Extended Mind vs. Enactive Cognition

For more than half a century, the cognitive sciences were dominated by a powerful, seductive orthodoxy: the computational-representational theory of mind. Under this Cartesian-inspired paradigm, thinking was conceived as an insular, intracranial operation in which the brain acts as an isolated biological computer, processing symbolic inputs transduced from sensory organs and generating motor outputs. Cognition was identified with algorithmic transformations over mental representations tucked securely within the biological boundaries of the skull. The world was merely an external source of data and a recipient of mechanical outputs, metaphysically severed from the cognitive machinery itself.

In recent decades, this foundational orthodoxy has faced systemic challenge through the emergence of post-Cartesian cognitive science, broadly coalescing under the moniker of 4E Cognition: embodied, embedded, enactive, and extended. Rather than treating mental phenomena as neuro-computational events sealed beneath the cranium, these emerging frameworks treat cognition as an active, environmentally situated, and physically distributed enterprise. However, within this shared repudiation of classic internalism, two theoretical juggernauts have developed along profoundly divergent trajectories: the Extended Mind Thesis (often referred to as the Hypothesis of Extended Cognition, or HEC) and Enactive Cognition (or Enactivism). While casual observers frequently conflate the two because both treat the skin-and-skull boundary as arbitrary, they arise from fundamentally opposed metaphysical foundations, utilize contradictory models of information and representation, and maintain radically distinct definitions of what constitutes a mind.

This treatise provides a systematic, exhaustive investigation into the conceptual architectures, historical geneses, metaphysical mechanics, points of divergence, and domains of convergence defining the Extended Mind Thesis and Enactive Cognition. By examining their handling of representation, functional parity, biological autonomy, agency, and the infamous coupling-constitution debate, this work elucidates the core philosophical questions of post-Cartesian thought: Is the mind an extended functional computer, or is it an autonomous, living process of sense-making?

1. Introduction to Extended Mind and Enactive Cognition

1.1 The Rise of Post-Cartesian Cognitive Science

The classical paradigm in cognitive science, championed by pioneers such as Jerry Fodor, Zenon Pylyshyn, and Allen Newell, was built upon the computational-representational model of mind. Within this framework, mental processes are realized through formal operations over symbolic structures possessing semantic and syntactic properties. The brain operates as a central processing unit, executing algorithmic routines over internal models of the external world. Perception is theorized as an inferential, input-driven reconstruction of environmental stimuli, while action is conceived as the downstream execution of motor commands computed within the central nervous system. This architecture reinforced a deep-seated Cartesian dualism: even where the classical paradigm was resolutely materialist, it retained the Cartesian structure by drawing an ontological boundary between the thinking subject (the intracranial computational system) and the external, mechanical realm of objects, environment, and physical affordances.

Toward the close of the twentieth century, this computational paradigm confronted intractable empirical and theoretical impasses. Chief among these was the infamous “frame problem” in artificial intelligence—the inability of purely formal, symbolic systems to adaptively determine which environmental features are relevant in an open-ended, dynamic world without incurring catastrophic computational overhead. Simultaneously, empirical discoveries in neurobiology, ecological psychology, dynamic systems modeling, and mobile robotics, notably spearheaded by figures like Rodney Brooks, demonstrated that complex adaptive behaviors could be achieved without centralized internal representations or elaborate world-models. Robots could exploit the physics of their own appendages and real-time environmental feedback loops to navigate intricate terrains, bypassing the need for intensive symbolic planning.

These developments catalyzed the emergence of the 4E cognition movement, which dismantled the insular, intracranial model of the mind by distributing cognitive processes across multiple ecological scales:

• Embodied Cognition: Posits that cognition is not merely executed by the brain, but is fundamentally structured and constrained by the physical morphology, sensorimotor capacities, and visceral physiology of the biological body.
• Embedded Cognition: Emphasizes that cognitive agents exploit environmental structures and heuristics to reduce internal computational loads, using external resources as scaffolding for biological processing.
• Enactive Cognition: Maintains that cognition is not the passive processing of pre-existing external information, but the active “bringing forth” or enacting of a world of significance through precarious, self-organizing sensorimotor interactions.
• Extended Cognition: Asserts that cognitive mechanisms do not simply interact with external artifacts; rather, physical, informational, and social components beyond the biological skin literally constitute subroutines of the cognitive system itself.

This shift from cranial internalism to dynamic, world-involving cognitive architectures required a complete reassessment of cognitive boundaries. The mind was no longer viewed as a static Cartesian inner sanctum, but as an open, relational, dynamic system intertwined with environmental affordances, physiological substrates, and artifactual technologies.

1.2 Historical Genesis of Both Frameworks

Although the Extended Mind Thesis and Enactive Cognition both emerged within the anti-Cartesian revolution of the 1990s, their historical and philosophical lineages are remarkably distinct. The Extended Mind Thesis was formally inaugurated in 1998 with the publication of the seminal paper “The Extended Mind” by analytic philosophers Andy Clark and David J. Chalmers in the journal Analysis. Clark and Chalmers advanced a radical form of “active externalism,” arguing that cognitive processes are not bounded by biological skin and skull, but can incorporate external components such as notebooks, calculators, digital devices, and linguistic symbols, provided these external elements play an appropriate functional role within the cognitive economy. The philosophical heritage of the Extended Mind Thesis lies firmly within the tradition of Anglo-American analytic philosophy of mind, specifically tracing its lineage through machine-state functionalism (Hilary Putnam), teleofunctionalism (Ruth Millikan), and naturalized epistemology.

In contrast, Enactive Cognition was established nearly a decade earlier with the publication of the 1991 foundational text The Embodied Mind: Cognitive Science and Human Experience by cognitive neuroscientist Francisco Varela, philosopher Evan Thompson, and psychologist Eleanor Rosch. The enactive approach emerged from a cross-pollination between theoretical biology, systems theory, cybernetics, and Continental phenomenology, specifically drawing upon the philosophical works of Maurice Merleau-Ponty, Edmund Husserl, and Martin Heidegger, alongside the biological theory of autopoiesis developed by Humberto Maturana and Francisco Varela in the 1970s.

The divergent genealogies of these frameworks dictate their theoretical preoccupations:

• Extended Mind (HEC): Inherits the analytic tradition’s preoccupation with computational functionalism, multiple realizability, mental state attribution, informational processing, and active causal externalism.
• Enactivism: Rooted in autopoietic biology, non-linear thermodynamics, phenomenological accounts of lived bodily experience, and the generative dynamics of self-producing physical unities.

Thus, while Clark and Chalmers approached the boundary question from an analytic functionalist perspective, asking whether external components satisfy computational definitions of belief and memory, Varela, Thompson, and Rosch approached it from an existential and biological perspective, investigating how living systems construct meaning and experience in an indifferent physical universe.

