; the book’s claims should be treated as indexed to findings as of the writing period, not as permanently established.

Last updated: 2026-04-16 | Sources: 2 (/dump/overall.md; Ch. 16 for Dupoux/LeCun/Malik 2026) — updated 2026-04-16