1.3 Conceptual Conflation and the Need for Demarcation

In contemporary philosophy of cognitive science, the Extended Mind Thesis and Enactive Cognition are often carelessly lumped together under the generic umbrella of “extended” or “externalist” cognitive frameworks. This conflation is superficially understandable: both frameworks reject intracranialism; both assert that cognitive science must look beyond the skull; both critique the traditional input-computation-output pipeline of classic cognitivism; and both emphasize agent-environment coupling. To the non-specialist, arguing that a notebook forms part of an agent’s memory (HEC) appears functionally aligned with arguing that a creature enacts its perceptual world through worldly engagement (Enactivism).

However, this apparent alignment masks deep, irreconcilable metaphysical tensions. The Extended Mind Thesis is fundamentally an extension of liberalized functionalism. It retains the core concepts of information processing, storage, retrieval, and symbolic manipulation, merely expanding the spatial boundary of where these functional processes can be physically realized. Under HEC, mental states can reside in silicon chips, ink on paper, or cloud servers, provided they fulfill the requisite functional role in driving behavior. The mind is extended because the computational circuit spreads into the environment.

Enactivism, conversely, rejects the foundational ontology of functionalism, information processing, and representationalism entirely. For the enactivist, cognition is not the processing of abstract information, nor is it realized by arbitrary functional components linked in a causal chain. Cognition is sense-making: the precarious, self-maintaining activity of a living, autonomous biological organization. An external artifact cannot simply become part of a cognitive system merely by acting as a passive functional substitute for a neural subroutine. For enactivism, the mind is not an extended computational device; it is a relational dynamic of living agency.

Demarcating these two frameworks is not a pedantic exercise in philosophical taxonomy; it has direct consequences for the empirical sciences of mind. Determining the boundary conditions of cognition directly dictates how cognitive scientists model dynamic systems, how roboticists design artificial agents, how neuroscientists interpret brain-body-environment coupling, and how philosophers evaluate the nature of consciousness, intentionality, and agency.

2. Foundational Overview of the Extended Mind Thesis (HEC)

2.1 Core Tenets and the Parity Principle

The core proposition of the Extended Mind Thesis is that the physical substrates of cognitive processes can, and regularly do, extend into an agent’s external environment. Clark and Chalmers do not merely argue that external resources aid, assist, or scaffold human thought—a trivial claim that virtually all classical cognitivists would accept—but rather that external components can constitute genuine parts of the cognitive process itself. This form of active externalism asserts that external objects and systems play an active, in-the-moment causal role in driving cognitive behavior, forming an integrated cognitive architecture whose overall operational integrity relies upon both biological and non-biological elements.

To defend this radical claim against internalist intuition, Clark and Chalmers formulated what has become known as the Parity Principle:

“If, as we push our way into the world, a part of the world functions as a process which, were it done in the head, we would have no hesitation in recognizing as part of the cognitive process, then that part of the world is (so we claim) part of the cognitive process.”

— Andy Clark and David J. Chalmers (1998)

The Parity Principle is an epistemic and metaphysical heuristic designed to neutralize neuro-centrism and biocentric chauvinism. It asks theorists to perform a cognitive “veil of ignorance”: if an external computational or storage process performs the precise functional task that we would unhesitatingly categorize as cognitive were it observed running on a biological neural circuit, we are philosophically obligated to categorize the external component as part of the cognitive system as well.

To illustrate the application of the Parity Principle, Clark and Chalmers presented their canonical thought experiment comparing two individuals: Inga and Otto.

• Inga: A biologically typical human who hears of an exhibition at the Museum of Modern Art (MoMA). She consults her biological memory, recalls that MoMA is located on 53rd Street, and walks to the museum. Her belief about the museum’s location was stored in her biological memory as a non-occurrent, dispositional belief prior to being retrieved into conscious awareness.
• Otto: A man suffering from an advanced stage of Alzheimer’s disease who relies upon a physical notebook that he carries everywhere. In this notebook, Otto carefully writes down new information and consults previously logged entries whenever needed. When Otto hears about the exhibition at MoMA, he consults his notebook, finds the entry stating that MoMA is on 53rd Street, and walks to the museum.

Clark and Chalmers argue that Otto’s notebook plays the exact same functional role for Otto that biological memory plays for Inga. Prior to consulting his notebook, Otto possessed the dispositional belief that MoMA is on 53rd Street, precisely as Inga did. The fact that Otto’s information is written in ink on cellulose paper outside his biological skull, whereas Inga’s is encoded in synaptic weights within her hippocampus and cerebral cortex, is an arbitrary geographic detail. Under the Parity Principle, Otto’s notebook is not an external cognitive prop; it is a physical constituent of his extended memory system.

2.2 Extended Functionalism and Cognitive Artifacts

The Extended Mind Thesis is deeply grounded in the logic of analytic functionalism. Classical functionalism, developed by philosophers such as Hilary Putnam, argues that mental states (such as beliefs, desires, pain, and perceptual states) are defined not by their internal physical realization (e.g., carbon-based neurons vs. silicon microchips), but by their functional and causal roles within a wider system. Mental states are defined by their causal relations to sensory inputs, their systemic relations to other mental states, and their causal contributions to behavioral outputs. This principle—known as multiple realizability—was originally deployed to demonstrate that artificial intelligences or extraterrestrial organisms with non-biological hardware could possess authentic mental states.

Clark and Chalmers took the multiple realizability thesis to its logical, spatial extreme. If mental states and cognitive operations are multiply realizable, there is no principled philosophical reason why their realization must be confined within the biological skin of the organism. Extended functionalism asserts that as long as an environmental artifact is causally wired into an agent’s cognitive loops in the appropriate functional manner, that artifact serves as a physical realization base for the agent’s mental states. The artifact is not merely an external tool; it is transformed into a cognitive artifact that performs active epistemic operations.

To conceptualize how humans exploit these environmental resources, Clark introduced the distinction between pragmatic action and epistemic action, drawing directly from the work of cognitive scientists David Kirsh and Paul Maglio. A pragmatic action is an action undertaken purely to achieve physical alterations in the environment to bring about a desired state of affairs (such as moving a chess piece to make a move). An epistemic action, by contrast, is a physical alteration of the environment executed specifically to simplify, accelerate, or facilitate a cognitive computation. Classic examples include:

• Rotating shapes physically in the video game Tetris to determine whether they fit into a slot, rather than rotating them via internal mental imagery.
• Rearranging letter tiles on a Scrabble rack to facilitate lexical retrieval and anagram solving.
• Using a pen and paper to perform long division, where the external symbols serve as active memory registers that hold intermediate computational results.

To guard against the absurd consequence that any tool or external library could be arbitrarily designated as part of someone’s mind, Clark and Chalmers established rigorous boundary conditions, colloquially known as the glue-and-trust criteria. For an external resource to count as a constitutive component of an agent’s extended cognitive architecture, it must satisfy four functional constraints:

1. Constant Availability: The external resource must be reliably and constantly available to the agent whenever it is required.
2. Direct Endorsement: Information retrieved from the external resource must be automatically and unreflectively endorsed by the agent as trustworthy, carrying the same epistemic authority as biological memory.
3. Easy Accessibility: The information contained within the resource must be readily accessible without complex, prohibitive retrieval operations.
4. Past Conscious Endorsement: The information in the resource must have been consciously endorsed by the agent in the past (though this criterion was later relaxed by Clark in subsequent writings).

When an external system fulfills these criteria, it becomes tightly coupled with the agent, functioning as a seamless extension of the agent’s cognitive architecture.

2.3 Evolution Across Theoretical Waves

Since its 1998 debut, the Extended Mind Thesis has undergone significant conceptual refinement, evolving through three distinct “waves” of philosophical development characterized by increasingly sophisticated accounts of artifactual distribution.

First-Wave Extended Mind (The Parity Wave):
The first wave was defined by Clark and Chalmers’ original formulations, anchored directly to the Parity Principle. First-wave HEC relied heavily on functional isomorphism and computational similarity between internal and external processes. It operated by demonstrating that an external artifact could mimic or match the functional dynamics of internal biological systems (such as Otto’s notebook mimicking Inga’s biological memory). While rhetorically powerful, first-wave HEC was criticized for being excessively neurocentric. By measuring the cognitive status of external tools against the benchmark of whether they replicate what the brain already does, the parity principle paradoxically retained the biological brain as the gold standard of mentality, downplaying the unique, non-isomorphic contributions made by external technologies.

Second-Wave Extended Mind (The Complementarity Wave):
In response to these criticisms, theorists such as John Sutton, Richard Menary, and Andy Clark himself articulated second-wave HEC, shifting the conceptual foundation from parity to complementarity. Second-wave HEC does not require an external artifact to function like a biological brain system. Instead, it recognizes that external artifacts are cognitively valuable precisely because they possess functional, structural, and computational properties that biological wetware lacks. Biological brains excel at pattern recognition, rapid associative retrieval, emotional valuation, and sensorimotor coordination, but are notoriously deficient at persistent data retention, precise mathematical calculations, and error-free serial symbolic processing. Second-wave HEC argues that true cognitive extension occurs when biological capacities are integrated with radically non-biological, complementary external formats (such as formal mathematical notation, digital databases, calendars, and spatial diagrams). The cognitive system is an asymmetric, hybrid assembly where heterogeneous biological and artifactual components cooperate dynamically to complete complex cognitive tasks that neither could achieve in isolation.

Third-Wave Extended Mind (The Distributed and Institutional Wave):
Pushed forward by theorists like Edwin Hutchins, Shaun Gallagher, and Lambros Malafouris, third-wave HEC moves past isolated individual agents coupled to specific tools, conceptualizing cognition as thoroughly distributed across socio-cultural scaffolding, socio-technical practices, and “cognitive institutions.” In this paradigm, cognition extends into legal systems, scientific laboratories, financial markets, and cultural traditions. A courtroom, complete with its codified legal statutes, formal procedural rules, evidentiary exhibits, and specialized social roles (judge, jury, prosecution, defense), functions as an extended cognitive system for determining societal culpability. Third-wave HEC converges closely with Material Engagement Theory, asserting that human mentality has always been co-constituted by material engagement with cultural artifacts, from prehistoric stone tools and cave art to contemporary digital networks.

3. Foundational Overview of Enactive Cognition (Enactivism)

3.1 Autopoiesis, Autonomy, and Sense-Making

Enactive Cognition diverges radically from the computational and functional paradigms of philosophy of mind. Rather than viewing cognition as an abstract computational function that can be arbitrarily mapped across biological or non-biological systems, enactivism grounds cognitive phenomena directly in the biological organization of living matter. The foundation of enactivism is rooted in the concept of autopoiesis (literally “self-creation” or “self-production”), formulated by Chilean biologists and cyberneticians Humberto Maturana and Francisco Varela.

An autopoietic system is an autonomous, self-generating network of biochemical processes that continuously produces and regenerates the very components that constitute the network, while simultaneously generating a bounded physical membrane (such as a cellular lipid bilayer) that delimits the system from its surrounding medium. This structural organization is defined by operational closure: the system’s processes recursively depend upon one another for their ongoing regeneration. Unlike machines or computers, which are organized allopoietically (producing something other than themselves, such as a factory producing an automobile or a program generating an output), an autopoietic system’s primary and continuous product is itself.

From this foundational biological ontology, enactivism derives its core tenets:

• Biological Precariousness: Living systems are fundamentally precarious. They exist far from thermodynamic equilibrium; their component processes are transient and will spontaneously decay, dissolve, or perish unless actively sustained by continuous metabolic work. Precariousness imparts an intrinsic existential vulnerability to the organism.
• Intrinsic Normativity: Because the organism is precarious and must actively maintain its own organization, encounters with the environment are never neutral. Environmental conditions are inherently significant, good or bad, dangerous or beneficial, depending on whether they support or threaten the organism’s autopoietic continuation.
• Sense-Making: Sense-making is the process whereby an organism, through its precarious self-maintenance, transforms an indifferent physical world into an environment of ecological significance and meaningful affordances. For a bacterium swimming in a medium, a chemical gradient is not an abstract string of computational data; it is enacted as “food” (to be approached) or “toxin” (to be avoided). The organism does not extract pre-existing meaning from the environment; it actively casts a web of significance over the world.
• The Life-Mind Continuity Thesis: Enactivism explicitly defends the view that life and mind share an identical core organizational logic. Cognition is not an advanced software routine that evolution superimposed onto sophisticated biological computers; rather, cognitive activity is co-extensive with life itself. To be alive is to engage in sense-making, and sense-making is the ancestral root of all cognition. Mind is life in the process of ongoing self-individuation.

3.2 Sensorimotor and Radical Enactive Formulations

Within the broader enactive church, two major theoretical developments have heavily influenced the discourse of cognitive science: Sensorimotor Enactivism and Radical Enactivism.

Sensorimotor Enactivism:
Formulated by J. Kevin O’Regan and Alva Noë in their landmark 2001 paper “A sensorimotor account of vision and visual consciousness”, this branch focuses heavily on perceptual experience. O’Regan and Noë challenged the classical view that visual perception consists in the brain generating an internal, rich, three-dimensional neural representation or “internal movie” of the external scene. Instead, they argued that perception is an exploratory physical activity: seeing is not something that happens in us; it is something we do. Perceptual experience is constituted by our practical mastery and implicit bodily knowledge of sensorimotor contingencies—the lawful, predictable patterns of change that occur in sensory stimulation as a direct result of bodily movement.

For example, when an agent tilts their head, shifts their gaze, or approaches an object, the visual inputs alter according to precise optical and geometric laws. When an agent grasps a spherical object, the tactile sensations shift in direct coordination with finger movements. To perceive an object as a three-dimensional sponge is not to reconstruct an internal model of a sponge in the occipital cortex; it is to possess practical bodily readiness to explore its texture, to know how its sensory profile will transform if squeezed, turned, or brought closer. The world is not represented internally; the world acts as its own external memory storage, accessed dynamically through ongoing sensorimotor interrogation.

Radical Enactive Cognition (REC):
Pioneered by Daniel Hutto and Erik Myin in their influential works Radicalizing Enactivism (2013) and Evolving Enactivism (2017), REC takes aim at the enduring representationalism lingering in cognitive science. Hutto and Myin propose the Hard Problem of Content: how could physical systems, operating strictly via natural laws, non-linear dynamics, and metabolic causal chains, generate mental representations that possess intrinsic semantic content—representations that are “about” something and possess truth-evaluable or accuracy conditions?

REC solves this problem by denying that basic cognition involves informational content, mental representations, or truth conditions whatsoever. While traditional cognitivists and extended mind theorists treat information as a substance that is encoded, processed, transferred, and stored, REC demonstrates that natural information is merely covariation (e.g., tree rings covarying with age, smoke covarying with fire). Covariation contains no semantic truth conditions; smoke does not “represent” fire, nor can tree rings be “false.” Hutto and Myin argue that basic cognition—such as perception, emotional response, athletic coordination, and dynamic navigation—consists entirely of contentless, direct, relational engagements with environmental affordances. Basic minds are not informational processors; they are dynamic, responsive bodies actively attuned to their worlds without representational intermediaries.

3.3 Participatory Sense-Making and Inter-Subjectivity

A frequent critique directed at early autopoietic theory was its solipsistic tendency: if an autopoietic system is defined by operational closure, how does it navigate meaningful, non-reductive social interactions? To address this challenge, enactive theorists Hanne De Jaegher and Ezequiel Di Paolo formulated the framework of participatory sense-making, extending enactive principles seamlessly into the realm of social cognition and intersubjectivity.

Mainstream cognitive science approached social cognition through either “Theory-Theory” (the agent uses an internal folk-psychological theoretical model to compute and infer the mental states of others) or “Simulation Theory” (the agent uses their own internal cognitive mechanisms to mentally simulate the experiences of others). Both paradigms are staunchly internalist, treating intersubjectivity as a spectator sport where isolated brains make detached computational guesses about hidden internal minds.

Participatory sense-making dismantles this spectator paradigm. De Jaegher and Di Paolo define social interaction as a dynamic, coupled process between two or more autonomous agents, wherein the coordination dynamics of the interaction itself generate emergent properties that cannot be reduced to the psychological states or intentions of the individual participants. When two people engage in conversation, dance, or negotiate a crowded sidewalk, their bodily movements, vocal cadences, facial expressions, and postural shifts become coupled through mutual sensorimotor feedback. This interaction can acquire its own operational momentum, sustaining itself even across misunderstandings or conversational shifts.

Within this interactive loop, individual sense-making is fundamentally altered: agents do not simply read each other’s pre-formed internal intentions; rather, new meanings, understandings, and joint intentions are enacted in and through the coordinated social dynamic itself. Intersubjectivity is thus experienced not as theoretical inference, but as lived, bodily, participatory synchronization—a shared dance of mutual affordance exploration without centralized representational modeling.

4. Key Differences Between Extended Mind and Enactive Cognition

While both the Extended Mind Thesis and Enactive Cognition reject cranial internalism, their divergence is absolute across nearly every core metaphysical and epistemological axis. The following comparative matrix outlines these primary dimensions of contrast:

Dimension / Axis	Extended Mind Thesis (HEC)	Enactive Cognition (Enactivism)
Metaphysical Foundation	Computational and Analytic Functionalism; Multiple Realizability; Active Causal Externalism.	Theoretical Biology; Autopoiesis; Operational Closure; Phenomenological Existentialism.
Definition of Mind	An integrated computational system executing functional tasks across biological and non-biological elements.	An autonomous, precarious living process actively bringing forth a meaningful world via sense-making.
Status of Representation	Representationalist and Informational; external artifacts store, manipulate, and encode semantic content.	Radically Anti-Representationalist for basic cognition; denies that minds operate via semantic internal models.
Role of the Body	Instrumental and contingent; biological wetware is simply one arbitrary medium capable of realizing cognitive functions.	Constitutive and irreplaceable; the physical morphology, precarious metabolism, and sensorimotor loops constitute mind.
Nature of Information	Objective, symbolic, or syntactic data that can be offloaded, stored, retrieved, and transmitted across media.	Relational, non-symbolic, enacted significance generated directly through the organism’s normative self-maintenance.
Boundary Criterion	Parity Principle and Complementarity; functional coupling satisfying glue-and-trust criteria.	Operational Closure and Autopoietic Autonomy; systemic boundary of the self-producing, precarious unity.
Locus of Agency	Centered on the human agent who instrumentalizes and co-opts external functional artifacts.	An emergent property of the self-individuating biological system striving to preserve its precarious existence.
Response to Coupling-Constitution	Emphasizes systemic functional integration, continuous feedback loops, and reciprocal causal coupling.	Rejects functional parity; relies on operational closure to establish intrinsic, non-arbitrary constitutive boundaries.
Treatment of Artifacts	Artifacts (e.g., notebooks, smartphones) literally constitute part of the physical machinery of mental states.	Artifacts act as dynamic scaffolds or incorporated bodily instruments; inert artifacts never constitute minds in themselves.

4.1 Metaphysical Foundations: Functionalism vs. Autopoietic Emergence

The primary metaphysical schism separating these paradigms concerns functionalism versus autopoietic emergence. The Extended Mind Thesis is fundamentally conservative with respect to the core ontology of twentieth-century philosophy of mind: it takes functionalism—the view that mental states are functional roles defined by input-output transitions and systemic causal networks—and simply liberalizes its geographic boundary. For HEC, whether a component is composed of carbon, silicon, or paper and graphite is irrelevant. What matters is the functional role: if a component plays the functional role of memory retrieval, it is memory. This functionalism is thoroughly medium-independent. The material substrate is decoupled from the cognitive identity of the process.

Enactivism rejects functionalism and medium-independence entirely as an artificial, Cartesian abstraction. To an enactivist, you cannot divorce cognitive operations from the physical, biochemical, and thermodynamic realities of the living organism. Cognition is not an abstract computational software running on arbitrary hardware; it is an emergent biological manifestation of a self-producing system navigating its thermodynamic vulnerability. Biological systems are not indifferent to their material medium—their cognition is tied to the fact that their tissues, cells, and organs must continuously consume energy and regenerate their own structures to forestall death.

Because HEC is grounded in computational functionalism, it easily extends cognitive states to passive physical tools (notebooks, smartwatches, hard drives). Enactivism, by contrast, restricts genuine cognitive agency to autonomous, self-individuating systems. A smartphone or a notebook can never be part of a cognitive system in the constitutive sense demanded by functional parity, because the notebook lacks operational closure, lacks precariousness, faces no existential vulnerability, and has no capacity for self-production. For the enactivist, the extended mind commits a profound category mistake: it mistakes the external tools used by a living cognitive agent for the living cognitive process itself.

4.2 The Status of Mental Representation and Information Processing

The second fundamental divergence lies in their handling of mental representation and the nature of information. Extended Mind theorists, including Clark and Chalmers, operate squarely within an informational and broadly representationalist architecture. While Clark frequently emphasizes non-linear dynamical systems and action-oriented representations, HEC fundamentally depends on the existence of portable, content-bearing representations. In the canonical Inga/Otto thought experiment, Otto’s notebook is designated as an extended belief because it stores symbolic, propositional representations (“MoMA is on 53rd Street”) that are retrieved, decoded, and translated into action. The entire architecture of HEC relies upon treating information as a quantifiable, communicable commodity that can reside internally in neural assemblies or externally in physical inscriptions.

Enactivists, especially Radical Enactivists such as Hutto and Myin, reject this computational-representational foundation entirely. Under Radical Enactive Cognition, the assumption that basic minds process, manipulate, or store internal representations is an untenable relic of Cartesianism, exacerbated by the computational metaphor of the twentieth century. Enactivism points out that representational theories of mind cannot solve the “Hard Problem of Content”: they cannot explain how neural states (or ink marks in a notebook) acquire intrinsic intentionality or semantic aboutness without appealing to an external interpretant or falling into homunculus fallacies.

Within enactive theory, information is not an objective substance residing in external storage banks or neural files. Information is relational, dynamical, and enacted. An organism does not “retrieve” a stored model of the world to act; it relies on its cultivated sensorimotor attunement to respond dynamically to environmental affordances. Where the Extended Mind theorist sees Otto retrieving an encoded symbolic representation from an external memory storage module, the enactivist sees a living human being interacting with a cultural artifact, executing an environmental action that scaffolds his current bodily orientation, without the notebook itself constituting a cognitive mind or possessing intrinsic mental states.

4.3 Boundaries, Coupling, and the Constitution Question

The debate over cognitive boundaries has generated one of the most contentious controversies in contemporary philosophy of mind: the Coupling-Constitution Fallacy, an objection originally formulated by internalist philosophers Fred Adams and Kenneth Aizawa (2001, 2008). Adams and Aizawa argue that extended mind theorists routinely commit a basic logical error: they conflate a tight, continuous causal coupling between an agent and an environmental tool with the metaphysical constitution of the cognitive agent itself.

To illustrate the fallacy, Adams and Aizawa use the analogy of a beating heart and its coupled environment: a human heart is causally coupled to the oxygen in the ambient air, the lungs, and the bloodstream; without oxygenated blood, the heart ceases to beat. Yet, no physiologist would assert that the ambient air or the lungs constitute part of the heart itself. Similarly, coupling an agent’s brain to a notebook, a smartphone, or a computer monitor does not make the artifact part of the mind; it simply means the mind is causally interacting with an external environmental object.

The Extended Mind Thesis and Enactive Cognition respond to this foundational challenge in radically contradictory ways:

• The Extended Mind Strategy: Extended mind theorists (like Clark) attempt to defeat the fallacy by demonstrating that the causal loops between agent and artifact are extraordinarily dense, non-linear, bidirectional, and temporally integrated. Clark argues that when external tools are bound to the agent through the glue-and-trust criteria, treating the biological boundary as an ontologically privileged cutoff is arbitrary neuro-chauvinism. The systemic whole acts as a unified functional unit that fulfills cognitive tasks.
• The Enactive Strategy: Enactivists side with the critics against HEC on this point, agreeing that HEC falls victim to the Coupling-Constitution Fallacy because it lacks an ontological criterion for where a system begins and ends. Because functionalism treats any causal relation that satisfies a task description as cognitive, HEC suffers from boundless cognitive bloat—potentially extending the mind across the entire World Wide Web.

Enactivism resolves the constitution question through operational closure and autopoietic autonomy. A system’s boundaries are not decided by arbitrary external task definitions or functional convenience; they are demarcated by the self-generated, topological boundary produced by the system’s own autonomous, precarious dynamics. The membrane of a cell, or the lived sensorimotor body of an animal, defines the cognitive agent because that boundary is actively generated and preserved by the system itself. External tools, no matter how tightly coupled, are not part of the operational closure of the autopoietic organism. Thus, enactivism avoids cognitive bloat by grounding constitution in biological organization, completely rejecting the functional parity that underwrites the extended mind.

4.4 Teleology and the Locus of Agency

The final foundational divergence between the two paradigms concerns the origin of intentionality, teleology, and the locus of agency. In the Extended Mind framework, artifacts are fundamentally passive, allopoietic instruments. A notebook, a personal computer, a smartphone, or an algorithm has no intrinsic goals, no survival instincts, no existential dread, and no intrinsic intentionality. The intentionality of Otto’s notebook is entirely derivative: it was authored by Otto, is interpreted by Otto, and relies upon Otto’s biological brain to read, evaluate, and act upon its contents. HEC essentially treats agency as an exclusively human property, which then co-opts external instruments into its extended computational orbit.

Enactivism, in stark contrast, grounds agency in the intrinsic teleology of self-maintaining life. Because an autopoietic organism is precarious, it must actively maintain its identity against thermodynamic entropy. This survival imperative generates intrinsic normativity: things in the world are fundamentally meaningful, beneficial, or hazardous for the organism itself. Agency is not an added computational capacity; it is the capacity of an autonomous unity to regulate its interactions with the environment to preserve its ongoing existence.

Consequently, enactivists point out that extended artifacts possess no agency whatsoever:

A notebook never acts on its own behalf. It does not care whether Otto forgets his appointment. It has no world, no norms, and enacts no sense.

By absorbing inert, non-autonomous artifacts into the definition of “mind,” the Extended Mind Thesis strips cognition of its most vital and unique characteristic: the capacity for intrinsic, normative sense-making. In enactive terms, the mind is never composed of lifeless tools; rather, a living agent actively incorporates tools into its lived sensorimotor repertoire to extend its practical reach within an already enacted world.

5. Similarities and Theoretical Convergence

5.1 Shared Anti-Cartesian and Anti-Neurocentric Ontologies

Despite their deep metaphysical battles, the Extended Mind Thesis and Enactive Cognition are deeply united in their mission to overthrow Cartesian internalism and intracranial reductionism. Both frameworks stand shoulder-to-shoulder in rejecting the mind-brain identity theory—the reductionist claim that cognitive states are completely reducible to localized neural events firing inside the cerebral cortex. Both frameworks vehemently oppose the image of the human mind as a detached, disembodied spectator observing the world from behind the biological screen of the eyes.

For both camps, cognition is inherently an active, world-involving, relational process. The traditional input-computation-output pipeline of classical cognitive science is recognized by both HEC and enactivism as an artificial laboratory construct that completely fails to describe how organisms function in natural settings. Both paradigms argue that the environment is not a passive domain from which data is scraped and internally modeled, but an essential arena of direct interaction. By taking cognitive science out of the head, both frameworks have paved the way for holistic, ecologically valid approaches to understanding the mind, forcing cognitive psychology, neurobiology, and philosophy to reckon with the situated reality of cognitive life.

5.2 Methodological Parallels and Dynamical Systems Theory

At the level of formal modeling and scientific methodology, both the Extended Mind Thesis and Enactive Cognition rely heavily upon Dynamical Systems Theory (DST). Rather than utilizing the discrete, serial, algorithmic flowcharts of classical cognitivism, both paradigms model cognitive phenomena using differential equations, continuous state spaces, attractors, trajectories, and coupled variables.

In classical cognitive modeling, an agent receives a discrete symbolic input $x$, computes a transformation $f(x)$ through internal representational registers, and outputs an action $y$. In dynamic systems modeling, the agent and the environment are represented as two coupled, non-linear dynamic systems whose variables mutually and continuously determine each other’s rate of change over time:

• The organism’s state variables $\mathbf{x}_O$ evolve as a function of their own internal dynamics and environmental parameters: $\dot{\mathbf{x}}_O = f(\mathbf{x}_O, \mathbf{x}_E)$.
• Concurrently, the environmental variables $\mathbf{x}_E$ evolve as a function of ecological dynamics and the organism’s bodily movements: $\dot{\mathbf{x}}_E = g(\mathbf{x}_E, \mathbf{x}_O)$.

Through this continuous feedback loop, the boundary between “input” and “output” collapses entirely. Action and perception become an unbroken, co-evolving trajectory through state space. Both paradigms embrace this mathematical formalism to articulate real-time situational action loops, finding empirical inspiration in the ecological psychology of J.J. Gibson, particularly the concept of affordances—action possibilities directly offered by the environment to an agent with specific bodily capacities.

5.3 Emergent Syntheses: Enactive Extended Mind

The philosophical gridlock between HEC and Enactivism has catalyzed creative attempts at theoretical synthesis, resulting in what several philosophers term the Enactive Extended Mind. Spearheaded by theorists such as Ezequiel Di Paolo (2009), Julian Kiverstein, and Mirko Farina, this syncretic movement seeks to preserve the enactive commitment to autopoiesis and sense-making while simultaneously accommodating the genuine material incorporation of external artifacts into cognitive architectures.

To achieve this synthesis, Di Paolo introduces the concept of habits as self-individuating, autonomous dynamic systems. Just as an autopoietic biological organism is composed of self-producing biochemical processes, human psychology is composed of self-sustaining, precarious networks of sensorimotor habits. Habits sustain their own recurrence through continuous enactments in the world; if an agent ceases to exercise a habit, that habit degrades and dissolves. Under this formulation, when an external artifact—such as a blind person’s white cane, an artist’s paintbrush, or a digital smartphone—becomes integrated into the sensorimotor loop of a habit, the artifact is no longer viewed as a passive, external functional data register (as first-wave HEC claimed). Instead, it is actively incorporated into the operational closure of the living habit itself.

This enactive extension discards the crude Parity Principle entirely. Extension is no longer understood as an external tool mimicking an internal brain state. Rather, cognitive extension is reframed as the dynamic incorporation of environmental structures into the ongoing, precarious sense-making of the living agent. The artifact participates directly in the lived body schema, expanding the organism’s horizon of affordances and transforming the landscape of possible sense-making.

6. Practical Applications and Framework Selection

6.1 Human-Computer Interaction (HCI) and Digital Augmentation

The choice between the Extended Mind Thesis and Enactive Cognition provides radically diverging theoretical blueprints for designers and software engineers working in Human-Computer Interaction (HCI), digital augmentation, and wearable computing.

The Extended Mind framework serves as the natural theoretical backbone for designers seeking to optimize cognitive efficiency, informational retrieval, and externalized data management. Because HEC views the smartphone or digital interface as a literal extension of the user’s functional architecture, design under this paradigm prioritizes:

• Minimizing interface friction to satisfy the glue-and-trust criteria.
• Designing transparent data structures that emulate the seamless availability of biological memory.
• Enabling intuitive digital offloading, allowing users to externalize complex serial tasks (e.g., GPS navigation, algorithmic scheduling, externalized knowledge graphs) to free up biological processing bandwidth.

Enactive HCI, by contrast, rejects the notion that computers should merely act as passive data storage vaults or external informational processors. Rooted in sensorimotor contingency theory, enactive interaction design prioritizes tangible computing, motor resonance, and rich sensorimotor feedback. Enactive design focuses on how physical engagement with a device transforms bodily experience, prioritizing haptic interfaces, expressive micro-gestures, and spatial computing environments that require active physical attunement. Under an enactive approach, a digital interface is successful not when it seamlessly feeds data into an agent’s cognitive pipeline, but when it enriches the user’s sensorimotor mastery, allowing novel forms of exploratory sense-making to emerge through embodied interaction.

6.2 Robotics, Artificial Intelligence, and Autonomous Agents

In the fields of robotics and artificial intelligence, the divide between HEC and Enactivism delineates two fundamentally incompatible architectural strategies for building artificial agents.

Extended Robotics:
Builds modular, distributed, functional systems that offload computational complexity into the physical environment through passive tags, structured infrastructure, and cloud computing. Extended robotic systems do not need to understand their environment holistically; they navigate and complete specialized tasks by distributing computational subroutines across external beacons, RFID systems, and coordinated multi-agent databases. The environment is engineered as an active, external component of the robot’s algorithmic processing loop.

Enactive Robotics:
Rooted in the seminal work of Rodney Brooks, Francisco Varela, and continuous cybernetic systems, enactive robotics rejects modular computational subroutines entirely. Enactive robotic architectures are designed around biological principles of homeostatic regulation, adaptive autonomy, and precarious self-maintenance. An enactive robot does not run functional software designed to optimize an external programmer’s task. Instead, it is equipped with simulated or physical metabolic processes that require it to seek out energetic resources, avoid environmental dangers, and dynamically adapt its sensorimotor control loops to ensure its own systemic continuity. Enactive robotics demonstrates that true autonomous agency, environmental meaning, and behavioral flexibility emerge only when an artificial agent possesses an existential stake in its own survival.

6.3 Psychiatry, Psychotherapy, and Neurodiversity

The therapeutic and psychiatric implications of these two post-Cartesian frameworks offer distinct, powerful tools for diagnosing and treating mental health conditions and conceptualizing neurodiversity.

Extended Psychiatry:
Examines mental disorders through the lens of broken, dysfunctional, or compromised external scaffolds. Viewed through HEC, conditions such as Alzheimer’s disease, severe ADHD, or executive dysfunction are not simply neurochemical defects confined within the brain tissue. They represent breakdowns in the agent-environment system. If an individual with Alzheimer’s loses their notebook, or an individual with ADHD is stripped of their smartphone timers, structured workspaces, and organizational apps, their cognitive impairment drastically increases. Therapeutic interventions grounded in HEC focus heavily on ecological restructuring: engineering resilient external environments, stabilizing domestic spaces, implementing digital prosthetic systems, and building socio-technical scaffolds that support and extend the individual’s compromised biological capacities.

Enactive Psychiatry:
Pioneered by philosophers and psychiatrists like Sanneke de Haan and Thomas Fuchs, enactive psychiatry conceptualizes mental illness not as an internal brain disease or an informational failure, but as a systematic perturbation of sense-making, bodily resonance, and participatory coordination. Depression, for instance, is diagnosed not merely as a serotonin deficit, but as a profound alteration of bodily attunement: the world loses its affordances; objects appear distant, lifeless, and devoid of practical allure; and the patient’s lived body becomes heavy, rigid, and resistant to environmental interaction.

Schizophrenia, under an enactive lens, is understood as a fundamental disruption of the basic sense of bodily selfhood (iper-reflexivity and diminished self-affection), while autism is conceptualized within participatory sense-making as a divergence in sensorimotor coordination styles that alters how intersubjective interactions are lived and negotiated. Enactive psychotherapy therefore shifts the clinical focus away from purely chemical manipulation or abstract cognitive reframing, targeting instead somatic attunement, embodied therapy, movement rehabilitation, and interactive participatory dynamics.

7. Frequently Asked Questions

7.1 Is the Extended Mind Thesis inherently representationalist?

In its standard, classical formulations, the Extended Mind Thesis is fundamentally representationalist. The original arguments presented by Andy Clark and David Chalmers rely upon external artifacts holding and manipulating informational content that possesses semantic, truth-evaluable conditions. In their canonical thought experiment, Otto’s notebook is designated as an extended cognitive component precisely because it stores non-occurrent, dispositional beliefs expressed in propositional, symbolic sentences (“The Museum of Modern Art is on 53rd Street”). The entire architecture of first-wave and much of second-wave HEC depends on the concept that internal mental representations can be seamlessly complemented or substituted by external, artifactual representations.

However, it is worth noting that some extended mind theorists have attempted to formulate non-representational varieties of cognitive extension by integrating aspects of ecological psychology and dynamical systems theory. Andy Clark himself has often defended the concept of “action-oriented representations,” which sit ambiguously between classical symbolic representations and pure sensorimotor engagement. Nevertheless, whenever HEC insists that an artifact constitutes a mental state (such as a belief, knowledge, or memory), it inevitably imports representational assumptions. This stands in stark, irreconcilable contrast to Radical Enactivism, which adamantly insists that all basic cognitive operations are entirely contentless and non-representational.

7.2 Can an enactivist acknowledge cognitive extension into tools?

Yes, an enactivist can readily acknowledge that tools and technologies alter, enrich, and restructure human cognition, but they reject the philosophical mechanism of “extension” as formulated by HEC. Enactivists do not believe that tools constitute cognitive systems through functional parity or computational substitution. An enactivist will never agree that an inert notebook or a passive hard drive is literally “a mind” or a “constitutive part of a memory state.”

Instead, enactivists explain tool use through the phenomenological concepts of bodily incorporation and sensorimotor augmentation. Drawing on Maurice Merleau-Ponty’s famous example of the blind person’s cane, an enactivist notes that with practice, the cane ceases to be an external object perceived by the hands; instead, it is incorporated into the agent’s lived body schema. The cane becomes a transparent medium through which the agent actively perceives the tactile affordances of the sidewalk. The cane does not store beliefs; rather, it extends the agent’s exploratory sensorimotor reach. Thus, for an enactivist, tools transform the agent’s dynamic capacities for sense-making, but the cognition remains firmly rooted in the living, bodily agent wielding the tool within an enacted environment.

7.3 How do both paradigms resolve the Coupling-Constitution Fallacy?

The Coupling-Constitution Fallacy, raised forcefully by critics such as Fred Adams and Kenneth Aizawa, asserts that showing an external object is causally coupled to an agent does not prove that the object is a metaphysical constituent of the agent’s mind. The Extended Mind Thesis and Enactive Cognition resolve this critique using completely different strategies:

The Extended Mind Thesis responds by denying that the boundary between the internal and external components is arbitrary, defending its position via systemic functionalism. Clark argues that when causal interactions between an agent and an external artifact exhibit dense, continuous, bidirectional, and non-linear interdependence—and when the artifact satisfies the strict glue-and-trust criteria—the entire integrated system acts as a single, unified cognitive engine. To demand that cognition reside solely in the biological wetware simply because it is biological is, according to HEC, an unprincipled form of biocentric chauvinism.

Enactive Cognition, conversely, concedes the validity of the critique against HEC, agreeing that mere functional coupling does not equal constitution. Enactivism avoids the fallacy by grounding cognitive constitution in the rigorous biological concept of operational closure. An entity is a cognitive system if and only if it generates and maintains its own physical and operational identity through autonomous, self-producing dynamics (autopoiesis). Because an autonomous system produces its own non-arbitrary boundary (such as the living body), enactivism possesses a natural, non-arbitrary principle of individuation that prevents the mind from bleeding indiscriminately into every casually coupled tool.

8. Conclusion: Navigating the Topography of 4E Cognition

8.1 Synthesis of Core Distinctions

The intellectual journey across the landscape of post-Cartesian cognitive science reveals that the Extended Mind Thesis and Enactive Cognition, despite their shared anti-Cartesian impulses, represent two radically divergent visions of what it means to possess a mind. Their points of divergence are profound, categorical, and metaphysical:

• The Extended Mind Thesis is the heroic culmination of analytic functionalism. It preserves the classic notions of computation, representation, information processing, and symbolic offloading, courageously pushing them beyond the biological boundary of the skull. It demonstrates that our technologies, artifacts, notebooks, and digital networks are not merely passive tools we observe, but active computational machinery structurally integrated into our cognitive operations. Mind, for HEC, is an open-ended computational network realized wherever functional tasks are efficiently executed.
• Enactive Cognition represents a profound ontological departure from the computational metaphor entirely. Grounded in theoretical biology, thermodynamics, and phenomenology, enactivism defines cognition as sense-making: the intrinsic, normative, precarious activity of a living organism continually striving to preserve its autonomous existence in an indifferent physical world. For the enactivist, mind is not an abstract computational process running indifferently across brains, paper, and silicon; mind is life itself in the ongoing, embodied act of bringing forth meaning.

The fundamental division between these frameworks ultimately reduces to a single question: Is cognition fundamentally about solving functional information-processing tasks, or is it fundamentally about the precarious biological living of an embodied agent enacting meaning? If one adopts the former view, the Extended Mind Thesis offers an unmatched, logically rigorous architecture. If one adopts the latter, Enactivism provides an ontologically deep, biologically grounded, and phenomenologically rich paradigm that restores the living body to the absolute center of cognitive science.

8.2 The Future of Post-Cartesian Philosophy of Mind

As cognitive science accelerates into the twenty-first century, the friction between the Extended Mind Thesis and Enactivism is generating fertile new theoretical territories. Nowhere is this more apparent than in the rapidly evolving dialogue between these frameworks and Predictive Processing / Active Inference, formulated by neuroscientists and philosophers such as Karl Friston, Andy Clark, and Jakob Hohwy.

Active inference posits that biological brains are fundamentally prediction engines, striving continuously to minimize variational free energy (a mathematical proxy for surprise and entropy) by generating top-down predictions of sensory inputs and acting on the world to fulfill those predictions. Theorists are currently divided over whether active inference validates an internalist, extended, or enactive view of the mind:

• Some, like Andy Clark, argue that active inference provides the ultimate computational mechanism for the Extended Mind, demonstrating how biological brains seamlessly incorporate external tools and cultural scaffolds to minimize long-term prediction errors.
• Others, like Ezequiel Di Paolo, Shaun Gallagher, and Sanneke de Haan, argue that active inference is fundamentally Enactive, as free-energy minimization is mathematically isomorphic to autopoietic self-maintenance, describing how precarious organisms preserve their homeostatic integrity through active worldly engagement.

Ultimately, the choice between the Extended Mind Thesis and Enactive Cognition is not a matter of crowning a singular victor; it is a matter of matching the appropriate philosophical framework to the specific explanatory scope of empirical inquiry. When analyzing distributed socio-technical systems, digital memory prosthetics, knowledge management, and human-computer interactions, the Extended Mind Thesis remains an indispensable, pragmatically powerful theoretical framework. When investigating the deep origins of agency, biological autonomy, emotional valence, perceptual consciousness, lived bodily experience, and psychiatric suffering, Enactive Cognition stands as the most profound, revolutionary paradigm of our time.

References

• Adams, F., & Aizawa, K. (2001). The bounds of cognition. Philosophical Psychology, 14(1), 43–64. https://doi.org/10.1080/09515080120033571
• Adams, F., & Aizawa, K. (2008). The Bounds of Cognition. Wiley-Blackwell.
• Clark, A. (2008). Supersizing the Mind: Embodiment, Action, and Cognitive Extension. Oxford University Press.
• Clark, A., & Chalmers, D. J. (1998). The extended mind. Analysis, 58(1), 7–19. https://doi.org/10.1093/analys/58.1.7
• De Haan, S. (2020). Enactive Psychiatry. Cambridge University Press. https://doi.org/10.1017/9781108685214
• De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making: An enactive approach to social cognition. Phenomenology and the Cognitive Sciences, 6(4), 485–507. https://doi.org/10.1007/s11097-007-9076-9
• Di Paolo, E. A. (2005). Autopoiesis, adaptivity, teleology, agency. Phenomenology and the Cognitive Sciences, 4(4), 429–452. https://doi.org/10.1007/s11097-005-9002-y
• Di Paolo, E. A. (2009). Extended life. Topoi, 28(1), 9–21. https://doi.org/10.1007/s11245-008-9042-3
• Fuchs, T. (2018). Ecology of the Brain: The Phenomenology and Biology of the Embodied Mind. Oxford University Press.
• Gallagher, S. (2017). Enactivist Interventions: Rethinking Mind. Oxford University Press.
• Hutchins, E. (1995). Cognition in the Wild. MIT Press.
• Hutto, D. D., & Myin, E. (2013). Radicalizing Enactivism: Basic Minds without Content. MIT Press.
• Hutto, D. D., & Myin, E. (2017). Evolving Enactivism: Basic Minds Meet Higher Content. MIT Press.
• Kirsh, D., & Maglio, P. (1994). On distinguishing epistemic from pragmatic action. Cognitive Science, 18(4), 513–549. https://doi.org/10.1207/s15516709cog1804_1
• Malafouris, L. (2013). How Things Shape the Mind: A Theory of Material Engagement. MIT Press.
• Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and Cognition: The Realization of the Living. D. Reidel Publishing Company. https://doi.org/10.1007/978-94-009-8947-4
• Menary, R. (Ed.). (2010). The Extended Mind. MIT Press.
• Noë, A. (2004). Action in Perception. MIT Press.
• O’Regan, J. K., & Noë, A. (2001). A sensorimotor account of vision and visual consciousness. Behavioral and Brain Sciences, 24(5), 939–973. https://doi.org/10.1017/s0140525x01000115
• Sutton, J. (2010). Exograms and interdisciplinarity: History, the extended mind, and the civilizing process. In R. Menary (Ed.), The Extended Mind (pp. 189–225). MIT Press.
• Thompson, E. (2007). Mind in Life: Biology, Phenomenology, and the Sciences of Mind. Harvard University Press.
• Varela, F. J., Thompson, E., & Rosch, E. (1991). The Embodied Mind: Cognitive Science and Human Experience. MIT Press.